{ "best_metric": 0.2568434178829193, "best_model_checkpoint": "./w2v-bert-2.0-hausa_100_400h/checkpoint-17000", "epoch": 15.460295151089248, "eval_steps": 1000, "global_step": 22000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0007027406886858749, "grad_norm": Infinity, "learning_rate": 0.0, "loss": 8.0003, "step": 1 }, { "epoch": 0.0014054813773717498, "grad_norm": 22.919769287109375, "learning_rate": 4.21644413211525e-09, "loss": 8.1116, "step": 2 }, { "epoch": 0.0021082220660576245, "grad_norm": 22.38802719116211, "learning_rate": 8.4328882642305e-09, "loss": 7.9554, "step": 3 }, { "epoch": 0.0028109627547434997, "grad_norm": 21.984024047851562, "learning_rate": 1.2649332396345749e-08, "loss": 7.8748, "step": 4 }, { "epoch": 0.0035137034434293743, "grad_norm": 22.405925750732422, "learning_rate": 1.6865776528461e-08, "loss": 7.9109, "step": 5 }, { "epoch": 0.004216444132115249, "grad_norm": 22.340869903564453, "learning_rate": 2.1082220660576247e-08, "loss": 7.8463, "step": 6 }, { "epoch": 0.004919184820801124, "grad_norm": 22.259456634521484, "learning_rate": 2.5298664792691498e-08, "loss": 7.8249, "step": 7 }, { "epoch": 0.005621925509486999, "grad_norm": 22.263946533203125, "learning_rate": 2.9515108924806746e-08, "loss": 7.8289, "step": 8 }, { "epoch": 0.006324666198172874, "grad_norm": 22.030807495117188, "learning_rate": 3.3731553056922e-08, "loss": 7.7524, "step": 9 }, { "epoch": 0.007027406886858749, "grad_norm": 22.278331756591797, "learning_rate": 3.7947997189037245e-08, "loss": 7.7915, "step": 10 }, { "epoch": 0.007730147575544624, "grad_norm": 22.038291931152344, "learning_rate": 4.2164441321152494e-08, "loss": 7.7311, "step": 11 }, { "epoch": 0.008432888264230498, "grad_norm": 22.195165634155273, "learning_rate": 4.638088545326775e-08, "loss": 7.7495, "step": 12 }, { "epoch": 0.009135628952916374, "grad_norm": 22.768667221069336, "learning_rate": 5.0597329585382996e-08, "loss": 7.8748, "step": 13 }, { "epoch": 0.009838369641602248, "grad_norm": 22.07448959350586, "learning_rate": 5.4813773717498244e-08, "loss": 7.6881, "step": 14 }, { "epoch": 0.010541110330288124, "grad_norm": 22.15041732788086, "learning_rate": 5.903021784961349e-08, "loss": 7.7006, "step": 15 }, { "epoch": 0.011243851018973999, "grad_norm": 21.80359649658203, "learning_rate": 6.324666198172873e-08, "loss": 7.6208, "step": 16 }, { "epoch": 0.011946591707659873, "grad_norm": 21.473072052001953, "learning_rate": 6.7463106113844e-08, "loss": 7.477, "step": 17 }, { "epoch": 0.012649332396345749, "grad_norm": 21.900619506835938, "learning_rate": 7.167955024595924e-08, "loss": 7.6062, "step": 18 }, { "epoch": 0.013352073085031623, "grad_norm": 22.025392532348633, "learning_rate": 7.589599437807449e-08, "loss": 7.5877, "step": 19 }, { "epoch": 0.014054813773717497, "grad_norm": 21.653242111206055, "learning_rate": 8.011243851018975e-08, "loss": 7.4625, "step": 20 }, { "epoch": 0.014757554462403373, "grad_norm": 21.541460037231445, "learning_rate": 8.432888264230499e-08, "loss": 7.5167, "step": 21 }, { "epoch": 0.015460295151089248, "grad_norm": 20.582595825195312, "learning_rate": 8.854532677442024e-08, "loss": 7.2581, "step": 22 }, { "epoch": 0.016163035839775124, "grad_norm": 19.82164764404297, "learning_rate": 9.27617709065355e-08, "loss": 7.0454, "step": 23 }, { "epoch": 0.016865776528460996, "grad_norm": 19.115880966186523, "learning_rate": 9.697821503865074e-08, "loss": 6.8229, "step": 24 }, { "epoch": 0.017568517217146872, "grad_norm": 19.196928024291992, "learning_rate": 1.0119465917076599e-07, "loss": 6.7085, "step": 25 }, { "epoch": 0.018271257905832748, "grad_norm": 22.31411361694336, "learning_rate": 1.0541110330288123e-07, "loss": 7.9851, "step": 26 }, { "epoch": 0.018973998594518624, "grad_norm": 22.894996643066406, "learning_rate": 1.0962754743499649e-07, "loss": 7.9619, "step": 27 }, { "epoch": 0.019676739283204497, "grad_norm": 22.653087615966797, "learning_rate": 1.1384399156711174e-07, "loss": 7.8654, "step": 28 }, { "epoch": 0.020379479971890373, "grad_norm": 22.68056297302246, "learning_rate": 1.1806043569922698e-07, "loss": 7.86, "step": 29 }, { "epoch": 0.02108222066057625, "grad_norm": 22.71889305114746, "learning_rate": 1.2227687983134223e-07, "loss": 7.8263, "step": 30 }, { "epoch": 0.02178496134926212, "grad_norm": 22.777402877807617, "learning_rate": 1.2649332396345747e-07, "loss": 7.791, "step": 31 }, { "epoch": 0.022487702037947997, "grad_norm": 23.078948974609375, "learning_rate": 1.3070976809557275e-07, "loss": 7.8163, "step": 32 }, { "epoch": 0.023190442726633873, "grad_norm": 22.728212356567383, "learning_rate": 1.34926212227688e-07, "loss": 7.7062, "step": 33 }, { "epoch": 0.023893183415319746, "grad_norm": 22.71955108642578, "learning_rate": 1.3914265635980322e-07, "loss": 7.7357, "step": 34 }, { "epoch": 0.024595924104005622, "grad_norm": 22.637060165405273, "learning_rate": 1.4335910049191849e-07, "loss": 7.6639, "step": 35 }, { "epoch": 0.025298664792691498, "grad_norm": 22.80999183654785, "learning_rate": 1.4757554462403372e-07, "loss": 7.663, "step": 36 }, { "epoch": 0.02600140548137737, "grad_norm": 22.76412010192871, "learning_rate": 1.5179198875614898e-07, "loss": 7.655, "step": 37 }, { "epoch": 0.026704146170063246, "grad_norm": 22.65781021118164, "learning_rate": 1.5600843288826424e-07, "loss": 7.6126, "step": 38 }, { "epoch": 0.027406886858749122, "grad_norm": 22.819217681884766, "learning_rate": 1.602248770203795e-07, "loss": 7.5513, "step": 39 }, { "epoch": 0.028109627547434995, "grad_norm": 22.34787940979004, "learning_rate": 1.6444132115249474e-07, "loss": 7.5054, "step": 40 }, { "epoch": 0.02881236823612087, "grad_norm": 22.360000610351562, "learning_rate": 1.6865776528460997e-07, "loss": 7.4627, "step": 41 }, { "epoch": 0.029515108924806747, "grad_norm": 22.289859771728516, "learning_rate": 1.7287420941672524e-07, "loss": 7.4175, "step": 42 }, { "epoch": 0.030217849613492623, "grad_norm": 22.60727882385254, "learning_rate": 1.7709065354884047e-07, "loss": 7.4269, "step": 43 }, { "epoch": 0.030920590302178495, "grad_norm": 22.18010711669922, "learning_rate": 1.8130709768095573e-07, "loss": 7.3605, "step": 44 }, { "epoch": 0.03162333099086437, "grad_norm": 22.806358337402344, "learning_rate": 1.85523541813071e-07, "loss": 7.4893, "step": 45 }, { "epoch": 0.03232607167955025, "grad_norm": 22.51970672607422, "learning_rate": 1.8973998594518623e-07, "loss": 7.3524, "step": 46 }, { "epoch": 0.03302881236823612, "grad_norm": 20.866682052612305, "learning_rate": 1.939564300773015e-07, "loss": 6.9901, "step": 47 }, { "epoch": 0.03373155305692199, "grad_norm": 20.13766098022461, "learning_rate": 1.9817287420941672e-07, "loss": 6.8399, "step": 48 }, { "epoch": 0.03443429374560787, "grad_norm": 19.587310791015625, "learning_rate": 2.0238931834153198e-07, "loss": 6.7016, "step": 49 }, { "epoch": 0.035137034434293744, "grad_norm": 18.478219985961914, "learning_rate": 2.0660576247364722e-07, "loss": 6.379, "step": 50 }, { "epoch": 0.035839775122979624, "grad_norm": 23.867103576660156, "learning_rate": 2.1082220660576245e-07, "loss": 7.9255, "step": 51 }, { "epoch": 0.036542515811665496, "grad_norm": 23.81924057006836, "learning_rate": 2.1503865073787774e-07, "loss": 7.7177, "step": 52 }, { "epoch": 0.03724525650035137, "grad_norm": 23.446256637573242, "learning_rate": 2.1925509486999298e-07, "loss": 7.5992, "step": 53 }, { "epoch": 0.03794799718903725, "grad_norm": 23.749704360961914, "learning_rate": 2.2347153900210824e-07, "loss": 7.6223, "step": 54 }, { "epoch": 0.03865073787772312, "grad_norm": 24.07349395751953, "learning_rate": 2.2768798313422347e-07, "loss": 7.634, "step": 55 }, { "epoch": 0.03935347856640899, "grad_norm": 23.98552131652832, "learning_rate": 2.319044272663387e-07, "loss": 7.5894, "step": 56 }, { "epoch": 0.04005621925509487, "grad_norm": 24.093767166137695, "learning_rate": 2.3612087139845397e-07, "loss": 7.5714, "step": 57 }, { "epoch": 0.040758959943780745, "grad_norm": 23.798551559448242, "learning_rate": 2.403373155305692e-07, "loss": 7.4369, "step": 58 }, { "epoch": 0.04146170063246662, "grad_norm": 24.05326271057129, "learning_rate": 2.4455375966268446e-07, "loss": 7.4634, "step": 59 }, { "epoch": 0.0421644413211525, "grad_norm": 23.97357940673828, "learning_rate": 2.487702037947997e-07, "loss": 7.4092, "step": 60 }, { "epoch": 0.04286718200983837, "grad_norm": 23.905834197998047, "learning_rate": 2.5298664792691493e-07, "loss": 7.3878, "step": 61 }, { "epoch": 0.04356992269852424, "grad_norm": 23.99234390258789, "learning_rate": 2.5720309205903025e-07, "loss": 7.3445, "step": 62 }, { "epoch": 0.04427266338721012, "grad_norm": 23.811290740966797, "learning_rate": 2.614195361911455e-07, "loss": 7.2697, "step": 63 }, { "epoch": 0.044975404075895994, "grad_norm": 24.165096282958984, "learning_rate": 2.656359803232607e-07, "loss": 7.3219, "step": 64 }, { "epoch": 0.04567814476458187, "grad_norm": 23.651212692260742, "learning_rate": 2.69852424455376e-07, "loss": 7.2143, "step": 65 }, { "epoch": 0.046380885453267746, "grad_norm": 24.055200576782227, "learning_rate": 2.7406886858749124e-07, "loss": 7.2247, "step": 66 }, { "epoch": 0.04708362614195362, "grad_norm": 24.115028381347656, "learning_rate": 2.7828531271960645e-07, "loss": 7.1572, "step": 67 }, { "epoch": 0.04778636683063949, "grad_norm": 23.602455139160156, "learning_rate": 2.825017568517217e-07, "loss": 7.063, "step": 68 }, { "epoch": 0.04848910751932537, "grad_norm": 24.072877883911133, "learning_rate": 2.8671820098383697e-07, "loss": 7.0995, "step": 69 }, { "epoch": 0.049191848208011243, "grad_norm": 23.683137893676758, "learning_rate": 2.9093464511595223e-07, "loss": 7.0151, "step": 70 }, { "epoch": 0.049894588896697116, "grad_norm": 23.171613693237305, "learning_rate": 2.9515108924806744e-07, "loss": 6.851, "step": 71 }, { "epoch": 0.050597329585382995, "grad_norm": 22.805517196655273, "learning_rate": 2.993675333801827e-07, "loss": 6.7633, "step": 72 }, { "epoch": 0.05130007027406887, "grad_norm": 21.827896118164062, "learning_rate": 3.0358397751229796e-07, "loss": 6.6072, "step": 73 }, { "epoch": 0.05200281096275474, "grad_norm": 20.811796188354492, "learning_rate": 3.0780042164441317e-07, "loss": 6.4031, "step": 74 }, { "epoch": 0.05270555165144062, "grad_norm": 20.080097198486328, "learning_rate": 3.120168657765285e-07, "loss": 6.2229, "step": 75 }, { "epoch": 0.05340829234012649, "grad_norm": 26.084402084350586, "learning_rate": 3.1623330990864375e-07, "loss": 7.4782, "step": 76 }, { "epoch": 0.054111033028812365, "grad_norm": 26.291929244995117, "learning_rate": 3.20449754040759e-07, "loss": 7.3807, "step": 77 }, { "epoch": 0.054813773717498245, "grad_norm": 26.255090713500977, "learning_rate": 3.246661981728742e-07, "loss": 7.2978, "step": 78 }, { "epoch": 0.05551651440618412, "grad_norm": 26.146202087402344, "learning_rate": 3.288826423049895e-07, "loss": 7.2333, "step": 79 }, { "epoch": 0.05621925509486999, "grad_norm": 26.073848724365234, "learning_rate": 3.3309908643710474e-07, "loss": 7.1494, "step": 80 }, { "epoch": 0.05692199578355587, "grad_norm": 26.309385299682617, "learning_rate": 3.3731553056921995e-07, "loss": 7.165, "step": 81 }, { "epoch": 0.05762473647224174, "grad_norm": 26.796436309814453, "learning_rate": 3.415319747013352e-07, "loss": 7.1492, "step": 82 }, { "epoch": 0.05832747716092762, "grad_norm": 26.332202911376953, "learning_rate": 3.4574841883345047e-07, "loss": 7.0785, "step": 83 }, { "epoch": 0.059030217849613494, "grad_norm": 26.579050064086914, "learning_rate": 3.499648629655657e-07, "loss": 7.0351, "step": 84 }, { "epoch": 0.059732958538299366, "grad_norm": 26.590696334838867, "learning_rate": 3.5418130709768094e-07, "loss": 6.982, "step": 85 }, { "epoch": 0.060435699226985246, "grad_norm": 26.50705909729004, "learning_rate": 3.583977512297962e-07, "loss": 6.88, "step": 86 }, { "epoch": 0.06113843991567112, "grad_norm": 27.41244125366211, "learning_rate": 3.6261419536191146e-07, "loss": 6.969, "step": 87 }, { "epoch": 0.06184118060435699, "grad_norm": 26.879751205444336, "learning_rate": 3.668306394940267e-07, "loss": 6.8475, "step": 88 }, { "epoch": 0.06254392129304287, "grad_norm": 26.739870071411133, "learning_rate": 3.71047083626142e-07, "loss": 6.7621, "step": 89 }, { "epoch": 0.06324666198172874, "grad_norm": 27.01898956298828, "learning_rate": 3.7526352775825725e-07, "loss": 6.7551, "step": 90 }, { "epoch": 0.06394940267041462, "grad_norm": 26.941919326782227, "learning_rate": 3.7947997189037245e-07, "loss": 6.7192, "step": 91 }, { "epoch": 0.0646521433591005, "grad_norm": 27.147714614868164, "learning_rate": 3.836964160224877e-07, "loss": 6.6641, "step": 92 }, { "epoch": 0.06535488404778636, "grad_norm": 27.087236404418945, "learning_rate": 3.87912860154603e-07, "loss": 6.6202, "step": 93 }, { "epoch": 0.06605762473647224, "grad_norm": 26.615001678466797, "learning_rate": 3.9212930428671824e-07, "loss": 6.5049, "step": 94 }, { "epoch": 0.06676036542515812, "grad_norm": 27.718456268310547, "learning_rate": 3.9634574841883345e-07, "loss": 6.5959, "step": 95 }, { "epoch": 0.06746310611384398, "grad_norm": 25.74755096435547, "learning_rate": 4.005621925509487e-07, "loss": 6.345, "step": 96 }, { "epoch": 0.06816584680252986, "grad_norm": 26.04132843017578, "learning_rate": 4.0477863668306397e-07, "loss": 6.3312, "step": 97 }, { "epoch": 0.06886858749121574, "grad_norm": 24.026260375976562, "learning_rate": 4.089950808151792e-07, "loss": 6.1032, "step": 98 }, { "epoch": 0.06957132817990162, "grad_norm": 22.9947566986084, "learning_rate": 4.1321152494729444e-07, "loss": 5.9496, "step": 99 }, { "epoch": 0.07027406886858749, "grad_norm": 21.277101516723633, "learning_rate": 4.174279690794097e-07, "loss": 5.6515, "step": 100 }, { "epoch": 0.07097680955727337, "grad_norm": 30.402116775512695, "learning_rate": 4.216444132115249e-07, "loss": 6.7769, "step": 101 }, { "epoch": 0.07167955024595925, "grad_norm": 31.20366096496582, "learning_rate": 4.258608573436402e-07, "loss": 6.6767, "step": 102 }, { "epoch": 0.07238229093464511, "grad_norm": Infinity, "learning_rate": 4.258608573436402e-07, "loss": 6.6667, "step": 103 }, { "epoch": 0.07308503162333099, "grad_norm": 30.808921813964844, "learning_rate": 4.300773014757555e-07, "loss": 6.575, "step": 104 }, { "epoch": 0.07378777231201687, "grad_norm": 31.7376766204834, "learning_rate": 4.3429374560787075e-07, "loss": 6.5848, "step": 105 }, { "epoch": 0.07449051300070274, "grad_norm": 31.723865509033203, "learning_rate": 4.3851018973998595e-07, "loss": 6.5085, "step": 106 }, { "epoch": 0.07519325368938862, "grad_norm": 30.89188575744629, "learning_rate": 4.427266338721012e-07, "loss": 6.3758, "step": 107 }, { "epoch": 0.0758959943780745, "grad_norm": 31.54078483581543, "learning_rate": 4.469430780042165e-07, "loss": 6.386, "step": 108 }, { "epoch": 0.07659873506676036, "grad_norm": 32.462955474853516, "learning_rate": 4.511595221363317e-07, "loss": 6.3895, "step": 109 }, { "epoch": 0.07730147575544624, "grad_norm": 32.329036712646484, "learning_rate": 4.5537596626844695e-07, "loss": 6.3135, "step": 110 }, { "epoch": 0.07800421644413212, "grad_norm": 31.76789093017578, "learning_rate": 4.595924104005622e-07, "loss": 6.1998, "step": 111 }, { "epoch": 0.07870695713281799, "grad_norm": 32.259071350097656, "learning_rate": 4.638088545326774e-07, "loss": 6.1739, "step": 112 }, { "epoch": 0.07940969782150387, "grad_norm": 32.11015701293945, "learning_rate": 4.680252986647927e-07, "loss": 6.1138, "step": 113 }, { "epoch": 0.08011243851018975, "grad_norm": 32.5121955871582, "learning_rate": 4.7224174279690794e-07, "loss": 6.0663, "step": 114 }, { "epoch": 0.08081517919887561, "grad_norm": 31.907299041748047, "learning_rate": 4.7645818692902325e-07, "loss": 5.9713, "step": 115 }, { "epoch": 0.08151791988756149, "grad_norm": 31.6020450592041, "learning_rate": 4.806746310611384e-07, "loss": 5.9082, "step": 116 }, { "epoch": 0.08222066057624737, "grad_norm": 32.872222900390625, "learning_rate": 4.848910751932538e-07, "loss": 5.9198, "step": 117 }, { "epoch": 0.08292340126493324, "grad_norm": 32.81608581542969, "learning_rate": 4.891075193253689e-07, "loss": 5.8999, "step": 118 }, { "epoch": 0.08362614195361912, "grad_norm": 33.003292083740234, "learning_rate": 4.933239634574842e-07, "loss": 5.8289, "step": 119 }, { "epoch": 0.084328882642305, "grad_norm": 32.16100311279297, "learning_rate": 4.975404075895995e-07, "loss": 5.7388, "step": 120 }, { "epoch": 0.08503162333099086, "grad_norm": 30.932737350463867, "learning_rate": 5.017568517217147e-07, "loss": 5.6582, "step": 121 }, { "epoch": 0.08573436401967674, "grad_norm": 29.414026260375977, "learning_rate": 5.059732958538299e-07, "loss": 5.5184, "step": 122 }, { "epoch": 0.08643710470836262, "grad_norm": 27.461894989013672, "learning_rate": 5.101897399859452e-07, "loss": 5.4031, "step": 123 }, { "epoch": 0.08713984539704848, "grad_norm": 25.7247371673584, "learning_rate": 5.144061841180605e-07, "loss": 5.2876, "step": 124 }, { "epoch": 0.08784258608573436, "grad_norm": 23.19124412536621, "learning_rate": 5.186226282501757e-07, "loss": 5.0634, "step": 125 }, { "epoch": 0.08854532677442024, "grad_norm": 35.40072250366211, "learning_rate": 5.22839072382291e-07, "loss": 5.7608, "step": 126 }, { "epoch": 0.08924806746310611, "grad_norm": 36.28236389160156, "learning_rate": 5.270555165144062e-07, "loss": 5.6489, "step": 127 }, { "epoch": 0.08995080815179199, "grad_norm": 36.10710144042969, "learning_rate": 5.312719606465214e-07, "loss": 5.5906, "step": 128 }, { "epoch": 0.09065354884047787, "grad_norm": 35.82858657836914, "learning_rate": 5.354884047786367e-07, "loss": 5.5152, "step": 129 }, { "epoch": 0.09135628952916373, "grad_norm": 35.915008544921875, "learning_rate": 5.39704848910752e-07, "loss": 5.4646, "step": 130 }, { "epoch": 0.09205903021784961, "grad_norm": 35.153663635253906, "learning_rate": 5.439212930428671e-07, "loss": 5.3623, "step": 131 }, { "epoch": 0.09276177090653549, "grad_norm": 34.89278030395508, "learning_rate": 5.481377371749825e-07, "loss": 5.3121, "step": 132 }, { "epoch": 0.09346451159522136, "grad_norm": 34.38637924194336, "learning_rate": 5.523541813070976e-07, "loss": 5.2354, "step": 133 }, { "epoch": 0.09416725228390724, "grad_norm": 34.49753189086914, "learning_rate": 5.565706254392129e-07, "loss": 5.1908, "step": 134 }, { "epoch": 0.09486999297259312, "grad_norm": 34.46569061279297, "learning_rate": 5.607870695713282e-07, "loss": 5.1307, "step": 135 }, { "epoch": 0.09557273366127898, "grad_norm": 33.18268585205078, "learning_rate": 5.650035137034434e-07, "loss": 5.0356, "step": 136 }, { "epoch": 0.09627547434996486, "grad_norm": 33.20500564575195, "learning_rate": 5.692199578355588e-07, "loss": 4.9891, "step": 137 }, { "epoch": 0.09697821503865074, "grad_norm": 32.49551773071289, "learning_rate": 5.734364019676739e-07, "loss": 4.9412, "step": 138 }, { "epoch": 0.09768095572733661, "grad_norm": 32.36682891845703, "learning_rate": 5.776528460997892e-07, "loss": 4.8834, "step": 139 }, { "epoch": 0.09838369641602249, "grad_norm": 31.85828971862793, "learning_rate": 5.818692902319045e-07, "loss": 4.8134, "step": 140 }, { "epoch": 0.09908643710470837, "grad_norm": 30.56289291381836, "learning_rate": 5.860857343640197e-07, "loss": 4.756, "step": 141 }, { "epoch": 0.09978917779339423, "grad_norm": 29.82435417175293, "learning_rate": 5.903021784961349e-07, "loss": 4.6747, "step": 142 }, { "epoch": 0.10049191848208011, "grad_norm": 29.03946304321289, "learning_rate": 5.945186226282502e-07, "loss": 4.6267, "step": 143 }, { "epoch": 0.10119465917076599, "grad_norm": 27.1345272064209, "learning_rate": 5.987350667603654e-07, "loss": 4.5513, "step": 144 }, { "epoch": 0.10189739985945186, "grad_norm": 27.128276824951172, "learning_rate": 6.029515108924807e-07, "loss": 4.5255, "step": 145 }, { "epoch": 0.10260014054813774, "grad_norm": 26.508808135986328, "learning_rate": 6.071679550245959e-07, "loss": 4.5033, "step": 146 }, { "epoch": 0.10330288123682362, "grad_norm": 23.233661651611328, "learning_rate": 6.113843991567112e-07, "loss": 4.421, "step": 147 }, { "epoch": 0.10400562192550948, "grad_norm": 22.066946029663086, "learning_rate": 6.156008432888263e-07, "loss": 4.4109, "step": 148 }, { "epoch": 0.10470836261419536, "grad_norm": 19.61279296875, "learning_rate": 6.198172874209417e-07, "loss": 4.3333, "step": 149 }, { "epoch": 0.10541110330288124, "grad_norm": 16.591697692871094, "learning_rate": 6.24033731553057e-07, "loss": 4.2538, "step": 150 }, { "epoch": 0.1061138439915671, "grad_norm": 25.406282424926758, "learning_rate": 6.282501756851721e-07, "loss": 4.3723, "step": 151 }, { "epoch": 0.10681658468025299, "grad_norm": 23.21747398376465, "learning_rate": 6.324666198172875e-07, "loss": 4.2608, "step": 152 }, { "epoch": 0.10751932536893886, "grad_norm": 21.72841453552246, "learning_rate": 6.366830639494027e-07, "loss": 4.1989, "step": 153 }, { "epoch": 0.10822206605762473, "grad_norm": 19.840099334716797, "learning_rate": 6.40899508081518e-07, "loss": 4.1418, "step": 154 }, { "epoch": 0.10892480674631061, "grad_norm": 18.264371871948242, "learning_rate": 6.451159522136332e-07, "loss": 4.0961, "step": 155 }, { "epoch": 0.10962754743499649, "grad_norm": 16.4364013671875, "learning_rate": 6.493323963457484e-07, "loss": 4.0471, "step": 156 }, { "epoch": 0.11033028812368235, "grad_norm": 15.090377807617188, "learning_rate": 6.535488404778637e-07, "loss": 4.0239, "step": 157 }, { "epoch": 0.11103302881236823, "grad_norm": 13.647170066833496, "learning_rate": 6.57765284609979e-07, "loss": 3.9726, "step": 158 }, { "epoch": 0.11173576950105411, "grad_norm": 12.84835147857666, "learning_rate": 6.619817287420941e-07, "loss": 3.9553, "step": 159 }, { "epoch": 0.11243851018973998, "grad_norm": 11.957242965698242, "learning_rate": 6.661981728742095e-07, "loss": 3.925, "step": 160 }, { "epoch": 0.11314125087842586, "grad_norm": 11.397250175476074, "learning_rate": 6.704146170063246e-07, "loss": 3.9135, "step": 161 }, { "epoch": 0.11384399156711174, "grad_norm": 11.2194242477417, "learning_rate": 6.746310611384399e-07, "loss": 3.8767, "step": 162 }, { "epoch": 0.11454673225579762, "grad_norm": 11.165748596191406, "learning_rate": 6.788475052705553e-07, "loss": 3.8598, "step": 163 }, { "epoch": 0.11524947294448348, "grad_norm": 11.327119827270508, "learning_rate": 6.830639494026704e-07, "loss": 3.8397, "step": 164 }, { "epoch": 0.11595221363316936, "grad_norm": 11.615961074829102, "learning_rate": 6.872803935347857e-07, "loss": 3.826, "step": 165 }, { "epoch": 0.11665495432185524, "grad_norm": 11.568402290344238, "learning_rate": 6.914968376669009e-07, "loss": 3.8115, "step": 166 }, { "epoch": 0.11735769501054111, "grad_norm": 12.170754432678223, "learning_rate": 6.957132817990162e-07, "loss": 3.7727, "step": 167 }, { "epoch": 0.11806043569922699, "grad_norm": 12.069080352783203, "learning_rate": 6.999297259311314e-07, "loss": 3.753, "step": 168 }, { "epoch": 0.11876317638791287, "grad_norm": 11.866168022155762, "learning_rate": 7.041461700632467e-07, "loss": 3.7309, "step": 169 }, { "epoch": 0.11946591707659873, "grad_norm": 11.62770938873291, "learning_rate": 7.083626141953619e-07, "loss": 3.7255, "step": 170 }, { "epoch": 0.12016865776528461, "grad_norm": 10.828988075256348, "learning_rate": 7.125790583274772e-07, "loss": 3.7194, "step": 171 }, { "epoch": 0.12087139845397049, "grad_norm": 10.418066024780273, "learning_rate": 7.167955024595924e-07, "loss": 3.6962, "step": 172 }, { "epoch": 0.12157413914265636, "grad_norm": 10.138426780700684, "learning_rate": 7.210119465917077e-07, "loss": 3.7244, "step": 173 }, { "epoch": 0.12227687983134224, "grad_norm": 9.82909870147705, "learning_rate": 7.252283907238229e-07, "loss": 3.7163, "step": 174 }, { "epoch": 0.12297962052002812, "grad_norm": 10.851799011230469, "learning_rate": 7.294448348559382e-07, "loss": 3.737, "step": 175 }, { "epoch": 0.12368236120871398, "grad_norm": 8.952454566955566, "learning_rate": 7.336612789880534e-07, "loss": 3.5801, "step": 176 }, { "epoch": 0.12438510189739986, "grad_norm": 8.995676040649414, "learning_rate": 7.378777231201687e-07, "loss": 3.535, "step": 177 }, { "epoch": 0.12508784258608574, "grad_norm": 8.65063190460205, "learning_rate": 7.42094167252284e-07, "loss": 3.51, "step": 178 }, { "epoch": 0.1257905832747716, "grad_norm": 8.312562942504883, "learning_rate": 7.463106113843991e-07, "loss": 3.4829, "step": 179 }, { "epoch": 0.12649332396345747, "grad_norm": 7.914120674133301, "learning_rate": 7.505270555165145e-07, "loss": 3.4671, "step": 180 }, { "epoch": 0.12719606465214336, "grad_norm": 7.723028659820557, "learning_rate": 7.547434996486296e-07, "loss": 3.425, "step": 181 }, { "epoch": 0.12789880534082923, "grad_norm": 7.325604438781738, "learning_rate": 7.589599437807449e-07, "loss": 3.405, "step": 182 }, { "epoch": 0.1286015460295151, "grad_norm": 6.962283134460449, "learning_rate": 7.631763879128602e-07, "loss": 3.3925, "step": 183 }, { "epoch": 0.129304286718201, "grad_norm": 6.660511016845703, "learning_rate": 7.673928320449754e-07, "loss": 3.3703, "step": 184 }, { "epoch": 0.13000702740688685, "grad_norm": 6.515542030334473, "learning_rate": 7.716092761770906e-07, "loss": 3.3478, "step": 185 }, { "epoch": 0.13070976809557272, "grad_norm": 6.119519233703613, "learning_rate": 7.75825720309206e-07, "loss": 3.3293, "step": 186 }, { "epoch": 0.1314125087842586, "grad_norm": 6.100374698638916, "learning_rate": 7.800421644413211e-07, "loss": 3.3072, "step": 187 }, { "epoch": 0.13211524947294448, "grad_norm": 6.190479755401611, "learning_rate": 7.842586085734365e-07, "loss": 3.298, "step": 188 }, { "epoch": 0.13281799016163034, "grad_norm": 5.764830112457275, "learning_rate": 7.884750527055517e-07, "loss": 3.2633, "step": 189 }, { "epoch": 0.13352073085031624, "grad_norm": 5.672957420349121, "learning_rate": 7.926914968376669e-07, "loss": 3.255, "step": 190 }, { "epoch": 0.1342234715390021, "grad_norm": 5.543680667877197, "learning_rate": 7.969079409697823e-07, "loss": 3.2459, "step": 191 }, { "epoch": 0.13492621222768797, "grad_norm": 5.207180976867676, "learning_rate": 8.011243851018974e-07, "loss": 3.2361, "step": 192 }, { "epoch": 0.13562895291637386, "grad_norm": 5.126894474029541, "learning_rate": 8.053408292340127e-07, "loss": 3.2017, "step": 193 }, { "epoch": 0.13633169360505973, "grad_norm": 4.93182897567749, "learning_rate": 8.095572733661279e-07, "loss": 3.1994, "step": 194 }, { "epoch": 0.13703443429374562, "grad_norm": 4.6422929763793945, "learning_rate": 8.137737174982432e-07, "loss": 3.1508, "step": 195 }, { "epoch": 0.1377371749824315, "grad_norm": 4.602615833282471, "learning_rate": 8.179901616303584e-07, "loss": 3.1767, "step": 196 }, { "epoch": 0.13843991567111735, "grad_norm": 5.328883647918701, "learning_rate": 8.222066057624737e-07, "loss": 3.185, "step": 197 }, { "epoch": 0.13914265635980325, "grad_norm": 5.486371040344238, "learning_rate": 8.264230498945889e-07, "loss": 3.1813, "step": 198 }, { "epoch": 0.1398453970484891, "grad_norm": 5.176172256469727, "learning_rate": 8.306394940267041e-07, "loss": 3.2025, "step": 199 }, { "epoch": 0.14054813773717498, "grad_norm": 6.593674659729004, "learning_rate": 8.348559381588194e-07, "loss": 3.1908, "step": 200 }, { "epoch": 0.14125087842586087, "grad_norm": 7.912206172943115, "learning_rate": 8.390723822909347e-07, "loss": 3.1277, "step": 201 }, { "epoch": 0.14195361911454674, "grad_norm": 8.42443561553955, "learning_rate": 8.432888264230498e-07, "loss": 3.0977, "step": 202 }, { "epoch": 0.1426563598032326, "grad_norm": 8.450773239135742, "learning_rate": 8.475052705551652e-07, "loss": 3.0783, "step": 203 }, { "epoch": 0.1433591004919185, "grad_norm": 7.129877090454102, "learning_rate": 8.517217146872804e-07, "loss": 3.0632, "step": 204 }, { "epoch": 0.14406184118060436, "grad_norm": 5.799768447875977, "learning_rate": 8.559381588193957e-07, "loss": 3.0387, "step": 205 }, { "epoch": 0.14476458186929023, "grad_norm": 4.3592376708984375, "learning_rate": 8.60154602951511e-07, "loss": 3.0335, "step": 206 }, { "epoch": 0.14546732255797612, "grad_norm": 3.8875036239624023, "learning_rate": 8.643710470836261e-07, "loss": 3.0248, "step": 207 }, { "epoch": 0.14617006324666199, "grad_norm": 4.538383483886719, "learning_rate": 8.685874912157415e-07, "loss": 3.0175, "step": 208 }, { "epoch": 0.14687280393534785, "grad_norm": 5.342519283294678, "learning_rate": 8.728039353478566e-07, "loss": 2.9978, "step": 209 }, { "epoch": 0.14757554462403374, "grad_norm": 5.308180332183838, "learning_rate": 8.770203794799719e-07, "loss": 2.9782, "step": 210 }, { "epoch": 0.1482782853127196, "grad_norm": 5.475630283355713, "learning_rate": 8.812368236120872e-07, "loss": 2.9746, "step": 211 }, { "epoch": 0.14898102600140548, "grad_norm": 4.252923965454102, "learning_rate": 8.854532677442024e-07, "loss": 2.9736, "step": 212 }, { "epoch": 0.14968376669009137, "grad_norm": 3.5029678344726562, "learning_rate": 8.896697118763176e-07, "loss": 2.9512, "step": 213 }, { "epoch": 0.15038650737877723, "grad_norm": 3.454364776611328, "learning_rate": 8.93886156008433e-07, "loss": 2.9429, "step": 214 }, { "epoch": 0.1510892480674631, "grad_norm": 3.4732162952423096, "learning_rate": 8.981026001405481e-07, "loss": 2.9225, "step": 215 }, { "epoch": 0.151791988756149, "grad_norm": 3.889315128326416, "learning_rate": 9.023190442726634e-07, "loss": 2.9318, "step": 216 }, { "epoch": 0.15249472944483486, "grad_norm": 3.9984395503997803, "learning_rate": 9.065354884047787e-07, "loss": 2.9027, "step": 217 }, { "epoch": 0.15319747013352072, "grad_norm": 3.5159666538238525, "learning_rate": 9.107519325368939e-07, "loss": 2.9072, "step": 218 }, { "epoch": 0.15390021082220662, "grad_norm": 3.0376241207122803, "learning_rate": 9.149683766690092e-07, "loss": 2.9125, "step": 219 }, { "epoch": 0.15460295151089248, "grad_norm": 2.6865127086639404, "learning_rate": 9.191848208011244e-07, "loss": 2.8905, "step": 220 }, { "epoch": 0.15530569219957835, "grad_norm": 3.2591402530670166, "learning_rate": 9.234012649332397e-07, "loss": 2.9025, "step": 221 }, { "epoch": 0.15600843288826424, "grad_norm": 3.903589963912964, "learning_rate": 9.276177090653548e-07, "loss": 2.9036, "step": 222 }, { "epoch": 0.1567111735769501, "grad_norm": 4.442297458648682, "learning_rate": 9.318341531974702e-07, "loss": 2.914, "step": 223 }, { "epoch": 0.15741391426563597, "grad_norm": 4.943921089172363, "learning_rate": 9.360505973295854e-07, "loss": 2.9277, "step": 224 }, { "epoch": 0.15811665495432187, "grad_norm": 6.33236837387085, "learning_rate": 9.402670414617007e-07, "loss": 2.9396, "step": 225 }, { "epoch": 0.15881939564300773, "grad_norm": 10.275463104248047, "learning_rate": 9.444834855938159e-07, "loss": 2.8715, "step": 226 }, { "epoch": 0.1595221363316936, "grad_norm": 11.454379081726074, "learning_rate": 9.48699929725931e-07, "loss": 2.8834, "step": 227 }, { "epoch": 0.1602248770203795, "grad_norm": 11.259066581726074, "learning_rate": 9.529163738580465e-07, "loss": 2.8575, "step": 228 }, { "epoch": 0.16092761770906536, "grad_norm": 10.187173843383789, "learning_rate": 9.571328179901617e-07, "loss": 2.8506, "step": 229 }, { "epoch": 0.16163035839775122, "grad_norm": 8.738292694091797, "learning_rate": 9.613492621222768e-07, "loss": 2.8369, "step": 230 }, { "epoch": 0.16233309908643712, "grad_norm": 6.34637975692749, "learning_rate": 9.655657062543922e-07, "loss": 2.8239, "step": 231 }, { "epoch": 0.16303583977512298, "grad_norm": 2.417870044708252, "learning_rate": 9.697821503865075e-07, "loss": 2.8175, "step": 232 }, { "epoch": 0.16373858046380885, "grad_norm": 2.637657880783081, "learning_rate": 9.739985945186227e-07, "loss": 2.8071, "step": 233 }, { "epoch": 0.16444132115249474, "grad_norm": 4.504966735839844, "learning_rate": 9.782150386507379e-07, "loss": 2.8022, "step": 234 }, { "epoch": 0.1651440618411806, "grad_norm": 5.247546195983887, "learning_rate": 9.824314827828532e-07, "loss": 2.8041, "step": 235 }, { "epoch": 0.16584680252986647, "grad_norm": 6.185456275939941, "learning_rate": 9.866479269149684e-07, "loss": 2.8129, "step": 236 }, { "epoch": 0.16654954321855236, "grad_norm": 4.863704204559326, "learning_rate": 9.908643710470835e-07, "loss": 2.7871, "step": 237 }, { "epoch": 0.16725228390723823, "grad_norm": 3.1628968715667725, "learning_rate": 9.95080815179199e-07, "loss": 2.7909, "step": 238 }, { "epoch": 0.1679550245959241, "grad_norm": 2.2929391860961914, "learning_rate": 9.992972593113143e-07, "loss": 2.7868, "step": 239 }, { "epoch": 0.16865776528461, "grad_norm": 2.7466249465942383, "learning_rate": 1.0035137034434294e-06, "loss": 2.7863, "step": 240 }, { "epoch": 0.16936050597329586, "grad_norm": 3.0940847396850586, "learning_rate": 1.0077301475755446e-06, "loss": 2.768, "step": 241 }, { "epoch": 0.17006324666198172, "grad_norm": 2.6783504486083984, "learning_rate": 1.0119465917076597e-06, "loss": 2.7754, "step": 242 }, { "epoch": 0.17076598735066761, "grad_norm": 2.6458840370178223, "learning_rate": 1.016163035839775e-06, "loss": 2.7836, "step": 243 }, { "epoch": 0.17146872803935348, "grad_norm": 1.756826400756836, "learning_rate": 1.0203794799718905e-06, "loss": 2.7671, "step": 244 }, { "epoch": 0.17217146872803935, "grad_norm": 1.9358131885528564, "learning_rate": 1.0245959241040056e-06, "loss": 2.7832, "step": 245 }, { "epoch": 0.17287420941672524, "grad_norm": 3.6234259605407715, "learning_rate": 1.028812368236121e-06, "loss": 2.7994, "step": 246 }, { "epoch": 0.1735769501054111, "grad_norm": 3.6348397731781006, "learning_rate": 1.0330288123682362e-06, "loss": 2.8128, "step": 247 }, { "epoch": 0.17427969079409697, "grad_norm": 3.023005247116089, "learning_rate": 1.0372452565003513e-06, "loss": 2.8363, "step": 248 }, { "epoch": 0.17498243148278286, "grad_norm": 4.683254241943359, "learning_rate": 1.0414617006324667e-06, "loss": 2.8464, "step": 249 }, { "epoch": 0.17568517217146873, "grad_norm": 4.756512641906738, "learning_rate": 1.045678144764582e-06, "loss": 2.8389, "step": 250 }, { "epoch": 0.1763879128601546, "grad_norm": 11.45846176147461, "learning_rate": 1.0498945888966972e-06, "loss": 2.8028, "step": 251 }, { "epoch": 0.1770906535488405, "grad_norm": 13.095087051391602, "learning_rate": 1.0541110330288124e-06, "loss": 2.8029, "step": 252 }, { "epoch": 0.17779339423752635, "grad_norm": 12.982743263244629, "learning_rate": 1.0583274771609275e-06, "loss": 2.7981, "step": 253 }, { "epoch": 0.17849613492621222, "grad_norm": 11.700976371765137, "learning_rate": 1.0625439212930429e-06, "loss": 2.7811, "step": 254 }, { "epoch": 0.1791988756148981, "grad_norm": 10.728096961975098, "learning_rate": 1.0667603654251582e-06, "loss": 2.7678, "step": 255 }, { "epoch": 0.17990161630358398, "grad_norm": 7.679455757141113, "learning_rate": 1.0709768095572734e-06, "loss": 2.7415, "step": 256 }, { "epoch": 0.18060435699226984, "grad_norm": 4.416462421417236, "learning_rate": 1.0751932536893888e-06, "loss": 2.7451, "step": 257 }, { "epoch": 0.18130709768095574, "grad_norm": 1.5232970714569092, "learning_rate": 1.079409697821504e-06, "loss": 2.7459, "step": 258 }, { "epoch": 0.1820098383696416, "grad_norm": 3.5106844902038574, "learning_rate": 1.083626141953619e-06, "loss": 2.7361, "step": 259 }, { "epoch": 0.18271257905832747, "grad_norm": 5.795731067657471, "learning_rate": 1.0878425860857342e-06, "loss": 2.7418, "step": 260 }, { "epoch": 0.18341531974701336, "grad_norm": 6.66754674911499, "learning_rate": 1.0920590302178498e-06, "loss": 2.728, "step": 261 }, { "epoch": 0.18411806043569923, "grad_norm": 5.713644981384277, "learning_rate": 1.096275474349965e-06, "loss": 2.7313, "step": 262 }, { "epoch": 0.1848208011243851, "grad_norm": 3.576798915863037, "learning_rate": 1.1004919184820801e-06, "loss": 2.7326, "step": 263 }, { "epoch": 0.18552354181307099, "grad_norm": 2.2506015300750732, "learning_rate": 1.1047083626141953e-06, "loss": 2.7305, "step": 264 }, { "epoch": 0.18622628250175685, "grad_norm": 2.254970073699951, "learning_rate": 1.1089248067463106e-06, "loss": 2.725, "step": 265 }, { "epoch": 0.18692902319044272, "grad_norm": 3.360278844833374, "learning_rate": 1.1131412508784258e-06, "loss": 2.7289, "step": 266 }, { "epoch": 0.1876317638791286, "grad_norm": 3.1316347122192383, "learning_rate": 1.1173576950105412e-06, "loss": 2.7122, "step": 267 }, { "epoch": 0.18833450456781448, "grad_norm": 3.5673177242279053, "learning_rate": 1.1215741391426563e-06, "loss": 2.7236, "step": 268 }, { "epoch": 0.18903724525650034, "grad_norm": 2.1814587116241455, "learning_rate": 1.1257905832747717e-06, "loss": 2.7361, "step": 269 }, { "epoch": 0.18973998594518623, "grad_norm": 1.499142050743103, "learning_rate": 1.1300070274068868e-06, "loss": 2.723, "step": 270 }, { "epoch": 0.1904427266338721, "grad_norm": 3.384652614593506, "learning_rate": 1.134223471539002e-06, "loss": 2.7287, "step": 271 }, { "epoch": 0.19114546732255797, "grad_norm": 3.9073235988616943, "learning_rate": 1.1384399156711176e-06, "loss": 2.7506, "step": 272 }, { "epoch": 0.19184820801124386, "grad_norm": 3.664435625076294, "learning_rate": 1.1426563598032327e-06, "loss": 2.7794, "step": 273 }, { "epoch": 0.19255094869992972, "grad_norm": 5.044891834259033, "learning_rate": 1.1468728039353479e-06, "loss": 2.8017, "step": 274 }, { "epoch": 0.1932536893886156, "grad_norm": 4.9747395515441895, "learning_rate": 1.151089248067463e-06, "loss": 2.795, "step": 275 }, { "epoch": 0.19395643007730148, "grad_norm": 9.520419120788574, "learning_rate": 1.1553056921995784e-06, "loss": 2.7627, "step": 276 }, { "epoch": 0.19465917076598735, "grad_norm": 10.973010063171387, "learning_rate": 1.1595221363316936e-06, "loss": 2.75, "step": 277 }, { "epoch": 0.19536191145467321, "grad_norm": 9.868484497070312, "learning_rate": 1.163738580463809e-06, "loss": 2.7208, "step": 278 }, { "epoch": 0.1960646521433591, "grad_norm": 9.080767631530762, "learning_rate": 1.167955024595924e-06, "loss": 2.7299, "step": 279 }, { "epoch": 0.19676739283204497, "grad_norm": 8.12785816192627, "learning_rate": 1.1721714687280395e-06, "loss": 2.7135, "step": 280 }, { "epoch": 0.19747013352073084, "grad_norm": 4.87448787689209, "learning_rate": 1.1763879128601546e-06, "loss": 2.706, "step": 281 }, { "epoch": 0.19817287420941673, "grad_norm": 1.7455897331237793, "learning_rate": 1.1806043569922698e-06, "loss": 2.6982, "step": 282 }, { "epoch": 0.1988756148981026, "grad_norm": 2.6731984615325928, "learning_rate": 1.1848208011243851e-06, "loss": 2.6864, "step": 283 }, { "epoch": 0.19957835558678846, "grad_norm": 5.625666618347168, "learning_rate": 1.1890372452565005e-06, "loss": 2.7011, "step": 284 }, { "epoch": 0.20028109627547436, "grad_norm": 6.4451751708984375, "learning_rate": 1.1932536893886157e-06, "loss": 2.7048, "step": 285 }, { "epoch": 0.20098383696416022, "grad_norm": 6.862765789031982, "learning_rate": 1.1974701335207308e-06, "loss": 2.7077, "step": 286 }, { "epoch": 0.2016865776528461, "grad_norm": 4.685825824737549, "learning_rate": 1.2016865776528462e-06, "loss": 2.6938, "step": 287 }, { "epoch": 0.20238931834153198, "grad_norm": 1.7673301696777344, "learning_rate": 1.2059030217849613e-06, "loss": 2.6989, "step": 288 }, { "epoch": 0.20309205903021785, "grad_norm": 2.0235536098480225, "learning_rate": 1.2101194659170767e-06, "loss": 2.6772, "step": 289 }, { "epoch": 0.2037947997189037, "grad_norm": 3.1267807483673096, "learning_rate": 1.2143359100491919e-06, "loss": 2.6906, "step": 290 }, { "epoch": 0.2044975404075896, "grad_norm": 2.9699289798736572, "learning_rate": 1.2185523541813072e-06, "loss": 2.6897, "step": 291 }, { "epoch": 0.20520028109627547, "grad_norm": 2.8106508255004883, "learning_rate": 1.2227687983134224e-06, "loss": 2.6853, "step": 292 }, { "epoch": 0.20590302178496134, "grad_norm": 1.7325483560562134, "learning_rate": 1.2269852424455375e-06, "loss": 2.6762, "step": 293 }, { "epoch": 0.20660576247364723, "grad_norm": 2.1156699657440186, "learning_rate": 1.2312016865776527e-06, "loss": 2.689, "step": 294 }, { "epoch": 0.2073085031623331, "grad_norm": 2.714932680130005, "learning_rate": 1.2354181307097683e-06, "loss": 2.6986, "step": 295 }, { "epoch": 0.20801124385101896, "grad_norm": 3.164423704147339, "learning_rate": 1.2396345748418834e-06, "loss": 2.7007, "step": 296 }, { "epoch": 0.20871398453970486, "grad_norm": 2.965991497039795, "learning_rate": 1.2438510189739986e-06, "loss": 2.7122, "step": 297 }, { "epoch": 0.20941672522839072, "grad_norm": 2.1417396068573, "learning_rate": 1.248067463106114e-06, "loss": 2.7518, "step": 298 }, { "epoch": 0.2101194659170766, "grad_norm": 2.2218518257141113, "learning_rate": 1.252283907238229e-06, "loss": 2.7631, "step": 299 }, { "epoch": 0.21082220660576248, "grad_norm": 2.725879192352295, "learning_rate": 1.2565003513703443e-06, "loss": 2.7727, "step": 300 }, { "epoch": 0.21152494729444835, "grad_norm": 10.552002906799316, "learning_rate": 1.2607167955024596e-06, "loss": 2.7081, "step": 301 }, { "epoch": 0.2122276879831342, "grad_norm": 10.737143516540527, "learning_rate": 1.264933239634575e-06, "loss": 2.6959, "step": 302 }, { "epoch": 0.2129304286718201, "grad_norm": 10.0125093460083, "learning_rate": 1.2691496837666901e-06, "loss": 2.6804, "step": 303 }, { "epoch": 0.21363316936050597, "grad_norm": 8.201614379882812, "learning_rate": 1.2733661278988053e-06, "loss": 2.6658, "step": 304 }, { "epoch": 0.21433591004919184, "grad_norm": 5.557940483093262, "learning_rate": 1.2775825720309205e-06, "loss": 2.661, "step": 305 }, { "epoch": 0.21503865073787773, "grad_norm": 2.569955825805664, "learning_rate": 1.281799016163036e-06, "loss": 2.6746, "step": 306 }, { "epoch": 0.2157413914265636, "grad_norm": 2.7154202461242676, "learning_rate": 1.2860154602951512e-06, "loss": 2.6603, "step": 307 }, { "epoch": 0.21644413211524946, "grad_norm": 5.555208683013916, "learning_rate": 1.2902319044272663e-06, "loss": 2.6572, "step": 308 }, { "epoch": 0.21714687280393535, "grad_norm": 6.036032676696777, "learning_rate": 1.2944483485593815e-06, "loss": 2.6524, "step": 309 }, { "epoch": 0.21784961349262122, "grad_norm": 4.9504547119140625, "learning_rate": 1.2986647926914969e-06, "loss": 2.645, "step": 310 }, { "epoch": 0.21855235418130708, "grad_norm": 3.6744093894958496, "learning_rate": 1.302881236823612e-06, "loss": 2.6419, "step": 311 }, { "epoch": 0.21925509486999298, "grad_norm": 1.5746314525604248, "learning_rate": 1.3070976809557274e-06, "loss": 2.626, "step": 312 }, { "epoch": 0.21995783555867884, "grad_norm": 2.750112771987915, "learning_rate": 1.3113141250878428e-06, "loss": 2.6392, "step": 313 }, { "epoch": 0.2206605762473647, "grad_norm": 3.165656566619873, "learning_rate": 1.315530569219958e-06, "loss": 2.6348, "step": 314 }, { "epoch": 0.2213633169360506, "grad_norm": 3.158315896987915, "learning_rate": 1.319747013352073e-06, "loss": 2.6181, "step": 315 }, { "epoch": 0.22206605762473647, "grad_norm": 1.4494050741195679, "learning_rate": 1.3239634574841882e-06, "loss": 2.6209, "step": 316 }, { "epoch": 0.22276879831342233, "grad_norm": 2.0821337699890137, "learning_rate": 1.3281799016163036e-06, "loss": 2.6038, "step": 317 }, { "epoch": 0.22347153900210823, "grad_norm": 2.2930285930633545, "learning_rate": 1.332396345748419e-06, "loss": 2.6097, "step": 318 }, { "epoch": 0.2241742796907941, "grad_norm": 1.864495038986206, "learning_rate": 1.3366127898805341e-06, "loss": 2.635, "step": 319 }, { "epoch": 0.22487702037947996, "grad_norm": 1.568659782409668, "learning_rate": 1.3408292340126493e-06, "loss": 2.5899, "step": 320 }, { "epoch": 0.22557976106816585, "grad_norm": 1.5544955730438232, "learning_rate": 1.3450456781447646e-06, "loss": 2.6388, "step": 321 }, { "epoch": 0.22628250175685172, "grad_norm": 2.409167766571045, "learning_rate": 1.3492621222768798e-06, "loss": 2.6425, "step": 322 }, { "epoch": 0.22698524244553758, "grad_norm": 2.1384332180023193, "learning_rate": 1.3534785664089952e-06, "loss": 2.6561, "step": 323 }, { "epoch": 0.22768798313422348, "grad_norm": 2.4919137954711914, "learning_rate": 1.3576950105411105e-06, "loss": 2.6699, "step": 324 }, { "epoch": 0.22839072382290934, "grad_norm": 4.000579833984375, "learning_rate": 1.3619114546732257e-06, "loss": 2.6941, "step": 325 }, { "epoch": 0.22909346451159524, "grad_norm": 6.706967353820801, "learning_rate": 1.3661278988053408e-06, "loss": 2.6022, "step": 326 }, { "epoch": 0.2297962052002811, "grad_norm": 5.899526596069336, "learning_rate": 1.370344342937456e-06, "loss": 2.5648, "step": 327 }, { "epoch": 0.23049894588896697, "grad_norm": 5.061286926269531, "learning_rate": 1.3745607870695714e-06, "loss": 2.547, "step": 328 }, { "epoch": 0.23120168657765286, "grad_norm": 3.157846450805664, "learning_rate": 1.3787772312016867e-06, "loss": 2.5223, "step": 329 }, { "epoch": 0.23190442726633873, "grad_norm": 2.325828790664673, "learning_rate": 1.3829936753338019e-06, "loss": 2.508, "step": 330 }, { "epoch": 0.2326071679550246, "grad_norm": 2.4226107597351074, "learning_rate": 1.387210119465917e-06, "loss": 2.4868, "step": 331 }, { "epoch": 0.23330990864371048, "grad_norm": 3.3619611263275146, "learning_rate": 1.3914265635980324e-06, "loss": 2.4808, "step": 332 }, { "epoch": 0.23401264933239635, "grad_norm": 2.7042014598846436, "learning_rate": 1.3956430077301476e-06, "loss": 2.4654, "step": 333 }, { "epoch": 0.23471539002108222, "grad_norm": 2.123812437057495, "learning_rate": 1.3998594518622627e-06, "loss": 2.463, "step": 334 }, { "epoch": 0.2354181307097681, "grad_norm": 2.296865224838257, "learning_rate": 1.404075895994378e-06, "loss": 2.4285, "step": 335 }, { "epoch": 0.23612087139845397, "grad_norm": 2.668534278869629, "learning_rate": 1.4082923401264934e-06, "loss": 2.4209, "step": 336 }, { "epoch": 0.23682361208713984, "grad_norm": 2.6332995891571045, "learning_rate": 1.4125087842586086e-06, "loss": 2.399, "step": 337 }, { "epoch": 0.23752635277582573, "grad_norm": 2.4135961532592773, "learning_rate": 1.4167252283907238e-06, "loss": 2.3853, "step": 338 }, { "epoch": 0.2382290934645116, "grad_norm": 2.1026835441589355, "learning_rate": 1.4209416725228391e-06, "loss": 2.343, "step": 339 }, { "epoch": 0.23893183415319746, "grad_norm": 2.4277498722076416, "learning_rate": 1.4251581166549545e-06, "loss": 2.3552, "step": 340 }, { "epoch": 0.23963457484188336, "grad_norm": 2.4805331230163574, "learning_rate": 1.4293745607870697e-06, "loss": 2.3448, "step": 341 }, { "epoch": 0.24033731553056922, "grad_norm": 2.5443501472473145, "learning_rate": 1.4335910049191848e-06, "loss": 2.2942, "step": 342 }, { "epoch": 0.2410400562192551, "grad_norm": 2.7923550605773926, "learning_rate": 1.4378074490513002e-06, "loss": 2.2925, "step": 343 }, { "epoch": 0.24174279690794098, "grad_norm": 2.243821382522583, "learning_rate": 1.4420238931834153e-06, "loss": 2.2848, "step": 344 }, { "epoch": 0.24244553759662685, "grad_norm": 2.417539119720459, "learning_rate": 1.4462403373155305e-06, "loss": 2.2711, "step": 345 }, { "epoch": 0.2431482782853127, "grad_norm": 2.310455083847046, "learning_rate": 1.4504567814476459e-06, "loss": 2.2763, "step": 346 }, { "epoch": 0.2438510189739986, "grad_norm": 2.3740668296813965, "learning_rate": 1.4546732255797612e-06, "loss": 2.3032, "step": 347 }, { "epoch": 0.24455375966268447, "grad_norm": 2.246673345565796, "learning_rate": 1.4588896697118764e-06, "loss": 2.3143, "step": 348 }, { "epoch": 0.24525650035137034, "grad_norm": 4.354970455169678, "learning_rate": 1.4631061138439915e-06, "loss": 2.3667, "step": 349 }, { "epoch": 0.24595924104005623, "grad_norm": 4.33468770980835, "learning_rate": 1.467322557976107e-06, "loss": 2.4161, "step": 350 }, { "epoch": 0.2466619817287421, "grad_norm": 3.3293185234069824, "learning_rate": 1.471539002108222e-06, "loss": 2.1682, "step": 351 }, { "epoch": 0.24736472241742796, "grad_norm": 3.073286294937134, "learning_rate": 1.4757554462403374e-06, "loss": 2.0836, "step": 352 }, { "epoch": 0.24806746310611386, "grad_norm": 2.7034711837768555, "learning_rate": 1.4799718903724526e-06, "loss": 2.0529, "step": 353 }, { "epoch": 0.24877020379479972, "grad_norm": 2.4827816486358643, "learning_rate": 1.484188334504568e-06, "loss": 2.0316, "step": 354 }, { "epoch": 0.2494729444834856, "grad_norm": 2.35432505607605, "learning_rate": 1.488404778636683e-06, "loss": 1.9818, "step": 355 }, { "epoch": 0.2501756851721715, "grad_norm": 2.3900139331817627, "learning_rate": 1.4926212227687983e-06, "loss": 1.9411, "step": 356 }, { "epoch": 0.25087842586085735, "grad_norm": 2.7027058601379395, "learning_rate": 1.4968376669009136e-06, "loss": 1.9296, "step": 357 }, { "epoch": 0.2515811665495432, "grad_norm": 2.3695342540740967, "learning_rate": 1.501054111033029e-06, "loss": 1.897, "step": 358 }, { "epoch": 0.2522839072382291, "grad_norm": 3.69183611869812, "learning_rate": 1.5052705551651441e-06, "loss": 1.8814, "step": 359 }, { "epoch": 0.25298664792691494, "grad_norm": 2.8227949142456055, "learning_rate": 1.5094869992972593e-06, "loss": 1.852, "step": 360 }, { "epoch": 0.25368938861560086, "grad_norm": 2.693683624267578, "learning_rate": 1.5137034434293745e-06, "loss": 1.8406, "step": 361 }, { "epoch": 0.25439212930428673, "grad_norm": 2.3703830242156982, "learning_rate": 1.5179198875614898e-06, "loss": 1.8018, "step": 362 }, { "epoch": 0.2550948699929726, "grad_norm": 2.5001821517944336, "learning_rate": 1.5221363316936052e-06, "loss": 1.7674, "step": 363 }, { "epoch": 0.25579761068165846, "grad_norm": 4.354287624359131, "learning_rate": 1.5263527758257203e-06, "loss": 1.763, "step": 364 }, { "epoch": 0.2565003513703443, "grad_norm": 7.734760761260986, "learning_rate": 1.5305692199578357e-06, "loss": 1.7424, "step": 365 }, { "epoch": 0.2572030920590302, "grad_norm": 3.237273693084717, "learning_rate": 1.5347856640899509e-06, "loss": 1.7281, "step": 366 }, { "epoch": 0.2579058327477161, "grad_norm": 2.975923776626587, "learning_rate": 1.539002108222066e-06, "loss": 1.6888, "step": 367 }, { "epoch": 0.258608573436402, "grad_norm": 2.408940553665161, "learning_rate": 1.5432185523541812e-06, "loss": 1.6894, "step": 368 }, { "epoch": 0.25931131412508784, "grad_norm": 2.540715217590332, "learning_rate": 1.5474349964862968e-06, "loss": 1.658, "step": 369 }, { "epoch": 0.2600140548137737, "grad_norm": 2.4437246322631836, "learning_rate": 1.551651440618412e-06, "loss": 1.6454, "step": 370 }, { "epoch": 0.2607167955024596, "grad_norm": 3.6408002376556396, "learning_rate": 1.555867884750527e-06, "loss": 1.6798, "step": 371 }, { "epoch": 0.26141953619114544, "grad_norm": 3.093285083770752, "learning_rate": 1.5600843288826422e-06, "loss": 1.7304, "step": 372 }, { "epoch": 0.26212227687983136, "grad_norm": 5.66248083114624, "learning_rate": 1.5643007730147576e-06, "loss": 1.7704, "step": 373 }, { "epoch": 0.2628250175685172, "grad_norm": 6.252695083618164, "learning_rate": 1.568517217146873e-06, "loss": 1.832, "step": 374 }, { "epoch": 0.2635277582572031, "grad_norm": 6.822600841522217, "learning_rate": 1.5727336612789881e-06, "loss": 2.0077, "step": 375 }, { "epoch": 0.26423049894588896, "grad_norm": 3.3857839107513428, "learning_rate": 1.5769501054111035e-06, "loss": 1.5637, "step": 376 }, { "epoch": 0.2649332396345748, "grad_norm": 2.6823689937591553, "learning_rate": 1.5811665495432186e-06, "loss": 1.4768, "step": 377 }, { "epoch": 0.2656359803232607, "grad_norm": 2.4833364486694336, "learning_rate": 1.5853829936753338e-06, "loss": 1.4363, "step": 378 }, { "epoch": 0.2663387210119466, "grad_norm": 2.3741424083709717, "learning_rate": 1.589599437807449e-06, "loss": 1.4201, "step": 379 }, { "epoch": 0.2670414617006325, "grad_norm": 2.4011003971099854, "learning_rate": 1.5938158819395645e-06, "loss": 1.3902, "step": 380 }, { "epoch": 0.26774420238931834, "grad_norm": 2.2490460872650146, "learning_rate": 1.5980323260716797e-06, "loss": 1.3334, "step": 381 }, { "epoch": 0.2684469430780042, "grad_norm": 2.2405993938446045, "learning_rate": 1.6022487702037948e-06, "loss": 1.3237, "step": 382 }, { "epoch": 0.2691496837666901, "grad_norm": 2.3804898262023926, "learning_rate": 1.60646521433591e-06, "loss": 1.3209, "step": 383 }, { "epoch": 0.26985242445537594, "grad_norm": 2.328848361968994, "learning_rate": 1.6106816584680254e-06, "loss": 1.3017, "step": 384 }, { "epoch": 0.27055516514406186, "grad_norm": 2.1923837661743164, "learning_rate": 1.6148981026001405e-06, "loss": 1.2884, "step": 385 }, { "epoch": 0.2712579058327477, "grad_norm": 2.1823854446411133, "learning_rate": 1.6191145467322559e-06, "loss": 1.2581, "step": 386 }, { "epoch": 0.2719606465214336, "grad_norm": 3.1224286556243896, "learning_rate": 1.623330990864371e-06, "loss": 1.2112, "step": 387 }, { "epoch": 0.27266338721011946, "grad_norm": 2.0502095222473145, "learning_rate": 1.6275474349964864e-06, "loss": 1.2005, "step": 388 }, { "epoch": 0.2733661278988053, "grad_norm": 2.2062015533447266, "learning_rate": 1.6317638791286016e-06, "loss": 1.1867, "step": 389 }, { "epoch": 0.27406886858749124, "grad_norm": 3.336336851119995, "learning_rate": 1.6359803232607167e-06, "loss": 1.1741, "step": 390 }, { "epoch": 0.2747716092761771, "grad_norm": 2.0874998569488525, "learning_rate": 1.640196767392832e-06, "loss": 1.1534, "step": 391 }, { "epoch": 0.275474349964863, "grad_norm": 2.344176769256592, "learning_rate": 1.6444132115249474e-06, "loss": 1.1157, "step": 392 }, { "epoch": 0.27617709065354884, "grad_norm": 2.151472568511963, "learning_rate": 1.6486296556570626e-06, "loss": 1.1331, "step": 393 }, { "epoch": 0.2768798313422347, "grad_norm": 2.200514316558838, "learning_rate": 1.6528460997891778e-06, "loss": 1.0849, "step": 394 }, { "epoch": 0.27758257203092057, "grad_norm": 2.115391492843628, "learning_rate": 1.6570625439212931e-06, "loss": 1.1006, "step": 395 }, { "epoch": 0.2782853127196065, "grad_norm": 2.1975715160369873, "learning_rate": 1.6612789880534083e-06, "loss": 1.1532, "step": 396 }, { "epoch": 0.27898805340829236, "grad_norm": 2.4197542667388916, "learning_rate": 1.6654954321855236e-06, "loss": 1.1792, "step": 397 }, { "epoch": 0.2796907940969782, "grad_norm": 2.639007329940796, "learning_rate": 1.6697118763176388e-06, "loss": 1.3042, "step": 398 }, { "epoch": 0.2803935347856641, "grad_norm": 2.9324352741241455, "learning_rate": 1.6739283204497542e-06, "loss": 1.3412, "step": 399 }, { "epoch": 0.28109627547434995, "grad_norm": 6.8926873207092285, "learning_rate": 1.6781447645818693e-06, "loss": 1.5464, "step": 400 }, { "epoch": 0.2817990161630358, "grad_norm": 2.8471100330352783, "learning_rate": 1.6823612087139845e-06, "loss": 1.0371, "step": 401 }, { "epoch": 0.28250175685172174, "grad_norm": 2.441263437271118, "learning_rate": 1.6865776528460996e-06, "loss": 0.9247, "step": 402 }, { "epoch": 0.2832044975404076, "grad_norm": 2.079697847366333, "learning_rate": 1.6907940969782152e-06, "loss": 0.8959, "step": 403 }, { "epoch": 0.2839072382290935, "grad_norm": 2.317962646484375, "learning_rate": 1.6950105411103304e-06, "loss": 0.8683, "step": 404 }, { "epoch": 0.28460997891777934, "grad_norm": 3.0359792709350586, "learning_rate": 1.6992269852424455e-06, "loss": 0.8498, "step": 405 }, { "epoch": 0.2853127196064652, "grad_norm": 1.9881319999694824, "learning_rate": 1.7034434293745609e-06, "loss": 0.817, "step": 406 }, { "epoch": 0.28601546029515107, "grad_norm": 2.0888872146606445, "learning_rate": 1.707659873506676e-06, "loss": 0.8223, "step": 407 }, { "epoch": 0.286718200983837, "grad_norm": 1.9823766946792603, "learning_rate": 1.7118763176387914e-06, "loss": 0.8154, "step": 408 }, { "epoch": 0.28742094167252286, "grad_norm": 2.1407370567321777, "learning_rate": 1.7160927617709066e-06, "loss": 0.7777, "step": 409 }, { "epoch": 0.2881236823612087, "grad_norm": 2.151531457901001, "learning_rate": 1.720309205903022e-06, "loss": 0.7454, "step": 410 }, { "epoch": 0.2888264230498946, "grad_norm": 1.6970189809799194, "learning_rate": 1.724525650035137e-06, "loss": 0.7663, "step": 411 }, { "epoch": 0.28952916373858045, "grad_norm": 1.7885417938232422, "learning_rate": 1.7287420941672522e-06, "loss": 0.7468, "step": 412 }, { "epoch": 0.2902319044272663, "grad_norm": 2.2772748470306396, "learning_rate": 1.7329585382993674e-06, "loss": 0.7411, "step": 413 }, { "epoch": 0.29093464511595224, "grad_norm": 1.7538447380065918, "learning_rate": 1.737174982431483e-06, "loss": 0.7036, "step": 414 }, { "epoch": 0.2916373858046381, "grad_norm": 2.068662166595459, "learning_rate": 1.7413914265635981e-06, "loss": 0.7402, "step": 415 }, { "epoch": 0.29234012649332397, "grad_norm": 1.812911868095398, "learning_rate": 1.7456078706957133e-06, "loss": 0.7429, "step": 416 }, { "epoch": 0.29304286718200984, "grad_norm": 2.1298677921295166, "learning_rate": 1.7498243148278287e-06, "loss": 0.6805, "step": 417 }, { "epoch": 0.2937456078706957, "grad_norm": 2.2732269763946533, "learning_rate": 1.7540407589599438e-06, "loss": 0.702, "step": 418 }, { "epoch": 0.29444834855938157, "grad_norm": 1.9730002880096436, "learning_rate": 1.758257203092059e-06, "loss": 0.7386, "step": 419 }, { "epoch": 0.2951510892480675, "grad_norm": 1.911069631576538, "learning_rate": 1.7624736472241743e-06, "loss": 0.7391, "step": 420 }, { "epoch": 0.29585382993675335, "grad_norm": 1.8663997650146484, "learning_rate": 1.7666900913562897e-06, "loss": 0.7584, "step": 421 }, { "epoch": 0.2965565706254392, "grad_norm": 2.867609977722168, "learning_rate": 1.7709065354884049e-06, "loss": 0.8612, "step": 422 }, { "epoch": 0.2972593113141251, "grad_norm": 2.718142032623291, "learning_rate": 1.77512297962052e-06, "loss": 0.9734, "step": 423 }, { "epoch": 0.29796205200281095, "grad_norm": 10.754460334777832, "learning_rate": 1.7793394237526352e-06, "loss": 1.1009, "step": 424 }, { "epoch": 0.2986647926914968, "grad_norm": 4.806275367736816, "learning_rate": 1.7835558678847505e-06, "loss": 1.2104, "step": 425 }, { "epoch": 0.29936753338018274, "grad_norm": 2.444930076599121, "learning_rate": 1.787772312016866e-06, "loss": 0.7023, "step": 426 }, { "epoch": 0.3000702740688686, "grad_norm": 1.9594329595565796, "learning_rate": 1.791988756148981e-06, "loss": 0.5919, "step": 427 }, { "epoch": 0.30077301475755447, "grad_norm": 1.57171630859375, "learning_rate": 1.7962052002810962e-06, "loss": 0.5712, "step": 428 }, { "epoch": 0.30147575544624033, "grad_norm": 1.4985835552215576, "learning_rate": 1.8004216444132116e-06, "loss": 0.5361, "step": 429 }, { "epoch": 0.3021784961349262, "grad_norm": 2.391857624053955, "learning_rate": 1.8046380885453267e-06, "loss": 0.5264, "step": 430 }, { "epoch": 0.30288123682361207, "grad_norm": 2.393780469894409, "learning_rate": 1.808854532677442e-06, "loss": 0.5095, "step": 431 }, { "epoch": 0.303583977512298, "grad_norm": 1.7763341665267944, "learning_rate": 1.8130709768095575e-06, "loss": 0.5208, "step": 432 }, { "epoch": 0.30428671820098385, "grad_norm": 1.7494877576828003, "learning_rate": 1.8172874209416726e-06, "loss": 0.5068, "step": 433 }, { "epoch": 0.3049894588896697, "grad_norm": 1.5650620460510254, "learning_rate": 1.8215038650737878e-06, "loss": 0.5259, "step": 434 }, { "epoch": 0.3056921995783556, "grad_norm": 1.7242295742034912, "learning_rate": 1.825720309205903e-06, "loss": 0.4748, "step": 435 }, { "epoch": 0.30639494026704145, "grad_norm": 1.5219738483428955, "learning_rate": 1.8299367533380183e-06, "loss": 0.5041, "step": 436 }, { "epoch": 0.3070976809557273, "grad_norm": 1.3446509838104248, "learning_rate": 1.8341531974701337e-06, "loss": 0.4695, "step": 437 }, { "epoch": 0.30780042164441324, "grad_norm": 2.0105528831481934, "learning_rate": 1.8383696416022488e-06, "loss": 0.5092, "step": 438 }, { "epoch": 0.3085031623330991, "grad_norm": 3.5247886180877686, "learning_rate": 1.842586085734364e-06, "loss": 0.4388, "step": 439 }, { "epoch": 0.30920590302178497, "grad_norm": 1.5043163299560547, "learning_rate": 1.8468025298664794e-06, "loss": 0.4796, "step": 440 }, { "epoch": 0.30990864371047083, "grad_norm": 1.4402377605438232, "learning_rate": 1.8510189739985945e-06, "loss": 0.4734, "step": 441 }, { "epoch": 0.3106113843991567, "grad_norm": 2.554149627685547, "learning_rate": 1.8552354181307097e-06, "loss": 0.4431, "step": 442 }, { "epoch": 0.31131412508784256, "grad_norm": 1.8278350830078125, "learning_rate": 1.8594518622628252e-06, "loss": 0.4744, "step": 443 }, { "epoch": 0.3120168657765285, "grad_norm": 3.834531784057617, "learning_rate": 1.8636683063949404e-06, "loss": 0.5172, "step": 444 }, { "epoch": 0.31271960646521435, "grad_norm": 1.7996066808700562, "learning_rate": 1.8678847505270556e-06, "loss": 0.5264, "step": 445 }, { "epoch": 0.3134223471539002, "grad_norm": 3.961960792541504, "learning_rate": 1.8721011946591707e-06, "loss": 0.5789, "step": 446 }, { "epoch": 0.3141250878425861, "grad_norm": 2.8717453479766846, "learning_rate": 1.876317638791286e-06, "loss": 0.6308, "step": 447 }, { "epoch": 0.31482782853127195, "grad_norm": 2.9254300594329834, "learning_rate": 1.8805340829234014e-06, "loss": 0.7499, "step": 448 }, { "epoch": 0.3155305692199578, "grad_norm": 2.798349142074585, "learning_rate": 1.8847505270555166e-06, "loss": 0.8648, "step": 449 }, { "epoch": 0.31623330990864373, "grad_norm": 12.662175178527832, "learning_rate": 1.8889669711876318e-06, "loss": 1.1163, "step": 450 }, { "epoch": 0.3169360505973296, "grad_norm": 1.8052743673324585, "learning_rate": 1.893183415319747e-06, "loss": 0.5373, "step": 451 }, { "epoch": 0.31763879128601546, "grad_norm": 1.3639907836914062, "learning_rate": 1.897399859451862e-06, "loss": 0.4492, "step": 452 }, { "epoch": 0.31834153197470133, "grad_norm": 3.2248454093933105, "learning_rate": 1.9016163035839776e-06, "loss": 0.3852, "step": 453 }, { "epoch": 0.3190442726633872, "grad_norm": 1.3404768705368042, "learning_rate": 1.905832747716093e-06, "loss": 0.3689, "step": 454 }, { "epoch": 0.31974701335207306, "grad_norm": 1.3756855726242065, "learning_rate": 1.910049191848208e-06, "loss": 0.3559, "step": 455 }, { "epoch": 0.320449754040759, "grad_norm": 1.2119132280349731, "learning_rate": 1.9142656359803233e-06, "loss": 0.3494, "step": 456 }, { "epoch": 0.32115249472944485, "grad_norm": 1.2227891683578491, "learning_rate": 1.9184820801124387e-06, "loss": 0.3687, "step": 457 }, { "epoch": 0.3218552354181307, "grad_norm": 1.2620881795883179, "learning_rate": 1.9226985242445536e-06, "loss": 0.348, "step": 458 }, { "epoch": 0.3225579761068166, "grad_norm": 1.0130692720413208, "learning_rate": 1.926914968376669e-06, "loss": 0.3628, "step": 459 }, { "epoch": 0.32326071679550245, "grad_norm": 1.4454174041748047, "learning_rate": 1.9311314125087844e-06, "loss": 0.3667, "step": 460 }, { "epoch": 0.3239634574841883, "grad_norm": 1.1626477241516113, "learning_rate": 1.9353478566408997e-06, "loss": 0.3584, "step": 461 }, { "epoch": 0.32466619817287423, "grad_norm": 1.2956666946411133, "learning_rate": 1.939564300773015e-06, "loss": 0.3294, "step": 462 }, { "epoch": 0.3253689388615601, "grad_norm": 1.3076034784317017, "learning_rate": 1.94378074490513e-06, "loss": 0.3616, "step": 463 }, { "epoch": 0.32607167955024596, "grad_norm": 1.7715984582901, "learning_rate": 1.9479971890372454e-06, "loss": 0.3468, "step": 464 }, { "epoch": 0.32677442023893183, "grad_norm": 1.3080239295959473, "learning_rate": 1.9522136331693604e-06, "loss": 0.3826, "step": 465 }, { "epoch": 0.3274771609276177, "grad_norm": 1.7617839574813843, "learning_rate": 1.9564300773014757e-06, "loss": 0.4026, "step": 466 }, { "epoch": 0.32817990161630356, "grad_norm": 2.2863857746124268, "learning_rate": 1.9606465214335907e-06, "loss": 0.3819, "step": 467 }, { "epoch": 0.3288826423049895, "grad_norm": 1.5719105005264282, "learning_rate": 1.9648629655657065e-06, "loss": 0.4, "step": 468 }, { "epoch": 0.32958538299367535, "grad_norm": 1.6085655689239502, "learning_rate": 1.969079409697822e-06, "loss": 0.387, "step": 469 }, { "epoch": 0.3302881236823612, "grad_norm": 1.4778159856796265, "learning_rate": 1.9732958538299368e-06, "loss": 0.4077, "step": 470 }, { "epoch": 0.3309908643710471, "grad_norm": 1.751297116279602, "learning_rate": 1.977512297962052e-06, "loss": 0.4374, "step": 471 }, { "epoch": 0.33169360505973294, "grad_norm": 1.8975909948349, "learning_rate": 1.981728742094167e-06, "loss": 0.5349, "step": 472 }, { "epoch": 0.3323963457484188, "grad_norm": 2.3533501625061035, "learning_rate": 1.9859451862262824e-06, "loss": 0.5859, "step": 473 }, { "epoch": 0.33309908643710473, "grad_norm": 2.86635160446167, "learning_rate": 1.990161630358398e-06, "loss": 0.7809, "step": 474 }, { "epoch": 0.3338018271257906, "grad_norm": 4.445220470428467, "learning_rate": 1.994378074490513e-06, "loss": 0.9345, "step": 475 }, { "epoch": 0.33450456781447646, "grad_norm": 2.0127522945404053, "learning_rate": 1.9985945186226285e-06, "loss": 0.4238, "step": 476 }, { "epoch": 0.3352073085031623, "grad_norm": 1.372094988822937, "learning_rate": 2.0028109627547435e-06, "loss": 0.3194, "step": 477 }, { "epoch": 0.3359100491918482, "grad_norm": 1.560210108757019, "learning_rate": 2.007027406886859e-06, "loss": 0.2927, "step": 478 }, { "epoch": 0.33661278988053406, "grad_norm": 3.0371389389038086, "learning_rate": 2.0112438510189742e-06, "loss": 0.2991, "step": 479 }, { "epoch": 0.33731553056922, "grad_norm": 0.955586850643158, "learning_rate": 2.015460295151089e-06, "loss": 0.2938, "step": 480 }, { "epoch": 0.33801827125790584, "grad_norm": 1.1672108173370361, "learning_rate": 2.0196767392832045e-06, "loss": 0.2825, "step": 481 }, { "epoch": 0.3387210119465917, "grad_norm": 6.097266674041748, "learning_rate": 2.0238931834153195e-06, "loss": 0.2913, "step": 482 }, { "epoch": 0.3394237526352776, "grad_norm": 0.9421703219413757, "learning_rate": 2.0281096275474353e-06, "loss": 0.281, "step": 483 }, { "epoch": 0.34012649332396344, "grad_norm": 1.0454938411712646, "learning_rate": 2.03232607167955e-06, "loss": 0.2891, "step": 484 }, { "epoch": 0.3408292340126493, "grad_norm": 0.9471732378005981, "learning_rate": 2.0365425158116656e-06, "loss": 0.2604, "step": 485 }, { "epoch": 0.34153197470133523, "grad_norm": 0.9274056553840637, "learning_rate": 2.040758959943781e-06, "loss": 0.2696, "step": 486 }, { "epoch": 0.3422347153900211, "grad_norm": 1.1382710933685303, "learning_rate": 2.044975404075896e-06, "loss": 0.2688, "step": 487 }, { "epoch": 0.34293745607870696, "grad_norm": 1.9433718919754028, "learning_rate": 2.0491918482080113e-06, "loss": 0.3229, "step": 488 }, { "epoch": 0.3436401967673928, "grad_norm": 1.4058388471603394, "learning_rate": 2.053408292340126e-06, "loss": 0.2696, "step": 489 }, { "epoch": 0.3443429374560787, "grad_norm": 1.8928142786026, "learning_rate": 2.057624736472242e-06, "loss": 0.3121, "step": 490 }, { "epoch": 0.34504567814476456, "grad_norm": 1.08924400806427, "learning_rate": 2.0618411806043574e-06, "loss": 0.3175, "step": 491 }, { "epoch": 0.3457484188334505, "grad_norm": 1.196168303489685, "learning_rate": 2.0660576247364723e-06, "loss": 0.282, "step": 492 }, { "epoch": 0.34645115952213634, "grad_norm": 1.1598204374313354, "learning_rate": 2.0702740688685877e-06, "loss": 0.2859, "step": 493 }, { "epoch": 0.3471539002108222, "grad_norm": 1.4071358442306519, "learning_rate": 2.0744905130007026e-06, "loss": 0.307, "step": 494 }, { "epoch": 0.3478566408995081, "grad_norm": 1.3945677280426025, "learning_rate": 2.078706957132818e-06, "loss": 0.3916, "step": 495 }, { "epoch": 0.34855938158819394, "grad_norm": 1.4980506896972656, "learning_rate": 2.0829234012649333e-06, "loss": 0.3778, "step": 496 }, { "epoch": 0.3492621222768798, "grad_norm": 3.403280019760132, "learning_rate": 2.0871398453970487e-06, "loss": 0.4972, "step": 497 }, { "epoch": 0.3499648629655657, "grad_norm": 2.909106492996216, "learning_rate": 2.091356289529164e-06, "loss": 0.6243, "step": 498 }, { "epoch": 0.3506676036542516, "grad_norm": 3.506939172744751, "learning_rate": 2.095572733661279e-06, "loss": 0.7489, "step": 499 }, { "epoch": 0.35137034434293746, "grad_norm": 5.518486499786377, "learning_rate": 2.0997891777933944e-06, "loss": 0.974, "step": 500 }, { "epoch": 0.3520730850316233, "grad_norm": 1.647868037223816, "learning_rate": 2.1040056219255093e-06, "loss": 0.3733, "step": 501 }, { "epoch": 0.3527758257203092, "grad_norm": 1.2882521152496338, "learning_rate": 2.1082220660576247e-06, "loss": 0.2593, "step": 502 }, { "epoch": 0.35347856640899505, "grad_norm": 1.0451805591583252, "learning_rate": 2.11243851018974e-06, "loss": 0.2461, "step": 503 }, { "epoch": 0.354181307097681, "grad_norm": 1.534587025642395, "learning_rate": 2.116654954321855e-06, "loss": 0.2367, "step": 504 }, { "epoch": 0.35488404778636684, "grad_norm": 1.2103618383407593, "learning_rate": 2.120871398453971e-06, "loss": 0.2419, "step": 505 }, { "epoch": 0.3555867884750527, "grad_norm": 1.031491756439209, "learning_rate": 2.1250878425860857e-06, "loss": 0.2386, "step": 506 }, { "epoch": 0.35628952916373857, "grad_norm": 1.638991117477417, "learning_rate": 2.129304286718201e-06, "loss": 0.2425, "step": 507 }, { "epoch": 0.35699226985242444, "grad_norm": 0.9783702492713928, "learning_rate": 2.1335207308503165e-06, "loss": 0.2435, "step": 508 }, { "epoch": 0.3576950105411103, "grad_norm": 0.892388641834259, "learning_rate": 2.1377371749824314e-06, "loss": 0.2363, "step": 509 }, { "epoch": 0.3583977512297962, "grad_norm": 1.1120541095733643, "learning_rate": 2.141953619114547e-06, "loss": 0.2387, "step": 510 }, { "epoch": 0.3591004919184821, "grad_norm": 1.2327326536178589, "learning_rate": 2.1461700632466617e-06, "loss": 0.254, "step": 511 }, { "epoch": 0.35980323260716796, "grad_norm": 0.8491106629371643, "learning_rate": 2.1503865073787775e-06, "loss": 0.2333, "step": 512 }, { "epoch": 0.3605059732958538, "grad_norm": 1.4187260866165161, "learning_rate": 2.1546029515108925e-06, "loss": 0.2584, "step": 513 }, { "epoch": 0.3612087139845397, "grad_norm": 1.1586085557937622, "learning_rate": 2.158819395643008e-06, "loss": 0.2187, "step": 514 }, { "epoch": 0.36191145467322555, "grad_norm": 1.173498272895813, "learning_rate": 2.163035839775123e-06, "loss": 0.2815, "step": 515 }, { "epoch": 0.3626141953619115, "grad_norm": 1.214332103729248, "learning_rate": 2.167252283907238e-06, "loss": 0.2915, "step": 516 }, { "epoch": 0.36331693605059734, "grad_norm": 1.9353739023208618, "learning_rate": 2.1714687280393535e-06, "loss": 0.2356, "step": 517 }, { "epoch": 0.3640196767392832, "grad_norm": 1.3802398443222046, "learning_rate": 2.1756851721714685e-06, "loss": 0.2457, "step": 518 }, { "epoch": 0.36472241742796907, "grad_norm": 1.520150065422058, "learning_rate": 2.179901616303584e-06, "loss": 0.3225, "step": 519 }, { "epoch": 0.36542515811665494, "grad_norm": 2.037950038909912, "learning_rate": 2.1841180604356996e-06, "loss": 0.3081, "step": 520 }, { "epoch": 0.36612789880534086, "grad_norm": 5.983019828796387, "learning_rate": 2.1883345045678146e-06, "loss": 0.4016, "step": 521 }, { "epoch": 0.3668306394940267, "grad_norm": 2.1199557781219482, "learning_rate": 2.19255094869993e-06, "loss": 0.4748, "step": 522 }, { "epoch": 0.3675333801827126, "grad_norm": 4.840566635131836, "learning_rate": 2.196767392832045e-06, "loss": 0.5863, "step": 523 }, { "epoch": 0.36823612087139845, "grad_norm": 3.1063895225524902, "learning_rate": 2.2009838369641602e-06, "loss": 0.7521, "step": 524 }, { "epoch": 0.3689388615600843, "grad_norm": 5.244124889373779, "learning_rate": 2.2052002810962756e-06, "loss": 0.9015, "step": 525 }, { "epoch": 0.3696416022487702, "grad_norm": 1.682153582572937, "learning_rate": 2.2094167252283905e-06, "loss": 0.3414, "step": 526 }, { "epoch": 0.3703443429374561, "grad_norm": 1.3409557342529297, "learning_rate": 2.2136331693605063e-06, "loss": 0.2246, "step": 527 }, { "epoch": 0.37104708362614197, "grad_norm": 1.0476983785629272, "learning_rate": 2.2178496134926213e-06, "loss": 0.2397, "step": 528 }, { "epoch": 0.37174982431482784, "grad_norm": 0.8534001708030701, "learning_rate": 2.2220660576247366e-06, "loss": 0.1973, "step": 529 }, { "epoch": 0.3724525650035137, "grad_norm": 1.2622076272964478, "learning_rate": 2.2262825017568516e-06, "loss": 0.2217, "step": 530 }, { "epoch": 0.37315530569219957, "grad_norm": 1.1692821979522705, "learning_rate": 2.230498945888967e-06, "loss": 0.206, "step": 531 }, { "epoch": 0.37385804638088543, "grad_norm": 1.0886540412902832, "learning_rate": 2.2347153900210823e-06, "loss": 0.212, "step": 532 }, { "epoch": 0.37456078706957135, "grad_norm": 1.025087833404541, "learning_rate": 2.2389318341531973e-06, "loss": 0.2101, "step": 533 }, { "epoch": 0.3752635277582572, "grad_norm": 1.2183738946914673, "learning_rate": 2.2431482782853126e-06, "loss": 0.2243, "step": 534 }, { "epoch": 0.3759662684469431, "grad_norm": 0.8914302587509155, "learning_rate": 2.247364722417428e-06, "loss": 0.1989, "step": 535 }, { "epoch": 0.37666900913562895, "grad_norm": 0.9912717342376709, "learning_rate": 2.2515811665495434e-06, "loss": 0.208, "step": 536 }, { "epoch": 0.3773717498243148, "grad_norm": 1.0159708261489868, "learning_rate": 2.2557976106816587e-06, "loss": 0.1998, "step": 537 }, { "epoch": 0.3780744905130007, "grad_norm": 1.5961366891860962, "learning_rate": 2.2600140548137737e-06, "loss": 0.2005, "step": 538 }, { "epoch": 0.3787772312016866, "grad_norm": 0.9320374131202698, "learning_rate": 2.264230498945889e-06, "loss": 0.1899, "step": 539 }, { "epoch": 0.37947997189037247, "grad_norm": 1.6641614437103271, "learning_rate": 2.268446943078004e-06, "loss": 0.2291, "step": 540 }, { "epoch": 0.38018271257905834, "grad_norm": 3.386259078979492, "learning_rate": 2.2726633872101194e-06, "loss": 0.2541, "step": 541 }, { "epoch": 0.3808854532677442, "grad_norm": 1.267541766166687, "learning_rate": 2.276879831342235e-06, "loss": 0.2158, "step": 542 }, { "epoch": 0.38158819395643007, "grad_norm": 3.5978894233703613, "learning_rate": 2.28109627547435e-06, "loss": 0.2499, "step": 543 }, { "epoch": 0.38229093464511593, "grad_norm": 1.1906989812850952, "learning_rate": 2.2853127196064655e-06, "loss": 0.2771, "step": 544 }, { "epoch": 0.38299367533380185, "grad_norm": 1.4570783376693726, "learning_rate": 2.2895291637385804e-06, "loss": 0.2973, "step": 545 }, { "epoch": 0.3836964160224877, "grad_norm": 1.4935015439987183, "learning_rate": 2.2937456078706958e-06, "loss": 0.3856, "step": 546 }, { "epoch": 0.3843991567111736, "grad_norm": 1.9042906761169434, "learning_rate": 2.2979620520028107e-06, "loss": 0.4561, "step": 547 }, { "epoch": 0.38510189739985945, "grad_norm": 2.109248161315918, "learning_rate": 2.302178496134926e-06, "loss": 0.5896, "step": 548 }, { "epoch": 0.3858046380885453, "grad_norm": 2.602776527404785, "learning_rate": 2.3063949402670415e-06, "loss": 0.6855, "step": 549 }, { "epoch": 0.3865073787772312, "grad_norm": 3.6300394535064697, "learning_rate": 2.310611384399157e-06, "loss": 0.8849, "step": 550 }, { "epoch": 0.3872101194659171, "grad_norm": 1.724196434020996, "learning_rate": 2.314827828531272e-06, "loss": 0.3389, "step": 551 }, { "epoch": 0.38791286015460297, "grad_norm": 1.0129896402359009, "learning_rate": 2.319044272663387e-06, "loss": 0.2135, "step": 552 }, { "epoch": 0.38861560084328883, "grad_norm": 1.2128665447235107, "learning_rate": 2.3232607167955025e-06, "loss": 0.2295, "step": 553 }, { "epoch": 0.3893183415319747, "grad_norm": 0.891626238822937, "learning_rate": 2.327477160927618e-06, "loss": 0.1956, "step": 554 }, { "epoch": 0.39002108222066056, "grad_norm": 0.9105867147445679, "learning_rate": 2.331693605059733e-06, "loss": 0.1791, "step": 555 }, { "epoch": 0.39072382290934643, "grad_norm": 1.3476903438568115, "learning_rate": 2.335910049191848e-06, "loss": 0.1855, "step": 556 }, { "epoch": 0.39142656359803235, "grad_norm": 1.2565908432006836, "learning_rate": 2.3401264933239635e-06, "loss": 0.2092, "step": 557 }, { "epoch": 0.3921293042867182, "grad_norm": 0.9535101652145386, "learning_rate": 2.344342937456079e-06, "loss": 0.191, "step": 558 }, { "epoch": 0.3928320449754041, "grad_norm": 0.9305250644683838, "learning_rate": 2.3485593815881943e-06, "loss": 0.1892, "step": 559 }, { "epoch": 0.39353478566408995, "grad_norm": 0.8684438467025757, "learning_rate": 2.3527758257203092e-06, "loss": 0.1628, "step": 560 }, { "epoch": 0.3942375263527758, "grad_norm": 1.5766963958740234, "learning_rate": 2.3569922698524246e-06, "loss": 0.2129, "step": 561 }, { "epoch": 0.3949402670414617, "grad_norm": 1.8655157089233398, "learning_rate": 2.3612087139845395e-06, "loss": 0.1853, "step": 562 }, { "epoch": 0.3956430077301476, "grad_norm": 1.3168303966522217, "learning_rate": 2.365425158116655e-06, "loss": 0.2098, "step": 563 }, { "epoch": 0.39634574841883347, "grad_norm": 0.9409052133560181, "learning_rate": 2.3696416022487703e-06, "loss": 0.1953, "step": 564 }, { "epoch": 0.39704848910751933, "grad_norm": 1.0346027612686157, "learning_rate": 2.3738580463808856e-06, "loss": 0.2301, "step": 565 }, { "epoch": 0.3977512297962052, "grad_norm": 1.4325051307678223, "learning_rate": 2.378074490513001e-06, "loss": 0.2188, "step": 566 }, { "epoch": 0.39845397048489106, "grad_norm": 1.0991088151931763, "learning_rate": 2.382290934645116e-06, "loss": 0.2193, "step": 567 }, { "epoch": 0.39915671117357693, "grad_norm": 1.1169688701629639, "learning_rate": 2.3865073787772313e-06, "loss": 0.239, "step": 568 }, { "epoch": 0.39985945186226285, "grad_norm": 1.0416103601455688, "learning_rate": 2.3907238229093463e-06, "loss": 0.2047, "step": 569 }, { "epoch": 0.4005621925509487, "grad_norm": 2.3417274951934814, "learning_rate": 2.3949402670414616e-06, "loss": 0.2461, "step": 570 }, { "epoch": 0.4012649332396346, "grad_norm": 1.4421395063400269, "learning_rate": 2.399156711173577e-06, "loss": 0.3073, "step": 571 }, { "epoch": 0.40196767392832045, "grad_norm": 2.496241807937622, "learning_rate": 2.4033731553056924e-06, "loss": 0.44, "step": 572 }, { "epoch": 0.4026704146170063, "grad_norm": 1.9836231470108032, "learning_rate": 2.4075895994378077e-06, "loss": 0.5579, "step": 573 }, { "epoch": 0.4033731553056922, "grad_norm": 2.51808500289917, "learning_rate": 2.4118060435699227e-06, "loss": 0.7048, "step": 574 }, { "epoch": 0.4040758959943781, "grad_norm": 4.166978359222412, "learning_rate": 2.416022487702038e-06, "loss": 0.9061, "step": 575 }, { "epoch": 0.40477863668306396, "grad_norm": 2.3950002193450928, "learning_rate": 2.4202389318341534e-06, "loss": 0.3513, "step": 576 }, { "epoch": 0.40548137737174983, "grad_norm": 1.004717469215393, "learning_rate": 2.4244553759662683e-06, "loss": 0.1871, "step": 577 }, { "epoch": 0.4061841180604357, "grad_norm": 1.8323557376861572, "learning_rate": 2.4286718200983837e-06, "loss": 0.1996, "step": 578 }, { "epoch": 0.40688685874912156, "grad_norm": 1.0559887886047363, "learning_rate": 2.432888264230499e-06, "loss": 0.1608, "step": 579 }, { "epoch": 0.4075895994378074, "grad_norm": 0.7766121625900269, "learning_rate": 2.4371047083626144e-06, "loss": 0.1792, "step": 580 }, { "epoch": 0.40829234012649335, "grad_norm": 0.8083125352859497, "learning_rate": 2.4413211524947294e-06, "loss": 0.1553, "step": 581 }, { "epoch": 0.4089950808151792, "grad_norm": 0.9437946081161499, "learning_rate": 2.4455375966268448e-06, "loss": 0.1744, "step": 582 }, { "epoch": 0.4096978215038651, "grad_norm": 1.2556443214416504, "learning_rate": 2.44975404075896e-06, "loss": 0.1961, "step": 583 }, { "epoch": 0.41040056219255094, "grad_norm": 1.0488351583480835, "learning_rate": 2.453970484891075e-06, "loss": 0.1684, "step": 584 }, { "epoch": 0.4111033028812368, "grad_norm": 1.1506224870681763, "learning_rate": 2.4581869290231904e-06, "loss": 0.1721, "step": 585 }, { "epoch": 0.4118060435699227, "grad_norm": 1.022011637687683, "learning_rate": 2.4624033731553054e-06, "loss": 0.1968, "step": 586 }, { "epoch": 0.4125087842586086, "grad_norm": 0.8221303224563599, "learning_rate": 2.466619817287421e-06, "loss": 0.185, "step": 587 }, { "epoch": 0.41321152494729446, "grad_norm": 1.2939884662628174, "learning_rate": 2.4708362614195365e-06, "loss": 0.1802, "step": 588 }, { "epoch": 0.4139142656359803, "grad_norm": 1.2740558385849, "learning_rate": 2.4750527055516515e-06, "loss": 0.1689, "step": 589 }, { "epoch": 0.4146170063246662, "grad_norm": 0.9549258351325989, "learning_rate": 2.479269149683767e-06, "loss": 0.1844, "step": 590 }, { "epoch": 0.41531974701335206, "grad_norm": 1.2278658151626587, "learning_rate": 2.483485593815882e-06, "loss": 0.2326, "step": 591 }, { "epoch": 0.4160224877020379, "grad_norm": 1.1770538091659546, "learning_rate": 2.487702037947997e-06, "loss": 0.1824, "step": 592 }, { "epoch": 0.41672522839072385, "grad_norm": 0.9687115550041199, "learning_rate": 2.4919184820801125e-06, "loss": 0.2324, "step": 593 }, { "epoch": 0.4174279690794097, "grad_norm": 1.02065908908844, "learning_rate": 2.496134926212228e-06, "loss": 0.2225, "step": 594 }, { "epoch": 0.4181307097680956, "grad_norm": 4.851129055023193, "learning_rate": 2.5003513703443433e-06, "loss": 0.3234, "step": 595 }, { "epoch": 0.41883345045678144, "grad_norm": 1.1960426568984985, "learning_rate": 2.504567814476458e-06, "loss": 0.2952, "step": 596 }, { "epoch": 0.4195361911454673, "grad_norm": 3.089956521987915, "learning_rate": 2.5087842586085736e-06, "loss": 0.3888, "step": 597 }, { "epoch": 0.4202389318341532, "grad_norm": 4.124706745147705, "learning_rate": 2.5130007027406885e-06, "loss": 0.5439, "step": 598 }, { "epoch": 0.4209416725228391, "grad_norm": 4.463648319244385, "learning_rate": 2.517217146872804e-06, "loss": 0.6672, "step": 599 }, { "epoch": 0.42164441321152496, "grad_norm": 5.114402770996094, "learning_rate": 2.5214335910049192e-06, "loss": 0.8404, "step": 600 }, { "epoch": 0.4223471539002108, "grad_norm": 1.465946078300476, "learning_rate": 2.525650035137034e-06, "loss": 0.2781, "step": 601 }, { "epoch": 0.4230498945888967, "grad_norm": 0.7893495559692383, "learning_rate": 2.52986647926915e-06, "loss": 0.1687, "step": 602 }, { "epoch": 0.42375263527758256, "grad_norm": 0.8495550751686096, "learning_rate": 2.534082923401265e-06, "loss": 0.1612, "step": 603 }, { "epoch": 0.4244553759662684, "grad_norm": 0.6918806433677673, "learning_rate": 2.5382993675333803e-06, "loss": 0.1572, "step": 604 }, { "epoch": 0.42515811665495434, "grad_norm": 0.7797026038169861, "learning_rate": 2.5425158116654957e-06, "loss": 0.1608, "step": 605 }, { "epoch": 0.4258608573436402, "grad_norm": 0.841573178768158, "learning_rate": 2.5467322557976106e-06, "loss": 0.1533, "step": 606 }, { "epoch": 0.4265635980323261, "grad_norm": 1.3865249156951904, "learning_rate": 2.550948699929726e-06, "loss": 0.1535, "step": 607 }, { "epoch": 0.42726633872101194, "grad_norm": 0.9154455065727234, "learning_rate": 2.555165144061841e-06, "loss": 0.1646, "step": 608 }, { "epoch": 0.4279690794096978, "grad_norm": 0.7452192306518555, "learning_rate": 2.5593815881939567e-06, "loss": 0.1604, "step": 609 }, { "epoch": 0.42867182009838367, "grad_norm": 0.8925802111625671, "learning_rate": 2.563598032326072e-06, "loss": 0.1637, "step": 610 }, { "epoch": 0.4293745607870696, "grad_norm": 1.4005306959152222, "learning_rate": 2.567814476458187e-06, "loss": 0.1584, "step": 611 }, { "epoch": 0.43007730147575546, "grad_norm": 1.0018773078918457, "learning_rate": 2.5720309205903024e-06, "loss": 0.1509, "step": 612 }, { "epoch": 0.4307800421644413, "grad_norm": 0.8847672939300537, "learning_rate": 2.5762473647224173e-06, "loss": 0.1956, "step": 613 }, { "epoch": 0.4314827828531272, "grad_norm": 0.9790658354759216, "learning_rate": 2.5804638088545327e-06, "loss": 0.1654, "step": 614 }, { "epoch": 0.43218552354181305, "grad_norm": 0.8484473824501038, "learning_rate": 2.5846802529866476e-06, "loss": 0.1948, "step": 615 }, { "epoch": 0.4328882642304989, "grad_norm": 1.325492262840271, "learning_rate": 2.588896697118763e-06, "loss": 0.1909, "step": 616 }, { "epoch": 0.43359100491918484, "grad_norm": 0.9703555703163147, "learning_rate": 2.593113141250879e-06, "loss": 0.1749, "step": 617 }, { "epoch": 0.4342937456078707, "grad_norm": 0.9667356610298157, "learning_rate": 2.5973295853829937e-06, "loss": 0.2039, "step": 618 }, { "epoch": 0.4349964862965566, "grad_norm": 1.708251953125, "learning_rate": 2.601546029515109e-06, "loss": 0.2165, "step": 619 }, { "epoch": 0.43569922698524244, "grad_norm": 2.2732722759246826, "learning_rate": 2.605762473647224e-06, "loss": 0.2299, "step": 620 }, { "epoch": 0.4364019676739283, "grad_norm": 1.3671408891677856, "learning_rate": 2.6099789177793394e-06, "loss": 0.2795, "step": 621 }, { "epoch": 0.43710470836261417, "grad_norm": 1.9376873970031738, "learning_rate": 2.6141953619114548e-06, "loss": 0.4397, "step": 622 }, { "epoch": 0.4378074490513001, "grad_norm": 2.2976346015930176, "learning_rate": 2.6184118060435697e-06, "loss": 0.4811, "step": 623 }, { "epoch": 0.43851018973998596, "grad_norm": 3.1815149784088135, "learning_rate": 2.6226282501756855e-06, "loss": 0.6684, "step": 624 }, { "epoch": 0.4392129304286718, "grad_norm": 3.7687277793884277, "learning_rate": 2.6268446943078005e-06, "loss": 0.7496, "step": 625 }, { "epoch": 0.4399156711173577, "grad_norm": 1.5753145217895508, "learning_rate": 2.631061138439916e-06, "loss": 0.3047, "step": 626 }, { "epoch": 0.44061841180604355, "grad_norm": 1.0281440019607544, "learning_rate": 2.635277582572031e-06, "loss": 0.1651, "step": 627 }, { "epoch": 0.4413211524947294, "grad_norm": 1.7528101205825806, "learning_rate": 2.639494026704146e-06, "loss": 0.179, "step": 628 }, { "epoch": 0.44202389318341534, "grad_norm": 0.7991352677345276, "learning_rate": 2.6437104708362615e-06, "loss": 0.1317, "step": 629 }, { "epoch": 0.4427266338721012, "grad_norm": 0.6937671303749084, "learning_rate": 2.6479269149683764e-06, "loss": 0.1506, "step": 630 }, { "epoch": 0.44342937456078707, "grad_norm": 0.8777299523353577, "learning_rate": 2.6521433591004922e-06, "loss": 0.1275, "step": 631 }, { "epoch": 0.44413211524947294, "grad_norm": 0.8150135278701782, "learning_rate": 2.656359803232607e-06, "loss": 0.1304, "step": 632 }, { "epoch": 0.4448348559381588, "grad_norm": 0.9041200280189514, "learning_rate": 2.6605762473647226e-06, "loss": 0.1601, "step": 633 }, { "epoch": 0.44553759662684467, "grad_norm": 0.9494258761405945, "learning_rate": 2.664792691496838e-06, "loss": 0.1494, "step": 634 }, { "epoch": 0.4462403373155306, "grad_norm": 0.7912641167640686, "learning_rate": 2.669009135628953e-06, "loss": 0.1516, "step": 635 }, { "epoch": 0.44694307800421645, "grad_norm": 1.0298572778701782, "learning_rate": 2.6732255797610682e-06, "loss": 0.1582, "step": 636 }, { "epoch": 0.4476458186929023, "grad_norm": 0.7855019569396973, "learning_rate": 2.677442023893183e-06, "loss": 0.1344, "step": 637 }, { "epoch": 0.4483485593815882, "grad_norm": 1.2771341800689697, "learning_rate": 2.6816584680252985e-06, "loss": 0.1821, "step": 638 }, { "epoch": 0.44905130007027405, "grad_norm": 0.8297885060310364, "learning_rate": 2.6858749121574143e-06, "loss": 0.1625, "step": 639 }, { "epoch": 0.4497540407589599, "grad_norm": 0.9754140973091125, "learning_rate": 2.6900913562895293e-06, "loss": 0.1601, "step": 640 }, { "epoch": 0.45045678144764584, "grad_norm": 2.096916913986206, "learning_rate": 2.6943078004216446e-06, "loss": 0.176, "step": 641 }, { "epoch": 0.4511595221363317, "grad_norm": 1.0051190853118896, "learning_rate": 2.6985242445537596e-06, "loss": 0.1523, "step": 642 }, { "epoch": 0.45186226282501757, "grad_norm": 3.177295684814453, "learning_rate": 2.702740688685875e-06, "loss": 0.2158, "step": 643 }, { "epoch": 0.45256500351370343, "grad_norm": 1.3347276449203491, "learning_rate": 2.7069571328179903e-06, "loss": 0.2417, "step": 644 }, { "epoch": 0.4532677442023893, "grad_norm": 1.1815751791000366, "learning_rate": 2.7111735769501053e-06, "loss": 0.244, "step": 645 }, { "epoch": 0.45397048489107517, "grad_norm": 1.3109254837036133, "learning_rate": 2.715390021082221e-06, "loss": 0.2925, "step": 646 }, { "epoch": 0.4546732255797611, "grad_norm": 2.4989492893218994, "learning_rate": 2.719606465214336e-06, "loss": 0.3517, "step": 647 }, { "epoch": 0.45537596626844695, "grad_norm": 2.7492482662200928, "learning_rate": 2.7238229093464514e-06, "loss": 0.475, "step": 648 }, { "epoch": 0.4560787069571328, "grad_norm": 6.691035747528076, "learning_rate": 2.7280393534785663e-06, "loss": 0.584, "step": 649 }, { "epoch": 0.4567814476458187, "grad_norm": 4.115723609924316, "learning_rate": 2.7322557976106817e-06, "loss": 0.7874, "step": 650 }, { "epoch": 0.45748418833450455, "grad_norm": 2.319666862487793, "learning_rate": 2.736472241742797e-06, "loss": 0.2005, "step": 651 }, { "epoch": 0.45818692902319047, "grad_norm": 0.9737160801887512, "learning_rate": 2.740688685874912e-06, "loss": 0.1767, "step": 652 }, { "epoch": 0.45888966971187634, "grad_norm": 0.8475993275642395, "learning_rate": 2.7449051300070274e-06, "loss": 0.1698, "step": 653 }, { "epoch": 0.4595924104005622, "grad_norm": 0.6247663497924805, "learning_rate": 2.7491215741391427e-06, "loss": 0.1268, "step": 654 }, { "epoch": 0.46029515108924807, "grad_norm": 0.6768312454223633, "learning_rate": 2.753338018271258e-06, "loss": 0.1475, "step": 655 }, { "epoch": 0.46099789177793393, "grad_norm": 0.6792594194412231, "learning_rate": 2.7575544624033735e-06, "loss": 0.1268, "step": 656 }, { "epoch": 0.4617006324666198, "grad_norm": 0.8406147360801697, "learning_rate": 2.7617709065354884e-06, "loss": 0.1425, "step": 657 }, { "epoch": 0.4624033731553057, "grad_norm": 1.201116681098938, "learning_rate": 2.7659873506676038e-06, "loss": 0.1425, "step": 658 }, { "epoch": 0.4631061138439916, "grad_norm": 1.224636435508728, "learning_rate": 2.7702037947997187e-06, "loss": 0.1246, "step": 659 }, { "epoch": 0.46380885453267745, "grad_norm": 0.7555993795394897, "learning_rate": 2.774420238931834e-06, "loss": 0.1297, "step": 660 }, { "epoch": 0.4645115952213633, "grad_norm": 2.518109083175659, "learning_rate": 2.77863668306395e-06, "loss": 0.185, "step": 661 }, { "epoch": 0.4652143359100492, "grad_norm": 1.0680265426635742, "learning_rate": 2.782853127196065e-06, "loss": 0.1524, "step": 662 }, { "epoch": 0.46591707659873505, "grad_norm": 1.0682122707366943, "learning_rate": 2.78706957132818e-06, "loss": 0.1842, "step": 663 }, { "epoch": 0.46661981728742097, "grad_norm": 1.2808083295822144, "learning_rate": 2.791286015460295e-06, "loss": 0.1445, "step": 664 }, { "epoch": 0.46732255797610683, "grad_norm": 0.9070841073989868, "learning_rate": 2.7955024595924105e-06, "loss": 0.1761, "step": 665 }, { "epoch": 0.4680252986647927, "grad_norm": 1.2612248659133911, "learning_rate": 2.7997189037245254e-06, "loss": 0.1611, "step": 666 }, { "epoch": 0.46872803935347856, "grad_norm": 1.0169768333435059, "learning_rate": 2.803935347856641e-06, "loss": 0.1442, "step": 667 }, { "epoch": 0.46943078004216443, "grad_norm": 0.9508179426193237, "learning_rate": 2.808151791988756e-06, "loss": 0.196, "step": 668 }, { "epoch": 0.4701335207308503, "grad_norm": 1.1681957244873047, "learning_rate": 2.8123682361208715e-06, "loss": 0.2177, "step": 669 }, { "epoch": 0.4708362614195362, "grad_norm": 2.154606819152832, "learning_rate": 2.816584680252987e-06, "loss": 0.2554, "step": 670 }, { "epoch": 0.4715390021082221, "grad_norm": 1.1213093996047974, "learning_rate": 2.820801124385102e-06, "loss": 0.2555, "step": 671 }, { "epoch": 0.47224174279690795, "grad_norm": 1.8086296319961548, "learning_rate": 2.825017568517217e-06, "loss": 0.3307, "step": 672 }, { "epoch": 0.4729444834855938, "grad_norm": 2.220069408416748, "learning_rate": 2.8292340126493326e-06, "loss": 0.4965, "step": 673 }, { "epoch": 0.4736472241742797, "grad_norm": 2.49110746383667, "learning_rate": 2.8334504567814475e-06, "loss": 0.5523, "step": 674 }, { "epoch": 0.47434996486296555, "grad_norm": 6.42268705368042, "learning_rate": 2.837666900913563e-06, "loss": 0.8182, "step": 675 }, { "epoch": 0.47505270555165147, "grad_norm": 1.0963033437728882, "learning_rate": 2.8418833450456783e-06, "loss": 0.2503, "step": 676 }, { "epoch": 0.47575544624033733, "grad_norm": 0.9607471227645874, "learning_rate": 2.8460997891777936e-06, "loss": 0.1774, "step": 677 }, { "epoch": 0.4764581869290232, "grad_norm": 0.852593183517456, "learning_rate": 2.850316233309909e-06, "loss": 0.1385, "step": 678 }, { "epoch": 0.47716092761770906, "grad_norm": 0.8270108103752136, "learning_rate": 2.854532677442024e-06, "loss": 0.1453, "step": 679 }, { "epoch": 0.47786366830639493, "grad_norm": 0.7422179579734802, "learning_rate": 2.8587491215741393e-06, "loss": 0.1331, "step": 680 }, { "epoch": 0.4785664089950808, "grad_norm": 0.6209441423416138, "learning_rate": 2.8629655657062542e-06, "loss": 0.1179, "step": 681 }, { "epoch": 0.4792691496837667, "grad_norm": 0.8487251400947571, "learning_rate": 2.8671820098383696e-06, "loss": 0.1161, "step": 682 }, { "epoch": 0.4799718903724526, "grad_norm": 0.8047341108322144, "learning_rate": 2.871398453970485e-06, "loss": 0.1282, "step": 683 }, { "epoch": 0.48067463106113845, "grad_norm": 0.898066520690918, "learning_rate": 2.8756148981026003e-06, "loss": 0.1671, "step": 684 }, { "epoch": 0.4813773717498243, "grad_norm": 0.661515474319458, "learning_rate": 2.8798313422347157e-06, "loss": 0.1201, "step": 685 }, { "epoch": 0.4820801124385102, "grad_norm": 0.9009406566619873, "learning_rate": 2.8840477863668307e-06, "loss": 0.1384, "step": 686 }, { "epoch": 0.48278285312719604, "grad_norm": 0.9405791759490967, "learning_rate": 2.888264230498946e-06, "loss": 0.1385, "step": 687 }, { "epoch": 0.48348559381588196, "grad_norm": 0.9587562084197998, "learning_rate": 2.892480674631061e-06, "loss": 0.1592, "step": 688 }, { "epoch": 0.48418833450456783, "grad_norm": 0.7887436151504517, "learning_rate": 2.8966971187631763e-06, "loss": 0.1361, "step": 689 }, { "epoch": 0.4848910751932537, "grad_norm": 2.3427436351776123, "learning_rate": 2.9009135628952917e-06, "loss": 0.1647, "step": 690 }, { "epoch": 0.48559381588193956, "grad_norm": 4.425685882568359, "learning_rate": 2.905130007027407e-06, "loss": 0.1615, "step": 691 }, { "epoch": 0.4862965565706254, "grad_norm": 1.177504539489746, "learning_rate": 2.9093464511595224e-06, "loss": 0.1539, "step": 692 }, { "epoch": 0.4869992972593113, "grad_norm": 1.0210630893707275, "learning_rate": 2.9135628952916374e-06, "loss": 0.1752, "step": 693 }, { "epoch": 0.4877020379479972, "grad_norm": 1.0634220838546753, "learning_rate": 2.9177793394237527e-06, "loss": 0.2265, "step": 694 }, { "epoch": 0.4884047786366831, "grad_norm": 1.1697547435760498, "learning_rate": 2.921995783555868e-06, "loss": 0.22, "step": 695 }, { "epoch": 0.48910751932536894, "grad_norm": 3.5280144214630127, "learning_rate": 2.926212227687983e-06, "loss": 0.3005, "step": 696 }, { "epoch": 0.4898102600140548, "grad_norm": 11.846404075622559, "learning_rate": 2.9304286718200984e-06, "loss": 0.3458, "step": 697 }, { "epoch": 0.4905130007027407, "grad_norm": 2.4963691234588623, "learning_rate": 2.934645115952214e-06, "loss": 0.503, "step": 698 }, { "epoch": 0.49121574139142654, "grad_norm": 2.6190454959869385, "learning_rate": 2.938861560084329e-06, "loss": 0.5531, "step": 699 }, { "epoch": 0.49191848208011246, "grad_norm": 4.857889175415039, "learning_rate": 2.943078004216444e-06, "loss": 0.8262, "step": 700 }, { "epoch": 0.49262122276879833, "grad_norm": 1.0258433818817139, "learning_rate": 2.9472944483485595e-06, "loss": 0.2167, "step": 701 }, { "epoch": 0.4933239634574842, "grad_norm": 0.6831320524215698, "learning_rate": 2.951510892480675e-06, "loss": 0.14, "step": 702 }, { "epoch": 0.49402670414617006, "grad_norm": 0.6878474354743958, "learning_rate": 2.9557273366127898e-06, "loss": 0.1419, "step": 703 }, { "epoch": 0.4947294448348559, "grad_norm": 0.7155750393867493, "learning_rate": 2.959943780744905e-06, "loss": 0.1233, "step": 704 }, { "epoch": 0.4954321855235418, "grad_norm": 0.7772009372711182, "learning_rate": 2.96416022487702e-06, "loss": 0.1237, "step": 705 }, { "epoch": 0.4961349262122277, "grad_norm": 0.696556806564331, "learning_rate": 2.968376669009136e-06, "loss": 0.1114, "step": 706 }, { "epoch": 0.4968376669009136, "grad_norm": 0.7684633731842041, "learning_rate": 2.9725931131412512e-06, "loss": 0.1196, "step": 707 }, { "epoch": 0.49754040758959944, "grad_norm": 0.7977069616317749, "learning_rate": 2.976809557273366e-06, "loss": 0.136, "step": 708 }, { "epoch": 0.4982431482782853, "grad_norm": 0.8873041272163391, "learning_rate": 2.9810260014054816e-06, "loss": 0.1255, "step": 709 }, { "epoch": 0.4989458889669712, "grad_norm": 0.793103039264679, "learning_rate": 2.9852424455375965e-06, "loss": 0.0975, "step": 710 }, { "epoch": 0.49964862965565704, "grad_norm": 1.326518177986145, "learning_rate": 2.989458889669712e-06, "loss": 0.1258, "step": 711 }, { "epoch": 0.500351370344343, "grad_norm": 1.3123624324798584, "learning_rate": 2.9936753338018272e-06, "loss": 0.1306, "step": 712 }, { "epoch": 0.5010541110330288, "grad_norm": 0.8508638739585876, "learning_rate": 2.9978917779339426e-06, "loss": 0.1445, "step": 713 }, { "epoch": 0.5017568517217147, "grad_norm": 0.7881466746330261, "learning_rate": 3.002108222066058e-06, "loss": 0.1172, "step": 714 }, { "epoch": 0.5024595924104006, "grad_norm": 1.2354596853256226, "learning_rate": 3.006324666198173e-06, "loss": 0.2277, "step": 715 }, { "epoch": 0.5031623330990864, "grad_norm": 1.1032758951187134, "learning_rate": 3.0105411103302883e-06, "loss": 0.1521, "step": 716 }, { "epoch": 0.5038650737877723, "grad_norm": 3.4896156787872314, "learning_rate": 3.0147575544624032e-06, "loss": 0.1416, "step": 717 }, { "epoch": 0.5045678144764582, "grad_norm": 1.2280385494232178, "learning_rate": 3.0189739985945186e-06, "loss": 0.2127, "step": 718 }, { "epoch": 0.5052705551651441, "grad_norm": 2.6671032905578613, "learning_rate": 3.023190442726634e-06, "loss": 0.1946, "step": 719 }, { "epoch": 0.5059732958538299, "grad_norm": 1.1215647459030151, "learning_rate": 3.027406886858749e-06, "loss": 0.2291, "step": 720 }, { "epoch": 0.5066760365425158, "grad_norm": 1.7654212713241577, "learning_rate": 3.0316233309908647e-06, "loss": 0.2995, "step": 721 }, { "epoch": 0.5073787772312017, "grad_norm": 2.730743408203125, "learning_rate": 3.0358397751229796e-06, "loss": 0.3752, "step": 722 }, { "epoch": 0.5080815179198875, "grad_norm": 2.1216726303100586, "learning_rate": 3.040056219255095e-06, "loss": 0.4455, "step": 723 }, { "epoch": 0.5087842586085735, "grad_norm": 2.4707770347595215, "learning_rate": 3.0442726633872104e-06, "loss": 0.606, "step": 724 }, { "epoch": 0.5094869992972593, "grad_norm": 5.531240940093994, "learning_rate": 3.0484891075193253e-06, "loss": 0.7824, "step": 725 }, { "epoch": 0.5101897399859452, "grad_norm": 1.4698859453201294, "learning_rate": 3.0527055516514407e-06, "loss": 0.2054, "step": 726 }, { "epoch": 0.5108924806746311, "grad_norm": 0.7632827162742615, "learning_rate": 3.0569219957835556e-06, "loss": 0.1254, "step": 727 }, { "epoch": 0.5115952213633169, "grad_norm": 0.8354719281196594, "learning_rate": 3.0611384399156714e-06, "loss": 0.1235, "step": 728 }, { "epoch": 0.5122979620520028, "grad_norm": 1.2796456813812256, "learning_rate": 3.0653548840477868e-06, "loss": 0.1317, "step": 729 }, { "epoch": 0.5130007027406887, "grad_norm": 1.3533183336257935, "learning_rate": 3.0695713281799017e-06, "loss": 0.1219, "step": 730 }, { "epoch": 0.5137034434293746, "grad_norm": 0.722769021987915, "learning_rate": 3.073787772312017e-06, "loss": 0.1233, "step": 731 }, { "epoch": 0.5144061841180604, "grad_norm": 0.9259893298149109, "learning_rate": 3.078004216444132e-06, "loss": 0.1499, "step": 732 }, { "epoch": 0.5151089248067463, "grad_norm": 0.7433314919471741, "learning_rate": 3.0822206605762474e-06, "loss": 0.1239, "step": 733 }, { "epoch": 0.5158116654954322, "grad_norm": 0.693092942237854, "learning_rate": 3.0864371047083623e-06, "loss": 0.1082, "step": 734 }, { "epoch": 0.516514406184118, "grad_norm": 0.666925311088562, "learning_rate": 3.0906535488404777e-06, "loss": 0.1035, "step": 735 }, { "epoch": 0.517217146872804, "grad_norm": 0.6531984806060791, "learning_rate": 3.0948699929725935e-06, "loss": 0.1173, "step": 736 }, { "epoch": 0.5179198875614898, "grad_norm": 0.7211617231369019, "learning_rate": 3.0990864371047085e-06, "loss": 0.1302, "step": 737 }, { "epoch": 0.5186226282501757, "grad_norm": 1.4570786952972412, "learning_rate": 3.103302881236824e-06, "loss": 0.1452, "step": 738 }, { "epoch": 0.5193253689388616, "grad_norm": 1.1994516849517822, "learning_rate": 3.1075193253689388e-06, "loss": 0.1478, "step": 739 }, { "epoch": 0.5200281096275474, "grad_norm": 0.9215148687362671, "learning_rate": 3.111735769501054e-06, "loss": 0.1508, "step": 740 }, { "epoch": 0.5207308503162333, "grad_norm": 0.7242789268493652, "learning_rate": 3.1159522136331695e-06, "loss": 0.1353, "step": 741 }, { "epoch": 0.5214335910049192, "grad_norm": 1.4342468976974487, "learning_rate": 3.1201686577652844e-06, "loss": 0.1355, "step": 742 }, { "epoch": 0.5221363316936051, "grad_norm": 0.8839058876037598, "learning_rate": 3.1243851018974002e-06, "loss": 0.1735, "step": 743 }, { "epoch": 0.5228390723822909, "grad_norm": 1.1299434900283813, "learning_rate": 3.128601546029515e-06, "loss": 0.1878, "step": 744 }, { "epoch": 0.5235418130709768, "grad_norm": 1.4848209619522095, "learning_rate": 3.1328179901616305e-06, "loss": 0.193, "step": 745 }, { "epoch": 0.5242445537596627, "grad_norm": 1.4934059381484985, "learning_rate": 3.137034434293746e-06, "loss": 0.2518, "step": 746 }, { "epoch": 0.5249472944483485, "grad_norm": 1.5765827894210815, "learning_rate": 3.141250878425861e-06, "loss": 0.3678, "step": 747 }, { "epoch": 0.5256500351370345, "grad_norm": 1.7463819980621338, "learning_rate": 3.1454673225579762e-06, "loss": 0.4051, "step": 748 }, { "epoch": 0.5263527758257203, "grad_norm": 2.4877164363861084, "learning_rate": 3.149683766690091e-06, "loss": 0.5429, "step": 749 }, { "epoch": 0.5270555165144062, "grad_norm": 3.1468663215637207, "learning_rate": 3.153900210822207e-06, "loss": 0.689, "step": 750 }, { "epoch": 0.5277582572030921, "grad_norm": 1.1370480060577393, "learning_rate": 3.158116654954322e-06, "loss": 0.2239, "step": 751 }, { "epoch": 0.5284609978917779, "grad_norm": 0.5513385534286499, "learning_rate": 3.1623330990864373e-06, "loss": 0.1175, "step": 752 }, { "epoch": 0.5291637385804638, "grad_norm": 0.7069596648216248, "learning_rate": 3.1665495432185526e-06, "loss": 0.1161, "step": 753 }, { "epoch": 0.5298664792691496, "grad_norm": 0.5105127096176147, "learning_rate": 3.1707659873506676e-06, "loss": 0.1198, "step": 754 }, { "epoch": 0.5305692199578356, "grad_norm": 0.6914659738540649, "learning_rate": 3.174982431482783e-06, "loss": 0.1009, "step": 755 }, { "epoch": 0.5312719606465214, "grad_norm": 0.6749956011772156, "learning_rate": 3.179198875614898e-06, "loss": 0.1385, "step": 756 }, { "epoch": 0.5319747013352073, "grad_norm": 0.6555359959602356, "learning_rate": 3.1834153197470133e-06, "loss": 0.1244, "step": 757 }, { "epoch": 0.5326774420238932, "grad_norm": 0.7730734944343567, "learning_rate": 3.187631763879129e-06, "loss": 0.1059, "step": 758 }, { "epoch": 0.533380182712579, "grad_norm": 0.8112127780914307, "learning_rate": 3.191848208011244e-06, "loss": 0.1257, "step": 759 }, { "epoch": 0.534082923401265, "grad_norm": 0.5587027072906494, "learning_rate": 3.1960646521433594e-06, "loss": 0.1167, "step": 760 }, { "epoch": 0.5347856640899508, "grad_norm": 0.6424028873443604, "learning_rate": 3.2002810962754743e-06, "loss": 0.118, "step": 761 }, { "epoch": 0.5354884047786367, "grad_norm": 0.8763124346733093, "learning_rate": 3.2044975404075897e-06, "loss": 0.1145, "step": 762 }, { "epoch": 0.5361911454673226, "grad_norm": 0.6753503680229187, "learning_rate": 3.208713984539705e-06, "loss": 0.1156, "step": 763 }, { "epoch": 0.5368938861560084, "grad_norm": 4.192592620849609, "learning_rate": 3.21293042867182e-06, "loss": 0.119, "step": 764 }, { "epoch": 0.5375966268446943, "grad_norm": 0.7214124798774719, "learning_rate": 3.2171468728039358e-06, "loss": 0.1357, "step": 765 }, { "epoch": 0.5382993675333801, "grad_norm": 0.8733885884284973, "learning_rate": 3.2213633169360507e-06, "loss": 0.1516, "step": 766 }, { "epoch": 0.5390021082220661, "grad_norm": 0.8330471515655518, "learning_rate": 3.225579761068166e-06, "loss": 0.1506, "step": 767 }, { "epoch": 0.5397048489107519, "grad_norm": 0.8574863076210022, "learning_rate": 3.229796205200281e-06, "loss": 0.1508, "step": 768 }, { "epoch": 0.5404075895994378, "grad_norm": 0.9528377056121826, "learning_rate": 3.2340126493323964e-06, "loss": 0.1555, "step": 769 }, { "epoch": 0.5411103302881237, "grad_norm": 0.8411085605621338, "learning_rate": 3.2382290934645118e-06, "loss": 0.181, "step": 770 }, { "epoch": 0.5418130709768095, "grad_norm": 1.642541527748108, "learning_rate": 3.2424455375966267e-06, "loss": 0.2611, "step": 771 }, { "epoch": 0.5425158116654955, "grad_norm": 1.1532258987426758, "learning_rate": 3.246661981728742e-06, "loss": 0.2992, "step": 772 }, { "epoch": 0.5432185523541813, "grad_norm": 2.0559256076812744, "learning_rate": 3.2508784258608574e-06, "loss": 0.4995, "step": 773 }, { "epoch": 0.5439212930428672, "grad_norm": 2.6020667552948, "learning_rate": 3.255094869992973e-06, "loss": 0.5374, "step": 774 }, { "epoch": 0.5446240337315531, "grad_norm": 5.6077070236206055, "learning_rate": 3.259311314125088e-06, "loss": 0.798, "step": 775 }, { "epoch": 0.5453267744202389, "grad_norm": 0.8314149975776672, "learning_rate": 3.263527758257203e-06, "loss": 0.188, "step": 776 }, { "epoch": 0.5460295151089248, "grad_norm": 0.6235130429267883, "learning_rate": 3.2677442023893185e-06, "loss": 0.1147, "step": 777 }, { "epoch": 0.5467322557976106, "grad_norm": 0.5394691228866577, "learning_rate": 3.2719606465214334e-06, "loss": 0.1051, "step": 778 }, { "epoch": 0.5474349964862966, "grad_norm": 0.5729040503501892, "learning_rate": 3.276177090653549e-06, "loss": 0.1213, "step": 779 }, { "epoch": 0.5481377371749825, "grad_norm": 0.6419472694396973, "learning_rate": 3.280393534785664e-06, "loss": 0.0933, "step": 780 }, { "epoch": 0.5488404778636683, "grad_norm": 0.7696680426597595, "learning_rate": 3.2846099789177795e-06, "loss": 0.104, "step": 781 }, { "epoch": 0.5495432185523542, "grad_norm": 0.5462688207626343, "learning_rate": 3.288826423049895e-06, "loss": 0.0891, "step": 782 }, { "epoch": 0.55024595924104, "grad_norm": 1.266863465309143, "learning_rate": 3.29304286718201e-06, "loss": 0.13, "step": 783 }, { "epoch": 0.550948699929726, "grad_norm": 0.7956530451774597, "learning_rate": 3.297259311314125e-06, "loss": 0.1028, "step": 784 }, { "epoch": 0.5516514406184118, "grad_norm": 0.496067076921463, "learning_rate": 3.30147575544624e-06, "loss": 0.0941, "step": 785 }, { "epoch": 0.5523541813070977, "grad_norm": 0.9968611598014832, "learning_rate": 3.3056921995783555e-06, "loss": 0.1375, "step": 786 }, { "epoch": 0.5530569219957836, "grad_norm": 0.6641092896461487, "learning_rate": 3.309908643710471e-06, "loss": 0.1049, "step": 787 }, { "epoch": 0.5537596626844694, "grad_norm": 0.8251298069953918, "learning_rate": 3.3141250878425862e-06, "loss": 0.137, "step": 788 }, { "epoch": 0.5544624033731553, "grad_norm": 0.5841253399848938, "learning_rate": 3.3183415319747016e-06, "loss": 0.101, "step": 789 }, { "epoch": 0.5551651440618411, "grad_norm": 0.6463255882263184, "learning_rate": 3.3225579761068166e-06, "loss": 0.1274, "step": 790 }, { "epoch": 0.5558678847505271, "grad_norm": 0.7590716481208801, "learning_rate": 3.326774420238932e-06, "loss": 0.1297, "step": 791 }, { "epoch": 0.556570625439213, "grad_norm": 0.5651037693023682, "learning_rate": 3.3309908643710473e-06, "loss": 0.0991, "step": 792 }, { "epoch": 0.5572733661278988, "grad_norm": 1.0757735967636108, "learning_rate": 3.3352073085031622e-06, "loss": 0.1368, "step": 793 }, { "epoch": 0.5579761068165847, "grad_norm": 1.0692675113677979, "learning_rate": 3.3394237526352776e-06, "loss": 0.1731, "step": 794 }, { "epoch": 0.5586788475052705, "grad_norm": 1.099554419517517, "learning_rate": 3.343640196767393e-06, "loss": 0.1756, "step": 795 }, { "epoch": 0.5593815881939564, "grad_norm": 1.1107561588287354, "learning_rate": 3.3478566408995083e-06, "loss": 0.2818, "step": 796 }, { "epoch": 0.5600843288826423, "grad_norm": 1.7432169914245605, "learning_rate": 3.3520730850316233e-06, "loss": 0.3025, "step": 797 }, { "epoch": 0.5607870695713282, "grad_norm": 2.307389974594116, "learning_rate": 3.3562895291637386e-06, "loss": 0.4207, "step": 798 }, { "epoch": 0.5614898102600141, "grad_norm": 2.3873517513275146, "learning_rate": 3.360505973295854e-06, "loss": 0.6081, "step": 799 }, { "epoch": 0.5621925509486999, "grad_norm": 10.166254043579102, "learning_rate": 3.364722417427969e-06, "loss": 0.6059, "step": 800 }, { "epoch": 0.5628952916373858, "grad_norm": 1.087896704673767, "learning_rate": 3.3689388615600843e-06, "loss": 0.206, "step": 801 }, { "epoch": 0.5635980323260716, "grad_norm": 0.5982365608215332, "learning_rate": 3.3731553056921993e-06, "loss": 0.1037, "step": 802 }, { "epoch": 0.5643007730147576, "grad_norm": 0.4677375555038452, "learning_rate": 3.377371749824315e-06, "loss": 0.0956, "step": 803 }, { "epoch": 0.5650035137034435, "grad_norm": 0.7486531734466553, "learning_rate": 3.3815881939564304e-06, "loss": 0.1183, "step": 804 }, { "epoch": 0.5657062543921293, "grad_norm": 0.6223856806755066, "learning_rate": 3.3858046380885454e-06, "loss": 0.0969, "step": 805 }, { "epoch": 0.5664089950808152, "grad_norm": 0.5974777340888977, "learning_rate": 3.3900210822206607e-06, "loss": 0.0978, "step": 806 }, { "epoch": 0.567111735769501, "grad_norm": 1.3215612173080444, "learning_rate": 3.3942375263527757e-06, "loss": 0.097, "step": 807 }, { "epoch": 0.567814476458187, "grad_norm": 0.9233817458152771, "learning_rate": 3.398453970484891e-06, "loss": 0.1094, "step": 808 }, { "epoch": 0.5685172171468728, "grad_norm": 0.7397370934486389, "learning_rate": 3.4026704146170064e-06, "loss": 0.1174, "step": 809 }, { "epoch": 0.5692199578355587, "grad_norm": 0.6862801909446716, "learning_rate": 3.4068868587491218e-06, "loss": 0.1183, "step": 810 }, { "epoch": 0.5699226985242446, "grad_norm": 0.9587000608444214, "learning_rate": 3.411103302881237e-06, "loss": 0.1241, "step": 811 }, { "epoch": 0.5706254392129304, "grad_norm": 0.8270710706710815, "learning_rate": 3.415319747013352e-06, "loss": 0.1086, "step": 812 }, { "epoch": 0.5713281799016163, "grad_norm": 0.6584486365318298, "learning_rate": 3.4195361911454675e-06, "loss": 0.1305, "step": 813 }, { "epoch": 0.5720309205903021, "grad_norm": 0.7942273616790771, "learning_rate": 3.423752635277583e-06, "loss": 0.1205, "step": 814 }, { "epoch": 0.5727336612789881, "grad_norm": 0.65423583984375, "learning_rate": 3.4279690794096978e-06, "loss": 0.1137, "step": 815 }, { "epoch": 0.573436401967674, "grad_norm": 0.9472881555557251, "learning_rate": 3.432185523541813e-06, "loss": 0.1398, "step": 816 }, { "epoch": 0.5741391426563598, "grad_norm": 0.8188420534133911, "learning_rate": 3.4364019676739285e-06, "loss": 0.1082, "step": 817 }, { "epoch": 0.5748418833450457, "grad_norm": 1.0203806161880493, "learning_rate": 3.440618411806044e-06, "loss": 0.1636, "step": 818 }, { "epoch": 0.5755446240337315, "grad_norm": 1.5333151817321777, "learning_rate": 3.444834855938159e-06, "loss": 0.1906, "step": 819 }, { "epoch": 0.5762473647224174, "grad_norm": 1.1490001678466797, "learning_rate": 3.449051300070274e-06, "loss": 0.1916, "step": 820 }, { "epoch": 0.5769501054111033, "grad_norm": 1.7868757247924805, "learning_rate": 3.4532677442023895e-06, "loss": 0.2134, "step": 821 }, { "epoch": 0.5776528460997892, "grad_norm": 2.2091543674468994, "learning_rate": 3.4574841883345045e-06, "loss": 0.3466, "step": 822 }, { "epoch": 0.5783555867884751, "grad_norm": 1.9156982898712158, "learning_rate": 3.46170063246662e-06, "loss": 0.4515, "step": 823 }, { "epoch": 0.5790583274771609, "grad_norm": 2.353231430053711, "learning_rate": 3.465917076598735e-06, "loss": 0.5273, "step": 824 }, { "epoch": 0.5797610681658468, "grad_norm": 3.2292516231536865, "learning_rate": 3.4701335207308506e-06, "loss": 0.7569, "step": 825 }, { "epoch": 0.5804638088545326, "grad_norm": 0.9536106586456299, "learning_rate": 3.474349964862966e-06, "loss": 0.215, "step": 826 }, { "epoch": 0.5811665495432186, "grad_norm": 0.5718256235122681, "learning_rate": 3.478566408995081e-06, "loss": 0.1297, "step": 827 }, { "epoch": 0.5818692902319045, "grad_norm": 0.6399689316749573, "learning_rate": 3.4827828531271963e-06, "loss": 0.106, "step": 828 }, { "epoch": 0.5825720309205903, "grad_norm": 0.5247586369514465, "learning_rate": 3.4869992972593112e-06, "loss": 0.1018, "step": 829 }, { "epoch": 0.5832747716092762, "grad_norm": 0.668315589427948, "learning_rate": 3.4912157413914266e-06, "loss": 0.102, "step": 830 }, { "epoch": 0.583977512297962, "grad_norm": 0.7566106915473938, "learning_rate": 3.495432185523542e-06, "loss": 0.0831, "step": 831 }, { "epoch": 0.5846802529866479, "grad_norm": 0.586768388748169, "learning_rate": 3.4996486296556573e-06, "loss": 0.0997, "step": 832 }, { "epoch": 0.5853829936753338, "grad_norm": 1.1252340078353882, "learning_rate": 3.5038650737877727e-06, "loss": 0.102, "step": 833 }, { "epoch": 0.5860857343640197, "grad_norm": 0.6338616013526917, "learning_rate": 3.5080815179198876e-06, "loss": 0.1104, "step": 834 }, { "epoch": 0.5867884750527056, "grad_norm": 0.836013913154602, "learning_rate": 3.512297962052003e-06, "loss": 0.0908, "step": 835 }, { "epoch": 0.5874912157413914, "grad_norm": 0.5660910606384277, "learning_rate": 3.516514406184118e-06, "loss": 0.1073, "step": 836 }, { "epoch": 0.5881939564300773, "grad_norm": 0.8270347118377686, "learning_rate": 3.5207308503162333e-06, "loss": 0.1231, "step": 837 }, { "epoch": 0.5888966971187631, "grad_norm": 0.7229700088500977, "learning_rate": 3.5249472944483487e-06, "loss": 0.1289, "step": 838 }, { "epoch": 0.5895994378074491, "grad_norm": 0.7222298383712769, "learning_rate": 3.5291637385804636e-06, "loss": 0.1032, "step": 839 }, { "epoch": 0.590302178496135, "grad_norm": 0.7270954251289368, "learning_rate": 3.5333801827125794e-06, "loss": 0.1066, "step": 840 }, { "epoch": 0.5910049191848208, "grad_norm": 0.7893152236938477, "learning_rate": 3.5375966268446944e-06, "loss": 0.1197, "step": 841 }, { "epoch": 0.5917076598735067, "grad_norm": 0.7092378735542297, "learning_rate": 3.5418130709768097e-06, "loss": 0.1026, "step": 842 }, { "epoch": 0.5924104005621925, "grad_norm": 0.905844509601593, "learning_rate": 3.546029515108925e-06, "loss": 0.1568, "step": 843 }, { "epoch": 0.5931131412508784, "grad_norm": 0.6640276312828064, "learning_rate": 3.55024595924104e-06, "loss": 0.145, "step": 844 }, { "epoch": 0.5938158819395642, "grad_norm": 1.3228414058685303, "learning_rate": 3.5544624033731554e-06, "loss": 0.197, "step": 845 }, { "epoch": 0.5945186226282502, "grad_norm": 1.4411686658859253, "learning_rate": 3.5586788475052703e-06, "loss": 0.2453, "step": 846 }, { "epoch": 0.5952213633169361, "grad_norm": 1.3694274425506592, "learning_rate": 3.562895291637386e-06, "loss": 0.3058, "step": 847 }, { "epoch": 0.5959241040056219, "grad_norm": 1.6159627437591553, "learning_rate": 3.567111735769501e-06, "loss": 0.4431, "step": 848 }, { "epoch": 0.5966268446943078, "grad_norm": 2.036775827407837, "learning_rate": 3.5713281799016164e-06, "loss": 0.5295, "step": 849 }, { "epoch": 0.5973295853829936, "grad_norm": 4.623501300811768, "learning_rate": 3.575544624033732e-06, "loss": 0.6629, "step": 850 }, { "epoch": 0.5980323260716796, "grad_norm": 0.9597324132919312, "learning_rate": 3.5797610681658468e-06, "loss": 0.1884, "step": 851 }, { "epoch": 0.5987350667603655, "grad_norm": 0.579407811164856, "learning_rate": 3.583977512297962e-06, "loss": 0.1171, "step": 852 }, { "epoch": 0.5994378074490513, "grad_norm": 0.5775747895240784, "learning_rate": 3.588193956430077e-06, "loss": 0.0762, "step": 853 }, { "epoch": 0.6001405481377372, "grad_norm": 0.4668199419975281, "learning_rate": 3.5924104005621924e-06, "loss": 0.097, "step": 854 }, { "epoch": 0.600843288826423, "grad_norm": 0.6868484020233154, "learning_rate": 3.5966268446943082e-06, "loss": 0.1037, "step": 855 }, { "epoch": 0.6015460295151089, "grad_norm": 0.641685962677002, "learning_rate": 3.600843288826423e-06, "loss": 0.0855, "step": 856 }, { "epoch": 0.6022487702037947, "grad_norm": 0.5195399522781372, "learning_rate": 3.6050597329585385e-06, "loss": 0.0972, "step": 857 }, { "epoch": 0.6029515108924807, "grad_norm": 0.5964308381080627, "learning_rate": 3.6092761770906535e-06, "loss": 0.1147, "step": 858 }, { "epoch": 0.6036542515811666, "grad_norm": 0.5862053036689758, "learning_rate": 3.613492621222769e-06, "loss": 0.1042, "step": 859 }, { "epoch": 0.6043569922698524, "grad_norm": 0.5032446384429932, "learning_rate": 3.617709065354884e-06, "loss": 0.097, "step": 860 }, { "epoch": 0.6050597329585383, "grad_norm": 0.6016923785209656, "learning_rate": 3.621925509486999e-06, "loss": 0.1149, "step": 861 }, { "epoch": 0.6057624736472241, "grad_norm": 0.571669340133667, "learning_rate": 3.626141953619115e-06, "loss": 0.0904, "step": 862 }, { "epoch": 0.60646521433591, "grad_norm": 0.5769730806350708, "learning_rate": 3.63035839775123e-06, "loss": 0.1172, "step": 863 }, { "epoch": 0.607167955024596, "grad_norm": 0.5782731175422668, "learning_rate": 3.6345748418833453e-06, "loss": 0.1005, "step": 864 }, { "epoch": 0.6078706957132818, "grad_norm": 0.7162313461303711, "learning_rate": 3.63879128601546e-06, "loss": 0.1263, "step": 865 }, { "epoch": 0.6085734364019677, "grad_norm": 0.9053728580474854, "learning_rate": 3.6430077301475756e-06, "loss": 0.1473, "step": 866 }, { "epoch": 0.6092761770906535, "grad_norm": 0.7687336802482605, "learning_rate": 3.647224174279691e-06, "loss": 0.1459, "step": 867 }, { "epoch": 0.6099789177793394, "grad_norm": 0.6630240678787231, "learning_rate": 3.651440618411806e-06, "loss": 0.109, "step": 868 }, { "epoch": 0.6106816584680252, "grad_norm": 0.758234977722168, "learning_rate": 3.6556570625439212e-06, "loss": 0.1492, "step": 869 }, { "epoch": 0.6113843991567112, "grad_norm": 1.2012248039245605, "learning_rate": 3.6598735066760366e-06, "loss": 0.2318, "step": 870 }, { "epoch": 0.6120871398453971, "grad_norm": 1.5392838716506958, "learning_rate": 3.664089950808152e-06, "loss": 0.2373, "step": 871 }, { "epoch": 0.6127898805340829, "grad_norm": 2.4313151836395264, "learning_rate": 3.6683063949402673e-06, "loss": 0.3378, "step": 872 }, { "epoch": 0.6134926212227688, "grad_norm": 2.1087145805358887, "learning_rate": 3.6725228390723823e-06, "loss": 0.3952, "step": 873 }, { "epoch": 0.6141953619114546, "grad_norm": 4.140604496002197, "learning_rate": 3.6767392832044977e-06, "loss": 0.5485, "step": 874 }, { "epoch": 0.6148981026001406, "grad_norm": 4.350051403045654, "learning_rate": 3.6809557273366126e-06, "loss": 0.6777, "step": 875 }, { "epoch": 0.6156008432888265, "grad_norm": 1.2418829202651978, "learning_rate": 3.685172171468728e-06, "loss": 0.2205, "step": 876 }, { "epoch": 0.6163035839775123, "grad_norm": 0.5173723697662354, "learning_rate": 3.6893886156008438e-06, "loss": 0.1207, "step": 877 }, { "epoch": 0.6170063246661982, "grad_norm": 0.5598371028900146, "learning_rate": 3.6936050597329587e-06, "loss": 0.1046, "step": 878 }, { "epoch": 0.617709065354884, "grad_norm": 0.6694813966751099, "learning_rate": 3.697821503865074e-06, "loss": 0.0938, "step": 879 }, { "epoch": 0.6184118060435699, "grad_norm": 0.6333475112915039, "learning_rate": 3.702037947997189e-06, "loss": 0.0934, "step": 880 }, { "epoch": 0.6191145467322557, "grad_norm": 0.5329383015632629, "learning_rate": 3.7062543921293044e-06, "loss": 0.0724, "step": 881 }, { "epoch": 0.6198172874209417, "grad_norm": 0.5226027965545654, "learning_rate": 3.7104708362614193e-06, "loss": 0.0914, "step": 882 }, { "epoch": 0.6205200281096276, "grad_norm": 0.4514259397983551, "learning_rate": 3.7146872803935347e-06, "loss": 0.0688, "step": 883 }, { "epoch": 0.6212227687983134, "grad_norm": 0.5910199880599976, "learning_rate": 3.7189037245256505e-06, "loss": 0.111, "step": 884 }, { "epoch": 0.6219255094869993, "grad_norm": 0.782221794128418, "learning_rate": 3.7231201686577654e-06, "loss": 0.0979, "step": 885 }, { "epoch": 0.6226282501756851, "grad_norm": 0.8024323582649231, "learning_rate": 3.727336612789881e-06, "loss": 0.11, "step": 886 }, { "epoch": 0.623330990864371, "grad_norm": 0.599218487739563, "learning_rate": 3.7315530569219957e-06, "loss": 0.1003, "step": 887 }, { "epoch": 0.624033731553057, "grad_norm": 0.6351887583732605, "learning_rate": 3.735769501054111e-06, "loss": 0.1308, "step": 888 }, { "epoch": 0.6247364722417428, "grad_norm": 0.6605542898178101, "learning_rate": 3.7399859451862265e-06, "loss": 0.109, "step": 889 }, { "epoch": 0.6254392129304287, "grad_norm": 0.7132645845413208, "learning_rate": 3.7442023893183414e-06, "loss": 0.1072, "step": 890 }, { "epoch": 0.6261419536191145, "grad_norm": 0.6852517127990723, "learning_rate": 3.7484188334504568e-06, "loss": 0.1095, "step": 891 }, { "epoch": 0.6268446943078004, "grad_norm": 0.6194265484809875, "learning_rate": 3.752635277582572e-06, "loss": 0.1168, "step": 892 }, { "epoch": 0.6275474349964862, "grad_norm": 0.8126351237297058, "learning_rate": 3.7568517217146875e-06, "loss": 0.1475, "step": 893 }, { "epoch": 0.6282501756851722, "grad_norm": 0.8092824816703796, "learning_rate": 3.761068165846803e-06, "loss": 0.1468, "step": 894 }, { "epoch": 0.6289529163738581, "grad_norm": 0.9678798317909241, "learning_rate": 3.765284609978918e-06, "loss": 0.1841, "step": 895 }, { "epoch": 0.6296556570625439, "grad_norm": 1.043414831161499, "learning_rate": 3.769501054111033e-06, "loss": 0.1552, "step": 896 }, { "epoch": 0.6303583977512298, "grad_norm": 1.6092442274093628, "learning_rate": 3.773717498243148e-06, "loss": 0.3142, "step": 897 }, { "epoch": 0.6310611384399156, "grad_norm": 1.831642985343933, "learning_rate": 3.7779339423752635e-06, "loss": 0.4097, "step": 898 }, { "epoch": 0.6317638791286015, "grad_norm": 1.8438252210617065, "learning_rate": 3.7821503865073784e-06, "loss": 0.4642, "step": 899 }, { "epoch": 0.6324666198172875, "grad_norm": 3.5867037773132324, "learning_rate": 3.786366830639494e-06, "loss": 0.6949, "step": 900 }, { "epoch": 0.6331693605059733, "grad_norm": 0.7101882696151733, "learning_rate": 3.790583274771609e-06, "loss": 0.1572, "step": 901 }, { "epoch": 0.6338721011946592, "grad_norm": 0.5811087489128113, "learning_rate": 3.794799718903724e-06, "loss": 0.0992, "step": 902 }, { "epoch": 0.634574841883345, "grad_norm": 0.8074566721916199, "learning_rate": 3.7990161630358403e-06, "loss": 0.1022, "step": 903 }, { "epoch": 0.6352775825720309, "grad_norm": 0.8254473805427551, "learning_rate": 3.8032326071679553e-06, "loss": 0.0999, "step": 904 }, { "epoch": 0.6359803232607167, "grad_norm": 0.4462158977985382, "learning_rate": 3.8074490513000706e-06, "loss": 0.0825, "step": 905 }, { "epoch": 0.6366830639494027, "grad_norm": 0.5543385148048401, "learning_rate": 3.811665495432186e-06, "loss": 0.0877, "step": 906 }, { "epoch": 0.6373858046380886, "grad_norm": 0.5234286785125732, "learning_rate": 3.815881939564301e-06, "loss": 0.0833, "step": 907 }, { "epoch": 0.6380885453267744, "grad_norm": 0.6360012888908386, "learning_rate": 3.820098383696416e-06, "loss": 0.0973, "step": 908 }, { "epoch": 0.6387912860154603, "grad_norm": 0.6018673777580261, "learning_rate": 3.824314827828532e-06, "loss": 0.0933, "step": 909 }, { "epoch": 0.6394940267041461, "grad_norm": 0.43278568983078003, "learning_rate": 3.828531271960647e-06, "loss": 0.0622, "step": 910 }, { "epoch": 0.640196767392832, "grad_norm": 0.8238375782966614, "learning_rate": 3.832747716092762e-06, "loss": 0.1094, "step": 911 }, { "epoch": 0.640899508081518, "grad_norm": 1.2755941152572632, "learning_rate": 3.836964160224877e-06, "loss": 0.1035, "step": 912 }, { "epoch": 0.6416022487702038, "grad_norm": 0.5857747197151184, "learning_rate": 3.841180604356992e-06, "loss": 0.1108, "step": 913 }, { "epoch": 0.6423049894588897, "grad_norm": 0.5381284952163696, "learning_rate": 3.845397048489107e-06, "loss": 0.087, "step": 914 }, { "epoch": 0.6430077301475755, "grad_norm": 0.6259009838104248, "learning_rate": 3.849613492621222e-06, "loss": 0.1057, "step": 915 }, { "epoch": 0.6437104708362614, "grad_norm": 0.9007967114448547, "learning_rate": 3.853829936753338e-06, "loss": 0.1205, "step": 916 }, { "epoch": 0.6444132115249473, "grad_norm": 0.9107803106307983, "learning_rate": 3.858046380885453e-06, "loss": 0.0963, "step": 917 }, { "epoch": 0.6451159522136332, "grad_norm": 0.6539586186408997, "learning_rate": 3.862262825017569e-06, "loss": 0.1362, "step": 918 }, { "epoch": 0.6458186929023191, "grad_norm": 1.4036654233932495, "learning_rate": 3.8664792691496845e-06, "loss": 0.1626, "step": 919 }, { "epoch": 0.6465214335910049, "grad_norm": 2.865107536315918, "learning_rate": 3.8706957132817995e-06, "loss": 0.1898, "step": 920 }, { "epoch": 0.6472241742796908, "grad_norm": 1.08432936668396, "learning_rate": 3.874912157413914e-06, "loss": 0.2258, "step": 921 }, { "epoch": 0.6479269149683766, "grad_norm": 1.758602261543274, "learning_rate": 3.87912860154603e-06, "loss": 0.2968, "step": 922 }, { "epoch": 0.6486296556570625, "grad_norm": 2.1801414489746094, "learning_rate": 3.883345045678145e-06, "loss": 0.4222, "step": 923 }, { "epoch": 0.6493323963457485, "grad_norm": 2.2438597679138184, "learning_rate": 3.88756148981026e-06, "loss": 0.5008, "step": 924 }, { "epoch": 0.6500351370344343, "grad_norm": 3.748741865158081, "learning_rate": 3.891777933942375e-06, "loss": 0.6925, "step": 925 }, { "epoch": 0.6507378777231202, "grad_norm": 0.9807621836662292, "learning_rate": 3.895994378074491e-06, "loss": 0.1705, "step": 926 }, { "epoch": 0.651440618411806, "grad_norm": 0.6357351541519165, "learning_rate": 3.900210822206606e-06, "loss": 0.1198, "step": 927 }, { "epoch": 0.6521433591004919, "grad_norm": 0.915944516658783, "learning_rate": 3.904427266338721e-06, "loss": 0.1053, "step": 928 }, { "epoch": 0.6528460997891778, "grad_norm": 0.48445063829421997, "learning_rate": 3.9086437104708365e-06, "loss": 0.0862, "step": 929 }, { "epoch": 0.6535488404778637, "grad_norm": 0.5260390043258667, "learning_rate": 3.9128601546029514e-06, "loss": 0.0861, "step": 930 }, { "epoch": 0.6542515811665496, "grad_norm": 0.4411722421646118, "learning_rate": 3.917076598735066e-06, "loss": 0.0769, "step": 931 }, { "epoch": 0.6549543218552354, "grad_norm": 0.5712600350379944, "learning_rate": 3.921293042867181e-06, "loss": 0.0806, "step": 932 }, { "epoch": 0.6556570625439213, "grad_norm": 0.569637656211853, "learning_rate": 3.925509486999298e-06, "loss": 0.0777, "step": 933 }, { "epoch": 0.6563598032326071, "grad_norm": 0.6132969856262207, "learning_rate": 3.929725931131413e-06, "loss": 0.0891, "step": 934 }, { "epoch": 0.657062543921293, "grad_norm": 0.5545106530189514, "learning_rate": 3.933942375263528e-06, "loss": 0.0784, "step": 935 }, { "epoch": 0.657765284609979, "grad_norm": 0.5719489455223083, "learning_rate": 3.938158819395644e-06, "loss": 0.1018, "step": 936 }, { "epoch": 0.6584680252986648, "grad_norm": 0.548778772354126, "learning_rate": 3.942375263527759e-06, "loss": 0.0943, "step": 937 }, { "epoch": 0.6591707659873507, "grad_norm": 0.6666469573974609, "learning_rate": 3.9465917076598735e-06, "loss": 0.1126, "step": 938 }, { "epoch": 0.6598735066760365, "grad_norm": 0.6869754195213318, "learning_rate": 3.950808151791989e-06, "loss": 0.0804, "step": 939 }, { "epoch": 0.6605762473647224, "grad_norm": 0.6707573533058167, "learning_rate": 3.955024595924104e-06, "loss": 0.1173, "step": 940 }, { "epoch": 0.6612789880534083, "grad_norm": 0.6482703685760498, "learning_rate": 3.959241040056219e-06, "loss": 0.1491, "step": 941 }, { "epoch": 0.6619817287420942, "grad_norm": 0.6001825928688049, "learning_rate": 3.963457484188334e-06, "loss": 0.093, "step": 942 }, { "epoch": 0.6626844694307801, "grad_norm": 0.8382095694541931, "learning_rate": 3.96767392832045e-06, "loss": 0.1272, "step": 943 }, { "epoch": 0.6633872101194659, "grad_norm": 0.7064149975776672, "learning_rate": 3.971890372452565e-06, "loss": 0.1364, "step": 944 }, { "epoch": 0.6640899508081518, "grad_norm": 1.0406051874160767, "learning_rate": 3.97610681658468e-06, "loss": 0.1846, "step": 945 }, { "epoch": 0.6647926914968376, "grad_norm": 1.1147974729537964, "learning_rate": 3.980323260716796e-06, "loss": 0.2473, "step": 946 }, { "epoch": 0.6654954321855235, "grad_norm": 1.5347576141357422, "learning_rate": 3.9845397048489106e-06, "loss": 0.2901, "step": 947 }, { "epoch": 0.6661981728742095, "grad_norm": 2.273149013519287, "learning_rate": 3.988756148981026e-06, "loss": 0.43, "step": 948 }, { "epoch": 0.6669009135628953, "grad_norm": 2.388969659805298, "learning_rate": 3.992972593113141e-06, "loss": 0.5106, "step": 949 }, { "epoch": 0.6676036542515812, "grad_norm": 2.6202728748321533, "learning_rate": 3.997189037245257e-06, "loss": 0.6424, "step": 950 }, { "epoch": 0.668306394940267, "grad_norm": 1.0269968509674072, "learning_rate": 4.001405481377372e-06, "loss": 0.2299, "step": 951 }, { "epoch": 0.6690091356289529, "grad_norm": 0.7305585145950317, "learning_rate": 4.005621925509487e-06, "loss": 0.113, "step": 952 }, { "epoch": 0.6697118763176388, "grad_norm": 0.524955153465271, "learning_rate": 4.009838369641603e-06, "loss": 0.1048, "step": 953 }, { "epoch": 0.6704146170063247, "grad_norm": 0.7749505639076233, "learning_rate": 4.014054813773718e-06, "loss": 0.106, "step": 954 }, { "epoch": 0.6711173576950106, "grad_norm": 0.5057756304740906, "learning_rate": 4.018271257905833e-06, "loss": 0.1098, "step": 955 }, { "epoch": 0.6718200983836964, "grad_norm": 0.5252646207809448, "learning_rate": 4.0224877020379484e-06, "loss": 0.0913, "step": 956 }, { "epoch": 0.6725228390723823, "grad_norm": 0.5941205620765686, "learning_rate": 4.026704146170063e-06, "loss": 0.0752, "step": 957 }, { "epoch": 0.6732255797610681, "grad_norm": 0.5559599995613098, "learning_rate": 4.030920590302178e-06, "loss": 0.0958, "step": 958 }, { "epoch": 0.673928320449754, "grad_norm": 0.6009759306907654, "learning_rate": 4.035137034434293e-06, "loss": 0.107, "step": 959 }, { "epoch": 0.67463106113844, "grad_norm": 0.484508752822876, "learning_rate": 4.039353478566409e-06, "loss": 0.0931, "step": 960 }, { "epoch": 0.6753338018271258, "grad_norm": 0.815963089466095, "learning_rate": 4.043569922698524e-06, "loss": 0.0878, "step": 961 }, { "epoch": 0.6760365425158117, "grad_norm": 0.6741816997528076, "learning_rate": 4.047786366830639e-06, "loss": 0.1029, "step": 962 }, { "epoch": 0.6767392832044975, "grad_norm": 0.5325302481651306, "learning_rate": 4.052002810962756e-06, "loss": 0.0911, "step": 963 }, { "epoch": 0.6774420238931834, "grad_norm": 0.7578814625740051, "learning_rate": 4.0562192550948705e-06, "loss": 0.0992, "step": 964 }, { "epoch": 0.6781447645818693, "grad_norm": 0.8661431074142456, "learning_rate": 4.0604356992269855e-06, "loss": 0.1071, "step": 965 }, { "epoch": 0.6788475052705552, "grad_norm": 1.1022727489471436, "learning_rate": 4.0646521433591e-06, "loss": 0.1244, "step": 966 }, { "epoch": 0.6795502459592411, "grad_norm": 1.5532499551773071, "learning_rate": 4.068868587491216e-06, "loss": 0.1315, "step": 967 }, { "epoch": 0.6802529866479269, "grad_norm": 0.6179517507553101, "learning_rate": 4.073085031623331e-06, "loss": 0.1131, "step": 968 }, { "epoch": 0.6809557273366128, "grad_norm": 0.8559219837188721, "learning_rate": 4.077301475755446e-06, "loss": 0.1859, "step": 969 }, { "epoch": 0.6816584680252986, "grad_norm": 0.6782516241073608, "learning_rate": 4.081517919887562e-06, "loss": 0.1393, "step": 970 }, { "epoch": 0.6823612087139845, "grad_norm": 1.109126091003418, "learning_rate": 4.085734364019677e-06, "loss": 0.2557, "step": 971 }, { "epoch": 0.6830639494026705, "grad_norm": 1.2585078477859497, "learning_rate": 4.089950808151792e-06, "loss": 0.2712, "step": 972 }, { "epoch": 0.6837666900913563, "grad_norm": 1.4959303140640259, "learning_rate": 4.0941672522839076e-06, "loss": 0.429, "step": 973 }, { "epoch": 0.6844694307800422, "grad_norm": 1.631069302558899, "learning_rate": 4.0983836964160225e-06, "loss": 0.487, "step": 974 }, { "epoch": 0.685172171468728, "grad_norm": 6.135562896728516, "learning_rate": 4.1026001405481375e-06, "loss": 0.7167, "step": 975 }, { "epoch": 0.6858749121574139, "grad_norm": 0.9974015951156616, "learning_rate": 4.106816584680252e-06, "loss": 0.1971, "step": 976 }, { "epoch": 0.6865776528460998, "grad_norm": 0.6748353242874146, "learning_rate": 4.111033028812369e-06, "loss": 0.0914, "step": 977 }, { "epoch": 0.6872803935347856, "grad_norm": 0.4709434509277344, "learning_rate": 4.115249472944484e-06, "loss": 0.0886, "step": 978 }, { "epoch": 0.6879831342234716, "grad_norm": 1.3228737115859985, "learning_rate": 4.119465917076599e-06, "loss": 0.0912, "step": 979 }, { "epoch": 0.6886858749121574, "grad_norm": 0.5257927775382996, "learning_rate": 4.123682361208715e-06, "loss": 0.0769, "step": 980 }, { "epoch": 0.6893886156008433, "grad_norm": 0.5499150156974792, "learning_rate": 4.12789880534083e-06, "loss": 0.0742, "step": 981 }, { "epoch": 0.6900913562895291, "grad_norm": 0.5136544704437256, "learning_rate": 4.132115249472945e-06, "loss": 0.0716, "step": 982 }, { "epoch": 0.690794096978215, "grad_norm": 0.5423275828361511, "learning_rate": 4.1363316936050595e-06, "loss": 0.0773, "step": 983 }, { "epoch": 0.691496837666901, "grad_norm": 0.5956935882568359, "learning_rate": 4.140548137737175e-06, "loss": 0.0945, "step": 984 }, { "epoch": 0.6921995783555868, "grad_norm": 0.6208829283714294, "learning_rate": 4.14476458186929e-06, "loss": 0.0657, "step": 985 }, { "epoch": 0.6929023190442727, "grad_norm": 0.6467328667640686, "learning_rate": 4.148981026001405e-06, "loss": 0.106, "step": 986 }, { "epoch": 0.6936050597329585, "grad_norm": 0.6153718829154968, "learning_rate": 4.153197470133521e-06, "loss": 0.0836, "step": 987 }, { "epoch": 0.6943078004216444, "grad_norm": 0.6138646602630615, "learning_rate": 4.157413914265636e-06, "loss": 0.1169, "step": 988 }, { "epoch": 0.6950105411103303, "grad_norm": 0.6355115175247192, "learning_rate": 4.161630358397751e-06, "loss": 0.1053, "step": 989 }, { "epoch": 0.6957132817990161, "grad_norm": 0.7153375148773193, "learning_rate": 4.165846802529867e-06, "loss": 0.1233, "step": 990 }, { "epoch": 0.6964160224877021, "grad_norm": 0.6181850433349609, "learning_rate": 4.170063246661982e-06, "loss": 0.1198, "step": 991 }, { "epoch": 0.6971187631763879, "grad_norm": 0.4887823164463043, "learning_rate": 4.1742796907940974e-06, "loss": 0.1181, "step": 992 }, { "epoch": 0.6978215038650738, "grad_norm": 0.7464542984962463, "learning_rate": 4.178496134926212e-06, "loss": 0.118, "step": 993 }, { "epoch": 0.6985242445537596, "grad_norm": 1.0940500497817993, "learning_rate": 4.182712579058328e-06, "loss": 0.1194, "step": 994 }, { "epoch": 0.6992269852424455, "grad_norm": 5.904663562774658, "learning_rate": 4.186929023190443e-06, "loss": 0.1653, "step": 995 }, { "epoch": 0.6999297259311315, "grad_norm": 1.4267209768295288, "learning_rate": 4.191145467322558e-06, "loss": 0.1849, "step": 996 }, { "epoch": 0.7006324666198173, "grad_norm": 2.6190969944000244, "learning_rate": 4.195361911454674e-06, "loss": 0.2615, "step": 997 }, { "epoch": 0.7013352073085032, "grad_norm": 1.513951063156128, "learning_rate": 4.199578355586789e-06, "loss": 0.3636, "step": 998 }, { "epoch": 0.702037947997189, "grad_norm": 2.3092925548553467, "learning_rate": 4.203794799718904e-06, "loss": 0.5086, "step": 999 }, { "epoch": 0.7027406886858749, "grad_norm": 2.780078887939453, "learning_rate": 4.208011243851019e-06, "loss": 0.6382, "step": 1000 }, { "epoch": 0.7027406886858749, "eval_cer": 0.21958364701646427, "eval_loss": 0.5261573791503906, "eval_runtime": 20.6017, "eval_samples_per_second": 220.273, "eval_steps_per_second": 0.728, "eval_wer": 0.44085360744113294, "step": 1000 }, { "epoch": 0.7034434293745608, "grad_norm": 1.0570772886276245, "learning_rate": 4.2122276879831345e-06, "loss": 0.2174, "step": 1001 }, { "epoch": 0.7041461700632466, "grad_norm": 0.5904741883277893, "learning_rate": 4.216444132115249e-06, "loss": 0.1006, "step": 1002 }, { "epoch": 0.7048489107519326, "grad_norm": 0.453570157289505, "learning_rate": 4.220660576247364e-06, "loss": 0.0942, "step": 1003 }, { "epoch": 0.7055516514406184, "grad_norm": 0.41151559352874756, "learning_rate": 4.22487702037948e-06, "loss": 0.0916, "step": 1004 }, { "epoch": 0.7062543921293043, "grad_norm": 0.43477386236190796, "learning_rate": 4.229093464511595e-06, "loss": 0.0986, "step": 1005 }, { "epoch": 0.7069571328179901, "grad_norm": 0.5181138515472412, "learning_rate": 4.23330990864371e-06, "loss": 0.0655, "step": 1006 }, { "epoch": 0.707659873506676, "grad_norm": 0.5549378395080566, "learning_rate": 4.237526352775826e-06, "loss": 0.0902, "step": 1007 }, { "epoch": 0.708362614195362, "grad_norm": 0.6114147305488586, "learning_rate": 4.241742796907942e-06, "loss": 0.1103, "step": 1008 }, { "epoch": 0.7090653548840478, "grad_norm": 0.5550990104675293, "learning_rate": 4.2459592410400565e-06, "loss": 0.1144, "step": 1009 }, { "epoch": 0.7097680955727337, "grad_norm": 0.4620648920536041, "learning_rate": 4.2501756851721715e-06, "loss": 0.0593, "step": 1010 }, { "epoch": 0.7104708362614195, "grad_norm": 0.48512718081474304, "learning_rate": 4.254392129304287e-06, "loss": 0.0832, "step": 1011 }, { "epoch": 0.7111735769501054, "grad_norm": 0.4333072304725647, "learning_rate": 4.258608573436402e-06, "loss": 0.0656, "step": 1012 }, { "epoch": 0.7118763176387913, "grad_norm": 0.7477991580963135, "learning_rate": 4.262825017568517e-06, "loss": 0.086, "step": 1013 }, { "epoch": 0.7125790583274771, "grad_norm": 0.6108619570732117, "learning_rate": 4.267041461700633e-06, "loss": 0.1048, "step": 1014 }, { "epoch": 0.7132817990161631, "grad_norm": 0.614913284778595, "learning_rate": 4.271257905832748e-06, "loss": 0.1086, "step": 1015 }, { "epoch": 0.7139845397048489, "grad_norm": 0.6813833713531494, "learning_rate": 4.275474349964863e-06, "loss": 0.0994, "step": 1016 }, { "epoch": 0.7146872803935348, "grad_norm": 0.6128227114677429, "learning_rate": 4.279690794096978e-06, "loss": 0.0931, "step": 1017 }, { "epoch": 0.7153900210822206, "grad_norm": 0.8403228521347046, "learning_rate": 4.283907238229094e-06, "loss": 0.1163, "step": 1018 }, { "epoch": 0.7160927617709065, "grad_norm": 0.6415989398956299, "learning_rate": 4.2881236823612085e-06, "loss": 0.1326, "step": 1019 }, { "epoch": 0.7167955024595924, "grad_norm": 0.7586535215377808, "learning_rate": 4.2923401264933235e-06, "loss": 0.1301, "step": 1020 }, { "epoch": 0.7174982431482783, "grad_norm": 1.030776858329773, "learning_rate": 4.296556570625439e-06, "loss": 0.1906, "step": 1021 }, { "epoch": 0.7182009838369642, "grad_norm": 1.2723350524902344, "learning_rate": 4.300773014757555e-06, "loss": 0.2896, "step": 1022 }, { "epoch": 0.71890372452565, "grad_norm": 1.6184358596801758, "learning_rate": 4.30498945888967e-06, "loss": 0.3962, "step": 1023 }, { "epoch": 0.7196064652143359, "grad_norm": 4.2391204833984375, "learning_rate": 4.309205903021785e-06, "loss": 0.4677, "step": 1024 }, { "epoch": 0.7203092059030218, "grad_norm": 4.069004058837891, "learning_rate": 4.313422347153901e-06, "loss": 0.6702, "step": 1025 }, { "epoch": 0.7210119465917076, "grad_norm": 1.9343633651733398, "learning_rate": 4.317638791286016e-06, "loss": 0.2122, "step": 1026 }, { "epoch": 0.7217146872803936, "grad_norm": 0.5260276794433594, "learning_rate": 4.321855235418131e-06, "loss": 0.113, "step": 1027 }, { "epoch": 0.7224174279690794, "grad_norm": 0.5599494576454163, "learning_rate": 4.326071679550246e-06, "loss": 0.0862, "step": 1028 }, { "epoch": 0.7231201686577653, "grad_norm": 0.48701417446136475, "learning_rate": 4.330288123682361e-06, "loss": 0.1065, "step": 1029 }, { "epoch": 0.7238229093464511, "grad_norm": 0.47223374247550964, "learning_rate": 4.334504567814476e-06, "loss": 0.0887, "step": 1030 }, { "epoch": 0.724525650035137, "grad_norm": 0.43776455521583557, "learning_rate": 4.338721011946592e-06, "loss": 0.0645, "step": 1031 }, { "epoch": 0.725228390723823, "grad_norm": 0.5906615257263184, "learning_rate": 4.342937456078707e-06, "loss": 0.0648, "step": 1032 }, { "epoch": 0.7259311314125088, "grad_norm": 0.48612791299819946, "learning_rate": 4.347153900210822e-06, "loss": 0.089, "step": 1033 }, { "epoch": 0.7266338721011947, "grad_norm": 0.41866251826286316, "learning_rate": 4.351370344342937e-06, "loss": 0.0988, "step": 1034 }, { "epoch": 0.7273366127898805, "grad_norm": 0.44549742341041565, "learning_rate": 4.355586788475053e-06, "loss": 0.0657, "step": 1035 }, { "epoch": 0.7280393534785664, "grad_norm": 0.5952585935592651, "learning_rate": 4.359803232607168e-06, "loss": 0.1095, "step": 1036 }, { "epoch": 0.7287420941672523, "grad_norm": 0.4282941222190857, "learning_rate": 4.3640196767392834e-06, "loss": 0.0732, "step": 1037 }, { "epoch": 0.7294448348559381, "grad_norm": 1.6695762872695923, "learning_rate": 4.368236120871399e-06, "loss": 0.0935, "step": 1038 }, { "epoch": 0.7301475755446241, "grad_norm": 0.5004169344902039, "learning_rate": 4.372452565003514e-06, "loss": 0.0669, "step": 1039 }, { "epoch": 0.7308503162333099, "grad_norm": 0.6803144812583923, "learning_rate": 4.376669009135629e-06, "loss": 0.1115, "step": 1040 }, { "epoch": 0.7315530569219958, "grad_norm": 0.9439788460731506, "learning_rate": 4.380885453267744e-06, "loss": 0.127, "step": 1041 }, { "epoch": 0.7322557976106817, "grad_norm": 1.8541450500488281, "learning_rate": 4.38510189739986e-06, "loss": 0.1191, "step": 1042 }, { "epoch": 0.7329585382993675, "grad_norm": 0.7401757836341858, "learning_rate": 4.389318341531975e-06, "loss": 0.1292, "step": 1043 }, { "epoch": 0.7336612789880534, "grad_norm": 1.2462983131408691, "learning_rate": 4.39353478566409e-06, "loss": 0.1348, "step": 1044 }, { "epoch": 0.7343640196767393, "grad_norm": 0.9159403443336487, "learning_rate": 4.3977512297962055e-06, "loss": 0.2218, "step": 1045 }, { "epoch": 0.7350667603654252, "grad_norm": 0.863340437412262, "learning_rate": 4.4019676739283205e-06, "loss": 0.1641, "step": 1046 }, { "epoch": 0.735769501054111, "grad_norm": 1.4918181896209717, "learning_rate": 4.406184118060435e-06, "loss": 0.2811, "step": 1047 }, { "epoch": 0.7364722417427969, "grad_norm": 1.8215034008026123, "learning_rate": 4.410400562192551e-06, "loss": 0.4118, "step": 1048 }, { "epoch": 0.7371749824314828, "grad_norm": 2.2793664932250977, "learning_rate": 4.414617006324666e-06, "loss": 0.4958, "step": 1049 }, { "epoch": 0.7378777231201686, "grad_norm": 4.362558841705322, "learning_rate": 4.418833450456781e-06, "loss": 0.6842, "step": 1050 }, { "epoch": 0.7385804638088546, "grad_norm": 0.6955499649047852, "learning_rate": 4.423049894588896e-06, "loss": 0.1609, "step": 1051 }, { "epoch": 0.7392832044975404, "grad_norm": 0.5710420608520508, "learning_rate": 4.427266338721013e-06, "loss": 0.0889, "step": 1052 }, { "epoch": 0.7399859451862263, "grad_norm": 0.39019277691841125, "learning_rate": 4.431482782853128e-06, "loss": 0.068, "step": 1053 }, { "epoch": 0.7406886858749122, "grad_norm": 0.5204043984413147, "learning_rate": 4.4356992269852426e-06, "loss": 0.0888, "step": 1054 }, { "epoch": 0.741391426563598, "grad_norm": 0.4207502007484436, "learning_rate": 4.439915671117358e-06, "loss": 0.071, "step": 1055 }, { "epoch": 0.7420941672522839, "grad_norm": 0.481734961271286, "learning_rate": 4.444132115249473e-06, "loss": 0.0783, "step": 1056 }, { "epoch": 0.7427969079409698, "grad_norm": 0.5175879597663879, "learning_rate": 4.448348559381588e-06, "loss": 0.0906, "step": 1057 }, { "epoch": 0.7434996486296557, "grad_norm": 0.44115522503852844, "learning_rate": 4.452565003513703e-06, "loss": 0.0588, "step": 1058 }, { "epoch": 0.7442023893183415, "grad_norm": 0.5209271311759949, "learning_rate": 4.456781447645819e-06, "loss": 0.0694, "step": 1059 }, { "epoch": 0.7449051300070274, "grad_norm": 0.5435191988945007, "learning_rate": 4.460997891777934e-06, "loss": 0.0801, "step": 1060 }, { "epoch": 0.7456078706957133, "grad_norm": 0.9624141454696655, "learning_rate": 4.465214335910049e-06, "loss": 0.0936, "step": 1061 }, { "epoch": 0.7463106113843991, "grad_norm": 0.47606056928634644, "learning_rate": 4.469430780042165e-06, "loss": 0.0772, "step": 1062 }, { "epoch": 0.7470133520730851, "grad_norm": 0.6489266753196716, "learning_rate": 4.47364722417428e-06, "loss": 0.0842, "step": 1063 }, { "epoch": 0.7477160927617709, "grad_norm": 0.47763878107070923, "learning_rate": 4.4778636683063945e-06, "loss": 0.0708, "step": 1064 }, { "epoch": 0.7484188334504568, "grad_norm": 0.6037685871124268, "learning_rate": 4.48208011243851e-06, "loss": 0.117, "step": 1065 }, { "epoch": 0.7491215741391427, "grad_norm": 0.7454593777656555, "learning_rate": 4.486296556570625e-06, "loss": 0.1311, "step": 1066 }, { "epoch": 0.7498243148278285, "grad_norm": 0.6869489550590515, "learning_rate": 4.490513000702741e-06, "loss": 0.0904, "step": 1067 }, { "epoch": 0.7505270555165144, "grad_norm": 0.7778688073158264, "learning_rate": 4.494729444834856e-06, "loss": 0.1197, "step": 1068 }, { "epoch": 0.7512297962052003, "grad_norm": 0.7346179485321045, "learning_rate": 4.498945888966972e-06, "loss": 0.1129, "step": 1069 }, { "epoch": 0.7519325368938862, "grad_norm": 0.8100537061691284, "learning_rate": 4.503162333099087e-06, "loss": 0.1912, "step": 1070 }, { "epoch": 0.752635277582572, "grad_norm": 0.9085155129432678, "learning_rate": 4.507378777231202e-06, "loss": 0.1715, "step": 1071 }, { "epoch": 0.7533380182712579, "grad_norm": 1.2592906951904297, "learning_rate": 4.5115952213633175e-06, "loss": 0.2487, "step": 1072 }, { "epoch": 0.7540407589599438, "grad_norm": 1.8794214725494385, "learning_rate": 4.515811665495432e-06, "loss": 0.4413, "step": 1073 }, { "epoch": 0.7547434996486296, "grad_norm": 1.9510105848312378, "learning_rate": 4.520028109627547e-06, "loss": 0.452, "step": 1074 }, { "epoch": 0.7554462403373156, "grad_norm": 2.9660634994506836, "learning_rate": 4.524244553759662e-06, "loss": 0.6473, "step": 1075 }, { "epoch": 0.7561489810260014, "grad_norm": 0.6992705464363098, "learning_rate": 4.528460997891778e-06, "loss": 0.1917, "step": 1076 }, { "epoch": 0.7568517217146873, "grad_norm": 0.6144616603851318, "learning_rate": 4.532677442023893e-06, "loss": 0.0792, "step": 1077 }, { "epoch": 0.7575544624033732, "grad_norm": 0.427329421043396, "learning_rate": 4.536893886156008e-06, "loss": 0.0925, "step": 1078 }, { "epoch": 0.758257203092059, "grad_norm": 0.5377605557441711, "learning_rate": 4.541110330288124e-06, "loss": 0.0692, "step": 1079 }, { "epoch": 0.7589599437807449, "grad_norm": 0.469825804233551, "learning_rate": 4.545326774420239e-06, "loss": 0.0698, "step": 1080 }, { "epoch": 0.7596626844694307, "grad_norm": 0.4230661988258362, "learning_rate": 4.549543218552354e-06, "loss": 0.067, "step": 1081 }, { "epoch": 0.7603654251581167, "grad_norm": 0.43673601746559143, "learning_rate": 4.55375966268447e-06, "loss": 0.0776, "step": 1082 }, { "epoch": 0.7610681658468025, "grad_norm": 0.5472078323364258, "learning_rate": 4.557976106816585e-06, "loss": 0.0826, "step": 1083 }, { "epoch": 0.7617709065354884, "grad_norm": 0.43809935450553894, "learning_rate": 4.5621925509487e-06, "loss": 0.0874, "step": 1084 }, { "epoch": 0.7624736472241743, "grad_norm": 0.3913225531578064, "learning_rate": 4.566408995080815e-06, "loss": 0.0625, "step": 1085 }, { "epoch": 0.7631763879128601, "grad_norm": 0.43462976813316345, "learning_rate": 4.570625439212931e-06, "loss": 0.0651, "step": 1086 }, { "epoch": 0.763879128601546, "grad_norm": 1.07741117477417, "learning_rate": 4.574841883345046e-06, "loss": 0.0783, "step": 1087 }, { "epoch": 0.7645818692902319, "grad_norm": 0.4895462393760681, "learning_rate": 4.579058327477161e-06, "loss": 0.0694, "step": 1088 }, { "epoch": 0.7652846099789178, "grad_norm": 0.5583158135414124, "learning_rate": 4.583274771609277e-06, "loss": 0.0801, "step": 1089 }, { "epoch": 0.7659873506676037, "grad_norm": 0.9765302538871765, "learning_rate": 4.5874912157413915e-06, "loss": 0.1024, "step": 1090 }, { "epoch": 0.7666900913562895, "grad_norm": 0.8028756380081177, "learning_rate": 4.5917076598735065e-06, "loss": 0.1099, "step": 1091 }, { "epoch": 0.7673928320449754, "grad_norm": 1.0039482116699219, "learning_rate": 4.5959241040056214e-06, "loss": 0.0745, "step": 1092 }, { "epoch": 0.7680955727336612, "grad_norm": 1.6164939403533936, "learning_rate": 4.600140548137737e-06, "loss": 0.0915, "step": 1093 }, { "epoch": 0.7687983134223472, "grad_norm": 0.7183098793029785, "learning_rate": 4.604356992269852e-06, "loss": 0.1131, "step": 1094 }, { "epoch": 0.769501054111033, "grad_norm": 0.7931156754493713, "learning_rate": 4.608573436401967e-06, "loss": 0.1543, "step": 1095 }, { "epoch": 0.7702037947997189, "grad_norm": 0.7038898468017578, "learning_rate": 4.612789880534083e-06, "loss": 0.1553, "step": 1096 }, { "epoch": 0.7709065354884048, "grad_norm": 1.260475754737854, "learning_rate": 4.617006324666199e-06, "loss": 0.2989, "step": 1097 }, { "epoch": 0.7716092761770906, "grad_norm": 2.1257245540618896, "learning_rate": 4.621222768798314e-06, "loss": 0.3984, "step": 1098 }, { "epoch": 0.7723120168657766, "grad_norm": 1.9621790647506714, "learning_rate": 4.6254392129304294e-06, "loss": 0.4687, "step": 1099 }, { "epoch": 0.7730147575544624, "grad_norm": 2.8995420932769775, "learning_rate": 4.629655657062544e-06, "loss": 0.6056, "step": 1100 }, { "epoch": 0.7737174982431483, "grad_norm": 0.6588873267173767, "learning_rate": 4.633872101194659e-06, "loss": 0.1924, "step": 1101 }, { "epoch": 0.7744202389318342, "grad_norm": 0.45358237624168396, "learning_rate": 4.638088545326774e-06, "loss": 0.1081, "step": 1102 }, { "epoch": 0.77512297962052, "grad_norm": 0.531358540058136, "learning_rate": 4.64230498945889e-06, "loss": 0.0994, "step": 1103 }, { "epoch": 0.7758257203092059, "grad_norm": 0.5143305659294128, "learning_rate": 4.646521433591005e-06, "loss": 0.1031, "step": 1104 }, { "epoch": 0.7765284609978917, "grad_norm": 0.5619973540306091, "learning_rate": 4.65073787772312e-06, "loss": 0.072, "step": 1105 }, { "epoch": 0.7772312016865777, "grad_norm": 0.44695422053337097, "learning_rate": 4.654954321855236e-06, "loss": 0.0659, "step": 1106 }, { "epoch": 0.7779339423752635, "grad_norm": 0.4808768630027771, "learning_rate": 4.659170765987351e-06, "loss": 0.093, "step": 1107 }, { "epoch": 0.7786366830639494, "grad_norm": 0.49408042430877686, "learning_rate": 4.663387210119466e-06, "loss": 0.0945, "step": 1108 }, { "epoch": 0.7793394237526353, "grad_norm": 0.4480993151664734, "learning_rate": 4.6676036542515806e-06, "loss": 0.0731, "step": 1109 }, { "epoch": 0.7800421644413211, "grad_norm": 0.5937999486923218, "learning_rate": 4.671820098383696e-06, "loss": 0.069, "step": 1110 }, { "epoch": 0.780744905130007, "grad_norm": 0.569385826587677, "learning_rate": 4.676036542515812e-06, "loss": 0.0984, "step": 1111 }, { "epoch": 0.7814476458186929, "grad_norm": 0.4831339418888092, "learning_rate": 4.680252986647927e-06, "loss": 0.0764, "step": 1112 }, { "epoch": 0.7821503865073788, "grad_norm": 0.6359565258026123, "learning_rate": 4.684469430780043e-06, "loss": 0.0906, "step": 1113 }, { "epoch": 0.7828531271960647, "grad_norm": 0.46480807662010193, "learning_rate": 4.688685874912158e-06, "loss": 0.0715, "step": 1114 }, { "epoch": 0.7835558678847505, "grad_norm": 0.9903660416603088, "learning_rate": 4.692902319044273e-06, "loss": 0.1101, "step": 1115 }, { "epoch": 0.7842586085734364, "grad_norm": 1.5666203498840332, "learning_rate": 4.6971187631763886e-06, "loss": 0.095, "step": 1116 }, { "epoch": 0.7849613492621222, "grad_norm": 0.5387712717056274, "learning_rate": 4.7013352073085035e-06, "loss": 0.0812, "step": 1117 }, { "epoch": 0.7856640899508082, "grad_norm": 1.4764059782028198, "learning_rate": 4.7055516514406184e-06, "loss": 0.1171, "step": 1118 }, { "epoch": 0.786366830639494, "grad_norm": 0.7848532795906067, "learning_rate": 4.709768095572733e-06, "loss": 0.1403, "step": 1119 }, { "epoch": 0.7870695713281799, "grad_norm": 0.7512660622596741, "learning_rate": 4.713984539704849e-06, "loss": 0.1523, "step": 1120 }, { "epoch": 0.7877723120168658, "grad_norm": 0.9318837523460388, "learning_rate": 4.718200983836964e-06, "loss": 0.2225, "step": 1121 }, { "epoch": 0.7884750527055516, "grad_norm": 1.7515285015106201, "learning_rate": 4.722417427969079e-06, "loss": 0.3004, "step": 1122 }, { "epoch": 0.7891777933942375, "grad_norm": 1.4756476879119873, "learning_rate": 4.726633872101195e-06, "loss": 0.3077, "step": 1123 }, { "epoch": 0.7898805340829234, "grad_norm": 1.931374192237854, "learning_rate": 4.73085031623331e-06, "loss": 0.4762, "step": 1124 }, { "epoch": 0.7905832747716093, "grad_norm": 3.512845516204834, "learning_rate": 4.735066760365425e-06, "loss": 0.6107, "step": 1125 }, { "epoch": 0.7912860154602952, "grad_norm": 0.691102921962738, "learning_rate": 4.7392832044975405e-06, "loss": 0.1713, "step": 1126 }, { "epoch": 0.791988756148981, "grad_norm": 0.42186906933784485, "learning_rate": 4.743499648629656e-06, "loss": 0.0813, "step": 1127 }, { "epoch": 0.7926914968376669, "grad_norm": 0.5086817741394043, "learning_rate": 4.747716092761771e-06, "loss": 0.0983, "step": 1128 }, { "epoch": 0.7933942375263527, "grad_norm": 0.4191987216472626, "learning_rate": 4.751932536893886e-06, "loss": 0.0749, "step": 1129 }, { "epoch": 0.7940969782150387, "grad_norm": 0.37019550800323486, "learning_rate": 4.756148981026002e-06, "loss": 0.0641, "step": 1130 }, { "epoch": 0.7947997189037245, "grad_norm": 0.37278786301612854, "learning_rate": 4.760365425158117e-06, "loss": 0.0658, "step": 1131 }, { "epoch": 0.7955024595924104, "grad_norm": 0.51566082239151, "learning_rate": 4.764581869290232e-06, "loss": 0.0746, "step": 1132 }, { "epoch": 0.7962052002810963, "grad_norm": 0.38276898860931396, "learning_rate": 4.768798313422348e-06, "loss": 0.08, "step": 1133 }, { "epoch": 0.7969079409697821, "grad_norm": 0.4269614815711975, "learning_rate": 4.773014757554463e-06, "loss": 0.081, "step": 1134 }, { "epoch": 0.797610681658468, "grad_norm": 0.5200303196907043, "learning_rate": 4.7772312016865776e-06, "loss": 0.0971, "step": 1135 }, { "epoch": 0.7983134223471539, "grad_norm": 0.634192705154419, "learning_rate": 4.7814476458186925e-06, "loss": 0.0982, "step": 1136 }, { "epoch": 0.7990161630358398, "grad_norm": 0.392656534910202, "learning_rate": 4.785664089950808e-06, "loss": 0.0628, "step": 1137 }, { "epoch": 0.7997189037245257, "grad_norm": 0.42863044142723083, "learning_rate": 4.789880534082923e-06, "loss": 0.0918, "step": 1138 }, { "epoch": 0.8004216444132115, "grad_norm": 0.5683818459510803, "learning_rate": 4.794096978215038e-06, "loss": 0.085, "step": 1139 }, { "epoch": 0.8011243851018974, "grad_norm": 0.5904935598373413, "learning_rate": 4.798313422347154e-06, "loss": 0.091, "step": 1140 }, { "epoch": 0.8018271257905832, "grad_norm": 0.6184195280075073, "learning_rate": 4.80252986647927e-06, "loss": 0.1335, "step": 1141 }, { "epoch": 0.8025298664792692, "grad_norm": 0.9794450998306274, "learning_rate": 4.806746310611385e-06, "loss": 0.1007, "step": 1142 }, { "epoch": 0.803232607167955, "grad_norm": 0.5830860137939453, "learning_rate": 4.8109627547435e-06, "loss": 0.1198, "step": 1143 }, { "epoch": 0.8039353478566409, "grad_norm": 0.630998969078064, "learning_rate": 4.8151791988756154e-06, "loss": 0.1146, "step": 1144 }, { "epoch": 0.8046380885453268, "grad_norm": 1.2683783769607544, "learning_rate": 4.81939564300773e-06, "loss": 0.1694, "step": 1145 }, { "epoch": 0.8053408292340126, "grad_norm": 1.239225149154663, "learning_rate": 4.823612087139845e-06, "loss": 0.1926, "step": 1146 }, { "epoch": 0.8060435699226985, "grad_norm": 1.0332289934158325, "learning_rate": 4.827828531271961e-06, "loss": 0.2367, "step": 1147 }, { "epoch": 0.8067463106113844, "grad_norm": 4.32489013671875, "learning_rate": 4.832044975404076e-06, "loss": 0.3335, "step": 1148 }, { "epoch": 0.8074490513000703, "grad_norm": 1.7695239782333374, "learning_rate": 4.836261419536191e-06, "loss": 0.4787, "step": 1149 }, { "epoch": 0.8081517919887562, "grad_norm": 2.8655316829681396, "learning_rate": 4.840477863668307e-06, "loss": 0.6793, "step": 1150 }, { "epoch": 0.808854532677442, "grad_norm": 0.850614607334137, "learning_rate": 4.844694307800422e-06, "loss": 0.2045, "step": 1151 }, { "epoch": 0.8095572733661279, "grad_norm": 0.4564149081707001, "learning_rate": 4.848910751932537e-06, "loss": 0.0878, "step": 1152 }, { "epoch": 0.8102600140548137, "grad_norm": 0.42790696024894714, "learning_rate": 4.853127196064652e-06, "loss": 0.0974, "step": 1153 }, { "epoch": 0.8109627547434997, "grad_norm": 0.44735702872276306, "learning_rate": 4.857343640196767e-06, "loss": 0.0633, "step": 1154 }, { "epoch": 0.8116654954321855, "grad_norm": 0.7085447311401367, "learning_rate": 4.861560084328882e-06, "loss": 0.094, "step": 1155 }, { "epoch": 0.8123682361208714, "grad_norm": 0.40218061208724976, "learning_rate": 4.865776528460998e-06, "loss": 0.0449, "step": 1156 }, { "epoch": 0.8130709768095573, "grad_norm": 0.4578198790550232, "learning_rate": 4.869992972593114e-06, "loss": 0.0708, "step": 1157 }, { "epoch": 0.8137737174982431, "grad_norm": 0.43399369716644287, "learning_rate": 4.874209416725229e-06, "loss": 0.0941, "step": 1158 }, { "epoch": 0.814476458186929, "grad_norm": 0.5096080899238586, "learning_rate": 4.878425860857344e-06, "loss": 0.0699, "step": 1159 }, { "epoch": 0.8151791988756149, "grad_norm": 0.4109979569911957, "learning_rate": 4.882642304989459e-06, "loss": 0.0528, "step": 1160 }, { "epoch": 0.8158819395643008, "grad_norm": 0.7273101806640625, "learning_rate": 4.8868587491215746e-06, "loss": 0.1151, "step": 1161 }, { "epoch": 0.8165846802529867, "grad_norm": 0.503545880317688, "learning_rate": 4.8910751932536895e-06, "loss": 0.0654, "step": 1162 }, { "epoch": 0.8172874209416725, "grad_norm": 0.48698610067367554, "learning_rate": 4.8952916373858045e-06, "loss": 0.0801, "step": 1163 }, { "epoch": 0.8179901616303584, "grad_norm": 0.7083731889724731, "learning_rate": 4.89950808151792e-06, "loss": 0.0777, "step": 1164 }, { "epoch": 0.8186929023190442, "grad_norm": 1.7980629205703735, "learning_rate": 4.903724525650035e-06, "loss": 0.1062, "step": 1165 }, { "epoch": 0.8193956430077302, "grad_norm": 0.8011090159416199, "learning_rate": 4.90794096978215e-06, "loss": 0.109, "step": 1166 }, { "epoch": 0.8200983836964161, "grad_norm": 0.7370602488517761, "learning_rate": 4.912157413914266e-06, "loss": 0.0938, "step": 1167 }, { "epoch": 0.8208011243851019, "grad_norm": 0.6493812203407288, "learning_rate": 4.916373858046381e-06, "loss": 0.1292, "step": 1168 }, { "epoch": 0.8215038650737878, "grad_norm": 0.9117781519889832, "learning_rate": 4.920590302178496e-06, "loss": 0.1309, "step": 1169 }, { "epoch": 0.8222066057624736, "grad_norm": 0.8840309977531433, "learning_rate": 4.924806746310611e-06, "loss": 0.1842, "step": 1170 }, { "epoch": 0.8229093464511595, "grad_norm": 0.9525498747825623, "learning_rate": 4.929023190442727e-06, "loss": 0.1624, "step": 1171 }, { "epoch": 0.8236120871398454, "grad_norm": 1.0127084255218506, "learning_rate": 4.933239634574842e-06, "loss": 0.2792, "step": 1172 }, { "epoch": 0.8243148278285313, "grad_norm": 1.5282233953475952, "learning_rate": 4.937456078706957e-06, "loss": 0.423, "step": 1173 }, { "epoch": 0.8250175685172172, "grad_norm": 4.333634376525879, "learning_rate": 4.941672522839073e-06, "loss": 0.4629, "step": 1174 }, { "epoch": 0.825720309205903, "grad_norm": 7.724267959594727, "learning_rate": 4.945888966971188e-06, "loss": 0.5784, "step": 1175 }, { "epoch": 0.8264230498945889, "grad_norm": 0.9572295546531677, "learning_rate": 4.950105411103303e-06, "loss": 0.1862, "step": 1176 }, { "epoch": 0.8271257905832747, "grad_norm": 0.38968056440353394, "learning_rate": 4.954321855235418e-06, "loss": 0.0719, "step": 1177 }, { "epoch": 0.8278285312719607, "grad_norm": 0.46957895159721375, "learning_rate": 4.958538299367534e-06, "loss": 0.0893, "step": 1178 }, { "epoch": 0.8285312719606466, "grad_norm": 0.4812454283237457, "learning_rate": 4.962754743499649e-06, "loss": 0.083, "step": 1179 }, { "epoch": 0.8292340126493324, "grad_norm": 0.39622899889945984, "learning_rate": 4.966971187631764e-06, "loss": 0.0692, "step": 1180 }, { "epoch": 0.8299367533380183, "grad_norm": 0.41332119703292847, "learning_rate": 4.971187631763879e-06, "loss": 0.0535, "step": 1181 }, { "epoch": 0.8306394940267041, "grad_norm": 0.4430185854434967, "learning_rate": 4.975404075895994e-06, "loss": 0.0733, "step": 1182 }, { "epoch": 0.83134223471539, "grad_norm": 0.4780099093914032, "learning_rate": 4.979620520028109e-06, "loss": 0.0726, "step": 1183 }, { "epoch": 0.8320449754040758, "grad_norm": 1.0255191326141357, "learning_rate": 4.983836964160225e-06, "loss": 0.0727, "step": 1184 }, { "epoch": 0.8327477160927618, "grad_norm": 0.498713880777359, "learning_rate": 4.98805340829234e-06, "loss": 0.0581, "step": 1185 }, { "epoch": 0.8334504567814477, "grad_norm": 0.5978479385375977, "learning_rate": 4.992269852424456e-06, "loss": 0.0812, "step": 1186 }, { "epoch": 0.8341531974701335, "grad_norm": 0.42356783151626587, "learning_rate": 4.996486296556571e-06, "loss": 0.0714, "step": 1187 }, { "epoch": 0.8348559381588194, "grad_norm": 0.619775116443634, "learning_rate": 5.0007027406886865e-06, "loss": 0.0894, "step": 1188 }, { "epoch": 0.8355586788475052, "grad_norm": 0.44371336698532104, "learning_rate": 5.0049191848208015e-06, "loss": 0.074, "step": 1189 }, { "epoch": 0.8362614195361912, "grad_norm": 0.6306453347206116, "learning_rate": 5.009135628952916e-06, "loss": 0.1016, "step": 1190 }, { "epoch": 0.8369641602248771, "grad_norm": 0.5846635699272156, "learning_rate": 5.013352073085032e-06, "loss": 0.1154, "step": 1191 }, { "epoch": 0.8376669009135629, "grad_norm": 0.555352509021759, "learning_rate": 5.017568517217147e-06, "loss": 0.0778, "step": 1192 }, { "epoch": 0.8383696416022488, "grad_norm": 0.6679520010948181, "learning_rate": 5.021784961349262e-06, "loss": 0.1251, "step": 1193 }, { "epoch": 0.8390723822909346, "grad_norm": 1.947561264038086, "learning_rate": 5.026001405481377e-06, "loss": 0.1551, "step": 1194 }, { "epoch": 0.8397751229796205, "grad_norm": 1.1122002601623535, "learning_rate": 5.030217849613493e-06, "loss": 0.148, "step": 1195 }, { "epoch": 0.8404778636683063, "grad_norm": 0.8267525434494019, "learning_rate": 5.034434293745608e-06, "loss": 0.1781, "step": 1196 }, { "epoch": 0.8411806043569923, "grad_norm": 1.0851391553878784, "learning_rate": 5.038650737877723e-06, "loss": 0.2989, "step": 1197 }, { "epoch": 0.8418833450456782, "grad_norm": 1.7518491744995117, "learning_rate": 5.0428671820098385e-06, "loss": 0.3572, "step": 1198 }, { "epoch": 0.842586085734364, "grad_norm": 1.8160192966461182, "learning_rate": 5.0470836261419534e-06, "loss": 0.4278, "step": 1199 }, { "epoch": 0.8432888264230499, "grad_norm": 4.666247367858887, "learning_rate": 5.051300070274068e-06, "loss": 0.5935, "step": 1200 }, { "epoch": 0.8439915671117357, "grad_norm": 0.681096076965332, "learning_rate": 5.055516514406185e-06, "loss": 0.1451, "step": 1201 }, { "epoch": 0.8446943078004217, "grad_norm": 0.5691401958465576, "learning_rate": 5.0597329585383e-06, "loss": 0.0967, "step": 1202 }, { "epoch": 0.8453970484891076, "grad_norm": 0.5097993612289429, "learning_rate": 5.063949402670415e-06, "loss": 0.0826, "step": 1203 }, { "epoch": 0.8460997891777934, "grad_norm": 0.49392056465148926, "learning_rate": 5.06816584680253e-06, "loss": 0.0638, "step": 1204 }, { "epoch": 0.8468025298664793, "grad_norm": 0.77614825963974, "learning_rate": 5.072382290934646e-06, "loss": 0.0764, "step": 1205 }, { "epoch": 0.8475052705551651, "grad_norm": 0.45939427614212036, "learning_rate": 5.076598735066761e-06, "loss": 0.068, "step": 1206 }, { "epoch": 0.848208011243851, "grad_norm": 0.5951583385467529, "learning_rate": 5.0808151791988755e-06, "loss": 0.0877, "step": 1207 }, { "epoch": 0.8489107519325368, "grad_norm": 0.43766680359840393, "learning_rate": 5.085031623330991e-06, "loss": 0.0868, "step": 1208 }, { "epoch": 0.8496134926212228, "grad_norm": 0.3859589695930481, "learning_rate": 5.089248067463106e-06, "loss": 0.0591, "step": 1209 }, { "epoch": 0.8503162333099087, "grad_norm": 0.480936735868454, "learning_rate": 5.093464511595221e-06, "loss": 0.0749, "step": 1210 }, { "epoch": 0.8510189739985945, "grad_norm": 0.5146955847740173, "learning_rate": 5.097680955727336e-06, "loss": 0.0753, "step": 1211 }, { "epoch": 0.8517217146872804, "grad_norm": 0.47993525862693787, "learning_rate": 5.101897399859452e-06, "loss": 0.0714, "step": 1212 }, { "epoch": 0.8524244553759662, "grad_norm": 0.6025767922401428, "learning_rate": 5.106113843991567e-06, "loss": 0.0801, "step": 1213 }, { "epoch": 0.8531271960646521, "grad_norm": 0.479079008102417, "learning_rate": 5.110330288123682e-06, "loss": 0.0655, "step": 1214 }, { "epoch": 0.8538299367533381, "grad_norm": 0.6060339212417603, "learning_rate": 5.114546732255798e-06, "loss": 0.0968, "step": 1215 }, { "epoch": 0.8545326774420239, "grad_norm": 0.5554654002189636, "learning_rate": 5.118763176387913e-06, "loss": 0.1005, "step": 1216 }, { "epoch": 0.8552354181307098, "grad_norm": 0.5918015837669373, "learning_rate": 5.122979620520028e-06, "loss": 0.1308, "step": 1217 }, { "epoch": 0.8559381588193956, "grad_norm": 0.5722882747650146, "learning_rate": 5.127196064652144e-06, "loss": 0.0831, "step": 1218 }, { "epoch": 0.8566408995080815, "grad_norm": 0.8069517016410828, "learning_rate": 5.131412508784259e-06, "loss": 0.1334, "step": 1219 }, { "epoch": 0.8573436401967673, "grad_norm": 0.7495511770248413, "learning_rate": 5.135628952916374e-06, "loss": 0.1296, "step": 1220 }, { "epoch": 0.8580463808854533, "grad_norm": 1.2389112710952759, "learning_rate": 5.139845397048489e-06, "loss": 0.173, "step": 1221 }, { "epoch": 0.8587491215741392, "grad_norm": 1.4261969327926636, "learning_rate": 5.144061841180605e-06, "loss": 0.2557, "step": 1222 }, { "epoch": 0.859451862262825, "grad_norm": 1.9844884872436523, "learning_rate": 5.14827828531272e-06, "loss": 0.3725, "step": 1223 }, { "epoch": 0.8601546029515109, "grad_norm": 3.3915112018585205, "learning_rate": 5.152494729444835e-06, "loss": 0.5385, "step": 1224 }, { "epoch": 0.8608573436401967, "grad_norm": 3.958629846572876, "learning_rate": 5.1567111735769504e-06, "loss": 0.6085, "step": 1225 }, { "epoch": 0.8615600843288826, "grad_norm": 0.6127060055732727, "learning_rate": 5.160927617709065e-06, "loss": 0.1576, "step": 1226 }, { "epoch": 0.8622628250175686, "grad_norm": 0.48735305666923523, "learning_rate": 5.16514406184118e-06, "loss": 0.0892, "step": 1227 }, { "epoch": 0.8629655657062544, "grad_norm": 0.3977656662464142, "learning_rate": 5.169360505973295e-06, "loss": 0.084, "step": 1228 }, { "epoch": 0.8636683063949403, "grad_norm": 0.625966489315033, "learning_rate": 5.173576950105411e-06, "loss": 0.077, "step": 1229 }, { "epoch": 0.8643710470836261, "grad_norm": 0.47247380018234253, "learning_rate": 5.177793394237526e-06, "loss": 0.0691, "step": 1230 }, { "epoch": 0.865073787772312, "grad_norm": 0.4537741243839264, "learning_rate": 5.182009838369642e-06, "loss": 0.0611, "step": 1231 }, { "epoch": 0.8657765284609978, "grad_norm": 0.5633171796798706, "learning_rate": 5.186226282501758e-06, "loss": 0.086, "step": 1232 }, { "epoch": 0.8664792691496838, "grad_norm": 0.46536383032798767, "learning_rate": 5.1904427266338725e-06, "loss": 0.0708, "step": 1233 }, { "epoch": 0.8671820098383697, "grad_norm": 0.471202552318573, "learning_rate": 5.1946591707659875e-06, "loss": 0.0878, "step": 1234 }, { "epoch": 0.8678847505270555, "grad_norm": 0.420541912317276, "learning_rate": 5.198875614898103e-06, "loss": 0.0622, "step": 1235 }, { "epoch": 0.8685874912157414, "grad_norm": 0.46385592222213745, "learning_rate": 5.203092059030218e-06, "loss": 0.0828, "step": 1236 }, { "epoch": 0.8692902319044272, "grad_norm": 0.484045147895813, "learning_rate": 5.207308503162333e-06, "loss": 0.0601, "step": 1237 }, { "epoch": 0.8699929725931131, "grad_norm": 0.726810097694397, "learning_rate": 5.211524947294448e-06, "loss": 0.079, "step": 1238 }, { "epoch": 0.8706957132817991, "grad_norm": 0.8518301844596863, "learning_rate": 5.215741391426564e-06, "loss": 0.0763, "step": 1239 }, { "epoch": 0.8713984539704849, "grad_norm": 0.5617519617080688, "learning_rate": 5.219957835558679e-06, "loss": 0.1044, "step": 1240 }, { "epoch": 0.8721011946591708, "grad_norm": 1.5282658338546753, "learning_rate": 5.224174279690794e-06, "loss": 0.1371, "step": 1241 }, { "epoch": 0.8728039353478566, "grad_norm": 0.46899619698524475, "learning_rate": 5.2283907238229096e-06, "loss": 0.0697, "step": 1242 }, { "epoch": 0.8735066760365425, "grad_norm": 0.7257707715034485, "learning_rate": 5.2326071679550245e-06, "loss": 0.1084, "step": 1243 }, { "epoch": 0.8742094167252283, "grad_norm": 0.8180670142173767, "learning_rate": 5.2368236120871395e-06, "loss": 0.1086, "step": 1244 }, { "epoch": 0.8749121574139143, "grad_norm": 0.7886786460876465, "learning_rate": 5.241040056219255e-06, "loss": 0.1709, "step": 1245 }, { "epoch": 0.8756148981026002, "grad_norm": 1.2745620012283325, "learning_rate": 5.245256500351371e-06, "loss": 0.1853, "step": 1246 }, { "epoch": 0.876317638791286, "grad_norm": 1.5694239139556885, "learning_rate": 5.249472944483486e-06, "loss": 0.2725, "step": 1247 }, { "epoch": 0.8770203794799719, "grad_norm": 1.3470216989517212, "learning_rate": 5.253689388615601e-06, "loss": 0.2897, "step": 1248 }, { "epoch": 0.8777231201686577, "grad_norm": 11.043054580688477, "learning_rate": 5.257905832747717e-06, "loss": 0.452, "step": 1249 }, { "epoch": 0.8784258608573436, "grad_norm": 2.872622013092041, "learning_rate": 5.262122276879832e-06, "loss": 0.5626, "step": 1250 }, { "epoch": 0.8791286015460296, "grad_norm": 0.6015369296073914, "learning_rate": 5.266338721011947e-06, "loss": 0.1354, "step": 1251 }, { "epoch": 0.8798313422347154, "grad_norm": 0.49007752537727356, "learning_rate": 5.270555165144062e-06, "loss": 0.0594, "step": 1252 }, { "epoch": 0.8805340829234013, "grad_norm": 2.1010236740112305, "learning_rate": 5.274771609276177e-06, "loss": 0.0975, "step": 1253 }, { "epoch": 0.8812368236120871, "grad_norm": 0.6334958076477051, "learning_rate": 5.278988053408292e-06, "loss": 0.0748, "step": 1254 }, { "epoch": 0.881939564300773, "grad_norm": 0.5345879793167114, "learning_rate": 5.283204497540407e-06, "loss": 0.0633, "step": 1255 }, { "epoch": 0.8826423049894588, "grad_norm": 0.37770527601242065, "learning_rate": 5.287420941672523e-06, "loss": 0.0563, "step": 1256 }, { "epoch": 0.8833450456781448, "grad_norm": 0.3538095951080322, "learning_rate": 5.291637385804638e-06, "loss": 0.0693, "step": 1257 }, { "epoch": 0.8840477863668307, "grad_norm": 0.3979608714580536, "learning_rate": 5.295853829936753e-06, "loss": 0.0548, "step": 1258 }, { "epoch": 0.8847505270555165, "grad_norm": 0.5348856449127197, "learning_rate": 5.300070274068869e-06, "loss": 0.1133, "step": 1259 }, { "epoch": 0.8854532677442024, "grad_norm": 0.9485612511634827, "learning_rate": 5.3042867182009845e-06, "loss": 0.0806, "step": 1260 }, { "epoch": 0.8861560084328882, "grad_norm": 0.44248899817466736, "learning_rate": 5.308503162333099e-06, "loss": 0.0697, "step": 1261 }, { "epoch": 0.8868587491215741, "grad_norm": 0.44639384746551514, "learning_rate": 5.312719606465214e-06, "loss": 0.0787, "step": 1262 }, { "epoch": 0.8875614898102601, "grad_norm": 0.4423125982284546, "learning_rate": 5.31693605059733e-06, "loss": 0.0982, "step": 1263 }, { "epoch": 0.8882642304989459, "grad_norm": 0.4814676344394684, "learning_rate": 5.321152494729445e-06, "loss": 0.062, "step": 1264 }, { "epoch": 0.8889669711876318, "grad_norm": 0.5403311848640442, "learning_rate": 5.32536893886156e-06, "loss": 0.1135, "step": 1265 }, { "epoch": 0.8896697118763176, "grad_norm": 0.5318147540092468, "learning_rate": 5.329585382993676e-06, "loss": 0.1073, "step": 1266 }, { "epoch": 0.8903724525650035, "grad_norm": 0.5262113809585571, "learning_rate": 5.333801827125791e-06, "loss": 0.0786, "step": 1267 }, { "epoch": 0.8910751932536893, "grad_norm": 0.6115972399711609, "learning_rate": 5.338018271257906e-06, "loss": 0.1055, "step": 1268 }, { "epoch": 0.8917779339423753, "grad_norm": 0.8191081285476685, "learning_rate": 5.3422347153900215e-06, "loss": 0.1031, "step": 1269 }, { "epoch": 0.8924806746310612, "grad_norm": 0.7566415071487427, "learning_rate": 5.3464511595221365e-06, "loss": 0.1668, "step": 1270 }, { "epoch": 0.893183415319747, "grad_norm": 0.8187085390090942, "learning_rate": 5.350667603654251e-06, "loss": 0.1999, "step": 1271 }, { "epoch": 0.8938861560084329, "grad_norm": 1.050538420677185, "learning_rate": 5.354884047786366e-06, "loss": 0.2396, "step": 1272 }, { "epoch": 0.8945888966971187, "grad_norm": 1.2962671518325806, "learning_rate": 5.359100491918482e-06, "loss": 0.3081, "step": 1273 }, { "epoch": 0.8952916373858046, "grad_norm": 1.9510174989700317, "learning_rate": 5.363316936050597e-06, "loss": 0.4128, "step": 1274 }, { "epoch": 0.8959943780744906, "grad_norm": 2.040647506713867, "learning_rate": 5.367533380182713e-06, "loss": 0.5413, "step": 1275 }, { "epoch": 0.8966971187631764, "grad_norm": 0.6382134556770325, "learning_rate": 5.371749824314829e-06, "loss": 0.1509, "step": 1276 }, { "epoch": 0.8973998594518623, "grad_norm": 0.448557049036026, "learning_rate": 5.375966268446944e-06, "loss": 0.0742, "step": 1277 }, { "epoch": 0.8981026001405481, "grad_norm": 0.7419982552528381, "learning_rate": 5.3801827125790585e-06, "loss": 0.0727, "step": 1278 }, { "epoch": 0.898805340829234, "grad_norm": 0.5711937546730042, "learning_rate": 5.3843991567111735e-06, "loss": 0.0921, "step": 1279 }, { "epoch": 0.8995080815179198, "grad_norm": 0.4111557602882385, "learning_rate": 5.388615600843289e-06, "loss": 0.0619, "step": 1280 }, { "epoch": 0.9002108222066058, "grad_norm": 0.397163987159729, "learning_rate": 5.392832044975404e-06, "loss": 0.0631, "step": 1281 }, { "epoch": 0.9009135628952917, "grad_norm": 0.41996610164642334, "learning_rate": 5.397048489107519e-06, "loss": 0.0698, "step": 1282 }, { "epoch": 0.9016163035839775, "grad_norm": 0.48643046617507935, "learning_rate": 5.401264933239635e-06, "loss": 0.0824, "step": 1283 }, { "epoch": 0.9023190442726634, "grad_norm": 0.41281092166900635, "learning_rate": 5.40548137737175e-06, "loss": 0.0612, "step": 1284 }, { "epoch": 0.9030217849613492, "grad_norm": 0.5477018356323242, "learning_rate": 5.409697821503865e-06, "loss": 0.0624, "step": 1285 }, { "epoch": 0.9037245256500351, "grad_norm": 0.631848931312561, "learning_rate": 5.413914265635981e-06, "loss": 0.089, "step": 1286 }, { "epoch": 0.9044272663387211, "grad_norm": 0.39692237973213196, "learning_rate": 5.418130709768096e-06, "loss": 0.053, "step": 1287 }, { "epoch": 0.9051300070274069, "grad_norm": 0.672526478767395, "learning_rate": 5.4223471539002105e-06, "loss": 0.0818, "step": 1288 }, { "epoch": 0.9058327477160928, "grad_norm": 0.406676322221756, "learning_rate": 5.4265635980323255e-06, "loss": 0.0645, "step": 1289 }, { "epoch": 0.9065354884047786, "grad_norm": 0.5050777196884155, "learning_rate": 5.430780042164442e-06, "loss": 0.0746, "step": 1290 }, { "epoch": 0.9072382290934645, "grad_norm": 0.5083892941474915, "learning_rate": 5.434996486296557e-06, "loss": 0.0983, "step": 1291 }, { "epoch": 0.9079409697821503, "grad_norm": 0.48875847458839417, "learning_rate": 5.439212930428672e-06, "loss": 0.0792, "step": 1292 }, { "epoch": 0.9086437104708363, "grad_norm": 0.5346589684486389, "learning_rate": 5.443429374560788e-06, "loss": 0.1039, "step": 1293 }, { "epoch": 0.9093464511595222, "grad_norm": 0.5811461210250854, "learning_rate": 5.447645818692903e-06, "loss": 0.1218, "step": 1294 }, { "epoch": 0.910049191848208, "grad_norm": 0.691685140132904, "learning_rate": 5.451862262825018e-06, "loss": 0.134, "step": 1295 }, { "epoch": 0.9107519325368939, "grad_norm": 0.8378998041152954, "learning_rate": 5.456078706957133e-06, "loss": 0.1406, "step": 1296 }, { "epoch": 0.9114546732255797, "grad_norm": 1.1371963024139404, "learning_rate": 5.460295151089248e-06, "loss": 0.2567, "step": 1297 }, { "epoch": 0.9121574139142656, "grad_norm": 1.5318646430969238, "learning_rate": 5.464511595221363e-06, "loss": 0.3614, "step": 1298 }, { "epoch": 0.9128601546029516, "grad_norm": 1.6692320108413696, "learning_rate": 5.468728039353478e-06, "loss": 0.4279, "step": 1299 }, { "epoch": 0.9135628952916374, "grad_norm": 3.0119264125823975, "learning_rate": 5.472944483485594e-06, "loss": 0.5849, "step": 1300 }, { "epoch": 0.9142656359803233, "grad_norm": 0.4887550175189972, "learning_rate": 5.477160927617709e-06, "loss": 0.1448, "step": 1301 }, { "epoch": 0.9149683766690091, "grad_norm": 0.5391324162483215, "learning_rate": 5.481377371749824e-06, "loss": 0.1089, "step": 1302 }, { "epoch": 0.915671117357695, "grad_norm": 0.36351478099823, "learning_rate": 5.48559381588194e-06, "loss": 0.0696, "step": 1303 }, { "epoch": 0.9163738580463809, "grad_norm": 0.37780648469924927, "learning_rate": 5.489810260014055e-06, "loss": 0.0698, "step": 1304 }, { "epoch": 0.9170765987350668, "grad_norm": 0.4764348864555359, "learning_rate": 5.4940267041461705e-06, "loss": 0.0659, "step": 1305 }, { "epoch": 0.9177793394237527, "grad_norm": 0.3792550265789032, "learning_rate": 5.4982431482782854e-06, "loss": 0.0641, "step": 1306 }, { "epoch": 0.9184820801124385, "grad_norm": 0.34496378898620605, "learning_rate": 5.502459592410401e-06, "loss": 0.0531, "step": 1307 }, { "epoch": 0.9191848208011244, "grad_norm": 0.42677634954452515, "learning_rate": 5.506676036542516e-06, "loss": 0.082, "step": 1308 }, { "epoch": 0.9198875614898102, "grad_norm": 0.7613552808761597, "learning_rate": 5.510892480674631e-06, "loss": 0.076, "step": 1309 }, { "epoch": 0.9205903021784961, "grad_norm": 0.8581016659736633, "learning_rate": 5.515108924806747e-06, "loss": 0.0743, "step": 1310 }, { "epoch": 0.921293042867182, "grad_norm": 0.4055659770965576, "learning_rate": 5.519325368938862e-06, "loss": 0.0694, "step": 1311 }, { "epoch": 0.9219957835558679, "grad_norm": 0.7093830704689026, "learning_rate": 5.523541813070977e-06, "loss": 0.0635, "step": 1312 }, { "epoch": 0.9226985242445538, "grad_norm": 0.6829875111579895, "learning_rate": 5.527758257203092e-06, "loss": 0.0886, "step": 1313 }, { "epoch": 0.9234012649332396, "grad_norm": 0.5044428706169128, "learning_rate": 5.5319747013352075e-06, "loss": 0.0647, "step": 1314 }, { "epoch": 0.9241040056219255, "grad_norm": 0.49848926067352295, "learning_rate": 5.5361911454673225e-06, "loss": 0.1115, "step": 1315 }, { "epoch": 0.9248067463106114, "grad_norm": 0.5784189105033875, "learning_rate": 5.540407589599437e-06, "loss": 0.0775, "step": 1316 }, { "epoch": 0.9255094869992972, "grad_norm": 0.46392640471458435, "learning_rate": 5.544624033731553e-06, "loss": 0.0819, "step": 1317 }, { "epoch": 0.9262122276879832, "grad_norm": 0.5460196733474731, "learning_rate": 5.548840477863668e-06, "loss": 0.0584, "step": 1318 }, { "epoch": 0.926914968376669, "grad_norm": 0.7532581686973572, "learning_rate": 5.553056921995783e-06, "loss": 0.1332, "step": 1319 }, { "epoch": 0.9276177090653549, "grad_norm": 0.8058597445487976, "learning_rate": 5.5572733661279e-06, "loss": 0.1523, "step": 1320 }, { "epoch": 0.9283204497540407, "grad_norm": 0.8502719402313232, "learning_rate": 5.561489810260015e-06, "loss": 0.1396, "step": 1321 }, { "epoch": 0.9290231904427266, "grad_norm": 1.1816110610961914, "learning_rate": 5.56570625439213e-06, "loss": 0.2586, "step": 1322 }, { "epoch": 0.9297259311314126, "grad_norm": 1.3864552974700928, "learning_rate": 5.5699226985242446e-06, "loss": 0.3156, "step": 1323 }, { "epoch": 0.9304286718200984, "grad_norm": 1.9504446983337402, "learning_rate": 5.57413914265636e-06, "loss": 0.4876, "step": 1324 }, { "epoch": 0.9311314125087843, "grad_norm": 4.63406229019165, "learning_rate": 5.578355586788475e-06, "loss": 0.5861, "step": 1325 }, { "epoch": 0.9318341531974701, "grad_norm": 0.8462967872619629, "learning_rate": 5.58257203092059e-06, "loss": 0.1831, "step": 1326 }, { "epoch": 0.932536893886156, "grad_norm": 0.3981171250343323, "learning_rate": 5.586788475052706e-06, "loss": 0.0763, "step": 1327 }, { "epoch": 0.9332396345748419, "grad_norm": 0.5957453846931458, "learning_rate": 5.591004919184821e-06, "loss": 0.0752, "step": 1328 }, { "epoch": 0.9339423752635277, "grad_norm": 0.36106953024864197, "learning_rate": 5.595221363316936e-06, "loss": 0.0706, "step": 1329 }, { "epoch": 0.9346451159522137, "grad_norm": 0.416818767786026, "learning_rate": 5.599437807449051e-06, "loss": 0.0657, "step": 1330 }, { "epoch": 0.9353478566408995, "grad_norm": 0.48974719643592834, "learning_rate": 5.603654251581167e-06, "loss": 0.0636, "step": 1331 }, { "epoch": 0.9360505973295854, "grad_norm": 0.3686610758304596, "learning_rate": 5.607870695713282e-06, "loss": 0.0628, "step": 1332 }, { "epoch": 0.9367533380182712, "grad_norm": 1.3382245302200317, "learning_rate": 5.6120871398453965e-06, "loss": 0.0945, "step": 1333 }, { "epoch": 0.9374560787069571, "grad_norm": 0.4256041347980499, "learning_rate": 5.616303583977512e-06, "loss": 0.062, "step": 1334 }, { "epoch": 0.938158819395643, "grad_norm": 0.597256064414978, "learning_rate": 5.620520028109628e-06, "loss": 0.0581, "step": 1335 }, { "epoch": 0.9388615600843289, "grad_norm": 0.4565102159976959, "learning_rate": 5.624736472241743e-06, "loss": 0.0639, "step": 1336 }, { "epoch": 0.9395643007730148, "grad_norm": 0.5530626177787781, "learning_rate": 5.628952916373859e-06, "loss": 0.0491, "step": 1337 }, { "epoch": 0.9402670414617006, "grad_norm": 0.501365065574646, "learning_rate": 5.633169360505974e-06, "loss": 0.0826, "step": 1338 }, { "epoch": 0.9409697821503865, "grad_norm": 0.43901586532592773, "learning_rate": 5.637385804638089e-06, "loss": 0.0754, "step": 1339 }, { "epoch": 0.9416725228390724, "grad_norm": 0.5563682913780212, "learning_rate": 5.641602248770204e-06, "loss": 0.0953, "step": 1340 }, { "epoch": 0.9423752635277582, "grad_norm": 0.5966431498527527, "learning_rate": 5.6458186929023195e-06, "loss": 0.0917, "step": 1341 }, { "epoch": 0.9430780042164442, "grad_norm": 0.5543459057807922, "learning_rate": 5.650035137034434e-06, "loss": 0.1013, "step": 1342 }, { "epoch": 0.94378074490513, "grad_norm": 0.6732911467552185, "learning_rate": 5.654251581166549e-06, "loss": 0.1206, "step": 1343 }, { "epoch": 0.9444834855938159, "grad_norm": 1.4810049533843994, "learning_rate": 5.658468025298665e-06, "loss": 0.1247, "step": 1344 }, { "epoch": 0.9451862262825017, "grad_norm": 0.656135082244873, "learning_rate": 5.66268446943078e-06, "loss": 0.1217, "step": 1345 }, { "epoch": 0.9458889669711876, "grad_norm": 1.120105504989624, "learning_rate": 5.666900913562895e-06, "loss": 0.1747, "step": 1346 }, { "epoch": 0.9465917076598735, "grad_norm": 1.4287022352218628, "learning_rate": 5.67111735769501e-06, "loss": 0.2768, "step": 1347 }, { "epoch": 0.9472944483485594, "grad_norm": 1.9381290674209595, "learning_rate": 5.675333801827126e-06, "loss": 0.3781, "step": 1348 }, { "epoch": 0.9479971890372453, "grad_norm": 3.47402286529541, "learning_rate": 5.679550245959241e-06, "loss": 0.391, "step": 1349 }, { "epoch": 0.9486999297259311, "grad_norm": 2.4652442932128906, "learning_rate": 5.6837666900913565e-06, "loss": 0.5234, "step": 1350 }, { "epoch": 0.949402670414617, "grad_norm": 0.948367714881897, "learning_rate": 5.687983134223472e-06, "loss": 0.1629, "step": 1351 }, { "epoch": 0.9501054111033029, "grad_norm": 0.4193437397480011, "learning_rate": 5.692199578355587e-06, "loss": 0.0742, "step": 1352 }, { "epoch": 0.9508081517919887, "grad_norm": 0.36522582173347473, "learning_rate": 5.696416022487702e-06, "loss": 0.0555, "step": 1353 }, { "epoch": 0.9515108924806747, "grad_norm": 0.40641719102859497, "learning_rate": 5.700632466619818e-06, "loss": 0.0516, "step": 1354 }, { "epoch": 0.9522136331693605, "grad_norm": 0.6382269263267517, "learning_rate": 5.704848910751933e-06, "loss": 0.0825, "step": 1355 }, { "epoch": 0.9529163738580464, "grad_norm": 0.45060399174690247, "learning_rate": 5.709065354884048e-06, "loss": 0.0743, "step": 1356 }, { "epoch": 0.9536191145467322, "grad_norm": 0.4931419789791107, "learning_rate": 5.713281799016163e-06, "loss": 0.0608, "step": 1357 }, { "epoch": 0.9543218552354181, "grad_norm": 0.5840076804161072, "learning_rate": 5.717498243148279e-06, "loss": 0.0591, "step": 1358 }, { "epoch": 0.955024595924104, "grad_norm": 0.4029836058616638, "learning_rate": 5.7217146872803935e-06, "loss": 0.0664, "step": 1359 }, { "epoch": 0.9557273366127899, "grad_norm": 0.3920617997646332, "learning_rate": 5.7259311314125085e-06, "loss": 0.0522, "step": 1360 }, { "epoch": 0.9564300773014758, "grad_norm": 0.48270320892333984, "learning_rate": 5.730147575544624e-06, "loss": 0.0798, "step": 1361 }, { "epoch": 0.9571328179901616, "grad_norm": 0.43584761023521423, "learning_rate": 5.734364019676739e-06, "loss": 0.0533, "step": 1362 }, { "epoch": 0.9578355586788475, "grad_norm": 0.6245052814483643, "learning_rate": 5.738580463808854e-06, "loss": 0.0784, "step": 1363 }, { "epoch": 0.9585382993675334, "grad_norm": 0.4683986306190491, "learning_rate": 5.74279690794097e-06, "loss": 0.0639, "step": 1364 }, { "epoch": 0.9592410400562192, "grad_norm": 0.5538116693496704, "learning_rate": 5.747013352073086e-06, "loss": 0.0839, "step": 1365 }, { "epoch": 0.9599437807449052, "grad_norm": 0.47376054525375366, "learning_rate": 5.751229796205201e-06, "loss": 0.0731, "step": 1366 }, { "epoch": 0.960646521433591, "grad_norm": 0.5152063965797424, "learning_rate": 5.755446240337316e-06, "loss": 0.0602, "step": 1367 }, { "epoch": 0.9613492621222769, "grad_norm": 0.5633213520050049, "learning_rate": 5.7596626844694314e-06, "loss": 0.1062, "step": 1368 }, { "epoch": 0.9620520028109627, "grad_norm": 0.632211446762085, "learning_rate": 5.763879128601546e-06, "loss": 0.0971, "step": 1369 }, { "epoch": 0.9627547434996486, "grad_norm": 0.7205144762992859, "learning_rate": 5.768095572733661e-06, "loss": 0.1222, "step": 1370 }, { "epoch": 0.9634574841883345, "grad_norm": 0.6457942128181458, "learning_rate": 5.772312016865777e-06, "loss": 0.164, "step": 1371 }, { "epoch": 0.9641602248770204, "grad_norm": 1.3624376058578491, "learning_rate": 5.776528460997892e-06, "loss": 0.291, "step": 1372 }, { "epoch": 0.9648629655657063, "grad_norm": 1.9237531423568726, "learning_rate": 5.780744905130007e-06, "loss": 0.3211, "step": 1373 }, { "epoch": 0.9655657062543921, "grad_norm": 1.8669722080230713, "learning_rate": 5.784961349262122e-06, "loss": 0.4046, "step": 1374 }, { "epoch": 0.966268446943078, "grad_norm": 2.69187068939209, "learning_rate": 5.789177793394238e-06, "loss": 0.5507, "step": 1375 }, { "epoch": 0.9669711876317639, "grad_norm": 0.6364705562591553, "learning_rate": 5.793394237526353e-06, "loss": 0.1577, "step": 1376 }, { "epoch": 0.9676739283204497, "grad_norm": 0.5007463097572327, "learning_rate": 5.797610681658468e-06, "loss": 0.0774, "step": 1377 }, { "epoch": 0.9683766690091357, "grad_norm": 0.7425020933151245, "learning_rate": 5.801827125790583e-06, "loss": 0.0993, "step": 1378 }, { "epoch": 0.9690794096978215, "grad_norm": 0.6189239621162415, "learning_rate": 5.806043569922699e-06, "loss": 0.0724, "step": 1379 }, { "epoch": 0.9697821503865074, "grad_norm": 0.49729299545288086, "learning_rate": 5.810260014054814e-06, "loss": 0.0623, "step": 1380 }, { "epoch": 0.9704848910751932, "grad_norm": 0.6362632513046265, "learning_rate": 5.814476458186929e-06, "loss": 0.0594, "step": 1381 }, { "epoch": 0.9711876317638791, "grad_norm": 0.5549564957618713, "learning_rate": 5.818692902319045e-06, "loss": 0.0541, "step": 1382 }, { "epoch": 0.971890372452565, "grad_norm": 0.37204471230506897, "learning_rate": 5.82290934645116e-06, "loss": 0.0605, "step": 1383 }, { "epoch": 0.9725931131412509, "grad_norm": 0.5011727213859558, "learning_rate": 5.827125790583275e-06, "loss": 0.0673, "step": 1384 }, { "epoch": 0.9732958538299368, "grad_norm": 0.5411353707313538, "learning_rate": 5.8313422347153905e-06, "loss": 0.0571, "step": 1385 }, { "epoch": 0.9739985945186226, "grad_norm": 0.8977087736129761, "learning_rate": 5.8355586788475055e-06, "loss": 0.0736, "step": 1386 }, { "epoch": 0.9747013352073085, "grad_norm": 0.40996649861335754, "learning_rate": 5.8397751229796204e-06, "loss": 0.043, "step": 1387 }, { "epoch": 0.9754040758959944, "grad_norm": 0.5832712650299072, "learning_rate": 5.843991567111736e-06, "loss": 0.0615, "step": 1388 }, { "epoch": 0.9761068165846802, "grad_norm": 0.6379785537719727, "learning_rate": 5.848208011243851e-06, "loss": 0.076, "step": 1389 }, { "epoch": 0.9768095572733662, "grad_norm": 0.4709866940975189, "learning_rate": 5.852424455375966e-06, "loss": 0.0574, "step": 1390 }, { "epoch": 0.977512297962052, "grad_norm": 1.9567046165466309, "learning_rate": 5.856640899508081e-06, "loss": 0.1226, "step": 1391 }, { "epoch": 0.9782150386507379, "grad_norm": 0.46722376346588135, "learning_rate": 5.860857343640197e-06, "loss": 0.0665, "step": 1392 }, { "epoch": 0.9789177793394237, "grad_norm": 0.7003970742225647, "learning_rate": 5.865073787772312e-06, "loss": 0.1063, "step": 1393 }, { "epoch": 0.9796205200281096, "grad_norm": 1.24379301071167, "learning_rate": 5.869290231904428e-06, "loss": 0.0969, "step": 1394 }, { "epoch": 0.9803232607167955, "grad_norm": 0.9607970118522644, "learning_rate": 5.873506676036543e-06, "loss": 0.1473, "step": 1395 }, { "epoch": 0.9810260014054814, "grad_norm": 0.7818043231964111, "learning_rate": 5.877723120168658e-06, "loss": 0.1817, "step": 1396 }, { "epoch": 0.9817287420941673, "grad_norm": 1.3393855094909668, "learning_rate": 5.881939564300773e-06, "loss": 0.2542, "step": 1397 }, { "epoch": 0.9824314827828531, "grad_norm": 1.364931344985962, "learning_rate": 5.886156008432888e-06, "loss": 0.3551, "step": 1398 }, { "epoch": 0.983134223471539, "grad_norm": 1.7364929914474487, "learning_rate": 5.890372452565004e-06, "loss": 0.3979, "step": 1399 }, { "epoch": 0.9838369641602249, "grad_norm": 2.4351165294647217, "learning_rate": 5.894588896697119e-06, "loss": 0.4753, "step": 1400 }, { "epoch": 0.9845397048489107, "grad_norm": 0.6694924235343933, "learning_rate": 5.898805340829234e-06, "loss": 0.1506, "step": 1401 }, { "epoch": 0.9852424455375967, "grad_norm": 0.876908004283905, "learning_rate": 5.90302178496135e-06, "loss": 0.111, "step": 1402 }, { "epoch": 0.9859451862262825, "grad_norm": 0.4496653974056244, "learning_rate": 5.907238229093465e-06, "loss": 0.0684, "step": 1403 }, { "epoch": 0.9866479269149684, "grad_norm": 0.531929075717926, "learning_rate": 5.9114546732255796e-06, "loss": 0.087, "step": 1404 }, { "epoch": 0.9873506676036542, "grad_norm": 0.9465158581733704, "learning_rate": 5.915671117357695e-06, "loss": 0.0507, "step": 1405 }, { "epoch": 0.9880534082923401, "grad_norm": 0.3567866086959839, "learning_rate": 5.91988756148981e-06, "loss": 0.0619, "step": 1406 }, { "epoch": 0.988756148981026, "grad_norm": 0.4727637767791748, "learning_rate": 5.924104005621925e-06, "loss": 0.0775, "step": 1407 }, { "epoch": 0.9894588896697118, "grad_norm": 0.44192877411842346, "learning_rate": 5.92832044975404e-06, "loss": 0.0641, "step": 1408 }, { "epoch": 0.9901616303583978, "grad_norm": 0.5233160853385925, "learning_rate": 5.932536893886157e-06, "loss": 0.054, "step": 1409 }, { "epoch": 0.9908643710470836, "grad_norm": 0.46769723296165466, "learning_rate": 5.936753338018272e-06, "loss": 0.0708, "step": 1410 }, { "epoch": 0.9915671117357695, "grad_norm": 0.431522399187088, "learning_rate": 5.940969782150387e-06, "loss": 0.085, "step": 1411 }, { "epoch": 0.9922698524244554, "grad_norm": 0.4207701086997986, "learning_rate": 5.9451862262825025e-06, "loss": 0.0636, "step": 1412 }, { "epoch": 0.9929725931131412, "grad_norm": 0.6088957786560059, "learning_rate": 5.9494026704146174e-06, "loss": 0.0553, "step": 1413 }, { "epoch": 0.9936753338018272, "grad_norm": 0.4999117851257324, "learning_rate": 5.953619114546732e-06, "loss": 0.0796, "step": 1414 }, { "epoch": 0.994378074490513, "grad_norm": 0.6087541580200195, "learning_rate": 5.957835558678847e-06, "loss": 0.0904, "step": 1415 }, { "epoch": 0.9950808151791989, "grad_norm": 1.0209853649139404, "learning_rate": 5.962052002810963e-06, "loss": 0.0842, "step": 1416 }, { "epoch": 0.9957835558678847, "grad_norm": 0.7810251712799072, "learning_rate": 5.966268446943078e-06, "loss": 0.1313, "step": 1417 }, { "epoch": 0.9964862965565706, "grad_norm": 3.278475046157837, "learning_rate": 5.970484891075193e-06, "loss": 0.1277, "step": 1418 }, { "epoch": 0.9971890372452565, "grad_norm": 1.1090415716171265, "learning_rate": 5.974701335207309e-06, "loss": 0.1844, "step": 1419 }, { "epoch": 0.9978917779339423, "grad_norm": 1.5161104202270508, "learning_rate": 5.978917779339424e-06, "loss": 0.2736, "step": 1420 }, { "epoch": 0.9985945186226283, "grad_norm": 3.19809627532959, "learning_rate": 5.983134223471539e-06, "loss": 0.3722, "step": 1421 }, { "epoch": 0.9992972593113141, "grad_norm": 2.7327051162719727, "learning_rate": 5.9873506676036545e-06, "loss": 0.4873, "step": 1422 }, { "epoch": 1.0, "grad_norm": 3.2787821292877197, "learning_rate": 5.991567111735769e-06, "loss": 0.3759, "step": 1423 }, { "epoch": 1.000702740688686, "grad_norm": 0.45697370171546936, "learning_rate": 5.995783555867885e-06, "loss": 0.1394, "step": 1424 }, { "epoch": 1.0014054813773718, "grad_norm": 0.36343511939048767, "learning_rate": 6e-06, "loss": 0.0778, "step": 1425 }, { "epoch": 1.0021082220660575, "grad_norm": 0.3707204759120941, "learning_rate": 6.004216444132116e-06, "loss": 0.0742, "step": 1426 }, { "epoch": 1.0028109627547435, "grad_norm": 0.3747798800468445, "learning_rate": 6.008432888264231e-06, "loss": 0.0759, "step": 1427 }, { "epoch": 1.0035137034434294, "grad_norm": 0.3834972679615021, "learning_rate": 6.012649332396346e-06, "loss": 0.0545, "step": 1428 }, { "epoch": 1.0042164441321153, "grad_norm": 0.33959025144577026, "learning_rate": 6.016865776528462e-06, "loss": 0.0475, "step": 1429 }, { "epoch": 1.0049191848208012, "grad_norm": 0.6041131019592285, "learning_rate": 6.0210822206605766e-06, "loss": 0.0712, "step": 1430 }, { "epoch": 1.005621925509487, "grad_norm": 0.3952963650226593, "learning_rate": 6.0252986647926915e-06, "loss": 0.0489, "step": 1431 }, { "epoch": 1.0063246661981728, "grad_norm": 0.3924621641635895, "learning_rate": 6.0295151089248065e-06, "loss": 0.0573, "step": 1432 }, { "epoch": 1.0070274068868588, "grad_norm": 0.49256590008735657, "learning_rate": 6.033731553056922e-06, "loss": 0.052, "step": 1433 }, { "epoch": 1.0077301475755447, "grad_norm": 1.1340930461883545, "learning_rate": 6.037947997189037e-06, "loss": 0.0894, "step": 1434 }, { "epoch": 1.0084328882642306, "grad_norm": 0.46295884251594543, "learning_rate": 6.042164441321152e-06, "loss": 0.0436, "step": 1435 }, { "epoch": 1.0091356289529163, "grad_norm": 0.5553387999534607, "learning_rate": 6.046380885453268e-06, "loss": 0.0843, "step": 1436 }, { "epoch": 1.0098383696416022, "grad_norm": 0.4986462891101837, "learning_rate": 6.050597329585383e-06, "loss": 0.0704, "step": 1437 }, { "epoch": 1.0105411103302882, "grad_norm": 1.752012014389038, "learning_rate": 6.054813773717498e-06, "loss": 0.0843, "step": 1438 }, { "epoch": 1.011243851018974, "grad_norm": 0.548039436340332, "learning_rate": 6.0590302178496144e-06, "loss": 0.0883, "step": 1439 }, { "epoch": 1.0119465917076598, "grad_norm": 0.6629095673561096, "learning_rate": 6.063246661981729e-06, "loss": 0.0736, "step": 1440 }, { "epoch": 1.0126493323963457, "grad_norm": 0.8315633535385132, "learning_rate": 6.067463106113844e-06, "loss": 0.0912, "step": 1441 }, { "epoch": 1.0133520730850316, "grad_norm": 0.7529453635215759, "learning_rate": 6.071679550245959e-06, "loss": 0.1075, "step": 1442 }, { "epoch": 1.0140548137737175, "grad_norm": 0.8003742694854736, "learning_rate": 6.075895994378075e-06, "loss": 0.1823, "step": 1443 }, { "epoch": 1.0147575544624035, "grad_norm": 0.8003247380256653, "learning_rate": 6.08011243851019e-06, "loss": 0.1606, "step": 1444 }, { "epoch": 1.0154602951510892, "grad_norm": 1.039345383644104, "learning_rate": 6.084328882642305e-06, "loss": 0.2533, "step": 1445 }, { "epoch": 1.016163035839775, "grad_norm": 1.7495557069778442, "learning_rate": 6.088545326774421e-06, "loss": 0.2853, "step": 1446 }, { "epoch": 1.016865776528461, "grad_norm": 1.8877333402633667, "learning_rate": 6.092761770906536e-06, "loss": 0.3501, "step": 1447 }, { "epoch": 1.017568517217147, "grad_norm": 2.663036346435547, "learning_rate": 6.096978215038651e-06, "loss": 0.5071, "step": 1448 }, { "epoch": 1.0182712579058328, "grad_norm": 0.4770645499229431, "learning_rate": 6.1011946591707656e-06, "loss": 0.126, "step": 1449 }, { "epoch": 1.0189739985945185, "grad_norm": 0.5533647537231445, "learning_rate": 6.105411103302881e-06, "loss": 0.0836, "step": 1450 }, { "epoch": 1.0196767392832045, "grad_norm": 0.5535093545913696, "learning_rate": 6.109627547434996e-06, "loss": 0.0841, "step": 1451 }, { "epoch": 1.0203794799718904, "grad_norm": 0.5357879996299744, "learning_rate": 6.113843991567111e-06, "loss": 0.0532, "step": 1452 }, { "epoch": 1.0210822206605763, "grad_norm": 0.3972303867340088, "learning_rate": 6.118060435699227e-06, "loss": 0.0598, "step": 1453 }, { "epoch": 1.0217849613492622, "grad_norm": 0.3923996388912201, "learning_rate": 6.122276879831343e-06, "loss": 0.0478, "step": 1454 }, { "epoch": 1.022487702037948, "grad_norm": 0.5883025527000427, "learning_rate": 6.126493323963458e-06, "loss": 0.0584, "step": 1455 }, { "epoch": 1.0231904427266338, "grad_norm": 0.42282333970069885, "learning_rate": 6.1307097680955736e-06, "loss": 0.06, "step": 1456 }, { "epoch": 1.0238931834153198, "grad_norm": 0.37129902839660645, "learning_rate": 6.1349262122276885e-06, "loss": 0.0556, "step": 1457 }, { "epoch": 1.0245959241040057, "grad_norm": 0.41598185896873474, "learning_rate": 6.1391426563598035e-06, "loss": 0.0528, "step": 1458 }, { "epoch": 1.0252986647926916, "grad_norm": 0.47424229979515076, "learning_rate": 6.143359100491918e-06, "loss": 0.0602, "step": 1459 }, { "epoch": 1.0260014054813773, "grad_norm": 0.37612298130989075, "learning_rate": 6.147575544624034e-06, "loss": 0.0562, "step": 1460 }, { "epoch": 1.0267041461700632, "grad_norm": 0.6137157678604126, "learning_rate": 6.151791988756149e-06, "loss": 0.0814, "step": 1461 }, { "epoch": 1.0274068868587491, "grad_norm": 0.39535894989967346, "learning_rate": 6.156008432888264e-06, "loss": 0.0539, "step": 1462 }, { "epoch": 1.028109627547435, "grad_norm": 0.5793585777282715, "learning_rate": 6.16022487702038e-06, "loss": 0.1161, "step": 1463 }, { "epoch": 1.0288123682361208, "grad_norm": 0.7385834455490112, "learning_rate": 6.164441321152495e-06, "loss": 0.1182, "step": 1464 }, { "epoch": 1.0295151089248067, "grad_norm": 0.3969365954399109, "learning_rate": 6.16865776528461e-06, "loss": 0.0666, "step": 1465 }, { "epoch": 1.0302178496134926, "grad_norm": 0.6131270527839661, "learning_rate": 6.172874209416725e-06, "loss": 0.0969, "step": 1466 }, { "epoch": 1.0309205903021785, "grad_norm": 0.665397584438324, "learning_rate": 6.1770906535488405e-06, "loss": 0.0903, "step": 1467 }, { "epoch": 1.0316233309908645, "grad_norm": 0.8435081243515015, "learning_rate": 6.1813070976809554e-06, "loss": 0.1465, "step": 1468 }, { "epoch": 1.0323260716795502, "grad_norm": 0.8177435398101807, "learning_rate": 6.185523541813071e-06, "loss": 0.1668, "step": 1469 }, { "epoch": 1.033028812368236, "grad_norm": 0.9415327310562134, "learning_rate": 6.189739985945187e-06, "loss": 0.2325, "step": 1470 }, { "epoch": 1.033731553056922, "grad_norm": 3.0599303245544434, "learning_rate": 6.193956430077302e-06, "loss": 0.2915, "step": 1471 }, { "epoch": 1.034434293745608, "grad_norm": 2.0106940269470215, "learning_rate": 6.198172874209417e-06, "loss": 0.3977, "step": 1472 }, { "epoch": 1.0351370344342938, "grad_norm": 4.902921676635742, "learning_rate": 6.202389318341533e-06, "loss": 0.5621, "step": 1473 }, { "epoch": 1.0358397751229795, "grad_norm": 0.5214247107505798, "learning_rate": 6.206605762473648e-06, "loss": 0.139, "step": 1474 }, { "epoch": 1.0365425158116655, "grad_norm": 0.5332420468330383, "learning_rate": 6.210822206605763e-06, "loss": 0.0808, "step": 1475 }, { "epoch": 1.0372452565003514, "grad_norm": 0.45051419734954834, "learning_rate": 6.2150386507378775e-06, "loss": 0.062, "step": 1476 }, { "epoch": 1.0379479971890373, "grad_norm": 0.44883519411087036, "learning_rate": 6.219255094869993e-06, "loss": 0.0577, "step": 1477 }, { "epoch": 1.0386507378777232, "grad_norm": 0.5620810985565186, "learning_rate": 6.223471539002108e-06, "loss": 0.0869, "step": 1478 }, { "epoch": 1.039353478566409, "grad_norm": 0.3897135853767395, "learning_rate": 6.227687983134223e-06, "loss": 0.0603, "step": 1479 }, { "epoch": 1.0400562192550948, "grad_norm": 0.7705488204956055, "learning_rate": 6.231904427266339e-06, "loss": 0.0376, "step": 1480 }, { "epoch": 1.0407589599437808, "grad_norm": 0.47137701511383057, "learning_rate": 6.236120871398454e-06, "loss": 0.0958, "step": 1481 }, { "epoch": 1.0414617006324667, "grad_norm": 0.43648502230644226, "learning_rate": 6.240337315530569e-06, "loss": 0.0736, "step": 1482 }, { "epoch": 1.0421644413211526, "grad_norm": 0.4624142050743103, "learning_rate": 6.244553759662684e-06, "loss": 0.0609, "step": 1483 }, { "epoch": 1.0428671820098383, "grad_norm": 0.5091442465782166, "learning_rate": 6.2487702037948005e-06, "loss": 0.0673, "step": 1484 }, { "epoch": 1.0435699226985242, "grad_norm": 0.38796114921569824, "learning_rate": 6.252986647926915e-06, "loss": 0.0635, "step": 1485 }, { "epoch": 1.0442726633872101, "grad_norm": 0.6586403846740723, "learning_rate": 6.25720309205903e-06, "loss": 0.0916, "step": 1486 }, { "epoch": 1.044975404075896, "grad_norm": 0.45383623242378235, "learning_rate": 6.261419536191146e-06, "loss": 0.0537, "step": 1487 }, { "epoch": 1.0456781447645818, "grad_norm": 0.40037187933921814, "learning_rate": 6.265635980323261e-06, "loss": 0.0677, "step": 1488 }, { "epoch": 1.0463808854532677, "grad_norm": 0.6651952862739563, "learning_rate": 6.269852424455376e-06, "loss": 0.0928, "step": 1489 }, { "epoch": 1.0470836261419536, "grad_norm": 0.5824512243270874, "learning_rate": 6.274068868587492e-06, "loss": 0.0658, "step": 1490 }, { "epoch": 1.0477863668306395, "grad_norm": 1.071931004524231, "learning_rate": 6.278285312719607e-06, "loss": 0.0916, "step": 1491 }, { "epoch": 1.0484891075193254, "grad_norm": 0.7158479690551758, "learning_rate": 6.282501756851722e-06, "loss": 0.0776, "step": 1492 }, { "epoch": 1.0491918482080111, "grad_norm": 0.7271090745925903, "learning_rate": 6.286718200983837e-06, "loss": 0.1295, "step": 1493 }, { "epoch": 1.049894588896697, "grad_norm": 0.7537633180618286, "learning_rate": 6.2909346451159524e-06, "loss": 0.1744, "step": 1494 }, { "epoch": 1.050597329585383, "grad_norm": 1.5890772342681885, "learning_rate": 6.295151089248067e-06, "loss": 0.2589, "step": 1495 }, { "epoch": 1.051300070274069, "grad_norm": 1.1826139688491821, "learning_rate": 6.299367533380182e-06, "loss": 0.3357, "step": 1496 }, { "epoch": 1.0520028109627548, "grad_norm": 2.1443357467651367, "learning_rate": 6.303583977512298e-06, "loss": 0.4158, "step": 1497 }, { "epoch": 1.0527055516514405, "grad_norm": 2.361624240875244, "learning_rate": 6.307800421644414e-06, "loss": 0.5599, "step": 1498 }, { "epoch": 1.0534082923401265, "grad_norm": 0.5916370749473572, "learning_rate": 6.312016865776529e-06, "loss": 0.151, "step": 1499 }, { "epoch": 1.0541110330288124, "grad_norm": 0.37359005212783813, "learning_rate": 6.316233309908644e-06, "loss": 0.0602, "step": 1500 }, { "epoch": 1.0548137737174983, "grad_norm": 0.470745712518692, "learning_rate": 6.32044975404076e-06, "loss": 0.0667, "step": 1501 }, { "epoch": 1.0555165144061842, "grad_norm": 0.43565037846565247, "learning_rate": 6.3246661981728745e-06, "loss": 0.0485, "step": 1502 }, { "epoch": 1.05621925509487, "grad_norm": 0.6414610743522644, "learning_rate": 6.3288826423049895e-06, "loss": 0.0778, "step": 1503 }, { "epoch": 1.0569219957835558, "grad_norm": 0.3401055335998535, "learning_rate": 6.333099086437105e-06, "loss": 0.0475, "step": 1504 }, { "epoch": 1.0576247364722418, "grad_norm": 0.37449365854263306, "learning_rate": 6.33731553056922e-06, "loss": 0.0517, "step": 1505 }, { "epoch": 1.0583274771609277, "grad_norm": 0.4353812336921692, "learning_rate": 6.341531974701335e-06, "loss": 0.0559, "step": 1506 }, { "epoch": 1.0590302178496136, "grad_norm": 0.5143307447433472, "learning_rate": 6.345748418833451e-06, "loss": 0.0487, "step": 1507 }, { "epoch": 1.0597329585382993, "grad_norm": 0.442687451839447, "learning_rate": 6.349964862965566e-06, "loss": 0.0529, "step": 1508 }, { "epoch": 1.0604356992269852, "grad_norm": 0.4616980254650116, "learning_rate": 6.354181307097681e-06, "loss": 0.0818, "step": 1509 }, { "epoch": 1.0611384399156711, "grad_norm": 0.37197059392929077, "learning_rate": 6.358397751229796e-06, "loss": 0.0655, "step": 1510 }, { "epoch": 1.061841180604357, "grad_norm": 0.578627347946167, "learning_rate": 6.3626141953619116e-06, "loss": 0.0728, "step": 1511 }, { "epoch": 1.062543921293043, "grad_norm": 0.49307045340538025, "learning_rate": 6.3668306394940265e-06, "loss": 0.0608, "step": 1512 }, { "epoch": 1.0632466619817287, "grad_norm": 0.5238860845565796, "learning_rate": 6.371047083626142e-06, "loss": 0.0885, "step": 1513 }, { "epoch": 1.0639494026704146, "grad_norm": 0.6770070195198059, "learning_rate": 6.375263527758258e-06, "loss": 0.1025, "step": 1514 }, { "epoch": 1.0646521433591005, "grad_norm": 0.506943941116333, "learning_rate": 6.379479971890373e-06, "loss": 0.0609, "step": 1515 }, { "epoch": 1.0653548840477864, "grad_norm": 0.6383757591247559, "learning_rate": 6.383696416022488e-06, "loss": 0.1018, "step": 1516 }, { "epoch": 1.0660576247364721, "grad_norm": 0.6648460030555725, "learning_rate": 6.387912860154603e-06, "loss": 0.0878, "step": 1517 }, { "epoch": 1.066760365425158, "grad_norm": 0.9293584227561951, "learning_rate": 6.392129304286719e-06, "loss": 0.1555, "step": 1518 }, { "epoch": 1.067463106113844, "grad_norm": 1.499895453453064, "learning_rate": 6.396345748418834e-06, "loss": 0.2082, "step": 1519 }, { "epoch": 1.06816584680253, "grad_norm": 1.3249156475067139, "learning_rate": 6.400562192550949e-06, "loss": 0.2074, "step": 1520 }, { "epoch": 1.0688685874912158, "grad_norm": 1.2726099491119385, "learning_rate": 6.404778636683064e-06, "loss": 0.3084, "step": 1521 }, { "epoch": 1.0695713281799015, "grad_norm": 3.240777015686035, "learning_rate": 6.408995080815179e-06, "loss": 0.3717, "step": 1522 }, { "epoch": 1.0702740688685874, "grad_norm": 1.8976423740386963, "learning_rate": 6.413211524947294e-06, "loss": 0.4456, "step": 1523 }, { "epoch": 1.0709768095572734, "grad_norm": 0.5520387291908264, "learning_rate": 6.41742796907941e-06, "loss": 0.1413, "step": 1524 }, { "epoch": 1.0716795502459593, "grad_norm": 0.4153732359409332, "learning_rate": 6.421644413211525e-06, "loss": 0.0663, "step": 1525 }, { "epoch": 1.0723822909346452, "grad_norm": 0.4837837517261505, "learning_rate": 6.42586085734364e-06, "loss": 0.0687, "step": 1526 }, { "epoch": 1.073085031623331, "grad_norm": 0.5458381772041321, "learning_rate": 6.430077301475755e-06, "loss": 0.0677, "step": 1527 }, { "epoch": 1.0737877723120168, "grad_norm": 0.6029407382011414, "learning_rate": 6.4342937456078715e-06, "loss": 0.078, "step": 1528 }, { "epoch": 1.0744905130007028, "grad_norm": 0.6689379811286926, "learning_rate": 6.4385101897399865e-06, "loss": 0.0468, "step": 1529 }, { "epoch": 1.0751932536893887, "grad_norm": 0.47726351022720337, "learning_rate": 6.442726633872101e-06, "loss": 0.0665, "step": 1530 }, { "epoch": 1.0758959943780746, "grad_norm": 0.3524337410926819, "learning_rate": 6.446943078004217e-06, "loss": 0.0556, "step": 1531 }, { "epoch": 1.0765987350667603, "grad_norm": 0.3781997263431549, "learning_rate": 6.451159522136332e-06, "loss": 0.0619, "step": 1532 }, { "epoch": 1.0773014757554462, "grad_norm": 0.3912004828453064, "learning_rate": 6.455375966268447e-06, "loss": 0.0632, "step": 1533 }, { "epoch": 1.0780042164441321, "grad_norm": 0.5569971203804016, "learning_rate": 6.459592410400562e-06, "loss": 0.084, "step": 1534 }, { "epoch": 1.078706957132818, "grad_norm": 0.2843516767024994, "learning_rate": 6.463808854532678e-06, "loss": 0.053, "step": 1535 }, { "epoch": 1.0794096978215038, "grad_norm": 0.4888468086719513, "learning_rate": 6.468025298664793e-06, "loss": 0.0935, "step": 1536 }, { "epoch": 1.0801124385101897, "grad_norm": 0.42813506722450256, "learning_rate": 6.472241742796908e-06, "loss": 0.0532, "step": 1537 }, { "epoch": 1.0808151791988756, "grad_norm": 0.6030667424201965, "learning_rate": 6.4764581869290235e-06, "loss": 0.0781, "step": 1538 }, { "epoch": 1.0815179198875615, "grad_norm": 0.45042529702186584, "learning_rate": 6.4806746310611385e-06, "loss": 0.076, "step": 1539 }, { "epoch": 1.0822206605762474, "grad_norm": 0.6704282760620117, "learning_rate": 6.484891075193253e-06, "loss": 0.0891, "step": 1540 }, { "epoch": 1.0829234012649331, "grad_norm": 0.5336280465126038, "learning_rate": 6.489107519325369e-06, "loss": 0.1125, "step": 1541 }, { "epoch": 1.083626141953619, "grad_norm": 0.5512068867683411, "learning_rate": 6.493323963457484e-06, "loss": 0.0755, "step": 1542 }, { "epoch": 1.084328882642305, "grad_norm": 0.7172701954841614, "learning_rate": 6.4975404075896e-06, "loss": 0.133, "step": 1543 }, { "epoch": 1.085031623330991, "grad_norm": 1.0922446250915527, "learning_rate": 6.501756851721715e-06, "loss": 0.1534, "step": 1544 }, { "epoch": 1.0857343640196768, "grad_norm": 0.8649986982345581, "learning_rate": 6.505973295853831e-06, "loss": 0.208, "step": 1545 }, { "epoch": 1.0864371047083625, "grad_norm": 1.8669407367706299, "learning_rate": 6.510189739985946e-06, "loss": 0.3113, "step": 1546 }, { "epoch": 1.0871398453970484, "grad_norm": 1.7221201658248901, "learning_rate": 6.5144061841180605e-06, "loss": 0.3842, "step": 1547 }, { "epoch": 1.0878425860857344, "grad_norm": 3.3071773052215576, "learning_rate": 6.518622628250176e-06, "loss": 0.5342, "step": 1548 }, { "epoch": 1.0885453267744203, "grad_norm": 1.1389280557632446, "learning_rate": 6.522839072382291e-06, "loss": 0.1832, "step": 1549 }, { "epoch": 1.0892480674631062, "grad_norm": 0.5553212761878967, "learning_rate": 6.527055516514406e-06, "loss": 0.0917, "step": 1550 }, { "epoch": 1.089950808151792, "grad_norm": 1.11082124710083, "learning_rate": 6.531271960646521e-06, "loss": 0.0678, "step": 1551 }, { "epoch": 1.0906535488404778, "grad_norm": 0.5675510168075562, "learning_rate": 6.535488404778637e-06, "loss": 0.0626, "step": 1552 }, { "epoch": 1.0913562895291637, "grad_norm": 0.3746297061443329, "learning_rate": 6.539704848910752e-06, "loss": 0.0576, "step": 1553 }, { "epoch": 1.0920590302178497, "grad_norm": 0.41739585995674133, "learning_rate": 6.543921293042867e-06, "loss": 0.049, "step": 1554 }, { "epoch": 1.0927617709065356, "grad_norm": 0.4867488741874695, "learning_rate": 6.548137737174983e-06, "loss": 0.0723, "step": 1555 }, { "epoch": 1.0934645115952213, "grad_norm": 0.31857338547706604, "learning_rate": 6.552354181307098e-06, "loss": 0.0337, "step": 1556 }, { "epoch": 1.0941672522839072, "grad_norm": 0.40387970209121704, "learning_rate": 6.5565706254392125e-06, "loss": 0.0619, "step": 1557 }, { "epoch": 1.0948699929725931, "grad_norm": 0.2992646098136902, "learning_rate": 6.560787069571328e-06, "loss": 0.0356, "step": 1558 }, { "epoch": 1.095572733661279, "grad_norm": 0.5410019159317017, "learning_rate": 6.565003513703444e-06, "loss": 0.0678, "step": 1559 }, { "epoch": 1.096275474349965, "grad_norm": 0.6178023815155029, "learning_rate": 6.569219957835559e-06, "loss": 0.0574, "step": 1560 }, { "epoch": 1.0969782150386507, "grad_norm": 0.45310619473457336, "learning_rate": 6.573436401967674e-06, "loss": 0.0556, "step": 1561 }, { "epoch": 1.0976809557273366, "grad_norm": 0.5689439177513123, "learning_rate": 6.57765284609979e-06, "loss": 0.077, "step": 1562 }, { "epoch": 1.0983836964160225, "grad_norm": 0.6811681389808655, "learning_rate": 6.581869290231905e-06, "loss": 0.0655, "step": 1563 }, { "epoch": 1.0990864371047084, "grad_norm": 0.48519185185432434, "learning_rate": 6.58608573436402e-06, "loss": 0.0729, "step": 1564 }, { "epoch": 1.0997891777933941, "grad_norm": 0.5935548543930054, "learning_rate": 6.5903021784961355e-06, "loss": 0.1044, "step": 1565 }, { "epoch": 1.10049191848208, "grad_norm": 0.5869494080543518, "learning_rate": 6.59451862262825e-06, "loss": 0.08, "step": 1566 }, { "epoch": 1.101194659170766, "grad_norm": 0.6179332137107849, "learning_rate": 6.598735066760365e-06, "loss": 0.0703, "step": 1567 }, { "epoch": 1.101897399859452, "grad_norm": 0.8535454273223877, "learning_rate": 6.60295151089248e-06, "loss": 0.1519, "step": 1568 }, { "epoch": 1.1026001405481378, "grad_norm": 0.8692838549613953, "learning_rate": 6.607167955024596e-06, "loss": 0.1513, "step": 1569 }, { "epoch": 1.1033028812368235, "grad_norm": 1.087809681892395, "learning_rate": 6.611384399156711e-06, "loss": 0.2451, "step": 1570 }, { "epoch": 1.1040056219255094, "grad_norm": 1.2239112854003906, "learning_rate": 6.615600843288826e-06, "loss": 0.2957, "step": 1571 }, { "epoch": 1.1047083626141954, "grad_norm": 1.4041328430175781, "learning_rate": 6.619817287420942e-06, "loss": 0.3982, "step": 1572 }, { "epoch": 1.1054111033028813, "grad_norm": 2.957044839859009, "learning_rate": 6.6240337315530575e-06, "loss": 0.5418, "step": 1573 }, { "epoch": 1.1061138439915672, "grad_norm": 0.8908143639564514, "learning_rate": 6.6282501756851725e-06, "loss": 0.1458, "step": 1574 }, { "epoch": 1.106816584680253, "grad_norm": 0.47511810064315796, "learning_rate": 6.6324666198172874e-06, "loss": 0.0795, "step": 1575 }, { "epoch": 1.1075193253689388, "grad_norm": 0.5243213176727295, "learning_rate": 6.636683063949403e-06, "loss": 0.082, "step": 1576 }, { "epoch": 1.1082220660576247, "grad_norm": 0.3880898058414459, "learning_rate": 6.640899508081518e-06, "loss": 0.0648, "step": 1577 }, { "epoch": 1.1089248067463107, "grad_norm": 0.34104999899864197, "learning_rate": 6.645115952213633e-06, "loss": 0.0481, "step": 1578 }, { "epoch": 1.1096275474349966, "grad_norm": 0.6646779179573059, "learning_rate": 6.649332396345749e-06, "loss": 0.0558, "step": 1579 }, { "epoch": 1.1103302881236823, "grad_norm": 0.34305599331855774, "learning_rate": 6.653548840477864e-06, "loss": 0.0613, "step": 1580 }, { "epoch": 1.1110330288123682, "grad_norm": 0.3385035991668701, "learning_rate": 6.657765284609979e-06, "loss": 0.0418, "step": 1581 }, { "epoch": 1.1117357695010541, "grad_norm": 0.663379430770874, "learning_rate": 6.661981728742095e-06, "loss": 0.0644, "step": 1582 }, { "epoch": 1.11243851018974, "grad_norm": 0.4698798358440399, "learning_rate": 6.6661981728742095e-06, "loss": 0.0471, "step": 1583 }, { "epoch": 1.1131412508784257, "grad_norm": 0.4718151092529297, "learning_rate": 6.6704146170063245e-06, "loss": 0.0924, "step": 1584 }, { "epoch": 1.1138439915671117, "grad_norm": 0.3425571024417877, "learning_rate": 6.674631061138439e-06, "loss": 0.0482, "step": 1585 }, { "epoch": 1.1145467322557976, "grad_norm": 0.5875996351242065, "learning_rate": 6.678847505270555e-06, "loss": 0.0781, "step": 1586 }, { "epoch": 1.1152494729444835, "grad_norm": 0.4455372989177704, "learning_rate": 6.68306394940267e-06, "loss": 0.0569, "step": 1587 }, { "epoch": 1.1159522136331694, "grad_norm": 0.5223621726036072, "learning_rate": 6.687280393534786e-06, "loss": 0.0649, "step": 1588 }, { "epoch": 1.1166549543218554, "grad_norm": 0.5068934559822083, "learning_rate": 6.691496837666902e-06, "loss": 0.0934, "step": 1589 }, { "epoch": 1.117357695010541, "grad_norm": 0.9669073224067688, "learning_rate": 6.695713281799017e-06, "loss": 0.0738, "step": 1590 }, { "epoch": 1.118060435699227, "grad_norm": 1.1322118043899536, "learning_rate": 6.699929725931132e-06, "loss": 0.0771, "step": 1591 }, { "epoch": 1.118763176387913, "grad_norm": 0.6330634951591492, "learning_rate": 6.7041461700632466e-06, "loss": 0.0821, "step": 1592 }, { "epoch": 1.1194659170765988, "grad_norm": 0.7128387689590454, "learning_rate": 6.708362614195362e-06, "loss": 0.1272, "step": 1593 }, { "epoch": 1.1201686577652845, "grad_norm": 0.9848220944404602, "learning_rate": 6.712579058327477e-06, "loss": 0.1604, "step": 1594 }, { "epoch": 1.1208713984539704, "grad_norm": 1.2414302825927734, "learning_rate": 6.716795502459592e-06, "loss": 0.2038, "step": 1595 }, { "epoch": 1.1215741391426564, "grad_norm": 1.9506969451904297, "learning_rate": 6.721011946591708e-06, "loss": 0.2803, "step": 1596 }, { "epoch": 1.1222768798313423, "grad_norm": 1.4168736934661865, "learning_rate": 6.725228390723823e-06, "loss": 0.4001, "step": 1597 }, { "epoch": 1.1229796205200282, "grad_norm": 2.5162105560302734, "learning_rate": 6.729444834855938e-06, "loss": 0.4421, "step": 1598 }, { "epoch": 1.123682361208714, "grad_norm": 0.4300132989883423, "learning_rate": 6.733661278988054e-06, "loss": 0.12, "step": 1599 }, { "epoch": 1.1243851018973998, "grad_norm": 0.44196802377700806, "learning_rate": 6.737877723120169e-06, "loss": 0.0702, "step": 1600 }, { "epoch": 1.1250878425860857, "grad_norm": 0.48803824186325073, "learning_rate": 6.742094167252284e-06, "loss": 0.0505, "step": 1601 }, { "epoch": 1.1257905832747717, "grad_norm": 0.4035335183143616, "learning_rate": 6.7463106113843985e-06, "loss": 0.0689, "step": 1602 }, { "epoch": 1.1264933239634574, "grad_norm": 1.0948880910873413, "learning_rate": 6.750527055516515e-06, "loss": 0.0532, "step": 1603 }, { "epoch": 1.1271960646521433, "grad_norm": 0.35979947447776794, "learning_rate": 6.75474349964863e-06, "loss": 0.0533, "step": 1604 }, { "epoch": 1.1278988053408292, "grad_norm": 0.4000673294067383, "learning_rate": 6.758959943780745e-06, "loss": 0.0514, "step": 1605 }, { "epoch": 1.1286015460295151, "grad_norm": 0.4128091335296631, "learning_rate": 6.763176387912861e-06, "loss": 0.0544, "step": 1606 }, { "epoch": 1.129304286718201, "grad_norm": 0.40461304783821106, "learning_rate": 6.767392832044976e-06, "loss": 0.0664, "step": 1607 }, { "epoch": 1.130007027406887, "grad_norm": 0.5278897285461426, "learning_rate": 6.771609276177091e-06, "loss": 0.0426, "step": 1608 }, { "epoch": 1.1307097680955727, "grad_norm": 0.4207574725151062, "learning_rate": 6.7758257203092065e-06, "loss": 0.0597, "step": 1609 }, { "epoch": 1.1314125087842586, "grad_norm": 0.7527398467063904, "learning_rate": 6.7800421644413215e-06, "loss": 0.0491, "step": 1610 }, { "epoch": 1.1321152494729445, "grad_norm": 0.5862365365028381, "learning_rate": 6.784258608573436e-06, "loss": 0.0704, "step": 1611 }, { "epoch": 1.1328179901616304, "grad_norm": 0.5362736582756042, "learning_rate": 6.788475052705551e-06, "loss": 0.0678, "step": 1612 }, { "epoch": 1.1335207308503161, "grad_norm": 0.47923362255096436, "learning_rate": 6.792691496837667e-06, "loss": 0.0712, "step": 1613 }, { "epoch": 1.134223471539002, "grad_norm": 0.6238951683044434, "learning_rate": 6.796907940969782e-06, "loss": 0.0681, "step": 1614 }, { "epoch": 1.134926212227688, "grad_norm": 0.5374364256858826, "learning_rate": 6.801124385101897e-06, "loss": 0.0724, "step": 1615 }, { "epoch": 1.135628952916374, "grad_norm": 0.5410443544387817, "learning_rate": 6.805340829234013e-06, "loss": 0.0888, "step": 1616 }, { "epoch": 1.1363316936050598, "grad_norm": 0.8166239857673645, "learning_rate": 6.809557273366128e-06, "loss": 0.0876, "step": 1617 }, { "epoch": 1.1370344342937457, "grad_norm": 0.6477342247962952, "learning_rate": 6.8137737174982436e-06, "loss": 0.1295, "step": 1618 }, { "epoch": 1.1377371749824314, "grad_norm": 0.7941052317619324, "learning_rate": 6.8179901616303585e-06, "loss": 0.1402, "step": 1619 }, { "epoch": 1.1384399156711174, "grad_norm": 1.4420382976531982, "learning_rate": 6.822206605762474e-06, "loss": 0.2508, "step": 1620 }, { "epoch": 1.1391426563598033, "grad_norm": 1.2545603513717651, "learning_rate": 6.826423049894589e-06, "loss": 0.2594, "step": 1621 }, { "epoch": 1.1398453970484892, "grad_norm": 1.7919037342071533, "learning_rate": 6.830639494026704e-06, "loss": 0.3762, "step": 1622 }, { "epoch": 1.140548137737175, "grad_norm": 2.0993337631225586, "learning_rate": 6.83485593815882e-06, "loss": 0.5138, "step": 1623 }, { "epoch": 1.1412508784258608, "grad_norm": 0.78038090467453, "learning_rate": 6.839072382290935e-06, "loss": 0.1953, "step": 1624 }, { "epoch": 1.1419536191145467, "grad_norm": 0.5084997415542603, "learning_rate": 6.84328882642305e-06, "loss": 0.0909, "step": 1625 }, { "epoch": 1.1426563598032327, "grad_norm": 0.5086880922317505, "learning_rate": 6.847505270555166e-06, "loss": 0.0917, "step": 1626 }, { "epoch": 1.1433591004919186, "grad_norm": 0.3951646685600281, "learning_rate": 6.851721714687281e-06, "loss": 0.0479, "step": 1627 }, { "epoch": 1.1440618411806043, "grad_norm": 0.3559173047542572, "learning_rate": 6.8559381588193955e-06, "loss": 0.0501, "step": 1628 }, { "epoch": 1.1447645818692902, "grad_norm": 0.394727498292923, "learning_rate": 6.8601546029515105e-06, "loss": 0.0463, "step": 1629 }, { "epoch": 1.1454673225579761, "grad_norm": 0.4396524131298065, "learning_rate": 6.864371047083626e-06, "loss": 0.0469, "step": 1630 }, { "epoch": 1.146170063246662, "grad_norm": 2.944401264190674, "learning_rate": 6.868587491215741e-06, "loss": 0.0569, "step": 1631 }, { "epoch": 1.1468728039353477, "grad_norm": 0.3655034303665161, "learning_rate": 6.872803935347857e-06, "loss": 0.058, "step": 1632 }, { "epoch": 1.1475755446240337, "grad_norm": 0.34862661361694336, "learning_rate": 6.877020379479973e-06, "loss": 0.0459, "step": 1633 }, { "epoch": 1.1482782853127196, "grad_norm": 0.3887476623058319, "learning_rate": 6.881236823612088e-06, "loss": 0.0813, "step": 1634 }, { "epoch": 1.1489810260014055, "grad_norm": 0.4134727120399475, "learning_rate": 6.885453267744203e-06, "loss": 0.0535, "step": 1635 }, { "epoch": 1.1496837666900914, "grad_norm": 0.4894808828830719, "learning_rate": 6.889669711876318e-06, "loss": 0.0539, "step": 1636 }, { "epoch": 1.1503865073787773, "grad_norm": 0.411670446395874, "learning_rate": 6.893886156008433e-06, "loss": 0.0547, "step": 1637 }, { "epoch": 1.151089248067463, "grad_norm": 0.5032523274421692, "learning_rate": 6.898102600140548e-06, "loss": 0.0736, "step": 1638 }, { "epoch": 1.151791988756149, "grad_norm": 0.5117863416671753, "learning_rate": 6.902319044272663e-06, "loss": 0.0783, "step": 1639 }, { "epoch": 1.1524947294448349, "grad_norm": 0.5338552594184875, "learning_rate": 6.906535488404779e-06, "loss": 0.0569, "step": 1640 }, { "epoch": 1.1531974701335208, "grad_norm": 0.749910295009613, "learning_rate": 6.910751932536894e-06, "loss": 0.088, "step": 1641 }, { "epoch": 1.1539002108222065, "grad_norm": 0.6284526586532593, "learning_rate": 6.914968376669009e-06, "loss": 0.1062, "step": 1642 }, { "epoch": 1.1546029515108924, "grad_norm": 0.5511365532875061, "learning_rate": 6.919184820801125e-06, "loss": 0.0949, "step": 1643 }, { "epoch": 1.1553056921995783, "grad_norm": 0.9166580438613892, "learning_rate": 6.92340126493324e-06, "loss": 0.1878, "step": 1644 }, { "epoch": 1.1560084328882643, "grad_norm": 1.0161123275756836, "learning_rate": 6.927617709065355e-06, "loss": 0.2224, "step": 1645 }, { "epoch": 1.1567111735769502, "grad_norm": 1.9421718120574951, "learning_rate": 6.93183415319747e-06, "loss": 0.2781, "step": 1646 }, { "epoch": 1.157413914265636, "grad_norm": 1.9898539781570435, "learning_rate": 6.936050597329586e-06, "loss": 0.3949, "step": 1647 }, { "epoch": 1.1581166549543218, "grad_norm": 2.975794553756714, "learning_rate": 6.940267041461701e-06, "loss": 0.5079, "step": 1648 }, { "epoch": 1.1588193956430077, "grad_norm": 0.47393161058425903, "learning_rate": 6.944483485593816e-06, "loss": 0.1268, "step": 1649 }, { "epoch": 1.1595221363316937, "grad_norm": 0.4242730736732483, "learning_rate": 6.948699929725932e-06, "loss": 0.0697, "step": 1650 }, { "epoch": 1.1602248770203796, "grad_norm": 0.3484107553958893, "learning_rate": 6.952916373858047e-06, "loss": 0.0648, "step": 1651 }, { "epoch": 1.1609276177090653, "grad_norm": 0.45166897773742676, "learning_rate": 6.957132817990162e-06, "loss": 0.0626, "step": 1652 }, { "epoch": 1.1616303583977512, "grad_norm": 0.35318616032600403, "learning_rate": 6.961349262122277e-06, "loss": 0.0551, "step": 1653 }, { "epoch": 1.1623330990864371, "grad_norm": 0.554347038269043, "learning_rate": 6.9655657062543925e-06, "loss": 0.0419, "step": 1654 }, { "epoch": 1.163035839775123, "grad_norm": 0.4155484735965729, "learning_rate": 6.9697821503865075e-06, "loss": 0.0635, "step": 1655 }, { "epoch": 1.163738580463809, "grad_norm": 0.6045804619789124, "learning_rate": 6.9739985945186224e-06, "loss": 0.0513, "step": 1656 }, { "epoch": 1.1644413211524947, "grad_norm": 0.49623754620552063, "learning_rate": 6.978215038650738e-06, "loss": 0.0719, "step": 1657 }, { "epoch": 1.1651440618411806, "grad_norm": 0.36543625593185425, "learning_rate": 6.982431482782853e-06, "loss": 0.0447, "step": 1658 }, { "epoch": 1.1658468025298665, "grad_norm": 0.4315451681613922, "learning_rate": 6.986647926914968e-06, "loss": 0.0528, "step": 1659 }, { "epoch": 1.1665495432185524, "grad_norm": 0.3484625816345215, "learning_rate": 6.990864371047084e-06, "loss": 0.0493, "step": 1660 }, { "epoch": 1.1672522839072381, "grad_norm": 0.3954762816429138, "learning_rate": 6.995080815179199e-06, "loss": 0.0798, "step": 1661 }, { "epoch": 1.167955024595924, "grad_norm": 0.4195518493652344, "learning_rate": 6.999297259311315e-06, "loss": 0.0635, "step": 1662 }, { "epoch": 1.16865776528461, "grad_norm": 0.5096548199653625, "learning_rate": 7.00351370344343e-06, "loss": 0.0627, "step": 1663 }, { "epoch": 1.1693605059732959, "grad_norm": 0.5761106610298157, "learning_rate": 7.007730147575545e-06, "loss": 0.0872, "step": 1664 }, { "epoch": 1.1700632466619818, "grad_norm": 0.7692916393280029, "learning_rate": 7.01194659170766e-06, "loss": 0.0859, "step": 1665 }, { "epoch": 1.1707659873506677, "grad_norm": 0.7655286192893982, "learning_rate": 7.016163035839775e-06, "loss": 0.0914, "step": 1666 }, { "epoch": 1.1714687280393534, "grad_norm": 0.4967603385448456, "learning_rate": 7.020379479971891e-06, "loss": 0.0846, "step": 1667 }, { "epoch": 1.1721714687280393, "grad_norm": 0.568595290184021, "learning_rate": 7.024595924104006e-06, "loss": 0.0852, "step": 1668 }, { "epoch": 1.1728742094167253, "grad_norm": 0.9700475931167603, "learning_rate": 7.028812368236121e-06, "loss": 0.1748, "step": 1669 }, { "epoch": 1.1735769501054112, "grad_norm": 1.265934944152832, "learning_rate": 7.033028812368236e-06, "loss": 0.1988, "step": 1670 }, { "epoch": 1.1742796907940969, "grad_norm": 1.132192611694336, "learning_rate": 7.037245256500352e-06, "loss": 0.293, "step": 1671 }, { "epoch": 1.1749824314827828, "grad_norm": 1.582732081413269, "learning_rate": 7.041461700632467e-06, "loss": 0.4131, "step": 1672 }, { "epoch": 1.1756851721714687, "grad_norm": 3.950300455093384, "learning_rate": 7.0456781447645816e-06, "loss": 0.4233, "step": 1673 }, { "epoch": 1.1763879128601546, "grad_norm": 0.9673948884010315, "learning_rate": 7.049894588896697e-06, "loss": 0.1453, "step": 1674 }, { "epoch": 1.1770906535488406, "grad_norm": 1.1929779052734375, "learning_rate": 7.054111033028812e-06, "loss": 0.0805, "step": 1675 }, { "epoch": 1.1777933942375263, "grad_norm": 0.4184363782405853, "learning_rate": 7.058327477160927e-06, "loss": 0.0596, "step": 1676 }, { "epoch": 1.1784961349262122, "grad_norm": 0.6300926208496094, "learning_rate": 7.062543921293043e-06, "loss": 0.0749, "step": 1677 }, { "epoch": 1.1791988756148981, "grad_norm": 0.4279938042163849, "learning_rate": 7.066760365425159e-06, "loss": 0.0559, "step": 1678 }, { "epoch": 1.179901616303584, "grad_norm": 0.5737128257751465, "learning_rate": 7.070976809557274e-06, "loss": 0.0632, "step": 1679 }, { "epoch": 1.1806043569922697, "grad_norm": 0.30075234174728394, "learning_rate": 7.075193253689389e-06, "loss": 0.053, "step": 1680 }, { "epoch": 1.1813070976809557, "grad_norm": 0.47351905703544617, "learning_rate": 7.0794096978215045e-06, "loss": 0.0661, "step": 1681 }, { "epoch": 1.1820098383696416, "grad_norm": 0.34077274799346924, "learning_rate": 7.0836261419536194e-06, "loss": 0.0457, "step": 1682 }, { "epoch": 1.1827125790583275, "grad_norm": 0.5511778593063354, "learning_rate": 7.087842586085734e-06, "loss": 0.058, "step": 1683 }, { "epoch": 1.1834153197470134, "grad_norm": 1.0603212118148804, "learning_rate": 7.09205903021785e-06, "loss": 0.0602, "step": 1684 }, { "epoch": 1.1841180604356993, "grad_norm": 0.38930612802505493, "learning_rate": 7.096275474349965e-06, "loss": 0.0541, "step": 1685 }, { "epoch": 1.184820801124385, "grad_norm": 0.41967493295669556, "learning_rate": 7.10049191848208e-06, "loss": 0.0696, "step": 1686 }, { "epoch": 1.185523541813071, "grad_norm": 0.4150473475456238, "learning_rate": 7.104708362614195e-06, "loss": 0.0667, "step": 1687 }, { "epoch": 1.1862262825017569, "grad_norm": 0.8459685444831848, "learning_rate": 7.108924806746311e-06, "loss": 0.0678, "step": 1688 }, { "epoch": 1.1869290231904428, "grad_norm": 1.048103928565979, "learning_rate": 7.113141250878426e-06, "loss": 0.0901, "step": 1689 }, { "epoch": 1.1876317638791285, "grad_norm": 0.39914482831954956, "learning_rate": 7.117357695010541e-06, "loss": 0.0516, "step": 1690 }, { "epoch": 1.1883345045678144, "grad_norm": 0.7164822220802307, "learning_rate": 7.1215741391426565e-06, "loss": 0.0853, "step": 1691 }, { "epoch": 1.1890372452565003, "grad_norm": 0.8420903086662292, "learning_rate": 7.125790583274772e-06, "loss": 0.08, "step": 1692 }, { "epoch": 1.1897399859451863, "grad_norm": 0.7224278450012207, "learning_rate": 7.130007027406887e-06, "loss": 0.1262, "step": 1693 }, { "epoch": 1.1904427266338722, "grad_norm": 0.6838920712471008, "learning_rate": 7.134223471539002e-06, "loss": 0.1697, "step": 1694 }, { "epoch": 1.1911454673225579, "grad_norm": 0.9871826171875, "learning_rate": 7.138439915671118e-06, "loss": 0.1811, "step": 1695 }, { "epoch": 1.1918482080112438, "grad_norm": 1.3814618587493896, "learning_rate": 7.142656359803233e-06, "loss": 0.3299, "step": 1696 }, { "epoch": 1.1925509486999297, "grad_norm": 1.351342797279358, "learning_rate": 7.146872803935348e-06, "loss": 0.3678, "step": 1697 }, { "epoch": 1.1932536893886156, "grad_norm": 3.812605381011963, "learning_rate": 7.151089248067464e-06, "loss": 0.5093, "step": 1698 }, { "epoch": 1.1939564300773016, "grad_norm": 0.5931066274642944, "learning_rate": 7.1553056921995786e-06, "loss": 0.1369, "step": 1699 }, { "epoch": 1.1946591707659873, "grad_norm": 0.3200526535511017, "learning_rate": 7.1595221363316935e-06, "loss": 0.0596, "step": 1700 }, { "epoch": 1.1953619114546732, "grad_norm": 0.3671976327896118, "learning_rate": 7.163738580463809e-06, "loss": 0.0585, "step": 1701 }, { "epoch": 1.196064652143359, "grad_norm": 0.3029539883136749, "learning_rate": 7.167955024595924e-06, "loss": 0.039, "step": 1702 }, { "epoch": 1.196767392832045, "grad_norm": 0.3351529836654663, "learning_rate": 7.172171468728039e-06, "loss": 0.0531, "step": 1703 }, { "epoch": 1.197470133520731, "grad_norm": 0.3827223479747772, "learning_rate": 7.176387912860154e-06, "loss": 0.0261, "step": 1704 }, { "epoch": 1.1981728742094166, "grad_norm": 0.3811368942260742, "learning_rate": 7.18060435699227e-06, "loss": 0.0431, "step": 1705 }, { "epoch": 1.1988756148981026, "grad_norm": 0.3108184337615967, "learning_rate": 7.184820801124385e-06, "loss": 0.0403, "step": 1706 }, { "epoch": 1.1995783555867885, "grad_norm": 0.4000326693058014, "learning_rate": 7.189037245256501e-06, "loss": 0.048, "step": 1707 }, { "epoch": 1.2002810962754744, "grad_norm": 0.3598572015762329, "learning_rate": 7.1932536893886164e-06, "loss": 0.0507, "step": 1708 }, { "epoch": 1.2009838369641601, "grad_norm": 0.5208864212036133, "learning_rate": 7.197470133520731e-06, "loss": 0.0798, "step": 1709 }, { "epoch": 1.201686577652846, "grad_norm": 0.40885797142982483, "learning_rate": 7.201686577652846e-06, "loss": 0.0419, "step": 1710 }, { "epoch": 1.202389318341532, "grad_norm": 0.6359207034111023, "learning_rate": 7.205903021784961e-06, "loss": 0.0654, "step": 1711 }, { "epoch": 1.2030920590302179, "grad_norm": 0.4181824326515198, "learning_rate": 7.210119465917077e-06, "loss": 0.0516, "step": 1712 }, { "epoch": 1.2037947997189038, "grad_norm": 0.6532586216926575, "learning_rate": 7.214335910049192e-06, "loss": 0.0715, "step": 1713 }, { "epoch": 1.2044975404075897, "grad_norm": 0.5594488382339478, "learning_rate": 7.218552354181307e-06, "loss": 0.0877, "step": 1714 }, { "epoch": 1.2052002810962754, "grad_norm": 0.6966838240623474, "learning_rate": 7.222768798313423e-06, "loss": 0.0777, "step": 1715 }, { "epoch": 1.2059030217849613, "grad_norm": 0.7368237972259521, "learning_rate": 7.226985242445538e-06, "loss": 0.0904, "step": 1716 }, { "epoch": 1.2066057624736473, "grad_norm": 0.49870815873146057, "learning_rate": 7.231201686577653e-06, "loss": 0.1184, "step": 1717 }, { "epoch": 1.2073085031623332, "grad_norm": 0.5415850281715393, "learning_rate": 7.235418130709768e-06, "loss": 0.1032, "step": 1718 }, { "epoch": 1.2080112438510189, "grad_norm": 0.909816324710846, "learning_rate": 7.239634574841883e-06, "loss": 0.129, "step": 1719 }, { "epoch": 1.2087139845397048, "grad_norm": 0.866576611995697, "learning_rate": 7.243851018973998e-06, "loss": 0.2477, "step": 1720 }, { "epoch": 1.2094167252283907, "grad_norm": 1.4351211786270142, "learning_rate": 7.248067463106113e-06, "loss": 0.3001, "step": 1721 }, { "epoch": 1.2101194659170766, "grad_norm": 1.4537887573242188, "learning_rate": 7.25228390723823e-06, "loss": 0.382, "step": 1722 }, { "epoch": 1.2108222066057626, "grad_norm": 4.530016899108887, "learning_rate": 7.256500351370345e-06, "loss": 0.4742, "step": 1723 }, { "epoch": 1.2115249472944483, "grad_norm": 0.5406470894813538, "learning_rate": 7.26071679550246e-06, "loss": 0.1291, "step": 1724 }, { "epoch": 1.2122276879831342, "grad_norm": 0.4911401569843292, "learning_rate": 7.2649332396345756e-06, "loss": 0.1015, "step": 1725 }, { "epoch": 1.21293042867182, "grad_norm": 0.3394670784473419, "learning_rate": 7.2691496837666905e-06, "loss": 0.0566, "step": 1726 }, { "epoch": 1.213633169360506, "grad_norm": 0.29292333126068115, "learning_rate": 7.2733661278988055e-06, "loss": 0.0449, "step": 1727 }, { "epoch": 1.2143359100491917, "grad_norm": 0.30881378054618835, "learning_rate": 7.27758257203092e-06, "loss": 0.0506, "step": 1728 }, { "epoch": 1.2150386507378776, "grad_norm": 0.3677496314048767, "learning_rate": 7.281799016163036e-06, "loss": 0.0473, "step": 1729 }, { "epoch": 1.2157413914265636, "grad_norm": 0.44390031695365906, "learning_rate": 7.286015460295151e-06, "loss": 0.0505, "step": 1730 }, { "epoch": 1.2164441321152495, "grad_norm": 0.3539511263370514, "learning_rate": 7.290231904427266e-06, "loss": 0.0475, "step": 1731 }, { "epoch": 1.2171468728039354, "grad_norm": 0.3118135631084442, "learning_rate": 7.294448348559382e-06, "loss": 0.0515, "step": 1732 }, { "epoch": 1.2178496134926213, "grad_norm": 0.3898327648639679, "learning_rate": 7.298664792691497e-06, "loss": 0.0471, "step": 1733 }, { "epoch": 1.218552354181307, "grad_norm": 0.4743826687335968, "learning_rate": 7.302881236823612e-06, "loss": 0.061, "step": 1734 }, { "epoch": 1.219255094869993, "grad_norm": 0.47018054127693176, "learning_rate": 7.3070976809557275e-06, "loss": 0.0673, "step": 1735 }, { "epoch": 1.2199578355586789, "grad_norm": 0.5339438319206238, "learning_rate": 7.3113141250878425e-06, "loss": 0.065, "step": 1736 }, { "epoch": 1.2206605762473648, "grad_norm": 0.4285909831523895, "learning_rate": 7.315530569219958e-06, "loss": 0.0692, "step": 1737 }, { "epoch": 1.2213633169360505, "grad_norm": 0.3894980549812317, "learning_rate": 7.319747013352073e-06, "loss": 0.0632, "step": 1738 }, { "epoch": 1.2220660576247364, "grad_norm": 0.49814993143081665, "learning_rate": 7.323963457484189e-06, "loss": 0.0993, "step": 1739 }, { "epoch": 1.2227687983134223, "grad_norm": 0.4550510048866272, "learning_rate": 7.328179901616304e-06, "loss": 0.082, "step": 1740 }, { "epoch": 1.2234715390021083, "grad_norm": 0.5249207019805908, "learning_rate": 7.332396345748419e-06, "loss": 0.0882, "step": 1741 }, { "epoch": 1.2241742796907942, "grad_norm": 0.6149151921272278, "learning_rate": 7.336612789880535e-06, "loss": 0.0911, "step": 1742 }, { "epoch": 1.2248770203794799, "grad_norm": 0.8145772814750671, "learning_rate": 7.34082923401265e-06, "loss": 0.1146, "step": 1743 }, { "epoch": 1.2255797610681658, "grad_norm": 0.729213535785675, "learning_rate": 7.345045678144765e-06, "loss": 0.1248, "step": 1744 }, { "epoch": 1.2262825017568517, "grad_norm": 3.271718740463257, "learning_rate": 7.3492621222768795e-06, "loss": 0.2566, "step": 1745 }, { "epoch": 1.2269852424455376, "grad_norm": 1.6083375215530396, "learning_rate": 7.353478566408995e-06, "loss": 0.3583, "step": 1746 }, { "epoch": 1.2276879831342236, "grad_norm": 2.0395472049713135, "learning_rate": 7.35769501054111e-06, "loss": 0.3538, "step": 1747 }, { "epoch": 1.2283907238229093, "grad_norm": 2.2485222816467285, "learning_rate": 7.361911454673225e-06, "loss": 0.5172, "step": 1748 }, { "epoch": 1.2290934645115952, "grad_norm": 0.6674479842185974, "learning_rate": 7.366127898805341e-06, "loss": 0.155, "step": 1749 }, { "epoch": 1.229796205200281, "grad_norm": 0.39962854981422424, "learning_rate": 7.370344342937456e-06, "loss": 0.0707, "step": 1750 }, { "epoch": 1.230498945888967, "grad_norm": 0.3623094856739044, "learning_rate": 7.374560787069572e-06, "loss": 0.0653, "step": 1751 }, { "epoch": 1.231201686577653, "grad_norm": 0.49771878123283386, "learning_rate": 7.3787772312016875e-06, "loss": 0.0703, "step": 1752 }, { "epoch": 1.2319044272663386, "grad_norm": 0.524122416973114, "learning_rate": 7.3829936753338025e-06, "loss": 0.0483, "step": 1753 }, { "epoch": 1.2326071679550246, "grad_norm": 0.41185131669044495, "learning_rate": 7.387210119465917e-06, "loss": 0.0556, "step": 1754 }, { "epoch": 1.2333099086437105, "grad_norm": 0.3403652310371399, "learning_rate": 7.391426563598032e-06, "loss": 0.0688, "step": 1755 }, { "epoch": 1.2340126493323964, "grad_norm": 0.7284820079803467, "learning_rate": 7.395643007730148e-06, "loss": 0.0528, "step": 1756 }, { "epoch": 1.234715390021082, "grad_norm": 0.38415801525115967, "learning_rate": 7.399859451862263e-06, "loss": 0.0514, "step": 1757 }, { "epoch": 1.235418130709768, "grad_norm": 0.33086806535720825, "learning_rate": 7.404075895994378e-06, "loss": 0.0453, "step": 1758 }, { "epoch": 1.236120871398454, "grad_norm": 0.7123534679412842, "learning_rate": 7.408292340126494e-06, "loss": 0.0535, "step": 1759 }, { "epoch": 1.2368236120871399, "grad_norm": 0.557234525680542, "learning_rate": 7.412508784258609e-06, "loss": 0.0617, "step": 1760 }, { "epoch": 1.2375263527758258, "grad_norm": 0.41220465302467346, "learning_rate": 7.416725228390724e-06, "loss": 0.0625, "step": 1761 }, { "epoch": 1.2382290934645117, "grad_norm": 0.3704998791217804, "learning_rate": 7.420941672522839e-06, "loss": 0.0416, "step": 1762 }, { "epoch": 1.2389318341531974, "grad_norm": 0.9693852066993713, "learning_rate": 7.4251581166549544e-06, "loss": 0.0968, "step": 1763 }, { "epoch": 1.2396345748418833, "grad_norm": 0.5331265330314636, "learning_rate": 7.429374560787069e-06, "loss": 0.0758, "step": 1764 }, { "epoch": 1.2403373155305693, "grad_norm": 0.4948807656764984, "learning_rate": 7.433591004919184e-06, "loss": 0.0423, "step": 1765 }, { "epoch": 1.2410400562192552, "grad_norm": 0.9848883152008057, "learning_rate": 7.437807449051301e-06, "loss": 0.0921, "step": 1766 }, { "epoch": 1.2417427969079409, "grad_norm": 0.7554680109024048, "learning_rate": 7.442023893183416e-06, "loss": 0.101, "step": 1767 }, { "epoch": 1.2424455375966268, "grad_norm": 0.8370440602302551, "learning_rate": 7.446240337315531e-06, "loss": 0.1193, "step": 1768 }, { "epoch": 1.2431482782853127, "grad_norm": 0.7099232077598572, "learning_rate": 7.450456781447647e-06, "loss": 0.1494, "step": 1769 }, { "epoch": 1.2438510189739986, "grad_norm": 1.8573052883148193, "learning_rate": 7.454673225579762e-06, "loss": 0.2216, "step": 1770 }, { "epoch": 1.2445537596626846, "grad_norm": 1.7719367742538452, "learning_rate": 7.4588896697118765e-06, "loss": 0.3012, "step": 1771 }, { "epoch": 1.2452565003513703, "grad_norm": 5.237682819366455, "learning_rate": 7.4631061138439915e-06, "loss": 0.4024, "step": 1772 }, { "epoch": 1.2459592410400562, "grad_norm": 3.704228401184082, "learning_rate": 7.467322557976107e-06, "loss": 0.5527, "step": 1773 }, { "epoch": 1.246661981728742, "grad_norm": 0.6378483772277832, "learning_rate": 7.471539002108222e-06, "loss": 0.1343, "step": 1774 }, { "epoch": 1.247364722417428, "grad_norm": 0.5224942564964294, "learning_rate": 7.475755446240337e-06, "loss": 0.0601, "step": 1775 }, { "epoch": 1.248067463106114, "grad_norm": 0.37004002928733826, "learning_rate": 7.479971890372453e-06, "loss": 0.0616, "step": 1776 }, { "epoch": 1.2487702037947996, "grad_norm": 0.3785662353038788, "learning_rate": 7.484188334504568e-06, "loss": 0.0549, "step": 1777 }, { "epoch": 1.2494729444834856, "grad_norm": 0.410144567489624, "learning_rate": 7.488404778636683e-06, "loss": 0.0515, "step": 1778 }, { "epoch": 1.2501756851721715, "grad_norm": 0.6038010120391846, "learning_rate": 7.492621222768798e-06, "loss": 0.0428, "step": 1779 }, { "epoch": 1.2508784258608574, "grad_norm": 0.4596497714519501, "learning_rate": 7.4968376669009136e-06, "loss": 0.0548, "step": 1780 }, { "epoch": 1.2515811665495433, "grad_norm": 0.389587938785553, "learning_rate": 7.501054111033029e-06, "loss": 0.0454, "step": 1781 }, { "epoch": 1.252283907238229, "grad_norm": 0.4561924338340759, "learning_rate": 7.505270555165144e-06, "loss": 0.0847, "step": 1782 }, { "epoch": 1.252986647926915, "grad_norm": 0.889938235282898, "learning_rate": 7.50948699929726e-06, "loss": 0.0429, "step": 1783 }, { "epoch": 1.2536893886156009, "grad_norm": 0.3921017050743103, "learning_rate": 7.513703443429375e-06, "loss": 0.0582, "step": 1784 }, { "epoch": 1.2543921293042868, "grad_norm": 0.36310529708862305, "learning_rate": 7.51791988756149e-06, "loss": 0.0436, "step": 1785 }, { "epoch": 1.2550948699929725, "grad_norm": 0.42675796151161194, "learning_rate": 7.522136331693606e-06, "loss": 0.0609, "step": 1786 }, { "epoch": 1.2557976106816584, "grad_norm": 0.4053211510181427, "learning_rate": 7.526352775825721e-06, "loss": 0.0581, "step": 1787 }, { "epoch": 1.2565003513703443, "grad_norm": 0.462252676486969, "learning_rate": 7.530569219957836e-06, "loss": 0.0731, "step": 1788 }, { "epoch": 1.2572030920590302, "grad_norm": 0.4773815870285034, "learning_rate": 7.534785664089951e-06, "loss": 0.0516, "step": 1789 }, { "epoch": 1.2579058327477162, "grad_norm": 0.44186264276504517, "learning_rate": 7.539002108222066e-06, "loss": 0.0598, "step": 1790 }, { "epoch": 1.258608573436402, "grad_norm": 0.852546215057373, "learning_rate": 7.543218552354181e-06, "loss": 0.0919, "step": 1791 }, { "epoch": 1.2593113141250878, "grad_norm": 0.6114134788513184, "learning_rate": 7.547434996486296e-06, "loss": 0.1126, "step": 1792 }, { "epoch": 1.2600140548137737, "grad_norm": 0.711297333240509, "learning_rate": 7.551651440618412e-06, "loss": 0.1394, "step": 1793 }, { "epoch": 1.2607167955024596, "grad_norm": 0.8734205365180969, "learning_rate": 7.555867884750527e-06, "loss": 0.1701, "step": 1794 }, { "epoch": 1.2614195361911453, "grad_norm": 2.1069087982177734, "learning_rate": 7.560084328882642e-06, "loss": 0.2474, "step": 1795 }, { "epoch": 1.2621222768798313, "grad_norm": 1.6516677141189575, "learning_rate": 7.564300773014757e-06, "loss": 0.3381, "step": 1796 }, { "epoch": 1.2628250175685172, "grad_norm": 2.004223108291626, "learning_rate": 7.568517217146873e-06, "loss": 0.4009, "step": 1797 }, { "epoch": 1.263527758257203, "grad_norm": 1.7674185037612915, "learning_rate": 7.572733661278988e-06, "loss": 0.4685, "step": 1798 }, { "epoch": 1.264230498945889, "grad_norm": 0.49753743410110474, "learning_rate": 7.5769501054111026e-06, "loss": 0.1302, "step": 1799 }, { "epoch": 1.264933239634575, "grad_norm": 0.34055909514427185, "learning_rate": 7.581166549543218e-06, "loss": 0.0593, "step": 1800 }, { "epoch": 1.2656359803232606, "grad_norm": 0.4385804533958435, "learning_rate": 7.585382993675333e-06, "loss": 0.0556, "step": 1801 }, { "epoch": 1.2663387210119466, "grad_norm": 0.42294561862945557, "learning_rate": 7.589599437807448e-06, "loss": 0.0607, "step": 1802 }, { "epoch": 1.2670414617006325, "grad_norm": 0.386763334274292, "learning_rate": 7.593815881939564e-06, "loss": 0.066, "step": 1803 }, { "epoch": 1.2677442023893184, "grad_norm": 0.30315595865249634, "learning_rate": 7.598032326071681e-06, "loss": 0.0378, "step": 1804 }, { "epoch": 1.268446943078004, "grad_norm": 0.3806140720844269, "learning_rate": 7.602248770203796e-06, "loss": 0.0388, "step": 1805 }, { "epoch": 1.26914968376669, "grad_norm": 1.1930323839187622, "learning_rate": 7.6064652143359106e-06, "loss": 0.0709, "step": 1806 }, { "epoch": 1.269852424455376, "grad_norm": 0.42012688517570496, "learning_rate": 7.610681658468026e-06, "loss": 0.0545, "step": 1807 }, { "epoch": 1.2705551651440619, "grad_norm": 0.3147648274898529, "learning_rate": 7.614898102600141e-06, "loss": 0.0393, "step": 1808 }, { "epoch": 1.2712579058327478, "grad_norm": 0.4556863009929657, "learning_rate": 7.619114546732256e-06, "loss": 0.0581, "step": 1809 }, { "epoch": 1.2719606465214337, "grad_norm": 0.4398593604564667, "learning_rate": 7.623330990864372e-06, "loss": 0.0405, "step": 1810 }, { "epoch": 1.2726633872101194, "grad_norm": 0.38097241520881653, "learning_rate": 7.627547434996487e-06, "loss": 0.0472, "step": 1811 }, { "epoch": 1.2733661278988053, "grad_norm": 0.6423070430755615, "learning_rate": 7.631763879128602e-06, "loss": 0.0459, "step": 1812 }, { "epoch": 1.2740688685874912, "grad_norm": 0.608688473701477, "learning_rate": 7.635980323260717e-06, "loss": 0.0866, "step": 1813 }, { "epoch": 1.2747716092761772, "grad_norm": 0.4009203016757965, "learning_rate": 7.640196767392832e-06, "loss": 0.0681, "step": 1814 }, { "epoch": 1.2754743499648629, "grad_norm": 0.3939235806465149, "learning_rate": 7.644413211524948e-06, "loss": 0.0641, "step": 1815 }, { "epoch": 1.2761770906535488, "grad_norm": 0.48910388350486755, "learning_rate": 7.648629655657063e-06, "loss": 0.0794, "step": 1816 }, { "epoch": 1.2768798313422347, "grad_norm": 0.9279074668884277, "learning_rate": 7.652846099789178e-06, "loss": 0.0839, "step": 1817 }, { "epoch": 1.2775825720309206, "grad_norm": 0.6556916832923889, "learning_rate": 7.657062543921293e-06, "loss": 0.1227, "step": 1818 }, { "epoch": 1.2782853127196065, "grad_norm": 0.8740938901901245, "learning_rate": 7.661278988053408e-06, "loss": 0.1465, "step": 1819 }, { "epoch": 1.2789880534082925, "grad_norm": 2.2857449054718018, "learning_rate": 7.665495432185523e-06, "loss": 0.2057, "step": 1820 }, { "epoch": 1.2796907940969782, "grad_norm": 3.1344246864318848, "learning_rate": 7.669711876317638e-06, "loss": 0.3334, "step": 1821 }, { "epoch": 1.280393534785664, "grad_norm": 1.9543510675430298, "learning_rate": 7.673928320449755e-06, "loss": 0.3883, "step": 1822 }, { "epoch": 1.28109627547435, "grad_norm": 2.2329835891723633, "learning_rate": 7.67814476458187e-06, "loss": 0.4641, "step": 1823 }, { "epoch": 1.2817990161630357, "grad_norm": 0.5007922053337097, "learning_rate": 7.682361208713985e-06, "loss": 0.126, "step": 1824 }, { "epoch": 1.2825017568517216, "grad_norm": 0.33694180846214294, "learning_rate": 7.6865776528461e-06, "loss": 0.0566, "step": 1825 }, { "epoch": 1.2832044975404076, "grad_norm": 0.5283912420272827, "learning_rate": 7.690794096978215e-06, "loss": 0.0848, "step": 1826 }, { "epoch": 1.2839072382290935, "grad_norm": 0.32579562067985535, "learning_rate": 7.69501054111033e-06, "loss": 0.0506, "step": 1827 }, { "epoch": 1.2846099789177794, "grad_norm": 0.38021120429039, "learning_rate": 7.699226985242444e-06, "loss": 0.0415, "step": 1828 }, { "epoch": 1.2853127196064653, "grad_norm": 0.3218313157558441, "learning_rate": 7.703443429374561e-06, "loss": 0.0342, "step": 1829 }, { "epoch": 1.286015460295151, "grad_norm": 0.3548140227794647, "learning_rate": 7.707659873506676e-06, "loss": 0.0352, "step": 1830 }, { "epoch": 1.286718200983837, "grad_norm": 0.33992066979408264, "learning_rate": 7.711876317638791e-06, "loss": 0.0494, "step": 1831 }, { "epoch": 1.2874209416725229, "grad_norm": 0.35891711711883545, "learning_rate": 7.716092761770906e-06, "loss": 0.0518, "step": 1832 }, { "epoch": 1.2881236823612088, "grad_norm": 0.37084901332855225, "learning_rate": 7.720309205903023e-06, "loss": 0.0427, "step": 1833 }, { "epoch": 1.2888264230498945, "grad_norm": 0.43618717789649963, "learning_rate": 7.724525650035137e-06, "loss": 0.0518, "step": 1834 }, { "epoch": 1.2895291637385804, "grad_norm": 0.49350279569625854, "learning_rate": 7.728742094167252e-06, "loss": 0.0538, "step": 1835 }, { "epoch": 1.2902319044272663, "grad_norm": 0.536106526851654, "learning_rate": 7.732958538299369e-06, "loss": 0.0641, "step": 1836 }, { "epoch": 1.2909346451159522, "grad_norm": 0.518599271774292, "learning_rate": 7.737174982431484e-06, "loss": 0.0475, "step": 1837 }, { "epoch": 1.2916373858046382, "grad_norm": 0.4648589789867401, "learning_rate": 7.741391426563599e-06, "loss": 0.0673, "step": 1838 }, { "epoch": 1.292340126493324, "grad_norm": 0.7036103010177612, "learning_rate": 7.745607870695714e-06, "loss": 0.0781, "step": 1839 }, { "epoch": 1.2930428671820098, "grad_norm": 0.4062550961971283, "learning_rate": 7.749824314827829e-06, "loss": 0.0573, "step": 1840 }, { "epoch": 1.2937456078706957, "grad_norm": 0.5973287224769592, "learning_rate": 7.754040758959944e-06, "loss": 0.089, "step": 1841 }, { "epoch": 1.2944483485593816, "grad_norm": 0.6265441179275513, "learning_rate": 7.75825720309206e-06, "loss": 0.0876, "step": 1842 }, { "epoch": 1.2951510892480675, "grad_norm": 0.6793888807296753, "learning_rate": 7.762473647224175e-06, "loss": 0.1152, "step": 1843 }, { "epoch": 1.2958538299367532, "grad_norm": 1.5898754596710205, "learning_rate": 7.76669009135629e-06, "loss": 0.123, "step": 1844 }, { "epoch": 1.2965565706254392, "grad_norm": 1.1152064800262451, "learning_rate": 7.770906535488405e-06, "loss": 0.2161, "step": 1845 }, { "epoch": 1.297259311314125, "grad_norm": 1.6625277996063232, "learning_rate": 7.77512297962052e-06, "loss": 0.2979, "step": 1846 }, { "epoch": 1.297962052002811, "grad_norm": 1.2316895723342896, "learning_rate": 7.779339423752635e-06, "loss": 0.2993, "step": 1847 }, { "epoch": 1.298664792691497, "grad_norm": 3.6157474517822266, "learning_rate": 7.78355586788475e-06, "loss": 0.4454, "step": 1848 }, { "epoch": 1.2993675333801828, "grad_norm": 0.5246172547340393, "learning_rate": 7.787772312016867e-06, "loss": 0.1329, "step": 1849 }, { "epoch": 1.3000702740688685, "grad_norm": 0.3069615364074707, "learning_rate": 7.791988756148982e-06, "loss": 0.057, "step": 1850 }, { "epoch": 1.3007730147575545, "grad_norm": 0.48566731810569763, "learning_rate": 7.796205200281097e-06, "loss": 0.0401, "step": 1851 }, { "epoch": 1.3014757554462404, "grad_norm": 0.4572603106498718, "learning_rate": 7.800421644413212e-06, "loss": 0.0637, "step": 1852 }, { "epoch": 1.302178496134926, "grad_norm": 0.3277852535247803, "learning_rate": 7.804638088545326e-06, "loss": 0.0536, "step": 1853 }, { "epoch": 1.302881236823612, "grad_norm": 0.2569284439086914, "learning_rate": 7.808854532677441e-06, "loss": 0.0272, "step": 1854 }, { "epoch": 1.303583977512298, "grad_norm": 0.298810750246048, "learning_rate": 7.813070976809556e-06, "loss": 0.0473, "step": 1855 }, { "epoch": 1.3042867182009839, "grad_norm": 0.5487832427024841, "learning_rate": 7.817287420941673e-06, "loss": 0.0539, "step": 1856 }, { "epoch": 1.3049894588896698, "grad_norm": 0.3420679569244385, "learning_rate": 7.821503865073788e-06, "loss": 0.0476, "step": 1857 }, { "epoch": 1.3056921995783557, "grad_norm": 0.5794788599014282, "learning_rate": 7.825720309205903e-06, "loss": 0.0582, "step": 1858 }, { "epoch": 1.3063949402670414, "grad_norm": 1.0464787483215332, "learning_rate": 7.829936753338018e-06, "loss": 0.1024, "step": 1859 }, { "epoch": 1.3070976809557273, "grad_norm": 0.3913733661174774, "learning_rate": 7.834153197470133e-06, "loss": 0.0294, "step": 1860 }, { "epoch": 1.3078004216444132, "grad_norm": 0.49574920535087585, "learning_rate": 7.838369641602248e-06, "loss": 0.0568, "step": 1861 }, { "epoch": 1.3085031623330992, "grad_norm": 0.39461806416511536, "learning_rate": 7.842586085734363e-06, "loss": 0.0468, "step": 1862 }, { "epoch": 1.3092059030217849, "grad_norm": 0.5222393274307251, "learning_rate": 7.846802529866481e-06, "loss": 0.0722, "step": 1863 }, { "epoch": 1.3099086437104708, "grad_norm": 0.41355660557746887, "learning_rate": 7.851018973998596e-06, "loss": 0.0633, "step": 1864 }, { "epoch": 1.3106113843991567, "grad_norm": 0.49647799134254456, "learning_rate": 7.855235418130711e-06, "loss": 0.0731, "step": 1865 }, { "epoch": 1.3113141250878426, "grad_norm": 0.517687976360321, "learning_rate": 7.859451862262826e-06, "loss": 0.0903, "step": 1866 }, { "epoch": 1.3120168657765285, "grad_norm": 0.8699033260345459, "learning_rate": 7.86366830639494e-06, "loss": 0.1415, "step": 1867 }, { "epoch": 1.3127196064652145, "grad_norm": 0.5866033434867859, "learning_rate": 7.867884750527056e-06, "loss": 0.0963, "step": 1868 }, { "epoch": 1.3134223471539002, "grad_norm": 0.7281909584999084, "learning_rate": 7.87210119465917e-06, "loss": 0.1842, "step": 1869 }, { "epoch": 1.314125087842586, "grad_norm": 1.0039759874343872, "learning_rate": 7.876317638791287e-06, "loss": 0.2067, "step": 1870 }, { "epoch": 1.314827828531272, "grad_norm": 1.295715093612671, "learning_rate": 7.880534082923402e-06, "loss": 0.285, "step": 1871 }, { "epoch": 1.3155305692199577, "grad_norm": 1.62386953830719, "learning_rate": 7.884750527055517e-06, "loss": 0.3862, "step": 1872 }, { "epoch": 1.3162333099086436, "grad_norm": 4.021759033203125, "learning_rate": 7.888966971187632e-06, "loss": 0.4978, "step": 1873 }, { "epoch": 1.3169360505973295, "grad_norm": 0.3850138485431671, "learning_rate": 7.893183415319747e-06, "loss": 0.112, "step": 1874 }, { "epoch": 1.3176387912860155, "grad_norm": 0.36281585693359375, "learning_rate": 7.897399859451862e-06, "loss": 0.0634, "step": 1875 }, { "epoch": 1.3183415319747014, "grad_norm": 0.37716731429100037, "learning_rate": 7.901616303583979e-06, "loss": 0.0769, "step": 1876 }, { "epoch": 1.3190442726633873, "grad_norm": 0.3677276074886322, "learning_rate": 7.905832747716094e-06, "loss": 0.0613, "step": 1877 }, { "epoch": 1.319747013352073, "grad_norm": 0.3234230577945709, "learning_rate": 7.910049191848209e-06, "loss": 0.043, "step": 1878 }, { "epoch": 1.320449754040759, "grad_norm": 0.2963528037071228, "learning_rate": 7.914265635980323e-06, "loss": 0.0444, "step": 1879 }, { "epoch": 1.3211524947294448, "grad_norm": 0.35609912872314453, "learning_rate": 7.918482080112438e-06, "loss": 0.051, "step": 1880 }, { "epoch": 1.3218552354181308, "grad_norm": 0.32773053646087646, "learning_rate": 7.922698524244553e-06, "loss": 0.0526, "step": 1881 }, { "epoch": 1.3225579761068165, "grad_norm": 0.3394498825073242, "learning_rate": 7.926914968376668e-06, "loss": 0.0468, "step": 1882 }, { "epoch": 1.3232607167955024, "grad_norm": 0.2849533259868622, "learning_rate": 7.931131412508785e-06, "loss": 0.0435, "step": 1883 }, { "epoch": 1.3239634574841883, "grad_norm": 0.43167051672935486, "learning_rate": 7.9353478566409e-06, "loss": 0.0627, "step": 1884 }, { "epoch": 1.3246661981728742, "grad_norm": 0.3043387234210968, "learning_rate": 7.939564300773015e-06, "loss": 0.0304, "step": 1885 }, { "epoch": 1.3253689388615602, "grad_norm": 0.5451861619949341, "learning_rate": 7.94378074490513e-06, "loss": 0.0586, "step": 1886 }, { "epoch": 1.326071679550246, "grad_norm": 0.5888434648513794, "learning_rate": 7.947997189037245e-06, "loss": 0.0684, "step": 1887 }, { "epoch": 1.3267744202389318, "grad_norm": 0.6769050359725952, "learning_rate": 7.95221363316936e-06, "loss": 0.0535, "step": 1888 }, { "epoch": 1.3274771609276177, "grad_norm": 0.5335603356361389, "learning_rate": 7.956430077301475e-06, "loss": 0.0768, "step": 1889 }, { "epoch": 1.3281799016163036, "grad_norm": 0.4112558960914612, "learning_rate": 7.960646521433591e-06, "loss": 0.0518, "step": 1890 }, { "epoch": 1.3288826423049895, "grad_norm": 0.585597038269043, "learning_rate": 7.964862965565706e-06, "loss": 0.0752, "step": 1891 }, { "epoch": 1.3295853829936752, "grad_norm": 0.4624848961830139, "learning_rate": 7.969079409697821e-06, "loss": 0.0938, "step": 1892 }, { "epoch": 1.3302881236823612, "grad_norm": 1.0511986017227173, "learning_rate": 7.973295853829938e-06, "loss": 0.0855, "step": 1893 }, { "epoch": 1.330990864371047, "grad_norm": 0.7803069353103638, "learning_rate": 7.977512297962053e-06, "loss": 0.1532, "step": 1894 }, { "epoch": 1.331693605059733, "grad_norm": 0.8537772297859192, "learning_rate": 7.981728742094168e-06, "loss": 0.2199, "step": 1895 }, { "epoch": 1.332396345748419, "grad_norm": 1.0424776077270508, "learning_rate": 7.985945186226283e-06, "loss": 0.2529, "step": 1896 }, { "epoch": 1.3330990864371048, "grad_norm": 1.9859482049942017, "learning_rate": 7.9901616303584e-06, "loss": 0.4178, "step": 1897 }, { "epoch": 1.3338018271257905, "grad_norm": 4.066018104553223, "learning_rate": 7.994378074490514e-06, "loss": 0.4789, "step": 1898 }, { "epoch": 1.3345045678144765, "grad_norm": 0.7526330947875977, "learning_rate": 7.998594518622629e-06, "loss": 0.1166, "step": 1899 }, { "epoch": 1.3352073085031624, "grad_norm": 0.44123154878616333, "learning_rate": 8.002810962754744e-06, "loss": 0.0546, "step": 1900 }, { "epoch": 1.335910049191848, "grad_norm": 0.309196412563324, "learning_rate": 8.007027406886859e-06, "loss": 0.0551, "step": 1901 }, { "epoch": 1.336612789880534, "grad_norm": 0.34697243571281433, "learning_rate": 8.011243851018974e-06, "loss": 0.0504, "step": 1902 }, { "epoch": 1.33731553056922, "grad_norm": 0.39438962936401367, "learning_rate": 8.015460295151089e-06, "loss": 0.0401, "step": 1903 }, { "epoch": 1.3380182712579058, "grad_norm": 0.32830363512039185, "learning_rate": 8.019676739283206e-06, "loss": 0.0531, "step": 1904 }, { "epoch": 1.3387210119465918, "grad_norm": 0.5112961530685425, "learning_rate": 8.02389318341532e-06, "loss": 0.0423, "step": 1905 }, { "epoch": 1.3394237526352777, "grad_norm": 0.4199411869049072, "learning_rate": 8.028109627547435e-06, "loss": 0.0527, "step": 1906 }, { "epoch": 1.3401264933239634, "grad_norm": 0.36185115575790405, "learning_rate": 8.03232607167955e-06, "loss": 0.0376, "step": 1907 }, { "epoch": 1.3408292340126493, "grad_norm": 0.34032297134399414, "learning_rate": 8.036542515811665e-06, "loss": 0.0349, "step": 1908 }, { "epoch": 1.3415319747013352, "grad_norm": 0.3027138113975525, "learning_rate": 8.04075895994378e-06, "loss": 0.042, "step": 1909 }, { "epoch": 1.3422347153900211, "grad_norm": 0.5202673673629761, "learning_rate": 8.044975404075897e-06, "loss": 0.0387, "step": 1910 }, { "epoch": 1.3429374560787068, "grad_norm": 0.38263288140296936, "learning_rate": 8.049191848208012e-06, "loss": 0.0694, "step": 1911 }, { "epoch": 1.3436401967673928, "grad_norm": 0.4831427037715912, "learning_rate": 8.053408292340127e-06, "loss": 0.0599, "step": 1912 }, { "epoch": 1.3443429374560787, "grad_norm": 0.5455146431922913, "learning_rate": 8.057624736472242e-06, "loss": 0.0522, "step": 1913 }, { "epoch": 1.3450456781447646, "grad_norm": 0.46650585532188416, "learning_rate": 8.061841180604357e-06, "loss": 0.0679, "step": 1914 }, { "epoch": 1.3457484188334505, "grad_norm": 0.4206324517726898, "learning_rate": 8.066057624736472e-06, "loss": 0.055, "step": 1915 }, { "epoch": 1.3464511595221365, "grad_norm": 0.9011726379394531, "learning_rate": 8.070274068868587e-06, "loss": 0.0867, "step": 1916 }, { "epoch": 1.3471539002108222, "grad_norm": 0.4965534210205078, "learning_rate": 8.074490513000703e-06, "loss": 0.0653, "step": 1917 }, { "epoch": 1.347856640899508, "grad_norm": 1.0609688758850098, "learning_rate": 8.078706957132818e-06, "loss": 0.1315, "step": 1918 }, { "epoch": 1.348559381588194, "grad_norm": 0.6498876214027405, "learning_rate": 8.082923401264933e-06, "loss": 0.1401, "step": 1919 }, { "epoch": 1.3492621222768797, "grad_norm": 0.932634174823761, "learning_rate": 8.087139845397048e-06, "loss": 0.2074, "step": 1920 }, { "epoch": 1.3499648629655656, "grad_norm": 1.4252897500991821, "learning_rate": 8.091356289529163e-06, "loss": 0.2487, "step": 1921 }, { "epoch": 1.3506676036542515, "grad_norm": 1.5642627477645874, "learning_rate": 8.095572733661278e-06, "loss": 0.366, "step": 1922 }, { "epoch": 1.3513703443429375, "grad_norm": 4.160162448883057, "learning_rate": 8.099789177793395e-06, "loss": 0.5602, "step": 1923 }, { "epoch": 1.3520730850316234, "grad_norm": 0.6678019762039185, "learning_rate": 8.104005621925511e-06, "loss": 0.1202, "step": 1924 }, { "epoch": 1.3527758257203093, "grad_norm": 0.3857235312461853, "learning_rate": 8.108222066057626e-06, "loss": 0.0631, "step": 1925 }, { "epoch": 1.353478566408995, "grad_norm": 0.30312082171440125, "learning_rate": 8.112438510189741e-06, "loss": 0.0473, "step": 1926 }, { "epoch": 1.354181307097681, "grad_norm": 0.40899771451950073, "learning_rate": 8.116654954321856e-06, "loss": 0.0559, "step": 1927 }, { "epoch": 1.3548840477863668, "grad_norm": 0.7078134417533875, "learning_rate": 8.120871398453971e-06, "loss": 0.0475, "step": 1928 }, { "epoch": 1.3555867884750528, "grad_norm": 0.31714698672294617, "learning_rate": 8.125087842586086e-06, "loss": 0.0416, "step": 1929 }, { "epoch": 1.3562895291637385, "grad_norm": 0.3323640525341034, "learning_rate": 8.1293042867182e-06, "loss": 0.05, "step": 1930 }, { "epoch": 1.3569922698524244, "grad_norm": 0.3263198435306549, "learning_rate": 8.133520730850317e-06, "loss": 0.0547, "step": 1931 }, { "epoch": 1.3576950105411103, "grad_norm": 0.333400160074234, "learning_rate": 8.137737174982432e-06, "loss": 0.0389, "step": 1932 }, { "epoch": 1.3583977512297962, "grad_norm": 0.37212294340133667, "learning_rate": 8.141953619114547e-06, "loss": 0.0439, "step": 1933 }, { "epoch": 1.3591004919184821, "grad_norm": 0.4957852363586426, "learning_rate": 8.146170063246662e-06, "loss": 0.0486, "step": 1934 }, { "epoch": 1.359803232607168, "grad_norm": 0.492192804813385, "learning_rate": 8.150386507378777e-06, "loss": 0.0446, "step": 1935 }, { "epoch": 1.3605059732958538, "grad_norm": 0.5900259613990784, "learning_rate": 8.154602951510892e-06, "loss": 0.0615, "step": 1936 }, { "epoch": 1.3612087139845397, "grad_norm": 0.47183290123939514, "learning_rate": 8.158819395643007e-06, "loss": 0.0396, "step": 1937 }, { "epoch": 1.3619114546732256, "grad_norm": 0.5474035143852234, "learning_rate": 8.163035839775124e-06, "loss": 0.0736, "step": 1938 }, { "epoch": 1.3626141953619115, "grad_norm": 0.5847229361534119, "learning_rate": 8.167252283907239e-06, "loss": 0.0832, "step": 1939 }, { "epoch": 1.3633169360505972, "grad_norm": 0.39998143911361694, "learning_rate": 8.171468728039354e-06, "loss": 0.0519, "step": 1940 }, { "epoch": 1.3640196767392831, "grad_norm": 0.6591793298721313, "learning_rate": 8.175685172171469e-06, "loss": 0.0681, "step": 1941 }, { "epoch": 1.364722417427969, "grad_norm": 0.5401723980903625, "learning_rate": 8.179901616303584e-06, "loss": 0.1013, "step": 1942 }, { "epoch": 1.365425158116655, "grad_norm": 0.5580028295516968, "learning_rate": 8.184118060435698e-06, "loss": 0.1033, "step": 1943 }, { "epoch": 1.366127898805341, "grad_norm": 0.6938269138336182, "learning_rate": 8.188334504567815e-06, "loss": 0.1547, "step": 1944 }, { "epoch": 1.3668306394940268, "grad_norm": 1.256926417350769, "learning_rate": 8.19255094869993e-06, "loss": 0.2634, "step": 1945 }, { "epoch": 1.3675333801827125, "grad_norm": 1.4912387132644653, "learning_rate": 8.196767392832045e-06, "loss": 0.2863, "step": 1946 }, { "epoch": 1.3682361208713985, "grad_norm": 4.16417121887207, "learning_rate": 8.20098383696416e-06, "loss": 0.4155, "step": 1947 }, { "epoch": 1.3689388615600844, "grad_norm": 3.3017194271087646, "learning_rate": 8.205200281096275e-06, "loss": 0.5225, "step": 1948 }, { "epoch": 1.36964160224877, "grad_norm": 0.43963396549224854, "learning_rate": 8.20941672522839e-06, "loss": 0.134, "step": 1949 }, { "epoch": 1.370344342937456, "grad_norm": 0.3519813120365143, "learning_rate": 8.213633169360505e-06, "loss": 0.0668, "step": 1950 }, { "epoch": 1.371047083626142, "grad_norm": 0.32229992747306824, "learning_rate": 8.217849613492621e-06, "loss": 0.0461, "step": 1951 }, { "epoch": 1.3717498243148278, "grad_norm": 0.37304162979125977, "learning_rate": 8.222066057624738e-06, "loss": 0.0451, "step": 1952 }, { "epoch": 1.3724525650035138, "grad_norm": 0.41719669103622437, "learning_rate": 8.226282501756853e-06, "loss": 0.0672, "step": 1953 }, { "epoch": 1.3731553056921997, "grad_norm": 0.34609442949295044, "learning_rate": 8.230498945888968e-06, "loss": 0.033, "step": 1954 }, { "epoch": 1.3738580463808854, "grad_norm": 0.31976839900016785, "learning_rate": 8.234715390021083e-06, "loss": 0.0344, "step": 1955 }, { "epoch": 1.3745607870695713, "grad_norm": 0.3228462338447571, "learning_rate": 8.238931834153198e-06, "loss": 0.0392, "step": 1956 }, { "epoch": 1.3752635277582572, "grad_norm": 0.3675471544265747, "learning_rate": 8.243148278285313e-06, "loss": 0.0639, "step": 1957 }, { "epoch": 1.3759662684469431, "grad_norm": 0.6102598309516907, "learning_rate": 8.24736472241743e-06, "loss": 0.0706, "step": 1958 }, { "epoch": 1.3766690091356288, "grad_norm": 0.4311094582080841, "learning_rate": 8.251581166549544e-06, "loss": 0.061, "step": 1959 }, { "epoch": 1.3773717498243148, "grad_norm": 0.34798887372016907, "learning_rate": 8.25579761068166e-06, "loss": 0.0482, "step": 1960 }, { "epoch": 1.3780744905130007, "grad_norm": 0.5011805295944214, "learning_rate": 8.260014054813774e-06, "loss": 0.0847, "step": 1961 }, { "epoch": 1.3787772312016866, "grad_norm": 0.47571712732315063, "learning_rate": 8.26423049894589e-06, "loss": 0.0576, "step": 1962 }, { "epoch": 1.3794799718903725, "grad_norm": 0.43116676807403564, "learning_rate": 8.268446943078004e-06, "loss": 0.0682, "step": 1963 }, { "epoch": 1.3801827125790584, "grad_norm": 0.4232073724269867, "learning_rate": 8.272663387210119e-06, "loss": 0.0606, "step": 1964 }, { "epoch": 1.3808854532677441, "grad_norm": 0.613056480884552, "learning_rate": 8.276879831342236e-06, "loss": 0.0891, "step": 1965 }, { "epoch": 1.38158819395643, "grad_norm": 0.4510515034198761, "learning_rate": 8.28109627547435e-06, "loss": 0.0528, "step": 1966 }, { "epoch": 1.382290934645116, "grad_norm": 1.044075608253479, "learning_rate": 8.285312719606466e-06, "loss": 0.1083, "step": 1967 }, { "epoch": 1.382993675333802, "grad_norm": 1.3115545511245728, "learning_rate": 8.28952916373858e-06, "loss": 0.0861, "step": 1968 }, { "epoch": 1.3836964160224876, "grad_norm": 1.0198887586593628, "learning_rate": 8.293745607870696e-06, "loss": 0.1779, "step": 1969 }, { "epoch": 1.3843991567111735, "grad_norm": 1.1592738628387451, "learning_rate": 8.29796205200281e-06, "loss": 0.2055, "step": 1970 }, { "epoch": 1.3851018973998594, "grad_norm": 1.1372270584106445, "learning_rate": 8.302178496134925e-06, "loss": 0.3261, "step": 1971 }, { "epoch": 1.3858046380885454, "grad_norm": 1.664182186126709, "learning_rate": 8.306394940267042e-06, "loss": 0.3166, "step": 1972 }, { "epoch": 1.3865073787772313, "grad_norm": 2.167663335800171, "learning_rate": 8.310611384399157e-06, "loss": 0.4688, "step": 1973 }, { "epoch": 1.3872101194659172, "grad_norm": 0.46878835558891296, "learning_rate": 8.314827828531272e-06, "loss": 0.1218, "step": 1974 }, { "epoch": 1.387912860154603, "grad_norm": 0.3894680142402649, "learning_rate": 8.319044272663387e-06, "loss": 0.0631, "step": 1975 }, { "epoch": 1.3886156008432888, "grad_norm": 0.3622167706489563, "learning_rate": 8.323260716795502e-06, "loss": 0.0625, "step": 1976 }, { "epoch": 1.3893183415319748, "grad_norm": 0.2937157452106476, "learning_rate": 8.327477160927617e-06, "loss": 0.0512, "step": 1977 }, { "epoch": 1.3900210822206605, "grad_norm": 0.37457048892974854, "learning_rate": 8.331693605059733e-06, "loss": 0.0396, "step": 1978 }, { "epoch": 1.3907238229093464, "grad_norm": 0.3293982744216919, "learning_rate": 8.335910049191848e-06, "loss": 0.0543, "step": 1979 }, { "epoch": 1.3914265635980323, "grad_norm": 0.23023836314678192, "learning_rate": 8.340126493323963e-06, "loss": 0.0256, "step": 1980 }, { "epoch": 1.3921293042867182, "grad_norm": 0.38776323199272156, "learning_rate": 8.344342937456078e-06, "loss": 0.0768, "step": 1981 }, { "epoch": 1.3928320449754041, "grad_norm": 0.6729966402053833, "learning_rate": 8.348559381588195e-06, "loss": 0.0685, "step": 1982 }, { "epoch": 1.39353478566409, "grad_norm": 0.3368460536003113, "learning_rate": 8.35277582572031e-06, "loss": 0.0443, "step": 1983 }, { "epoch": 1.3942375263527758, "grad_norm": 0.3955339193344116, "learning_rate": 8.356992269852425e-06, "loss": 0.0568, "step": 1984 }, { "epoch": 1.3949402670414617, "grad_norm": 0.32576417922973633, "learning_rate": 8.361208713984541e-06, "loss": 0.0474, "step": 1985 }, { "epoch": 1.3956430077301476, "grad_norm": 0.4490896761417389, "learning_rate": 8.365425158116656e-06, "loss": 0.0529, "step": 1986 }, { "epoch": 1.3963457484188335, "grad_norm": 0.38832998275756836, "learning_rate": 8.369641602248771e-06, "loss": 0.0478, "step": 1987 }, { "epoch": 1.3970484891075192, "grad_norm": 0.35930949449539185, "learning_rate": 8.373858046380886e-06, "loss": 0.0812, "step": 1988 }, { "epoch": 1.3977512297962051, "grad_norm": 0.42851078510284424, "learning_rate": 8.378074490513001e-06, "loss": 0.0728, "step": 1989 }, { "epoch": 1.398453970484891, "grad_norm": 0.435326486825943, "learning_rate": 8.382290934645116e-06, "loss": 0.0582, "step": 1990 }, { "epoch": 1.399156711173577, "grad_norm": 0.756625771522522, "learning_rate": 8.386507378777231e-06, "loss": 0.0928, "step": 1991 }, { "epoch": 1.399859451862263, "grad_norm": 0.7468721866607666, "learning_rate": 8.390723822909348e-06, "loss": 0.1379, "step": 1992 }, { "epoch": 1.4005621925509488, "grad_norm": 0.5820426940917969, "learning_rate": 8.394940267041463e-06, "loss": 0.1035, "step": 1993 }, { "epoch": 1.4012649332396345, "grad_norm": 0.6460005044937134, "learning_rate": 8.399156711173578e-06, "loss": 0.1575, "step": 1994 }, { "epoch": 1.4019676739283204, "grad_norm": 0.9855660200119019, "learning_rate": 8.403373155305693e-06, "loss": 0.1979, "step": 1995 }, { "epoch": 1.4026704146170064, "grad_norm": 1.3345650434494019, "learning_rate": 8.407589599437807e-06, "loss": 0.314, "step": 1996 }, { "epoch": 1.403373155305692, "grad_norm": 1.8971565961837769, "learning_rate": 8.411806043569922e-06, "loss": 0.3236, "step": 1997 }, { "epoch": 1.404075895994378, "grad_norm": 1.8048278093338013, "learning_rate": 8.416022487702037e-06, "loss": 0.428, "step": 1998 }, { "epoch": 1.404778636683064, "grad_norm": 0.3945331275463104, "learning_rate": 8.420238931834154e-06, "loss": 0.1019, "step": 1999 }, { "epoch": 1.4054813773717498, "grad_norm": 0.35126790404319763, "learning_rate": 8.424455375966269e-06, "loss": 0.0711, "step": 2000 }, { "epoch": 1.4054813773717498, "eval_cer": 0.2064580537461949, "eval_loss": 0.3726554811000824, "eval_runtime": 18.24, "eval_samples_per_second": 248.794, "eval_steps_per_second": 0.822, "eval_wer": 0.39476550934331445, "step": 2000 }, { "epoch": 1.4061841180604358, "grad_norm": 0.3225651681423187, "learning_rate": 8.428671820098384e-06, "loss": 0.064, "step": 2001 }, { "epoch": 1.4068868587491217, "grad_norm": 0.26798567175865173, "learning_rate": 8.432888264230499e-06, "loss": 0.0427, "step": 2002 }, { "epoch": 1.4075895994378074, "grad_norm": 0.3277446925640106, "learning_rate": 8.437104708362614e-06, "loss": 0.0595, "step": 2003 }, { "epoch": 1.4082923401264933, "grad_norm": 0.3261849582195282, "learning_rate": 8.441321152494729e-06, "loss": 0.0387, "step": 2004 }, { "epoch": 1.4089950808151792, "grad_norm": 0.28085052967071533, "learning_rate": 8.445537596626844e-06, "loss": 0.0434, "step": 2005 }, { "epoch": 1.4096978215038651, "grad_norm": 1.9512635469436646, "learning_rate": 8.44975404075896e-06, "loss": 0.0453, "step": 2006 }, { "epoch": 1.4104005621925508, "grad_norm": 0.32613152265548706, "learning_rate": 8.453970484891075e-06, "loss": 0.0597, "step": 2007 }, { "epoch": 1.4111033028812368, "grad_norm": 0.4199449419975281, "learning_rate": 8.45818692902319e-06, "loss": 0.0618, "step": 2008 }, { "epoch": 1.4118060435699227, "grad_norm": 0.413608580827713, "learning_rate": 8.462403373155305e-06, "loss": 0.0602, "step": 2009 }, { "epoch": 1.4125087842586086, "grad_norm": 0.3147572875022888, "learning_rate": 8.46661981728742e-06, "loss": 0.035, "step": 2010 }, { "epoch": 1.4132115249472945, "grad_norm": 0.5361464023590088, "learning_rate": 8.470836261419535e-06, "loss": 0.0733, "step": 2011 }, { "epoch": 1.4139142656359804, "grad_norm": 0.46828773617744446, "learning_rate": 8.475052705551652e-06, "loss": 0.0428, "step": 2012 }, { "epoch": 1.4146170063246661, "grad_norm": 0.5347342491149902, "learning_rate": 8.479269149683768e-06, "loss": 0.0864, "step": 2013 }, { "epoch": 1.415319747013352, "grad_norm": 0.7439039945602417, "learning_rate": 8.483485593815883e-06, "loss": 0.0904, "step": 2014 }, { "epoch": 1.416022487702038, "grad_norm": 5.246702671051025, "learning_rate": 8.487702037947998e-06, "loss": 0.0556, "step": 2015 }, { "epoch": 1.416725228390724, "grad_norm": 0.3961849808692932, "learning_rate": 8.491918482080113e-06, "loss": 0.0769, "step": 2016 }, { "epoch": 1.4174279690794096, "grad_norm": 8.245238304138184, "learning_rate": 8.496134926212228e-06, "loss": 0.1113, "step": 2017 }, { "epoch": 1.4181307097680955, "grad_norm": 0.8775112628936768, "learning_rate": 8.500351370344343e-06, "loss": 0.105, "step": 2018 }, { "epoch": 1.4188334504567814, "grad_norm": 3.450849771499634, "learning_rate": 8.50456781447646e-06, "loss": 0.1464, "step": 2019 }, { "epoch": 1.4195361911454674, "grad_norm": 2.501375436782837, "learning_rate": 8.508784258608575e-06, "loss": 0.2016, "step": 2020 }, { "epoch": 1.4202389318341533, "grad_norm": 3.8121774196624756, "learning_rate": 8.51300070274069e-06, "loss": 0.2941, "step": 2021 }, { "epoch": 1.4209416725228392, "grad_norm": 2.54646635055542, "learning_rate": 8.517217146872804e-06, "loss": 0.3864, "step": 2022 }, { "epoch": 1.421644413211525, "grad_norm": 2.905965805053711, "learning_rate": 8.52143359100492e-06, "loss": 0.4677, "step": 2023 }, { "epoch": 1.4223471539002108, "grad_norm": 0.449541300535202, "learning_rate": 8.525650035137034e-06, "loss": 0.123, "step": 2024 }, { "epoch": 1.4230498945888967, "grad_norm": 0.3810023069381714, "learning_rate": 8.52986647926915e-06, "loss": 0.0635, "step": 2025 }, { "epoch": 1.4237526352775824, "grad_norm": 0.31627053022384644, "learning_rate": 8.534082923401266e-06, "loss": 0.044, "step": 2026 }, { "epoch": 1.4244553759662684, "grad_norm": 0.24705040454864502, "learning_rate": 8.538299367533381e-06, "loss": 0.0533, "step": 2027 }, { "epoch": 1.4251581166549543, "grad_norm": 0.40092733502388, "learning_rate": 8.542515811665496e-06, "loss": 0.0465, "step": 2028 }, { "epoch": 1.4258608573436402, "grad_norm": 0.33267322182655334, "learning_rate": 8.54673225579761e-06, "loss": 0.0443, "step": 2029 }, { "epoch": 1.4265635980323261, "grad_norm": 0.42778220772743225, "learning_rate": 8.550948699929726e-06, "loss": 0.0395, "step": 2030 }, { "epoch": 1.427266338721012, "grad_norm": 0.5904695391654968, "learning_rate": 8.55516514406184e-06, "loss": 0.0574, "step": 2031 }, { "epoch": 1.4279690794096978, "grad_norm": 1.5449165105819702, "learning_rate": 8.559381588193956e-06, "loss": 0.0455, "step": 2032 }, { "epoch": 1.4286718200983837, "grad_norm": 0.35670751333236694, "learning_rate": 8.563598032326072e-06, "loss": 0.0497, "step": 2033 }, { "epoch": 1.4293745607870696, "grad_norm": 0.4312773048877716, "learning_rate": 8.567814476458187e-06, "loss": 0.0675, "step": 2034 }, { "epoch": 1.4300773014757555, "grad_norm": 0.24667397141456604, "learning_rate": 8.572030920590302e-06, "loss": 0.0301, "step": 2035 }, { "epoch": 1.4307800421644412, "grad_norm": 0.4389149844646454, "learning_rate": 8.576247364722417e-06, "loss": 0.0601, "step": 2036 }, { "epoch": 1.4314827828531271, "grad_norm": 0.3980671763420105, "learning_rate": 8.580463808854532e-06, "loss": 0.0472, "step": 2037 }, { "epoch": 1.432185523541813, "grad_norm": 0.49274829030036926, "learning_rate": 8.584680252986647e-06, "loss": 0.0563, "step": 2038 }, { "epoch": 1.432888264230499, "grad_norm": 0.3479017913341522, "learning_rate": 8.588896697118762e-06, "loss": 0.053, "step": 2039 }, { "epoch": 1.433591004919185, "grad_norm": 1.5282518863677979, "learning_rate": 8.593113141250879e-06, "loss": 0.077, "step": 2040 }, { "epoch": 1.4342937456078708, "grad_norm": 2.104172706604004, "learning_rate": 8.597329585382993e-06, "loss": 0.0629, "step": 2041 }, { "epoch": 1.4349964862965565, "grad_norm": 0.6912804245948792, "learning_rate": 8.60154602951511e-06, "loss": 0.0972, "step": 2042 }, { "epoch": 1.4356992269852424, "grad_norm": 0.9755844473838806, "learning_rate": 8.605762473647225e-06, "loss": 0.1088, "step": 2043 }, { "epoch": 1.4364019676739284, "grad_norm": 0.9396519660949707, "learning_rate": 8.60997891777934e-06, "loss": 0.1554, "step": 2044 }, { "epoch": 1.437104708362614, "grad_norm": 1.5714160203933716, "learning_rate": 8.614195361911455e-06, "loss": 0.1959, "step": 2045 }, { "epoch": 1.4378074490513, "grad_norm": 2.225090980529785, "learning_rate": 8.61841180604357e-06, "loss": 0.3428, "step": 2046 }, { "epoch": 1.438510189739986, "grad_norm": 1.5837452411651611, "learning_rate": 8.622628250175687e-06, "loss": 0.3268, "step": 2047 }, { "epoch": 1.4392129304286718, "grad_norm": 5.8981242179870605, "learning_rate": 8.626844694307801e-06, "loss": 0.4539, "step": 2048 }, { "epoch": 1.4399156711173577, "grad_norm": 0.4206879436969757, "learning_rate": 8.631061138439916e-06, "loss": 0.1129, "step": 2049 }, { "epoch": 1.4406184118060437, "grad_norm": 0.3272029161453247, "learning_rate": 8.635277582572031e-06, "loss": 0.0634, "step": 2050 }, { "epoch": 1.4413211524947294, "grad_norm": 0.42877939343452454, "learning_rate": 8.639494026704146e-06, "loss": 0.0508, "step": 2051 }, { "epoch": 1.4420238931834153, "grad_norm": 0.41791993379592896, "learning_rate": 8.643710470836261e-06, "loss": 0.0541, "step": 2052 }, { "epoch": 1.4427266338721012, "grad_norm": 0.35781416296958923, "learning_rate": 8.647926914968378e-06, "loss": 0.0543, "step": 2053 }, { "epoch": 1.4434293745607871, "grad_norm": 0.2601347267627716, "learning_rate": 8.652143359100493e-06, "loss": 0.0336, "step": 2054 }, { "epoch": 1.4441321152494728, "grad_norm": 0.3055557310581207, "learning_rate": 8.656359803232608e-06, "loss": 0.0516, "step": 2055 }, { "epoch": 1.4448348559381587, "grad_norm": 0.3025050759315491, "learning_rate": 8.660576247364723e-06, "loss": 0.041, "step": 2056 }, { "epoch": 1.4455375966268447, "grad_norm": 0.4100925326347351, "learning_rate": 8.664792691496838e-06, "loss": 0.0477, "step": 2057 }, { "epoch": 1.4462403373155306, "grad_norm": 0.3574170470237732, "learning_rate": 8.669009135628953e-06, "loss": 0.0376, "step": 2058 }, { "epoch": 1.4469430780042165, "grad_norm": 0.3426695466041565, "learning_rate": 8.673225579761068e-06, "loss": 0.0569, "step": 2059 }, { "epoch": 1.4476458186929024, "grad_norm": 0.42406120896339417, "learning_rate": 8.677442023893184e-06, "loss": 0.0358, "step": 2060 }, { "epoch": 1.4483485593815881, "grad_norm": 0.35917043685913086, "learning_rate": 8.681658468025299e-06, "loss": 0.0611, "step": 2061 }, { "epoch": 1.449051300070274, "grad_norm": 0.40947648882865906, "learning_rate": 8.685874912157414e-06, "loss": 0.0605, "step": 2062 }, { "epoch": 1.44975404075896, "grad_norm": 0.6926882863044739, "learning_rate": 8.690091356289529e-06, "loss": 0.063, "step": 2063 }, { "epoch": 1.450456781447646, "grad_norm": 0.40158727765083313, "learning_rate": 8.694307800421644e-06, "loss": 0.0747, "step": 2064 }, { "epoch": 1.4511595221363316, "grad_norm": 0.46441560983657837, "learning_rate": 8.698524244553759e-06, "loss": 0.0456, "step": 2065 }, { "epoch": 1.4518622628250175, "grad_norm": 0.4968129098415375, "learning_rate": 8.702740688685874e-06, "loss": 0.073, "step": 2066 }, { "epoch": 1.4525650035137034, "grad_norm": 0.5037140846252441, "learning_rate": 8.70695713281799e-06, "loss": 0.0924, "step": 2067 }, { "epoch": 1.4532677442023894, "grad_norm": 0.7547125220298767, "learning_rate": 8.711173576950105e-06, "loss": 0.0976, "step": 2068 }, { "epoch": 1.4539704848910753, "grad_norm": 0.8771512508392334, "learning_rate": 8.71539002108222e-06, "loss": 0.1559, "step": 2069 }, { "epoch": 1.4546732255797612, "grad_norm": 1.7205219268798828, "learning_rate": 8.719606465214335e-06, "loss": 0.1958, "step": 2070 }, { "epoch": 1.455375966268447, "grad_norm": 1.1724374294281006, "learning_rate": 8.72382290934645e-06, "loss": 0.2335, "step": 2071 }, { "epoch": 1.4560787069571328, "grad_norm": 1.9600718021392822, "learning_rate": 8.728039353478567e-06, "loss": 0.3206, "step": 2072 }, { "epoch": 1.4567814476458187, "grad_norm": 3.0187323093414307, "learning_rate": 8.732255797610682e-06, "loss": 0.4668, "step": 2073 }, { "epoch": 1.4574841883345044, "grad_norm": 7.648752689361572, "learning_rate": 8.736472241742798e-06, "loss": 0.1502, "step": 2074 }, { "epoch": 1.4581869290231904, "grad_norm": 0.4127943217754364, "learning_rate": 8.740688685874913e-06, "loss": 0.0804, "step": 2075 }, { "epoch": 1.4588896697118763, "grad_norm": 0.3496015667915344, "learning_rate": 8.744905130007028e-06, "loss": 0.0542, "step": 2076 }, { "epoch": 1.4595924104005622, "grad_norm": 0.2955314517021179, "learning_rate": 8.749121574139143e-06, "loss": 0.0401, "step": 2077 }, { "epoch": 1.4602951510892481, "grad_norm": 0.34555572271347046, "learning_rate": 8.753338018271258e-06, "loss": 0.0348, "step": 2078 }, { "epoch": 1.460997891777934, "grad_norm": 0.2574281096458435, "learning_rate": 8.757554462403373e-06, "loss": 0.0375, "step": 2079 }, { "epoch": 1.4617006324666197, "grad_norm": 0.38188451528549194, "learning_rate": 8.761770906535488e-06, "loss": 0.0525, "step": 2080 }, { "epoch": 1.4624033731553057, "grad_norm": 0.2776380479335785, "learning_rate": 8.765987350667605e-06, "loss": 0.0438, "step": 2081 }, { "epoch": 1.4631061138439916, "grad_norm": 0.3443364202976227, "learning_rate": 8.77020379479972e-06, "loss": 0.0462, "step": 2082 }, { "epoch": 1.4638088545326775, "grad_norm": 0.3771331310272217, "learning_rate": 8.774420238931835e-06, "loss": 0.0536, "step": 2083 }, { "epoch": 1.4645115952213632, "grad_norm": 0.4215662181377411, "learning_rate": 8.77863668306395e-06, "loss": 0.0502, "step": 2084 }, { "epoch": 1.4652143359100491, "grad_norm": 0.6799123287200928, "learning_rate": 8.782853127196065e-06, "loss": 0.0439, "step": 2085 }, { "epoch": 1.465917076598735, "grad_norm": 0.3780827820301056, "learning_rate": 8.78706957132818e-06, "loss": 0.0615, "step": 2086 }, { "epoch": 1.466619817287421, "grad_norm": 0.47903987765312195, "learning_rate": 8.791286015460296e-06, "loss": 0.0421, "step": 2087 }, { "epoch": 1.467322557976107, "grad_norm": 0.799750030040741, "learning_rate": 8.795502459592411e-06, "loss": 0.0612, "step": 2088 }, { "epoch": 1.4680252986647928, "grad_norm": 0.40671923756599426, "learning_rate": 8.799718903724526e-06, "loss": 0.0676, "step": 2089 }, { "epoch": 1.4687280393534785, "grad_norm": 0.48862677812576294, "learning_rate": 8.803935347856641e-06, "loss": 0.0684, "step": 2090 }, { "epoch": 1.4694307800421644, "grad_norm": 0.39744436740875244, "learning_rate": 8.808151791988756e-06, "loss": 0.0627, "step": 2091 }, { "epoch": 1.4701335207308504, "grad_norm": 0.48678526282310486, "learning_rate": 8.81236823612087e-06, "loss": 0.0782, "step": 2092 }, { "epoch": 1.4708362614195363, "grad_norm": 0.6143102049827576, "learning_rate": 8.816584680252986e-06, "loss": 0.1194, "step": 2093 }, { "epoch": 1.471539002108222, "grad_norm": 0.4866878092288971, "learning_rate": 8.820801124385102e-06, "loss": 0.1325, "step": 2094 }, { "epoch": 1.472241742796908, "grad_norm": 0.9179343581199646, "learning_rate": 8.825017568517217e-06, "loss": 0.1949, "step": 2095 }, { "epoch": 1.4729444834855938, "grad_norm": 1.068195104598999, "learning_rate": 8.829234012649332e-06, "loss": 0.2869, "step": 2096 }, { "epoch": 1.4736472241742797, "grad_norm": 4.319606304168701, "learning_rate": 8.833450456781447e-06, "loss": 0.3301, "step": 2097 }, { "epoch": 1.4743499648629657, "grad_norm": 4.608095645904541, "learning_rate": 8.837666900913562e-06, "loss": 0.4418, "step": 2098 }, { "epoch": 1.4750527055516516, "grad_norm": 0.5901064872741699, "learning_rate": 8.841883345045677e-06, "loss": 0.1629, "step": 2099 }, { "epoch": 1.4757554462403373, "grad_norm": 0.33502957224845886, "learning_rate": 8.846099789177792e-06, "loss": 0.057, "step": 2100 }, { "epoch": 1.4764581869290232, "grad_norm": 0.38462865352630615, "learning_rate": 8.85031623330991e-06, "loss": 0.0491, "step": 2101 }, { "epoch": 1.4771609276177091, "grad_norm": 0.4312857389450073, "learning_rate": 8.854532677442025e-06, "loss": 0.0364, "step": 2102 }, { "epoch": 1.4778636683063948, "grad_norm": 0.2623125910758972, "learning_rate": 8.85874912157414e-06, "loss": 0.0447, "step": 2103 }, { "epoch": 1.4785664089950807, "grad_norm": 0.4241899847984314, "learning_rate": 8.862965565706255e-06, "loss": 0.0425, "step": 2104 }, { "epoch": 1.4792691496837667, "grad_norm": 0.3167382478713989, "learning_rate": 8.86718200983837e-06, "loss": 0.0523, "step": 2105 }, { "epoch": 1.4799718903724526, "grad_norm": 0.31003135442733765, "learning_rate": 8.871398453970485e-06, "loss": 0.0473, "step": 2106 }, { "epoch": 1.4806746310611385, "grad_norm": 0.41070255637168884, "learning_rate": 8.8756148981026e-06, "loss": 0.0599, "step": 2107 }, { "epoch": 1.4813773717498244, "grad_norm": 0.29911932349205017, "learning_rate": 8.879831342234717e-06, "loss": 0.0273, "step": 2108 }, { "epoch": 1.4820801124385101, "grad_norm": 0.4342108368873596, "learning_rate": 8.884047786366832e-06, "loss": 0.054, "step": 2109 }, { "epoch": 1.482782853127196, "grad_norm": 0.29500812292099, "learning_rate": 8.888264230498947e-06, "loss": 0.0347, "step": 2110 }, { "epoch": 1.483485593815882, "grad_norm": 0.6018549203872681, "learning_rate": 8.892480674631062e-06, "loss": 0.0579, "step": 2111 }, { "epoch": 1.4841883345045679, "grad_norm": 0.5543568730354309, "learning_rate": 8.896697118763176e-06, "loss": 0.0946, "step": 2112 }, { "epoch": 1.4848910751932536, "grad_norm": 0.4599972069263458, "learning_rate": 8.900913562895291e-06, "loss": 0.0704, "step": 2113 }, { "epoch": 1.4855938158819395, "grad_norm": 0.6271045804023743, "learning_rate": 8.905130007027406e-06, "loss": 0.0902, "step": 2114 }, { "epoch": 1.4862965565706254, "grad_norm": 0.3051409125328064, "learning_rate": 8.909346451159523e-06, "loss": 0.0473, "step": 2115 }, { "epoch": 1.4869992972593113, "grad_norm": 0.5212839841842651, "learning_rate": 8.913562895291638e-06, "loss": 0.0711, "step": 2116 }, { "epoch": 1.4877020379479973, "grad_norm": 0.5753639936447144, "learning_rate": 8.917779339423753e-06, "loss": 0.0849, "step": 2117 }, { "epoch": 1.4884047786366832, "grad_norm": 0.8092588782310486, "learning_rate": 8.921995783555868e-06, "loss": 0.1105, "step": 2118 }, { "epoch": 1.489107519325369, "grad_norm": 0.7407414317131042, "learning_rate": 8.926212227687983e-06, "loss": 0.1435, "step": 2119 }, { "epoch": 1.4898102600140548, "grad_norm": 0.987558126449585, "learning_rate": 8.930428671820098e-06, "loss": 0.1805, "step": 2120 }, { "epoch": 1.4905130007027407, "grad_norm": 1.0957424640655518, "learning_rate": 8.934645115952214e-06, "loss": 0.2672, "step": 2121 }, { "epoch": 1.4912157413914264, "grad_norm": 1.8759512901306152, "learning_rate": 8.93886156008433e-06, "loss": 0.3312, "step": 2122 }, { "epoch": 1.4919184820801124, "grad_norm": 3.1942615509033203, "learning_rate": 8.943078004216444e-06, "loss": 0.4719, "step": 2123 }, { "epoch": 1.4926212227687983, "grad_norm": 0.7232943773269653, "learning_rate": 8.94729444834856e-06, "loss": 0.1201, "step": 2124 }, { "epoch": 1.4933239634574842, "grad_norm": 0.3365641236305237, "learning_rate": 8.951510892480674e-06, "loss": 0.0568, "step": 2125 }, { "epoch": 1.4940267041461701, "grad_norm": 0.36959749460220337, "learning_rate": 8.955727336612789e-06, "loss": 0.0427, "step": 2126 }, { "epoch": 1.494729444834856, "grad_norm": 0.2582738697528839, "learning_rate": 8.959943780744904e-06, "loss": 0.0318, "step": 2127 }, { "epoch": 1.4954321855235417, "grad_norm": 0.21944835782051086, "learning_rate": 8.96416022487702e-06, "loss": 0.0269, "step": 2128 }, { "epoch": 1.4961349262122277, "grad_norm": 0.5986309051513672, "learning_rate": 8.968376669009136e-06, "loss": 0.0371, "step": 2129 }, { "epoch": 1.4968376669009136, "grad_norm": 0.32638195157051086, "learning_rate": 8.97259311314125e-06, "loss": 0.0326, "step": 2130 }, { "epoch": 1.4975404075895995, "grad_norm": 0.6583226323127747, "learning_rate": 8.976809557273367e-06, "loss": 0.0641, "step": 2131 }, { "epoch": 1.4982431482782852, "grad_norm": 0.35937830805778503, "learning_rate": 8.981026001405482e-06, "loss": 0.0515, "step": 2132 }, { "epoch": 1.4989458889669711, "grad_norm": 0.29215720295906067, "learning_rate": 8.985242445537597e-06, "loss": 0.0344, "step": 2133 }, { "epoch": 1.499648629655657, "grad_norm": 0.3751777708530426, "learning_rate": 8.989458889669712e-06, "loss": 0.0504, "step": 2134 }, { "epoch": 1.500351370344343, "grad_norm": 0.314755916595459, "learning_rate": 8.993675333801829e-06, "loss": 0.0319, "step": 2135 }, { "epoch": 1.5010541110330289, "grad_norm": 0.444108784198761, "learning_rate": 8.997891777933944e-06, "loss": 0.0579, "step": 2136 }, { "epoch": 1.5017568517217148, "grad_norm": 0.4060017168521881, "learning_rate": 9.002108222066059e-06, "loss": 0.0514, "step": 2137 }, { "epoch": 1.5024595924104007, "grad_norm": 0.6636412739753723, "learning_rate": 9.006324666198173e-06, "loss": 0.0858, "step": 2138 }, { "epoch": 1.5031623330990864, "grad_norm": 0.7419934272766113, "learning_rate": 9.010541110330288e-06, "loss": 0.0712, "step": 2139 }, { "epoch": 1.5038650737877723, "grad_norm": 0.4745856523513794, "learning_rate": 9.014757554462403e-06, "loss": 0.0547, "step": 2140 }, { "epoch": 1.504567814476458, "grad_norm": 0.6197842359542847, "learning_rate": 9.018973998594518e-06, "loss": 0.0768, "step": 2141 }, { "epoch": 1.505270555165144, "grad_norm": 0.4682728946208954, "learning_rate": 9.023190442726635e-06, "loss": 0.0742, "step": 2142 }, { "epoch": 1.5059732958538299, "grad_norm": 0.6048453450202942, "learning_rate": 9.02740688685875e-06, "loss": 0.1115, "step": 2143 }, { "epoch": 1.5066760365425158, "grad_norm": 1.3938137292861938, "learning_rate": 9.031623330990865e-06, "loss": 0.1314, "step": 2144 }, { "epoch": 1.5073787772312017, "grad_norm": 0.8846178650856018, "learning_rate": 9.03583977512298e-06, "loss": 0.2085, "step": 2145 }, { "epoch": 1.5080815179198876, "grad_norm": 1.7023957967758179, "learning_rate": 9.040056219255095e-06, "loss": 0.2437, "step": 2146 }, { "epoch": 1.5087842586085736, "grad_norm": 1.3564094305038452, "learning_rate": 9.04427266338721e-06, "loss": 0.3752, "step": 2147 }, { "epoch": 1.5094869992972593, "grad_norm": 2.1019437313079834, "learning_rate": 9.048489107519325e-06, "loss": 0.4493, "step": 2148 }, { "epoch": 1.5101897399859452, "grad_norm": 0.5076905488967896, "learning_rate": 9.052705551651441e-06, "loss": 0.1236, "step": 2149 }, { "epoch": 1.510892480674631, "grad_norm": 0.3182792365550995, "learning_rate": 9.056921995783556e-06, "loss": 0.0508, "step": 2150 }, { "epoch": 1.5115952213633168, "grad_norm": 0.5601587891578674, "learning_rate": 9.061138439915671e-06, "loss": 0.054, "step": 2151 }, { "epoch": 1.5122979620520027, "grad_norm": 0.30031004548072815, "learning_rate": 9.065354884047786e-06, "loss": 0.0525, "step": 2152 }, { "epoch": 1.5130007027406887, "grad_norm": 0.39166757464408875, "learning_rate": 9.069571328179901e-06, "loss": 0.046, "step": 2153 }, { "epoch": 1.5137034434293746, "grad_norm": 0.3999174237251282, "learning_rate": 9.073787772312016e-06, "loss": 0.0475, "step": 2154 }, { "epoch": 1.5144061841180605, "grad_norm": 0.422025591135025, "learning_rate": 9.078004216444133e-06, "loss": 0.0428, "step": 2155 }, { "epoch": 1.5151089248067464, "grad_norm": 0.30940476059913635, "learning_rate": 9.082220660576248e-06, "loss": 0.0406, "step": 2156 }, { "epoch": 1.5158116654954323, "grad_norm": 0.3954017460346222, "learning_rate": 9.086437104708363e-06, "loss": 0.0549, "step": 2157 }, { "epoch": 1.516514406184118, "grad_norm": 0.5691171884536743, "learning_rate": 9.090653548840477e-06, "loss": 0.0485, "step": 2158 }, { "epoch": 1.517217146872804, "grad_norm": 0.40144556760787964, "learning_rate": 9.094869992972592e-06, "loss": 0.0708, "step": 2159 }, { "epoch": 1.5179198875614897, "grad_norm": 0.655035674571991, "learning_rate": 9.099086437104707e-06, "loss": 0.0324, "step": 2160 }, { "epoch": 1.5186226282501756, "grad_norm": 0.3210919201374054, "learning_rate": 9.103302881236824e-06, "loss": 0.0455, "step": 2161 }, { "epoch": 1.5193253689388615, "grad_norm": 0.3182774484157562, "learning_rate": 9.10751932536894e-06, "loss": 0.0345, "step": 2162 }, { "epoch": 1.5200281096275474, "grad_norm": 0.4749791920185089, "learning_rate": 9.111735769501056e-06, "loss": 0.0608, "step": 2163 }, { "epoch": 1.5207308503162333, "grad_norm": 0.41083747148513794, "learning_rate": 9.11595221363317e-06, "loss": 0.057, "step": 2164 }, { "epoch": 1.5214335910049193, "grad_norm": 0.6808415651321411, "learning_rate": 9.120168657765285e-06, "loss": 0.0571, "step": 2165 }, { "epoch": 1.5221363316936052, "grad_norm": 0.5027060508728027, "learning_rate": 9.1243851018974e-06, "loss": 0.0869, "step": 2166 }, { "epoch": 1.5228390723822909, "grad_norm": 0.6857293248176575, "learning_rate": 9.128601546029515e-06, "loss": 0.1028, "step": 2167 }, { "epoch": 1.5235418130709768, "grad_norm": 0.5452907085418701, "learning_rate": 9.13281799016163e-06, "loss": 0.0917, "step": 2168 }, { "epoch": 1.5242445537596627, "grad_norm": 0.6192089319229126, "learning_rate": 9.137034434293747e-06, "loss": 0.1521, "step": 2169 }, { "epoch": 1.5249472944483484, "grad_norm": 1.9168404340744019, "learning_rate": 9.141250878425862e-06, "loss": 0.2073, "step": 2170 }, { "epoch": 1.5256500351370343, "grad_norm": 0.9075291752815247, "learning_rate": 9.145467322557977e-06, "loss": 0.2581, "step": 2171 }, { "epoch": 1.5263527758257203, "grad_norm": 1.6204578876495361, "learning_rate": 9.149683766690092e-06, "loss": 0.3437, "step": 2172 }, { "epoch": 1.5270555165144062, "grad_norm": 2.852888822555542, "learning_rate": 9.153900210822207e-06, "loss": 0.4266, "step": 2173 }, { "epoch": 1.527758257203092, "grad_norm": 0.49970394372940063, "learning_rate": 9.158116654954322e-06, "loss": 0.1203, "step": 2174 }, { "epoch": 1.528460997891778, "grad_norm": 0.30500051379203796, "learning_rate": 9.162333099086437e-06, "loss": 0.0641, "step": 2175 }, { "epoch": 1.529163738580464, "grad_norm": 0.5080010294914246, "learning_rate": 9.166549543218553e-06, "loss": 0.0636, "step": 2176 }, { "epoch": 1.5298664792691496, "grad_norm": 0.4458600878715515, "learning_rate": 9.170765987350668e-06, "loss": 0.0507, "step": 2177 }, { "epoch": 1.5305692199578356, "grad_norm": 0.32133033871650696, "learning_rate": 9.174982431482783e-06, "loss": 0.0521, "step": 2178 }, { "epoch": 1.5312719606465213, "grad_norm": 0.479478657245636, "learning_rate": 9.179198875614898e-06, "loss": 0.044, "step": 2179 }, { "epoch": 1.5319747013352072, "grad_norm": 0.37278813123703003, "learning_rate": 9.183415319747013e-06, "loss": 0.0574, "step": 2180 }, { "epoch": 1.532677442023893, "grad_norm": 1.8330934047698975, "learning_rate": 9.187631763879128e-06, "loss": 0.0406, "step": 2181 }, { "epoch": 1.533380182712579, "grad_norm": 0.5089655518531799, "learning_rate": 9.191848208011243e-06, "loss": 0.0576, "step": 2182 }, { "epoch": 1.534082923401265, "grad_norm": 0.27563491463661194, "learning_rate": 9.19606465214336e-06, "loss": 0.0361, "step": 2183 }, { "epoch": 1.5347856640899509, "grad_norm": 0.4390771985054016, "learning_rate": 9.200281096275474e-06, "loss": 0.0693, "step": 2184 }, { "epoch": 1.5354884047786368, "grad_norm": 0.40847569704055786, "learning_rate": 9.20449754040759e-06, "loss": 0.047, "step": 2185 }, { "epoch": 1.5361911454673227, "grad_norm": 0.5500102639198303, "learning_rate": 9.208713984539704e-06, "loss": 0.0432, "step": 2186 }, { "epoch": 1.5368938861560084, "grad_norm": 0.40332937240600586, "learning_rate": 9.21293042867182e-06, "loss": 0.059, "step": 2187 }, { "epoch": 1.5375966268446943, "grad_norm": 0.4678916335105896, "learning_rate": 9.217146872803934e-06, "loss": 0.0622, "step": 2188 }, { "epoch": 1.53829936753338, "grad_norm": 0.8048183917999268, "learning_rate": 9.221363316936051e-06, "loss": 0.0706, "step": 2189 }, { "epoch": 1.539002108222066, "grad_norm": 0.4324797987937927, "learning_rate": 9.225579761068166e-06, "loss": 0.0455, "step": 2190 }, { "epoch": 1.5397048489107519, "grad_norm": 0.39914581179618835, "learning_rate": 9.229796205200282e-06, "loss": 0.0747, "step": 2191 }, { "epoch": 1.5404075895994378, "grad_norm": 0.38946494460105896, "learning_rate": 9.234012649332397e-06, "loss": 0.0638, "step": 2192 }, { "epoch": 1.5411103302881237, "grad_norm": 0.5820009112358093, "learning_rate": 9.238229093464512e-06, "loss": 0.0967, "step": 2193 }, { "epoch": 1.5418130709768096, "grad_norm": 0.7359590530395508, "learning_rate": 9.242445537596627e-06, "loss": 0.1416, "step": 2194 }, { "epoch": 1.5425158116654956, "grad_norm": 1.3679194450378418, "learning_rate": 9.246661981728742e-06, "loss": 0.2085, "step": 2195 }, { "epoch": 1.5432185523541813, "grad_norm": 2.5721635818481445, "learning_rate": 9.250878425860859e-06, "loss": 0.2526, "step": 2196 }, { "epoch": 1.5439212930428672, "grad_norm": 2.4111199378967285, "learning_rate": 9.255094869992974e-06, "loss": 0.3574, "step": 2197 }, { "epoch": 1.544624033731553, "grad_norm": 2.5203821659088135, "learning_rate": 9.259311314125089e-06, "loss": 0.3972, "step": 2198 }, { "epoch": 1.5453267744202388, "grad_norm": 0.432273805141449, "learning_rate": 9.263527758257204e-06, "loss": 0.1224, "step": 2199 }, { "epoch": 1.5460295151089247, "grad_norm": 0.3837645947933197, "learning_rate": 9.267744202389319e-06, "loss": 0.0664, "step": 2200 }, { "epoch": 1.5467322557976106, "grad_norm": 0.9327458143234253, "learning_rate": 9.271960646521434e-06, "loss": 0.0485, "step": 2201 }, { "epoch": 1.5474349964862966, "grad_norm": 0.323978990316391, "learning_rate": 9.276177090653549e-06, "loss": 0.0432, "step": 2202 }, { "epoch": 1.5481377371749825, "grad_norm": 0.27428919076919556, "learning_rate": 9.280393534785665e-06, "loss": 0.0424, "step": 2203 }, { "epoch": 1.5488404778636684, "grad_norm": 0.3257524371147156, "learning_rate": 9.28460997891778e-06, "loss": 0.0367, "step": 2204 }, { "epoch": 1.5495432185523543, "grad_norm": 0.25092631578445435, "learning_rate": 9.288826423049895e-06, "loss": 0.0274, "step": 2205 }, { "epoch": 1.55024595924104, "grad_norm": 0.3534005284309387, "learning_rate": 9.29304286718201e-06, "loss": 0.046, "step": 2206 }, { "epoch": 1.550948699929726, "grad_norm": 0.2819417715072632, "learning_rate": 9.297259311314125e-06, "loss": 0.0474, "step": 2207 }, { "epoch": 1.5516514406184116, "grad_norm": 0.26270222663879395, "learning_rate": 9.30147575544624e-06, "loss": 0.0277, "step": 2208 }, { "epoch": 1.5523541813070976, "grad_norm": 0.4513947665691376, "learning_rate": 9.305692199578355e-06, "loss": 0.0692, "step": 2209 }, { "epoch": 1.5530569219957835, "grad_norm": 0.3210453987121582, "learning_rate": 9.309908643710471e-06, "loss": 0.0297, "step": 2210 }, { "epoch": 1.5537596626844694, "grad_norm": 0.4956079125404358, "learning_rate": 9.314125087842586e-06, "loss": 0.0536, "step": 2211 }, { "epoch": 1.5544624033731553, "grad_norm": 0.27581772208213806, "learning_rate": 9.318341531974701e-06, "loss": 0.037, "step": 2212 }, { "epoch": 1.5551651440618413, "grad_norm": 0.4297178089618683, "learning_rate": 9.322557976106816e-06, "loss": 0.068, "step": 2213 }, { "epoch": 1.5558678847505272, "grad_norm": 0.6059157848358154, "learning_rate": 9.326774420238931e-06, "loss": 0.0615, "step": 2214 }, { "epoch": 1.556570625439213, "grad_norm": 0.2760978043079376, "learning_rate": 9.330990864371046e-06, "loss": 0.0367, "step": 2215 }, { "epoch": 1.5572733661278988, "grad_norm": 0.5047537684440613, "learning_rate": 9.335207308503161e-06, "loss": 0.0791, "step": 2216 }, { "epoch": 1.5579761068165847, "grad_norm": 0.48780956864356995, "learning_rate": 9.339423752635278e-06, "loss": 0.0804, "step": 2217 }, { "epoch": 1.5586788475052704, "grad_norm": 1.4314321279525757, "learning_rate": 9.343640196767393e-06, "loss": 0.1224, "step": 2218 }, { "epoch": 1.5593815881939563, "grad_norm": 0.7894369959831238, "learning_rate": 9.347856640899508e-06, "loss": 0.1148, "step": 2219 }, { "epoch": 1.5600843288826423, "grad_norm": 0.8547379374504089, "learning_rate": 9.352073085031624e-06, "loss": 0.1928, "step": 2220 }, { "epoch": 1.5607870695713282, "grad_norm": 3.673957109451294, "learning_rate": 9.35628952916374e-06, "loss": 0.251, "step": 2221 }, { "epoch": 1.561489810260014, "grad_norm": 2.7004830837249756, "learning_rate": 9.360505973295854e-06, "loss": 0.3309, "step": 2222 }, { "epoch": 1.5621925509487, "grad_norm": 2.1498124599456787, "learning_rate": 9.36472241742797e-06, "loss": 0.3767, "step": 2223 }, { "epoch": 1.562895291637386, "grad_norm": 0.8746214509010315, "learning_rate": 9.368938861560086e-06, "loss": 0.1483, "step": 2224 }, { "epoch": 1.5635980323260716, "grad_norm": 0.3066241443157196, "learning_rate": 9.3731553056922e-06, "loss": 0.0454, "step": 2225 }, { "epoch": 1.5643007730147576, "grad_norm": 0.3306412994861603, "learning_rate": 9.377371749824316e-06, "loss": 0.0505, "step": 2226 }, { "epoch": 1.5650035137034435, "grad_norm": 0.35988563299179077, "learning_rate": 9.38158819395643e-06, "loss": 0.0424, "step": 2227 }, { "epoch": 1.5657062543921292, "grad_norm": 0.27263015508651733, "learning_rate": 9.385804638088546e-06, "loss": 0.0291, "step": 2228 }, { "epoch": 1.566408995080815, "grad_norm": 0.3241138458251953, "learning_rate": 9.39002108222066e-06, "loss": 0.0333, "step": 2229 }, { "epoch": 1.567111735769501, "grad_norm": 0.3760855495929718, "learning_rate": 9.394237526352777e-06, "loss": 0.0468, "step": 2230 }, { "epoch": 1.567814476458187, "grad_norm": 0.3457091450691223, "learning_rate": 9.398453970484892e-06, "loss": 0.0542, "step": 2231 }, { "epoch": 1.5685172171468729, "grad_norm": 0.3419874310493469, "learning_rate": 9.402670414617007e-06, "loss": 0.0383, "step": 2232 }, { "epoch": 1.5692199578355588, "grad_norm": 0.5195604562759399, "learning_rate": 9.406886858749122e-06, "loss": 0.0399, "step": 2233 }, { "epoch": 1.5699226985242447, "grad_norm": 0.616249680519104, "learning_rate": 9.411103302881237e-06, "loss": 0.065, "step": 2234 }, { "epoch": 1.5706254392129304, "grad_norm": 0.28111618757247925, "learning_rate": 9.415319747013352e-06, "loss": 0.0362, "step": 2235 }, { "epoch": 1.5713281799016163, "grad_norm": 0.456143856048584, "learning_rate": 9.419536191145467e-06, "loss": 0.0562, "step": 2236 }, { "epoch": 1.572030920590302, "grad_norm": 0.7108597755432129, "learning_rate": 9.423752635277583e-06, "loss": 0.0447, "step": 2237 }, { "epoch": 1.572733661278988, "grad_norm": 0.5260196924209595, "learning_rate": 9.427969079409698e-06, "loss": 0.0622, "step": 2238 }, { "epoch": 1.5734364019676739, "grad_norm": 0.6279994249343872, "learning_rate": 9.432185523541813e-06, "loss": 0.0787, "step": 2239 }, { "epoch": 1.5741391426563598, "grad_norm": 0.34897977113723755, "learning_rate": 9.436401967673928e-06, "loss": 0.0626, "step": 2240 }, { "epoch": 1.5748418833450457, "grad_norm": 0.5488333106040955, "learning_rate": 9.440618411806043e-06, "loss": 0.073, "step": 2241 }, { "epoch": 1.5755446240337316, "grad_norm": 0.6671362519264221, "learning_rate": 9.444834855938158e-06, "loss": 0.1096, "step": 2242 }, { "epoch": 1.5762473647224176, "grad_norm": 0.4169994294643402, "learning_rate": 9.449051300070273e-06, "loss": 0.0817, "step": 2243 }, { "epoch": 1.5769501054111033, "grad_norm": 0.9853678345680237, "learning_rate": 9.45326774420239e-06, "loss": 0.1248, "step": 2244 }, { "epoch": 1.5776528460997892, "grad_norm": 1.1433223485946655, "learning_rate": 9.457484188334505e-06, "loss": 0.2072, "step": 2245 }, { "epoch": 1.578355586788475, "grad_norm": 1.0319116115570068, "learning_rate": 9.46170063246662e-06, "loss": 0.2598, "step": 2246 }, { "epoch": 1.5790583274771608, "grad_norm": 1.941306471824646, "learning_rate": 9.465917076598735e-06, "loss": 0.3701, "step": 2247 }, { "epoch": 1.5797610681658467, "grad_norm": 3.363394260406494, "learning_rate": 9.47013352073085e-06, "loss": 0.431, "step": 2248 }, { "epoch": 1.5804638088545326, "grad_norm": 0.6459230184555054, "learning_rate": 9.474349964862964e-06, "loss": 0.1123, "step": 2249 }, { "epoch": 1.5811665495432186, "grad_norm": 0.31479352712631226, "learning_rate": 9.478566408995081e-06, "loss": 0.0706, "step": 2250 }, { "epoch": 1.5818692902319045, "grad_norm": 0.42535996437072754, "learning_rate": 9.482782853127198e-06, "loss": 0.0403, "step": 2251 }, { "epoch": 1.5825720309205904, "grad_norm": 0.3386077284812927, "learning_rate": 9.486999297259313e-06, "loss": 0.0507, "step": 2252 }, { "epoch": 1.5832747716092763, "grad_norm": 0.33816638588905334, "learning_rate": 9.491215741391428e-06, "loss": 0.0408, "step": 2253 }, { "epoch": 1.583977512297962, "grad_norm": 0.3666364252567291, "learning_rate": 9.495432185523543e-06, "loss": 0.0427, "step": 2254 }, { "epoch": 1.584680252986648, "grad_norm": 0.29750221967697144, "learning_rate": 9.499648629655657e-06, "loss": 0.0522, "step": 2255 }, { "epoch": 1.5853829936753336, "grad_norm": 0.28753042221069336, "learning_rate": 9.503865073787772e-06, "loss": 0.0481, "step": 2256 }, { "epoch": 1.5860857343640196, "grad_norm": 0.2929779589176178, "learning_rate": 9.508081517919889e-06, "loss": 0.0601, "step": 2257 }, { "epoch": 1.5867884750527055, "grad_norm": 0.36397650837898254, "learning_rate": 9.512297962052004e-06, "loss": 0.0514, "step": 2258 }, { "epoch": 1.5874912157413914, "grad_norm": 0.5520057678222656, "learning_rate": 9.516514406184119e-06, "loss": 0.0438, "step": 2259 }, { "epoch": 1.5881939564300773, "grad_norm": 0.29571104049682617, "learning_rate": 9.520730850316234e-06, "loss": 0.0474, "step": 2260 }, { "epoch": 1.5888966971187632, "grad_norm": 0.34707286953926086, "learning_rate": 9.524947294448349e-06, "loss": 0.0496, "step": 2261 }, { "epoch": 1.5895994378074492, "grad_norm": 0.39686569571495056, "learning_rate": 9.529163738580464e-06, "loss": 0.0347, "step": 2262 }, { "epoch": 1.590302178496135, "grad_norm": 0.4931889474391937, "learning_rate": 9.533380182712579e-06, "loss": 0.0404, "step": 2263 }, { "epoch": 1.5910049191848208, "grad_norm": 0.4768810272216797, "learning_rate": 9.537596626844695e-06, "loss": 0.0792, "step": 2264 }, { "epoch": 1.5917076598735067, "grad_norm": 0.3598458170890808, "learning_rate": 9.54181307097681e-06, "loss": 0.0439, "step": 2265 }, { "epoch": 1.5924104005621924, "grad_norm": 0.46644556522369385, "learning_rate": 9.546029515108925e-06, "loss": 0.0858, "step": 2266 }, { "epoch": 1.5931131412508783, "grad_norm": 0.49387937784194946, "learning_rate": 9.55024595924104e-06, "loss": 0.0914, "step": 2267 }, { "epoch": 1.5938158819395642, "grad_norm": 0.4644274115562439, "learning_rate": 9.554462403373155e-06, "loss": 0.0796, "step": 2268 }, { "epoch": 1.5945186226282502, "grad_norm": 0.5547526478767395, "learning_rate": 9.55867884750527e-06, "loss": 0.13, "step": 2269 }, { "epoch": 1.595221363316936, "grad_norm": 0.8215802311897278, "learning_rate": 9.562895291637385e-06, "loss": 0.224, "step": 2270 }, { "epoch": 1.595924104005622, "grad_norm": 1.0428423881530762, "learning_rate": 9.567111735769502e-06, "loss": 0.2628, "step": 2271 }, { "epoch": 1.596626844694308, "grad_norm": 1.4070113897323608, "learning_rate": 9.571328179901617e-06, "loss": 0.3513, "step": 2272 }, { "epoch": 1.5973295853829936, "grad_norm": 1.9876995086669922, "learning_rate": 9.575544624033732e-06, "loss": 0.4591, "step": 2273 }, { "epoch": 1.5980323260716796, "grad_norm": 0.49623119831085205, "learning_rate": 9.579761068165846e-06, "loss": 0.1422, "step": 2274 }, { "epoch": 1.5987350667603655, "grad_norm": 0.3660385012626648, "learning_rate": 9.583977512297961e-06, "loss": 0.0521, "step": 2275 }, { "epoch": 1.5994378074490512, "grad_norm": 0.26432541012763977, "learning_rate": 9.588193956430076e-06, "loss": 0.0417, "step": 2276 }, { "epoch": 1.600140548137737, "grad_norm": 0.2718830406665802, "learning_rate": 9.592410400562191e-06, "loss": 0.0333, "step": 2277 }, { "epoch": 1.600843288826423, "grad_norm": 0.3729459345340729, "learning_rate": 9.596626844694308e-06, "loss": 0.0571, "step": 2278 }, { "epoch": 1.601546029515109, "grad_norm": 0.25856679677963257, "learning_rate": 9.600843288826423e-06, "loss": 0.0305, "step": 2279 }, { "epoch": 1.6022487702037949, "grad_norm": 0.2828661799430847, "learning_rate": 9.60505973295854e-06, "loss": 0.0472, "step": 2280 }, { "epoch": 1.6029515108924808, "grad_norm": 0.3034546375274658, "learning_rate": 9.609276177090654e-06, "loss": 0.0478, "step": 2281 }, { "epoch": 1.6036542515811667, "grad_norm": 0.29609209299087524, "learning_rate": 9.61349262122277e-06, "loss": 0.0569, "step": 2282 }, { "epoch": 1.6043569922698524, "grad_norm": 0.3191763162612915, "learning_rate": 9.617709065354884e-06, "loss": 0.0376, "step": 2283 }, { "epoch": 1.6050597329585383, "grad_norm": 0.3342810273170471, "learning_rate": 9.621925509487e-06, "loss": 0.063, "step": 2284 }, { "epoch": 1.605762473647224, "grad_norm": 0.27293860912323, "learning_rate": 9.626141953619116e-06, "loss": 0.0315, "step": 2285 }, { "epoch": 1.60646521433591, "grad_norm": 0.3787219822406769, "learning_rate": 9.630358397751231e-06, "loss": 0.0658, "step": 2286 }, { "epoch": 1.6071679550245959, "grad_norm": 0.32614070177078247, "learning_rate": 9.634574841883346e-06, "loss": 0.0358, "step": 2287 }, { "epoch": 1.6078706957132818, "grad_norm": 0.517367959022522, "learning_rate": 9.63879128601546e-06, "loss": 0.0528, "step": 2288 }, { "epoch": 1.6085734364019677, "grad_norm": 0.4944280982017517, "learning_rate": 9.643007730147576e-06, "loss": 0.0639, "step": 2289 }, { "epoch": 1.6092761770906536, "grad_norm": 0.4098902642726898, "learning_rate": 9.64722417427969e-06, "loss": 0.0472, "step": 2290 }, { "epoch": 1.6099789177793395, "grad_norm": 0.5256981253623962, "learning_rate": 9.651440618411807e-06, "loss": 0.0917, "step": 2291 }, { "epoch": 1.6106816584680252, "grad_norm": 1.1634222269058228, "learning_rate": 9.655657062543922e-06, "loss": 0.1168, "step": 2292 }, { "epoch": 1.6113843991567112, "grad_norm": 0.5505526661872864, "learning_rate": 9.659873506676037e-06, "loss": 0.0785, "step": 2293 }, { "epoch": 1.612087139845397, "grad_norm": 0.7068050503730774, "learning_rate": 9.664089950808152e-06, "loss": 0.1693, "step": 2294 }, { "epoch": 1.6127898805340828, "grad_norm": 1.617652177810669, "learning_rate": 9.668306394940267e-06, "loss": 0.2048, "step": 2295 }, { "epoch": 1.6134926212227687, "grad_norm": 1.724687933921814, "learning_rate": 9.672522839072382e-06, "loss": 0.2632, "step": 2296 }, { "epoch": 1.6141953619114546, "grad_norm": 1.4271513223648071, "learning_rate": 9.676739283204497e-06, "loss": 0.3518, "step": 2297 }, { "epoch": 1.6148981026001406, "grad_norm": 8.949830055236816, "learning_rate": 9.680955727336614e-06, "loss": 0.4569, "step": 2298 }, { "epoch": 1.6156008432888265, "grad_norm": 1.1306602954864502, "learning_rate": 9.685172171468729e-06, "loss": 0.1133, "step": 2299 }, { "epoch": 1.6163035839775124, "grad_norm": 0.32704201340675354, "learning_rate": 9.689388615600843e-06, "loss": 0.0388, "step": 2300 }, { "epoch": 1.6170063246661983, "grad_norm": 0.2437024861574173, "learning_rate": 9.693605059732958e-06, "loss": 0.0477, "step": 2301 }, { "epoch": 1.617709065354884, "grad_norm": 0.39245596528053284, "learning_rate": 9.697821503865073e-06, "loss": 0.0402, "step": 2302 }, { "epoch": 1.61841180604357, "grad_norm": 0.30389904975891113, "learning_rate": 9.702037947997188e-06, "loss": 0.0526, "step": 2303 }, { "epoch": 1.6191145467322556, "grad_norm": 0.2878871560096741, "learning_rate": 9.706254392129303e-06, "loss": 0.038, "step": 2304 }, { "epoch": 1.6198172874209416, "grad_norm": 0.24491877853870392, "learning_rate": 9.71047083626142e-06, "loss": 0.0292, "step": 2305 }, { "epoch": 1.6205200281096275, "grad_norm": 0.30778032541275024, "learning_rate": 9.714687280393535e-06, "loss": 0.0489, "step": 2306 }, { "epoch": 1.6212227687983134, "grad_norm": 0.29773733019828796, "learning_rate": 9.71890372452565e-06, "loss": 0.0353, "step": 2307 }, { "epoch": 1.6219255094869993, "grad_norm": 0.4294052720069885, "learning_rate": 9.723120168657765e-06, "loss": 0.0602, "step": 2308 }, { "epoch": 1.6226282501756852, "grad_norm": 0.34495335817337036, "learning_rate": 9.72733661278988e-06, "loss": 0.0514, "step": 2309 }, { "epoch": 1.6233309908643712, "grad_norm": 0.3215411901473999, "learning_rate": 9.731553056921996e-06, "loss": 0.0348, "step": 2310 }, { "epoch": 1.624033731553057, "grad_norm": 0.3282582461833954, "learning_rate": 9.735769501054111e-06, "loss": 0.0476, "step": 2311 }, { "epoch": 1.6247364722417428, "grad_norm": 0.2733859121799469, "learning_rate": 9.739985945186228e-06, "loss": 0.0311, "step": 2312 }, { "epoch": 1.6254392129304287, "grad_norm": 0.3500750660896301, "learning_rate": 9.744202389318343e-06, "loss": 0.0486, "step": 2313 }, { "epoch": 1.6261419536191144, "grad_norm": 0.33439770340919495, "learning_rate": 9.748418833450458e-06, "loss": 0.0741, "step": 2314 }, { "epoch": 1.6268446943078003, "grad_norm": 0.4434448778629303, "learning_rate": 9.752635277582573e-06, "loss": 0.0538, "step": 2315 }, { "epoch": 1.6275474349964862, "grad_norm": 0.41478675603866577, "learning_rate": 9.756851721714688e-06, "loss": 0.0714, "step": 2316 }, { "epoch": 1.6282501756851722, "grad_norm": 0.5734182000160217, "learning_rate": 9.761068165846803e-06, "loss": 0.1101, "step": 2317 }, { "epoch": 1.628952916373858, "grad_norm": 0.46348902583122253, "learning_rate": 9.765284609978918e-06, "loss": 0.1024, "step": 2318 }, { "epoch": 1.629655657062544, "grad_norm": 0.6244013905525208, "learning_rate": 9.769501054111034e-06, "loss": 0.1451, "step": 2319 }, { "epoch": 1.63035839775123, "grad_norm": 0.8217160701751709, "learning_rate": 9.773717498243149e-06, "loss": 0.1693, "step": 2320 }, { "epoch": 1.6310611384399156, "grad_norm": 1.1300098896026611, "learning_rate": 9.777933942375264e-06, "loss": 0.3049, "step": 2321 }, { "epoch": 1.6317638791286015, "grad_norm": 1.7022916078567505, "learning_rate": 9.782150386507379e-06, "loss": 0.357, "step": 2322 }, { "epoch": 1.6324666198172875, "grad_norm": 2.7635691165924072, "learning_rate": 9.786366830639494e-06, "loss": 0.4174, "step": 2323 }, { "epoch": 1.6331693605059732, "grad_norm": 0.3550402522087097, "learning_rate": 9.790583274771609e-06, "loss": 0.1202, "step": 2324 }, { "epoch": 1.633872101194659, "grad_norm": 0.3362025320529938, "learning_rate": 9.794799718903726e-06, "loss": 0.0532, "step": 2325 }, { "epoch": 1.634574841883345, "grad_norm": 0.7163921594619751, "learning_rate": 9.79901616303584e-06, "loss": 0.0741, "step": 2326 }, { "epoch": 1.635277582572031, "grad_norm": 0.23763345181941986, "learning_rate": 9.803232607167955e-06, "loss": 0.0346, "step": 2327 }, { "epoch": 1.6359803232607169, "grad_norm": 0.26402711868286133, "learning_rate": 9.80744905130007e-06, "loss": 0.0395, "step": 2328 }, { "epoch": 1.6366830639494028, "grad_norm": 0.2895848751068115, "learning_rate": 9.811665495432185e-06, "loss": 0.0325, "step": 2329 }, { "epoch": 1.6373858046380887, "grad_norm": 0.3143823444843292, "learning_rate": 9.8158819395643e-06, "loss": 0.0367, "step": 2330 }, { "epoch": 1.6380885453267744, "grad_norm": 0.33342620730400085, "learning_rate": 9.820098383696415e-06, "loss": 0.0509, "step": 2331 }, { "epoch": 1.6387912860154603, "grad_norm": 0.233085036277771, "learning_rate": 9.824314827828532e-06, "loss": 0.0351, "step": 2332 }, { "epoch": 1.639494026704146, "grad_norm": 0.27923575043678284, "learning_rate": 9.828531271960647e-06, "loss": 0.0345, "step": 2333 }, { "epoch": 1.640196767392832, "grad_norm": 0.3363879919052124, "learning_rate": 9.832747716092762e-06, "loss": 0.0503, "step": 2334 }, { "epoch": 1.6408995080815179, "grad_norm": 0.3173399567604065, "learning_rate": 9.836964160224877e-06, "loss": 0.0389, "step": 2335 }, { "epoch": 1.6416022487702038, "grad_norm": 0.4623333215713501, "learning_rate": 9.841180604356992e-06, "loss": 0.0808, "step": 2336 }, { "epoch": 1.6423049894588897, "grad_norm": 0.3285117745399475, "learning_rate": 9.845397048489107e-06, "loss": 0.0358, "step": 2337 }, { "epoch": 1.6430077301475756, "grad_norm": 0.48931437730789185, "learning_rate": 9.849613492621222e-06, "loss": 0.0654, "step": 2338 }, { "epoch": 1.6437104708362615, "grad_norm": 0.3704455494880676, "learning_rate": 9.85382993675334e-06, "loss": 0.0606, "step": 2339 }, { "epoch": 1.6444132115249475, "grad_norm": 0.30026212334632874, "learning_rate": 9.858046380885455e-06, "loss": 0.0444, "step": 2340 }, { "epoch": 1.6451159522136332, "grad_norm": 0.7778007388114929, "learning_rate": 9.86226282501757e-06, "loss": 0.0861, "step": 2341 }, { "epoch": 1.645818692902319, "grad_norm": 0.5129053592681885, "learning_rate": 9.866479269149685e-06, "loss": 0.0652, "step": 2342 }, { "epoch": 1.6465214335910048, "grad_norm": 0.5133492350578308, "learning_rate": 9.8706957132818e-06, "loss": 0.1004, "step": 2343 }, { "epoch": 1.6472241742796907, "grad_norm": 0.5785870552062988, "learning_rate": 9.874912157413915e-06, "loss": 0.1712, "step": 2344 }, { "epoch": 1.6479269149683766, "grad_norm": 0.8429995775222778, "learning_rate": 9.87912860154603e-06, "loss": 0.2012, "step": 2345 }, { "epoch": 1.6486296556570625, "grad_norm": 1.1904630661010742, "learning_rate": 9.883345045678146e-06, "loss": 0.2443, "step": 2346 }, { "epoch": 1.6493323963457485, "grad_norm": 1.449828863143921, "learning_rate": 9.887561489810261e-06, "loss": 0.3231, "step": 2347 }, { "epoch": 1.6500351370344344, "grad_norm": 2.0776867866516113, "learning_rate": 9.891777933942376e-06, "loss": 0.3762, "step": 2348 }, { "epoch": 1.6507378777231203, "grad_norm": 1.4547040462493896, "learning_rate": 9.895994378074491e-06, "loss": 0.1398, "step": 2349 }, { "epoch": 1.651440618411806, "grad_norm": 0.2808661460876465, "learning_rate": 9.900210822206606e-06, "loss": 0.0453, "step": 2350 }, { "epoch": 1.652143359100492, "grad_norm": 0.3266894519329071, "learning_rate": 9.904427266338721e-06, "loss": 0.0539, "step": 2351 }, { "epoch": 1.6528460997891778, "grad_norm": 0.2754737138748169, "learning_rate": 9.908643710470836e-06, "loss": 0.0341, "step": 2352 }, { "epoch": 1.6535488404778635, "grad_norm": 0.3195238709449768, "learning_rate": 9.912860154602952e-06, "loss": 0.0304, "step": 2353 }, { "epoch": 1.6542515811665495, "grad_norm": 0.2696624994277954, "learning_rate": 9.917076598735067e-06, "loss": 0.0306, "step": 2354 }, { "epoch": 1.6549543218552354, "grad_norm": 0.27993547916412354, "learning_rate": 9.921293042867182e-06, "loss": 0.0496, "step": 2355 }, { "epoch": 1.6556570625439213, "grad_norm": 0.28418102860450745, "learning_rate": 9.925509486999297e-06, "loss": 0.0251, "step": 2356 }, { "epoch": 1.6563598032326072, "grad_norm": 1.635558009147644, "learning_rate": 9.929725931131412e-06, "loss": 0.0737, "step": 2357 }, { "epoch": 1.6570625439212932, "grad_norm": 0.4060306251049042, "learning_rate": 9.933942375263527e-06, "loss": 0.0516, "step": 2358 }, { "epoch": 1.657765284609979, "grad_norm": 0.33933642506599426, "learning_rate": 9.938158819395644e-06, "loss": 0.0386, "step": 2359 }, { "epoch": 1.6584680252986648, "grad_norm": 0.42188647389411926, "learning_rate": 9.942375263527759e-06, "loss": 0.0504, "step": 2360 }, { "epoch": 1.6591707659873507, "grad_norm": 0.28463613986968994, "learning_rate": 9.946591707659874e-06, "loss": 0.0571, "step": 2361 }, { "epoch": 1.6598735066760364, "grad_norm": 0.32127270102500916, "learning_rate": 9.950808151791989e-06, "loss": 0.0397, "step": 2362 }, { "epoch": 1.6605762473647223, "grad_norm": 0.6323956847190857, "learning_rate": 9.955024595924104e-06, "loss": 0.0901, "step": 2363 }, { "epoch": 1.6612789880534082, "grad_norm": 0.4140136241912842, "learning_rate": 9.959241040056219e-06, "loss": 0.0654, "step": 2364 }, { "epoch": 1.6619817287420942, "grad_norm": 0.5065699219703674, "learning_rate": 9.963457484188333e-06, "loss": 0.0504, "step": 2365 }, { "epoch": 1.66268446943078, "grad_norm": 0.6936134696006775, "learning_rate": 9.96767392832045e-06, "loss": 0.0873, "step": 2366 }, { "epoch": 1.663387210119466, "grad_norm": 0.990585207939148, "learning_rate": 9.971890372452565e-06, "loss": 0.0941, "step": 2367 }, { "epoch": 1.664089950808152, "grad_norm": 0.8099033236503601, "learning_rate": 9.97610681658468e-06, "loss": 0.0872, "step": 2368 }, { "epoch": 1.6647926914968376, "grad_norm": 0.668873131275177, "learning_rate": 9.980323260716797e-06, "loss": 0.1361, "step": 2369 }, { "epoch": 1.6654954321855235, "grad_norm": 0.8701562881469727, "learning_rate": 9.984539704848912e-06, "loss": 0.1897, "step": 2370 }, { "epoch": 1.6661981728742095, "grad_norm": 1.1482300758361816, "learning_rate": 9.988756148981027e-06, "loss": 0.2472, "step": 2371 }, { "epoch": 1.6669009135628952, "grad_norm": 2.1140034198760986, "learning_rate": 9.992972593113141e-06, "loss": 0.3394, "step": 2372 }, { "epoch": 1.667603654251581, "grad_norm": NaN, "learning_rate": 9.992972593113141e-06, "loss": 0.4715, "step": 2373 }, { "epoch": 1.668306394940267, "grad_norm": 0.32306596636772156, "learning_rate": 9.997189037245258e-06, "loss": 0.0862, "step": 2374 }, { "epoch": 1.669009135628953, "grad_norm": 0.31155267357826233, "learning_rate": 1.0001405481377373e-05, "loss": 0.0675, "step": 2375 }, { "epoch": 1.6697118763176388, "grad_norm": 0.29579734802246094, "learning_rate": 1.0005621925509488e-05, "loss": 0.0378, "step": 2376 }, { "epoch": 1.6704146170063248, "grad_norm": 0.3139916956424713, "learning_rate": 1.0009838369641603e-05, "loss": 0.0355, "step": 2377 }, { "epoch": 1.6711173576950107, "grad_norm": 0.33311033248901367, "learning_rate": 1.0014054813773718e-05, "loss": 0.0354, "step": 2378 }, { "epoch": 1.6718200983836964, "grad_norm": 0.28172627091407776, "learning_rate": 1.0018271257905833e-05, "loss": 0.0236, "step": 2379 }, { "epoch": 1.6725228390723823, "grad_norm": 0.3273710608482361, "learning_rate": 1.0022487702037948e-05, "loss": 0.0287, "step": 2380 }, { "epoch": 1.673225579761068, "grad_norm": 0.37985819578170776, "learning_rate": 1.0026704146170064e-05, "loss": 0.0404, "step": 2381 }, { "epoch": 1.673928320449754, "grad_norm": 0.33835679292678833, "learning_rate": 1.003092059030218e-05, "loss": 0.0535, "step": 2382 }, { "epoch": 1.6746310611384398, "grad_norm": 0.3759658634662628, "learning_rate": 1.0035137034434294e-05, "loss": 0.0323, "step": 2383 }, { "epoch": 1.6753338018271258, "grad_norm": 0.5387056469917297, "learning_rate": 1.003935347856641e-05, "loss": 0.0377, "step": 2384 }, { "epoch": 1.6760365425158117, "grad_norm": 0.3007372319698334, "learning_rate": 1.0043569922698524e-05, "loss": 0.0419, "step": 2385 }, { "epoch": 1.6767392832044976, "grad_norm": 0.4655447006225586, "learning_rate": 1.0047786366830639e-05, "loss": 0.0752, "step": 2386 }, { "epoch": 1.6774420238931835, "grad_norm": 0.34377720952033997, "learning_rate": 1.0052002810962754e-05, "loss": 0.0372, "step": 2387 }, { "epoch": 1.6781447645818695, "grad_norm": 0.47116100788116455, "learning_rate": 1.005621925509487e-05, "loss": 0.093, "step": 2388 }, { "epoch": 1.6788475052705552, "grad_norm": 0.47002699971199036, "learning_rate": 1.0060435699226986e-05, "loss": 0.0768, "step": 2389 }, { "epoch": 1.679550245959241, "grad_norm": 0.38646939396858215, "learning_rate": 1.00646521433591e-05, "loss": 0.0551, "step": 2390 }, { "epoch": 1.6802529866479268, "grad_norm": 0.457620233297348, "learning_rate": 1.0068868587491216e-05, "loss": 0.074, "step": 2391 }, { "epoch": 1.6809557273366127, "grad_norm": 0.5724887251853943, "learning_rate": 1.007308503162333e-05, "loss": 0.0745, "step": 2392 }, { "epoch": 1.6816584680252986, "grad_norm": 0.6311786770820618, "learning_rate": 1.0077301475755445e-05, "loss": 0.0843, "step": 2393 }, { "epoch": 1.6823612087139845, "grad_norm": 0.6112505197525024, "learning_rate": 1.0081517919887562e-05, "loss": 0.1138, "step": 2394 }, { "epoch": 1.6830639494026705, "grad_norm": 1.0983120203018188, "learning_rate": 1.0085734364019677e-05, "loss": 0.1965, "step": 2395 }, { "epoch": 1.6837666900913564, "grad_norm": 1.3785783052444458, "learning_rate": 1.0089950808151792e-05, "loss": 0.2386, "step": 2396 }, { "epoch": 1.6844694307800423, "grad_norm": 1.3016537427902222, "learning_rate": 1.0094167252283907e-05, "loss": 0.3416, "step": 2397 }, { "epoch": 1.685172171468728, "grad_norm": 4.0396809577941895, "learning_rate": 1.0098383696416022e-05, "loss": 0.4487, "step": 2398 }, { "epoch": 1.685874912157414, "grad_norm": 0.48933789134025574, "learning_rate": 1.0102600140548137e-05, "loss": 0.0792, "step": 2399 }, { "epoch": 1.6865776528460998, "grad_norm": 0.358649879693985, "learning_rate": 1.0106816584680253e-05, "loss": 0.0595, "step": 2400 }, { "epoch": 1.6872803935347855, "grad_norm": 0.2587299346923828, "learning_rate": 1.011103302881237e-05, "loss": 0.0417, "step": 2401 }, { "epoch": 1.6879831342234715, "grad_norm": 0.29219475388526917, "learning_rate": 1.0115249472944485e-05, "loss": 0.0476, "step": 2402 }, { "epoch": 1.6886858749121574, "grad_norm": 0.3765694797039032, "learning_rate": 1.01194659170766e-05, "loss": 0.0593, "step": 2403 }, { "epoch": 1.6893886156008433, "grad_norm": 0.30684107542037964, "learning_rate": 1.0123682361208715e-05, "loss": 0.0348, "step": 2404 }, { "epoch": 1.6900913562895292, "grad_norm": 0.30698472261428833, "learning_rate": 1.012789880534083e-05, "loss": 0.0366, "step": 2405 }, { "epoch": 1.6907940969782151, "grad_norm": 0.4952695369720459, "learning_rate": 1.0132115249472945e-05, "loss": 0.0368, "step": 2406 }, { "epoch": 1.691496837666901, "grad_norm": 0.3110910952091217, "learning_rate": 1.013633169360506e-05, "loss": 0.0563, "step": 2407 }, { "epoch": 1.6921995783555868, "grad_norm": 0.3653528690338135, "learning_rate": 1.0140548137737176e-05, "loss": 0.0273, "step": 2408 }, { "epoch": 1.6929023190442727, "grad_norm": 0.4210033416748047, "learning_rate": 1.0144764581869291e-05, "loss": 0.0502, "step": 2409 }, { "epoch": 1.6936050597329584, "grad_norm": 0.3352878987789154, "learning_rate": 1.0148981026001406e-05, "loss": 0.027, "step": 2410 }, { "epoch": 1.6943078004216443, "grad_norm": 0.3951065242290497, "learning_rate": 1.0153197470133521e-05, "loss": 0.054, "step": 2411 }, { "epoch": 1.6950105411103302, "grad_norm": 0.3412156403064728, "learning_rate": 1.0157413914265636e-05, "loss": 0.0596, "step": 2412 }, { "epoch": 1.6957132817990161, "grad_norm": 0.461490660905838, "learning_rate": 1.0161630358397751e-05, "loss": 0.0592, "step": 2413 }, { "epoch": 1.696416022487702, "grad_norm": 0.4040648937225342, "learning_rate": 1.0165846802529866e-05, "loss": 0.0697, "step": 2414 }, { "epoch": 1.697118763176388, "grad_norm": 0.3458699882030487, "learning_rate": 1.0170063246661983e-05, "loss": 0.0475, "step": 2415 }, { "epoch": 1.697821503865074, "grad_norm": 0.5576385855674744, "learning_rate": 1.0174279690794098e-05, "loss": 0.0584, "step": 2416 }, { "epoch": 1.6985242445537596, "grad_norm": 0.4183370769023895, "learning_rate": 1.0178496134926213e-05, "loss": 0.1021, "step": 2417 }, { "epoch": 1.6992269852424455, "grad_norm": 0.7516841888427734, "learning_rate": 1.0182712579058327e-05, "loss": 0.0698, "step": 2418 }, { "epoch": 1.6999297259311315, "grad_norm": 0.8438791632652283, "learning_rate": 1.0186929023190442e-05, "loss": 0.1182, "step": 2419 }, { "epoch": 1.7006324666198172, "grad_norm": 0.9985702037811279, "learning_rate": 1.0191145467322557e-05, "loss": 0.2073, "step": 2420 }, { "epoch": 1.701335207308503, "grad_norm": 1.0349565744400024, "learning_rate": 1.0195361911454672e-05, "loss": 0.2968, "step": 2421 }, { "epoch": 1.702037947997189, "grad_norm": 2.1582212448120117, "learning_rate": 1.0199578355586789e-05, "loss": 0.3322, "step": 2422 }, { "epoch": 1.702740688685875, "grad_norm": 3.7225656509399414, "learning_rate": 1.0203794799718904e-05, "loss": 0.4069, "step": 2423 }, { "epoch": 1.7034434293745608, "grad_norm": 0.8318036198616028, "learning_rate": 1.0208011243851019e-05, "loss": 0.1752, "step": 2424 }, { "epoch": 1.7041461700632468, "grad_norm": 0.3504549264907837, "learning_rate": 1.0212227687983134e-05, "loss": 0.0571, "step": 2425 }, { "epoch": 1.7048489107519327, "grad_norm": 0.2728981673717499, "learning_rate": 1.0216444132115249e-05, "loss": 0.0445, "step": 2426 }, { "epoch": 1.7055516514406184, "grad_norm": 0.3572444021701813, "learning_rate": 1.0220660576247364e-05, "loss": 0.0353, "step": 2427 }, { "epoch": 1.7062543921293043, "grad_norm": 0.2793242633342743, "learning_rate": 1.022487702037948e-05, "loss": 0.0425, "step": 2428 }, { "epoch": 1.70695713281799, "grad_norm": 0.3179875612258911, "learning_rate": 1.0229093464511595e-05, "loss": 0.0402, "step": 2429 }, { "epoch": 1.707659873506676, "grad_norm": 0.32603558897972107, "learning_rate": 1.0233309908643712e-05, "loss": 0.0582, "step": 2430 }, { "epoch": 1.7083626141953618, "grad_norm": 0.3556661307811737, "learning_rate": 1.0237526352775827e-05, "loss": 0.0615, "step": 2431 }, { "epoch": 1.7090653548840478, "grad_norm": 0.3346932828426361, "learning_rate": 1.0241742796907942e-05, "loss": 0.0413, "step": 2432 }, { "epoch": 1.7097680955727337, "grad_norm": 0.3400801122188568, "learning_rate": 1.0245959241040057e-05, "loss": 0.0461, "step": 2433 }, { "epoch": 1.7104708362614196, "grad_norm": 0.3480308949947357, "learning_rate": 1.0250175685172172e-05, "loss": 0.0509, "step": 2434 }, { "epoch": 1.7111735769501055, "grad_norm": 0.6330466866493225, "learning_rate": 1.0254392129304288e-05, "loss": 0.037, "step": 2435 }, { "epoch": 1.7118763176387914, "grad_norm": 1.6964247226715088, "learning_rate": 1.0258608573436403e-05, "loss": 0.057, "step": 2436 }, { "epoch": 1.7125790583274771, "grad_norm": 0.4528255760669708, "learning_rate": 1.0262825017568518e-05, "loss": 0.0319, "step": 2437 }, { "epoch": 1.713281799016163, "grad_norm": 0.3557951748371124, "learning_rate": 1.0267041461700633e-05, "loss": 0.0552, "step": 2438 }, { "epoch": 1.7139845397048488, "grad_norm": 0.8996099233627319, "learning_rate": 1.0271257905832748e-05, "loss": 0.0692, "step": 2439 }, { "epoch": 1.7146872803935347, "grad_norm": 0.4295980930328369, "learning_rate": 1.0275474349964863e-05, "loss": 0.0351, "step": 2440 }, { "epoch": 1.7153900210822206, "grad_norm": 0.3852604031562805, "learning_rate": 1.0279690794096978e-05, "loss": 0.0609, "step": 2441 }, { "epoch": 1.7160927617709065, "grad_norm": 0.4187116026878357, "learning_rate": 1.0283907238229095e-05, "loss": 0.0619, "step": 2442 }, { "epoch": 1.7167955024595924, "grad_norm": 1.768725037574768, "learning_rate": 1.028812368236121e-05, "loss": 0.0942, "step": 2443 }, { "epoch": 1.7174982431482784, "grad_norm": 1.5121904611587524, "learning_rate": 1.0292340126493324e-05, "loss": 0.0978, "step": 2444 }, { "epoch": 1.7182009838369643, "grad_norm": 0.9418482184410095, "learning_rate": 1.029655657062544e-05, "loss": 0.1579, "step": 2445 }, { "epoch": 1.71890372452565, "grad_norm": 1.4327528476715088, "learning_rate": 1.0300773014757554e-05, "loss": 0.2984, "step": 2446 }, { "epoch": 1.719606465214336, "grad_norm": 1.3864867687225342, "learning_rate": 1.030498945888967e-05, "loss": 0.3047, "step": 2447 }, { "epoch": 1.7203092059030218, "grad_norm": 2.7728841304779053, "learning_rate": 1.0309205903021784e-05, "loss": 0.4314, "step": 2448 }, { "epoch": 1.7210119465917075, "grad_norm": 0.6198247671127319, "learning_rate": 1.0313422347153901e-05, "loss": 0.1456, "step": 2449 }, { "epoch": 1.7217146872803935, "grad_norm": 0.22350414097309113, "learning_rate": 1.0317638791286016e-05, "loss": 0.0497, "step": 2450 }, { "epoch": 1.7224174279690794, "grad_norm": 0.31565961241722107, "learning_rate": 1.032185523541813e-05, "loss": 0.0429, "step": 2451 }, { "epoch": 1.7231201686577653, "grad_norm": 0.32583940029144287, "learning_rate": 1.0326071679550246e-05, "loss": 0.0363, "step": 2452 }, { "epoch": 1.7238229093464512, "grad_norm": 0.33970436453819275, "learning_rate": 1.033028812368236e-05, "loss": 0.0806, "step": 2453 }, { "epoch": 1.7245256500351371, "grad_norm": 0.23261384665966034, "learning_rate": 1.0334504567814476e-05, "loss": 0.0326, "step": 2454 }, { "epoch": 1.725228390723823, "grad_norm": 0.23049034178256989, "learning_rate": 1.033872101194659e-05, "loss": 0.0289, "step": 2455 }, { "epoch": 1.7259311314125088, "grad_norm": 0.3782047927379608, "learning_rate": 1.0342937456078707e-05, "loss": 0.0518, "step": 2456 }, { "epoch": 1.7266338721011947, "grad_norm": 0.2800024449825287, "learning_rate": 1.0347153900210822e-05, "loss": 0.046, "step": 2457 }, { "epoch": 1.7273366127898804, "grad_norm": 0.2582035958766937, "learning_rate": 1.0351370344342937e-05, "loss": 0.0315, "step": 2458 }, { "epoch": 1.7280393534785663, "grad_norm": 0.26767992973327637, "learning_rate": 1.0355586788475052e-05, "loss": 0.0458, "step": 2459 }, { "epoch": 1.7287420941672522, "grad_norm": 0.2963234782218933, "learning_rate": 1.0359803232607169e-05, "loss": 0.0449, "step": 2460 }, { "epoch": 1.7294448348559381, "grad_norm": 0.34722769260406494, "learning_rate": 1.0364019676739284e-05, "loss": 0.0573, "step": 2461 }, { "epoch": 1.730147575544624, "grad_norm": 0.3983774483203888, "learning_rate": 1.0368236120871399e-05, "loss": 0.0525, "step": 2462 }, { "epoch": 1.73085031623331, "grad_norm": 0.3601188361644745, "learning_rate": 1.0372452565003515e-05, "loss": 0.0476, "step": 2463 }, { "epoch": 1.731553056921996, "grad_norm": 0.48253169655799866, "learning_rate": 1.037666900913563e-05, "loss": 0.0694, "step": 2464 }, { "epoch": 1.7322557976106818, "grad_norm": 0.437761515378952, "learning_rate": 1.0380885453267745e-05, "loss": 0.0539, "step": 2465 }, { "epoch": 1.7329585382993675, "grad_norm": 0.58209228515625, "learning_rate": 1.038510189739986e-05, "loss": 0.054, "step": 2466 }, { "epoch": 1.7336612789880534, "grad_norm": 0.47017914056777954, "learning_rate": 1.0389318341531975e-05, "loss": 0.088, "step": 2467 }, { "epoch": 1.7343640196767391, "grad_norm": 0.4940333068370819, "learning_rate": 1.039353478566409e-05, "loss": 0.1222, "step": 2468 }, { "epoch": 1.735066760365425, "grad_norm": 0.7430750131607056, "learning_rate": 1.0397751229796207e-05, "loss": 0.1984, "step": 2469 }, { "epoch": 1.735769501054111, "grad_norm": 1.055141568183899, "learning_rate": 1.0401967673928321e-05, "loss": 0.1922, "step": 2470 }, { "epoch": 1.736472241742797, "grad_norm": 1.3528021574020386, "learning_rate": 1.0406184118060436e-05, "loss": 0.312, "step": 2471 }, { "epoch": 1.7371749824314828, "grad_norm": 1.5921498537063599, "learning_rate": 1.0410400562192551e-05, "loss": 0.3328, "step": 2472 }, { "epoch": 1.7378777231201687, "grad_norm": 2.0333163738250732, "learning_rate": 1.0414617006324666e-05, "loss": 0.4344, "step": 2473 }, { "epoch": 1.7385804638088547, "grad_norm": 0.4564625024795532, "learning_rate": 1.0418833450456781e-05, "loss": 0.1297, "step": 2474 }, { "epoch": 1.7392832044975404, "grad_norm": 0.358733594417572, "learning_rate": 1.0423049894588896e-05, "loss": 0.0597, "step": 2475 }, { "epoch": 1.7399859451862263, "grad_norm": 0.49166861176490784, "learning_rate": 1.0427266338721013e-05, "loss": 0.0648, "step": 2476 }, { "epoch": 1.7406886858749122, "grad_norm": 0.42376700043678284, "learning_rate": 1.0431482782853128e-05, "loss": 0.0413, "step": 2477 }, { "epoch": 1.741391426563598, "grad_norm": 0.30762404203414917, "learning_rate": 1.0435699226985243e-05, "loss": 0.0464, "step": 2478 }, { "epoch": 1.7420941672522838, "grad_norm": 0.3043760657310486, "learning_rate": 1.0439915671117358e-05, "loss": 0.0372, "step": 2479 }, { "epoch": 1.7427969079409698, "grad_norm": 0.33543112874031067, "learning_rate": 1.0444132115249473e-05, "loss": 0.0452, "step": 2480 }, { "epoch": 1.7434996486296557, "grad_norm": 0.4036495089530945, "learning_rate": 1.0448348559381588e-05, "loss": 0.0433, "step": 2481 }, { "epoch": 1.7442023893183416, "grad_norm": 0.42963579297065735, "learning_rate": 1.0452565003513702e-05, "loss": 0.0436, "step": 2482 }, { "epoch": 1.7449051300070275, "grad_norm": 0.2324797362089157, "learning_rate": 1.0456781447645819e-05, "loss": 0.0429, "step": 2483 }, { "epoch": 1.7456078706957134, "grad_norm": 0.27055513858795166, "learning_rate": 1.0460997891777934e-05, "loss": 0.0417, "step": 2484 }, { "epoch": 1.7463106113843991, "grad_norm": 0.34675925970077515, "learning_rate": 1.0465214335910049e-05, "loss": 0.0474, "step": 2485 }, { "epoch": 1.747013352073085, "grad_norm": 0.2779361605644226, "learning_rate": 1.0469430780042164e-05, "loss": 0.0487, "step": 2486 }, { "epoch": 1.7477160927617708, "grad_norm": 0.41462576389312744, "learning_rate": 1.0473647224174279e-05, "loss": 0.0292, "step": 2487 }, { "epoch": 1.7484188334504567, "grad_norm": 0.32337483763694763, "learning_rate": 1.0477863668306394e-05, "loss": 0.0505, "step": 2488 }, { "epoch": 1.7491215741391426, "grad_norm": 0.4799526631832123, "learning_rate": 1.048208011243851e-05, "loss": 0.0489, "step": 2489 }, { "epoch": 1.7498243148278285, "grad_norm": 0.925146222114563, "learning_rate": 1.0486296556570627e-05, "loss": 0.0415, "step": 2490 }, { "epoch": 1.7505270555165144, "grad_norm": 0.6579524874687195, "learning_rate": 1.0490513000702742e-05, "loss": 0.0934, "step": 2491 }, { "epoch": 1.7512297962052004, "grad_norm": 0.37134668231010437, "learning_rate": 1.0494729444834857e-05, "loss": 0.0701, "step": 2492 }, { "epoch": 1.7519325368938863, "grad_norm": 0.662701427936554, "learning_rate": 1.0498945888966972e-05, "loss": 0.0997, "step": 2493 }, { "epoch": 1.752635277582572, "grad_norm": 1.517327070236206, "learning_rate": 1.0503162333099087e-05, "loss": 0.1169, "step": 2494 }, { "epoch": 1.753338018271258, "grad_norm": 1.2717468738555908, "learning_rate": 1.0507378777231202e-05, "loss": 0.1632, "step": 2495 }, { "epoch": 1.7540407589599438, "grad_norm": 1.0754601955413818, "learning_rate": 1.0511595221363317e-05, "loss": 0.2422, "step": 2496 }, { "epoch": 1.7547434996486295, "grad_norm": 1.396762728691101, "learning_rate": 1.0515811665495433e-05, "loss": 0.3532, "step": 2497 }, { "epoch": 1.7554462403373154, "grad_norm": 2.0016722679138184, "learning_rate": 1.0520028109627548e-05, "loss": 0.422, "step": 2498 }, { "epoch": 1.7561489810260014, "grad_norm": 0.42459821701049805, "learning_rate": 1.0524244553759663e-05, "loss": 0.1119, "step": 2499 }, { "epoch": 1.7568517217146873, "grad_norm": 0.21275348961353302, "learning_rate": 1.0528460997891778e-05, "loss": 0.0316, "step": 2500 }, { "epoch": 1.7575544624033732, "grad_norm": 0.5689807534217834, "learning_rate": 1.0532677442023893e-05, "loss": 0.0524, "step": 2501 }, { "epoch": 1.7582572030920591, "grad_norm": 0.3150360584259033, "learning_rate": 1.0536893886156008e-05, "loss": 0.0384, "step": 2502 }, { "epoch": 1.758959943780745, "grad_norm": 0.2995699644088745, "learning_rate": 1.0541110330288125e-05, "loss": 0.048, "step": 2503 }, { "epoch": 1.7596626844694307, "grad_norm": 0.2823265790939331, "learning_rate": 1.054532677442024e-05, "loss": 0.0407, "step": 2504 }, { "epoch": 1.7603654251581167, "grad_norm": 0.26134777069091797, "learning_rate": 1.0549543218552355e-05, "loss": 0.0349, "step": 2505 }, { "epoch": 1.7610681658468024, "grad_norm": 0.3012194037437439, "learning_rate": 1.055375966268447e-05, "loss": 0.0364, "step": 2506 }, { "epoch": 1.7617709065354883, "grad_norm": 0.3045694828033447, "learning_rate": 1.0557976106816585e-05, "loss": 0.0389, "step": 2507 }, { "epoch": 1.7624736472241742, "grad_norm": 0.32508644461631775, "learning_rate": 1.05621925509487e-05, "loss": 0.0425, "step": 2508 }, { "epoch": 1.7631763879128601, "grad_norm": 0.45031923055648804, "learning_rate": 1.0566408995080814e-05, "loss": 0.0575, "step": 2509 }, { "epoch": 1.763879128601546, "grad_norm": 0.30115947127342224, "learning_rate": 1.0570625439212931e-05, "loss": 0.0301, "step": 2510 }, { "epoch": 1.764581869290232, "grad_norm": 0.3486940264701843, "learning_rate": 1.0574841883345046e-05, "loss": 0.0589, "step": 2511 }, { "epoch": 1.765284609978918, "grad_norm": 0.34902864694595337, "learning_rate": 1.0579058327477161e-05, "loss": 0.0446, "step": 2512 }, { "epoch": 1.7659873506676038, "grad_norm": 0.49735724925994873, "learning_rate": 1.0583274771609276e-05, "loss": 0.0748, "step": 2513 }, { "epoch": 1.7666900913562895, "grad_norm": 0.5040584206581116, "learning_rate": 1.0587491215741391e-05, "loss": 0.0989, "step": 2514 }, { "epoch": 1.7673928320449754, "grad_norm": 0.4535643756389618, "learning_rate": 1.0591707659873506e-05, "loss": 0.0697, "step": 2515 }, { "epoch": 1.7680955727336611, "grad_norm": 1.2153338193893433, "learning_rate": 1.059592410400562e-05, "loss": 0.0598, "step": 2516 }, { "epoch": 1.768798313422347, "grad_norm": 0.5277150273323059, "learning_rate": 1.0600140548137737e-05, "loss": 0.123, "step": 2517 }, { "epoch": 1.769501054111033, "grad_norm": 0.7551487684249878, "learning_rate": 1.0604356992269852e-05, "loss": 0.0947, "step": 2518 }, { "epoch": 1.770203794799719, "grad_norm": 0.6921147704124451, "learning_rate": 1.0608573436401969e-05, "loss": 0.1216, "step": 2519 }, { "epoch": 1.7709065354884048, "grad_norm": 0.7946833372116089, "learning_rate": 1.0612789880534084e-05, "loss": 0.2053, "step": 2520 }, { "epoch": 1.7716092761770907, "grad_norm": 1.0003910064697266, "learning_rate": 1.0617006324666199e-05, "loss": 0.2473, "step": 2521 }, { "epoch": 1.7723120168657767, "grad_norm": 2.324859380722046, "learning_rate": 1.0621222768798314e-05, "loss": 0.3487, "step": 2522 }, { "epoch": 1.7730147575544624, "grad_norm": 2.2645626068115234, "learning_rate": 1.0625439212930429e-05, "loss": 0.3817, "step": 2523 }, { "epoch": 1.7737174982431483, "grad_norm": 0.7366374731063843, "learning_rate": 1.0629655657062545e-05, "loss": 0.0963, "step": 2524 }, { "epoch": 1.7744202389318342, "grad_norm": 0.30049675703048706, "learning_rate": 1.063387210119466e-05, "loss": 0.0555, "step": 2525 }, { "epoch": 1.77512297962052, "grad_norm": 0.3415667414665222, "learning_rate": 1.0638088545326775e-05, "loss": 0.0542, "step": 2526 }, { "epoch": 1.7758257203092058, "grad_norm": 0.2534957528114319, "learning_rate": 1.064230498945889e-05, "loss": 0.0429, "step": 2527 }, { "epoch": 1.7765284609978917, "grad_norm": 0.4160566031932831, "learning_rate": 1.0646521433591005e-05, "loss": 0.0501, "step": 2528 }, { "epoch": 1.7772312016865777, "grad_norm": 0.3432844579219818, "learning_rate": 1.065073787772312e-05, "loss": 0.0458, "step": 2529 }, { "epoch": 1.7779339423752636, "grad_norm": 0.2837897837162018, "learning_rate": 1.0654954321855237e-05, "loss": 0.0494, "step": 2530 }, { "epoch": 1.7786366830639495, "grad_norm": 0.2817327380180359, "learning_rate": 1.0659170765987352e-05, "loss": 0.0409, "step": 2531 }, { "epoch": 1.7793394237526354, "grad_norm": 0.3672429323196411, "learning_rate": 1.0663387210119467e-05, "loss": 0.055, "step": 2532 }, { "epoch": 1.7800421644413211, "grad_norm": 0.2471207231283188, "learning_rate": 1.0667603654251582e-05, "loss": 0.0422, "step": 2533 }, { "epoch": 1.780744905130007, "grad_norm": 0.3838813602924347, "learning_rate": 1.0671820098383697e-05, "loss": 0.0716, "step": 2534 }, { "epoch": 1.7814476458186927, "grad_norm": 0.3070583641529083, "learning_rate": 1.0676036542515811e-05, "loss": 0.0444, "step": 2535 }, { "epoch": 1.7821503865073787, "grad_norm": 0.33328357338905334, "learning_rate": 1.0680252986647926e-05, "loss": 0.0486, "step": 2536 }, { "epoch": 1.7828531271960646, "grad_norm": 0.3030230700969696, "learning_rate": 1.0684469430780043e-05, "loss": 0.0288, "step": 2537 }, { "epoch": 1.7835558678847505, "grad_norm": 0.4062303304672241, "learning_rate": 1.0688685874912158e-05, "loss": 0.0494, "step": 2538 }, { "epoch": 1.7842586085734364, "grad_norm": 0.8553524613380432, "learning_rate": 1.0692902319044273e-05, "loss": 0.0704, "step": 2539 }, { "epoch": 1.7849613492621224, "grad_norm": 0.3942453861236572, "learning_rate": 1.0697118763176388e-05, "loss": 0.0601, "step": 2540 }, { "epoch": 1.7856640899508083, "grad_norm": 0.43916043639183044, "learning_rate": 1.0701335207308503e-05, "loss": 0.0672, "step": 2541 }, { "epoch": 1.786366830639494, "grad_norm": 0.4140225052833557, "learning_rate": 1.0705551651440618e-05, "loss": 0.09, "step": 2542 }, { "epoch": 1.78706957132818, "grad_norm": 0.408452570438385, "learning_rate": 1.0709768095572733e-05, "loss": 0.0835, "step": 2543 }, { "epoch": 1.7877723120168658, "grad_norm": 0.7337705492973328, "learning_rate": 1.071398453970485e-05, "loss": 0.1752, "step": 2544 }, { "epoch": 1.7884750527055515, "grad_norm": 0.7109704613685608, "learning_rate": 1.0718200983836964e-05, "loss": 0.2179, "step": 2545 }, { "epoch": 1.7891777933942374, "grad_norm": 1.1693638563156128, "learning_rate": 1.072241742796908e-05, "loss": 0.2589, "step": 2546 }, { "epoch": 1.7898805340829234, "grad_norm": 1.360999584197998, "learning_rate": 1.0726633872101194e-05, "loss": 0.3006, "step": 2547 }, { "epoch": 1.7905832747716093, "grad_norm": 1.9392530918121338, "learning_rate": 1.0730850316233309e-05, "loss": 0.4229, "step": 2548 }, { "epoch": 1.7912860154602952, "grad_norm": 0.357143759727478, "learning_rate": 1.0735066760365426e-05, "loss": 0.0995, "step": 2549 }, { "epoch": 1.7919887561489811, "grad_norm": 0.34192025661468506, "learning_rate": 1.073928320449754e-05, "loss": 0.0531, "step": 2550 }, { "epoch": 1.792691496837667, "grad_norm": 0.2661254107952118, "learning_rate": 1.0743499648629657e-05, "loss": 0.0439, "step": 2551 }, { "epoch": 1.7933942375263527, "grad_norm": 0.30766546726226807, "learning_rate": 1.0747716092761772e-05, "loss": 0.0428, "step": 2552 }, { "epoch": 1.7940969782150387, "grad_norm": 0.5032153129577637, "learning_rate": 1.0751932536893887e-05, "loss": 0.0393, "step": 2553 }, { "epoch": 1.7947997189037244, "grad_norm": 0.27441465854644775, "learning_rate": 1.0756148981026002e-05, "loss": 0.0349, "step": 2554 }, { "epoch": 1.7955024595924103, "grad_norm": 0.2844305634498596, "learning_rate": 1.0760365425158117e-05, "loss": 0.0389, "step": 2555 }, { "epoch": 1.7962052002810962, "grad_norm": 0.2721550464630127, "learning_rate": 1.0764581869290232e-05, "loss": 0.0419, "step": 2556 }, { "epoch": 1.7969079409697821, "grad_norm": 0.3115713894367218, "learning_rate": 1.0768798313422347e-05, "loss": 0.0529, "step": 2557 }, { "epoch": 1.797610681658468, "grad_norm": 0.3191829323768616, "learning_rate": 1.0773014757554464e-05, "loss": 0.0413, "step": 2558 }, { "epoch": 1.798313422347154, "grad_norm": 0.34222033619880676, "learning_rate": 1.0777231201686579e-05, "loss": 0.0566, "step": 2559 }, { "epoch": 1.7990161630358399, "grad_norm": 0.4512309432029724, "learning_rate": 1.0781447645818694e-05, "loss": 0.0444, "step": 2560 }, { "epoch": 1.7997189037245258, "grad_norm": 0.4011073410511017, "learning_rate": 1.0785664089950808e-05, "loss": 0.069, "step": 2561 }, { "epoch": 1.8004216444132115, "grad_norm": 0.33960673213005066, "learning_rate": 1.0789880534082923e-05, "loss": 0.0307, "step": 2562 }, { "epoch": 1.8011243851018974, "grad_norm": 0.4244003891944885, "learning_rate": 1.0794096978215038e-05, "loss": 0.0694, "step": 2563 }, { "epoch": 1.8018271257905831, "grad_norm": 0.36801251769065857, "learning_rate": 1.0798313422347155e-05, "loss": 0.051, "step": 2564 }, { "epoch": 1.802529866479269, "grad_norm": 0.29661667346954346, "learning_rate": 1.080252986647927e-05, "loss": 0.0442, "step": 2565 }, { "epoch": 1.803232607167955, "grad_norm": 0.3474145829677582, "learning_rate": 1.0806746310611385e-05, "loss": 0.0664, "step": 2566 }, { "epoch": 1.803935347856641, "grad_norm": 0.5556042194366455, "learning_rate": 1.08109627547435e-05, "loss": 0.0708, "step": 2567 }, { "epoch": 1.8046380885453268, "grad_norm": 0.6603772044181824, "learning_rate": 1.0815179198875615e-05, "loss": 0.1053, "step": 2568 }, { "epoch": 1.8053408292340127, "grad_norm": 0.6094714403152466, "learning_rate": 1.081939564300773e-05, "loss": 0.1401, "step": 2569 }, { "epoch": 1.8060435699226987, "grad_norm": 2.7774107456207275, "learning_rate": 1.0823612087139845e-05, "loss": 0.175, "step": 2570 }, { "epoch": 1.8067463106113844, "grad_norm": 1.3908392190933228, "learning_rate": 1.0827828531271961e-05, "loss": 0.3278, "step": 2571 }, { "epoch": 1.8074490513000703, "grad_norm": 1.3071283102035522, "learning_rate": 1.0832044975404076e-05, "loss": 0.3118, "step": 2572 }, { "epoch": 1.8081517919887562, "grad_norm": 2.7549386024475098, "learning_rate": 1.0836261419536191e-05, "loss": 0.4199, "step": 2573 }, { "epoch": 1.808854532677442, "grad_norm": 0.527688205242157, "learning_rate": 1.0840477863668306e-05, "loss": 0.1135, "step": 2574 }, { "epoch": 1.8095572733661278, "grad_norm": 0.27304336428642273, "learning_rate": 1.0844694307800421e-05, "loss": 0.0441, "step": 2575 }, { "epoch": 1.8102600140548137, "grad_norm": 0.6782130599021912, "learning_rate": 1.0848910751932536e-05, "loss": 0.0428, "step": 2576 }, { "epoch": 1.8109627547434997, "grad_norm": 0.5486721992492676, "learning_rate": 1.0853127196064651e-05, "loss": 0.0428, "step": 2577 }, { "epoch": 1.8116654954321856, "grad_norm": 0.4764484465122223, "learning_rate": 1.0857343640196768e-05, "loss": 0.0331, "step": 2578 }, { "epoch": 1.8123682361208715, "grad_norm": 0.3536878526210785, "learning_rate": 1.0861560084328884e-05, "loss": 0.0406, "step": 2579 }, { "epoch": 1.8130709768095574, "grad_norm": 0.25890904664993286, "learning_rate": 1.0865776528461e-05, "loss": 0.0265, "step": 2580 }, { "epoch": 1.8137737174982431, "grad_norm": 0.3313944935798645, "learning_rate": 1.0869992972593114e-05, "loss": 0.0517, "step": 2581 }, { "epoch": 1.814476458186929, "grad_norm": 0.3427583873271942, "learning_rate": 1.0874209416725229e-05, "loss": 0.0521, "step": 2582 }, { "epoch": 1.8151791988756147, "grad_norm": 0.24371705949306488, "learning_rate": 1.0878425860857344e-05, "loss": 0.0258, "step": 2583 }, { "epoch": 1.8158819395643007, "grad_norm": 0.3171294033527374, "learning_rate": 1.0882642304989459e-05, "loss": 0.0596, "step": 2584 }, { "epoch": 1.8165846802529866, "grad_norm": 0.280688613653183, "learning_rate": 1.0886858749121576e-05, "loss": 0.0482, "step": 2585 }, { "epoch": 1.8172874209416725, "grad_norm": 0.3754027485847473, "learning_rate": 1.089107519325369e-05, "loss": 0.0564, "step": 2586 }, { "epoch": 1.8179901616303584, "grad_norm": 0.38200175762176514, "learning_rate": 1.0895291637385805e-05, "loss": 0.0433, "step": 2587 }, { "epoch": 1.8186929023190443, "grad_norm": 1.6382025480270386, "learning_rate": 1.089950808151792e-05, "loss": 0.0707, "step": 2588 }, { "epoch": 1.8193956430077303, "grad_norm": 0.5464630722999573, "learning_rate": 1.0903724525650035e-05, "loss": 0.0665, "step": 2589 }, { "epoch": 1.8200983836964162, "grad_norm": 0.36381053924560547, "learning_rate": 1.090794096978215e-05, "loss": 0.0376, "step": 2590 }, { "epoch": 1.8208011243851019, "grad_norm": 0.390597939491272, "learning_rate": 1.0912157413914265e-05, "loss": 0.0786, "step": 2591 }, { "epoch": 1.8215038650737878, "grad_norm": 0.4894365668296814, "learning_rate": 1.0916373858046382e-05, "loss": 0.0836, "step": 2592 }, { "epoch": 1.8222066057624735, "grad_norm": 0.5517553091049194, "learning_rate": 1.0920590302178497e-05, "loss": 0.1129, "step": 2593 }, { "epoch": 1.8229093464511594, "grad_norm": 0.5100826025009155, "learning_rate": 1.0924806746310612e-05, "loss": 0.1106, "step": 2594 }, { "epoch": 1.8236120871398454, "grad_norm": 0.8460251092910767, "learning_rate": 1.0929023190442727e-05, "loss": 0.1879, "step": 2595 }, { "epoch": 1.8243148278285313, "grad_norm": 1.4863345623016357, "learning_rate": 1.0933239634574842e-05, "loss": 0.2357, "step": 2596 }, { "epoch": 1.8250175685172172, "grad_norm": 1.4985778331756592, "learning_rate": 1.0937456078706957e-05, "loss": 0.2848, "step": 2597 }, { "epoch": 1.8257203092059031, "grad_norm": 7.242557048797607, "learning_rate": 1.0941672522839073e-05, "loss": 0.433, "step": 2598 }, { "epoch": 1.826423049894589, "grad_norm": 0.9964762330055237, "learning_rate": 1.0945888966971188e-05, "loss": 0.1242, "step": 2599 }, { "epoch": 1.8271257905832747, "grad_norm": 0.7420437932014465, "learning_rate": 1.0950105411103303e-05, "loss": 0.0467, "step": 2600 }, { "epoch": 1.8278285312719607, "grad_norm": 0.5754193663597107, "learning_rate": 1.0954321855235418e-05, "loss": 0.1021, "step": 2601 }, { "epoch": 1.8285312719606466, "grad_norm": 0.3862690329551697, "learning_rate": 1.0958538299367533e-05, "loss": 0.0558, "step": 2602 }, { "epoch": 1.8292340126493323, "grad_norm": 0.2985682487487793, "learning_rate": 1.0962754743499648e-05, "loss": 0.0344, "step": 2603 }, { "epoch": 1.8299367533380182, "grad_norm": 0.18603822588920593, "learning_rate": 1.0966971187631763e-05, "loss": 0.0217, "step": 2604 }, { "epoch": 1.8306394940267041, "grad_norm": 0.3263090252876282, "learning_rate": 1.097118763176388e-05, "loss": 0.0776, "step": 2605 }, { "epoch": 1.83134223471539, "grad_norm": 0.5617772340774536, "learning_rate": 1.0975404075895994e-05, "loss": 0.0515, "step": 2606 }, { "epoch": 1.832044975404076, "grad_norm": 0.24161536991596222, "learning_rate": 1.097962052002811e-05, "loss": 0.0379, "step": 2607 }, { "epoch": 1.8327477160927619, "grad_norm": 0.447704941034317, "learning_rate": 1.0983836964160226e-05, "loss": 0.0328, "step": 2608 }, { "epoch": 1.8334504567814478, "grad_norm": 0.47482624650001526, "learning_rate": 1.0988053408292341e-05, "loss": 0.0537, "step": 2609 }, { "epoch": 1.8341531974701335, "grad_norm": 0.35123971104621887, "learning_rate": 1.0992269852424456e-05, "loss": 0.0383, "step": 2610 }, { "epoch": 1.8348559381588194, "grad_norm": 0.3272206783294678, "learning_rate": 1.0996486296556571e-05, "loss": 0.0511, "step": 2611 }, { "epoch": 1.8355586788475051, "grad_norm": 0.4663461446762085, "learning_rate": 1.1000702740688688e-05, "loss": 0.0456, "step": 2612 }, { "epoch": 1.836261419536191, "grad_norm": 0.3224484920501709, "learning_rate": 1.1004919184820802e-05, "loss": 0.0545, "step": 2613 }, { "epoch": 1.836964160224877, "grad_norm": 0.5105805397033691, "learning_rate": 1.1009135628952917e-05, "loss": 0.0542, "step": 2614 }, { "epoch": 1.8376669009135629, "grad_norm": 0.4295184016227722, "learning_rate": 1.1013352073085032e-05, "loss": 0.0757, "step": 2615 }, { "epoch": 1.8383696416022488, "grad_norm": 0.35557499527931213, "learning_rate": 1.1017568517217147e-05, "loss": 0.0653, "step": 2616 }, { "epoch": 1.8390723822909347, "grad_norm": 0.44660353660583496, "learning_rate": 1.1021784961349262e-05, "loss": 0.0967, "step": 2617 }, { "epoch": 1.8397751229796206, "grad_norm": 0.3875942826271057, "learning_rate": 1.1026001405481377e-05, "loss": 0.0712, "step": 2618 }, { "epoch": 1.8404778636683063, "grad_norm": 0.7217148542404175, "learning_rate": 1.1030217849613494e-05, "loss": 0.1127, "step": 2619 }, { "epoch": 1.8411806043569923, "grad_norm": 0.8268424272537231, "learning_rate": 1.1034434293745609e-05, "loss": 0.182, "step": 2620 }, { "epoch": 1.8418833450456782, "grad_norm": 1.8099079132080078, "learning_rate": 1.1038650737877724e-05, "loss": 0.2996, "step": 2621 }, { "epoch": 1.8425860857343639, "grad_norm": 1.3447811603546143, "learning_rate": 1.1042867182009839e-05, "loss": 0.4047, "step": 2622 }, { "epoch": 1.8432888264230498, "grad_norm": 2.157015323638916, "learning_rate": 1.1047083626141954e-05, "loss": 0.4217, "step": 2623 }, { "epoch": 1.8439915671117357, "grad_norm": 0.6651638746261597, "learning_rate": 1.1051300070274069e-05, "loss": 0.1378, "step": 2624 }, { "epoch": 1.8446943078004217, "grad_norm": 0.36129266023635864, "learning_rate": 1.1055516514406183e-05, "loss": 0.0711, "step": 2625 }, { "epoch": 1.8453970484891076, "grad_norm": 0.29891207814216614, "learning_rate": 1.10597329585383e-05, "loss": 0.0389, "step": 2626 }, { "epoch": 1.8460997891777935, "grad_norm": 0.48659011721611023, "learning_rate": 1.1063949402670415e-05, "loss": 0.0356, "step": 2627 }, { "epoch": 1.8468025298664794, "grad_norm": 0.2420504242181778, "learning_rate": 1.106816584680253e-05, "loss": 0.0286, "step": 2628 }, { "epoch": 1.8475052705551651, "grad_norm": 0.2604161202907562, "learning_rate": 1.1072382290934645e-05, "loss": 0.0315, "step": 2629 }, { "epoch": 1.848208011243851, "grad_norm": 0.2847982943058014, "learning_rate": 1.107659873506676e-05, "loss": 0.0421, "step": 2630 }, { "epoch": 1.8489107519325367, "grad_norm": 0.48598483204841614, "learning_rate": 1.1080815179198875e-05, "loss": 0.064, "step": 2631 }, { "epoch": 1.8496134926212227, "grad_norm": 0.32283589243888855, "learning_rate": 1.1085031623330991e-05, "loss": 0.0448, "step": 2632 }, { "epoch": 1.8503162333099086, "grad_norm": 0.24825486540794373, "learning_rate": 1.1089248067463106e-05, "loss": 0.0335, "step": 2633 }, { "epoch": 1.8510189739985945, "grad_norm": 0.46552345156669617, "learning_rate": 1.1093464511595221e-05, "loss": 0.0621, "step": 2634 }, { "epoch": 1.8517217146872804, "grad_norm": 0.4760400652885437, "learning_rate": 1.1097680955727336e-05, "loss": 0.0374, "step": 2635 }, { "epoch": 1.8524244553759663, "grad_norm": 0.38712188601493835, "learning_rate": 1.1101897399859451e-05, "loss": 0.0545, "step": 2636 }, { "epoch": 1.8531271960646523, "grad_norm": 0.49138784408569336, "learning_rate": 1.1106113843991566e-05, "loss": 0.0307, "step": 2637 }, { "epoch": 1.8538299367533382, "grad_norm": 0.424345999956131, "learning_rate": 1.1110330288123683e-05, "loss": 0.0605, "step": 2638 }, { "epoch": 1.8545326774420239, "grad_norm": 0.4681400954723358, "learning_rate": 1.11145467322558e-05, "loss": 0.0599, "step": 2639 }, { "epoch": 1.8552354181307098, "grad_norm": 0.4001332223415375, "learning_rate": 1.1118763176387914e-05, "loss": 0.0284, "step": 2640 }, { "epoch": 1.8559381588193955, "grad_norm": 0.4061988592147827, "learning_rate": 1.112297962052003e-05, "loss": 0.0875, "step": 2641 }, { "epoch": 1.8566408995080814, "grad_norm": 0.6574118733406067, "learning_rate": 1.1127196064652144e-05, "loss": 0.0862, "step": 2642 }, { "epoch": 1.8573436401967673, "grad_norm": 0.4436493217945099, "learning_rate": 1.113141250878426e-05, "loss": 0.0806, "step": 2643 }, { "epoch": 1.8580463808854533, "grad_norm": 0.7919248342514038, "learning_rate": 1.1135628952916374e-05, "loss": 0.1541, "step": 2644 }, { "epoch": 1.8587491215741392, "grad_norm": 0.8380445241928101, "learning_rate": 1.1139845397048489e-05, "loss": 0.1946, "step": 2645 }, { "epoch": 1.859451862262825, "grad_norm": 0.9281836748123169, "learning_rate": 1.1144061841180606e-05, "loss": 0.2288, "step": 2646 }, { "epoch": 1.860154602951511, "grad_norm": 2.3465445041656494, "learning_rate": 1.114827828531272e-05, "loss": 0.327, "step": 2647 }, { "epoch": 1.8608573436401967, "grad_norm": 2.194774866104126, "learning_rate": 1.1152494729444836e-05, "loss": 0.3899, "step": 2648 }, { "epoch": 1.8615600843288826, "grad_norm": 0.4341872036457062, "learning_rate": 1.115671117357695e-05, "loss": 0.1103, "step": 2649 }, { "epoch": 1.8622628250175686, "grad_norm": 0.26044631004333496, "learning_rate": 1.1160927617709066e-05, "loss": 0.0534, "step": 2650 }, { "epoch": 1.8629655657062543, "grad_norm": 0.45357412099838257, "learning_rate": 1.116514406184118e-05, "loss": 0.0462, "step": 2651 }, { "epoch": 1.8636683063949402, "grad_norm": 0.26189133524894714, "learning_rate": 1.1169360505973295e-05, "loss": 0.0458, "step": 2652 }, { "epoch": 1.864371047083626, "grad_norm": 0.25680071115493774, "learning_rate": 1.1173576950105412e-05, "loss": 0.0361, "step": 2653 }, { "epoch": 1.865073787772312, "grad_norm": 0.4669685661792755, "learning_rate": 1.1177793394237527e-05, "loss": 0.035, "step": 2654 }, { "epoch": 1.865776528460998, "grad_norm": 0.2602902054786682, "learning_rate": 1.1182009838369642e-05, "loss": 0.042, "step": 2655 }, { "epoch": 1.8664792691496839, "grad_norm": 0.35433515906333923, "learning_rate": 1.1186226282501757e-05, "loss": 0.0419, "step": 2656 }, { "epoch": 1.8671820098383698, "grad_norm": 0.2838897705078125, "learning_rate": 1.1190442726633872e-05, "loss": 0.0408, "step": 2657 }, { "epoch": 1.8678847505270555, "grad_norm": 0.369048148393631, "learning_rate": 1.1194659170765987e-05, "loss": 0.0466, "step": 2658 }, { "epoch": 1.8685874912157414, "grad_norm": 0.3295706808567047, "learning_rate": 1.1198875614898102e-05, "loss": 0.0467, "step": 2659 }, { "epoch": 1.8692902319044271, "grad_norm": 0.30252644419670105, "learning_rate": 1.1203092059030218e-05, "loss": 0.0361, "step": 2660 }, { "epoch": 1.869992972593113, "grad_norm": 0.40953052043914795, "learning_rate": 1.1207308503162333e-05, "loss": 0.0427, "step": 2661 }, { "epoch": 1.870695713281799, "grad_norm": 0.3670688569545746, "learning_rate": 1.1211524947294448e-05, "loss": 0.0404, "step": 2662 }, { "epoch": 1.8713984539704849, "grad_norm": 0.45844390988349915, "learning_rate": 1.1215741391426563e-05, "loss": 0.0479, "step": 2663 }, { "epoch": 1.8721011946591708, "grad_norm": 0.4243421256542206, "learning_rate": 1.1219957835558678e-05, "loss": 0.0758, "step": 2664 }, { "epoch": 1.8728039353478567, "grad_norm": 0.42642804980278015, "learning_rate": 1.1224174279690793e-05, "loss": 0.0302, "step": 2665 }, { "epoch": 1.8735066760365426, "grad_norm": 0.7948978543281555, "learning_rate": 1.122839072382291e-05, "loss": 0.0727, "step": 2666 }, { "epoch": 1.8742094167252283, "grad_norm": 0.3988829255104065, "learning_rate": 1.1232607167955025e-05, "loss": 0.0639, "step": 2667 }, { "epoch": 1.8749121574139143, "grad_norm": 0.6451216340065002, "learning_rate": 1.1236823612087141e-05, "loss": 0.101, "step": 2668 }, { "epoch": 1.8756148981026002, "grad_norm": 0.7585822343826294, "learning_rate": 1.1241040056219256e-05, "loss": 0.1504, "step": 2669 }, { "epoch": 1.8763176387912859, "grad_norm": 0.7349403500556946, "learning_rate": 1.1245256500351371e-05, "loss": 0.1807, "step": 2670 }, { "epoch": 1.8770203794799718, "grad_norm": 1.0743119716644287, "learning_rate": 1.1249472944483486e-05, "loss": 0.2615, "step": 2671 }, { "epoch": 1.8777231201686577, "grad_norm": 1.7319248914718628, "learning_rate": 1.1253689388615601e-05, "loss": 0.2807, "step": 2672 }, { "epoch": 1.8784258608573436, "grad_norm": 1.9071550369262695, "learning_rate": 1.1257905832747718e-05, "loss": 0.4516, "step": 2673 }, { "epoch": 1.8791286015460296, "grad_norm": 0.39626094698905945, "learning_rate": 1.1262122276879833e-05, "loss": 0.1231, "step": 2674 }, { "epoch": 1.8798313422347155, "grad_norm": 0.240386500954628, "learning_rate": 1.1266338721011948e-05, "loss": 0.039, "step": 2675 }, { "epoch": 1.8805340829234014, "grad_norm": 0.4159679412841797, "learning_rate": 1.1270555165144063e-05, "loss": 0.0488, "step": 2676 }, { "epoch": 1.881236823612087, "grad_norm": 0.3797987699508667, "learning_rate": 1.1274771609276177e-05, "loss": 0.0571, "step": 2677 }, { "epoch": 1.881939564300773, "grad_norm": 0.2744351029396057, "learning_rate": 1.1278988053408292e-05, "loss": 0.0387, "step": 2678 }, { "epoch": 1.8826423049894587, "grad_norm": 0.3548046350479126, "learning_rate": 1.1283204497540407e-05, "loss": 0.0388, "step": 2679 }, { "epoch": 1.8833450456781446, "grad_norm": 0.2262018918991089, "learning_rate": 1.1287420941672524e-05, "loss": 0.0271, "step": 2680 }, { "epoch": 1.8840477863668306, "grad_norm": 0.27224108576774597, "learning_rate": 1.1291637385804639e-05, "loss": 0.046, "step": 2681 }, { "epoch": 1.8847505270555165, "grad_norm": 0.29539501667022705, "learning_rate": 1.1295853829936754e-05, "loss": 0.0482, "step": 2682 }, { "epoch": 1.8854532677442024, "grad_norm": 0.3618180453777313, "learning_rate": 1.1300070274068869e-05, "loss": 0.0339, "step": 2683 }, { "epoch": 1.8861560084328883, "grad_norm": 0.3487306833267212, "learning_rate": 1.1304286718200984e-05, "loss": 0.0474, "step": 2684 }, { "epoch": 1.8868587491215743, "grad_norm": 0.2591104507446289, "learning_rate": 1.1308503162333099e-05, "loss": 0.0242, "step": 2685 }, { "epoch": 1.8875614898102602, "grad_norm": 0.44691917300224304, "learning_rate": 1.1312719606465214e-05, "loss": 0.0627, "step": 2686 }, { "epoch": 1.8882642304989459, "grad_norm": 0.23462088406085968, "learning_rate": 1.131693605059733e-05, "loss": 0.033, "step": 2687 }, { "epoch": 1.8889669711876318, "grad_norm": 0.783885657787323, "learning_rate": 1.1321152494729445e-05, "loss": 0.0789, "step": 2688 }, { "epoch": 1.8896697118763175, "grad_norm": 0.48431143164634705, "learning_rate": 1.132536893886156e-05, "loss": 0.0618, "step": 2689 }, { "epoch": 1.8903724525650034, "grad_norm": 0.3406674265861511, "learning_rate": 1.1329585382993675e-05, "loss": 0.0473, "step": 2690 }, { "epoch": 1.8910751932536893, "grad_norm": 0.36314380168914795, "learning_rate": 1.133380182712579e-05, "loss": 0.0571, "step": 2691 }, { "epoch": 1.8917779339423753, "grad_norm": 0.47034087777137756, "learning_rate": 1.1338018271257905e-05, "loss": 0.0884, "step": 2692 }, { "epoch": 1.8924806746310612, "grad_norm": 0.5475443005561829, "learning_rate": 1.134223471539002e-05, "loss": 0.0803, "step": 2693 }, { "epoch": 1.893183415319747, "grad_norm": 0.6375974416732788, "learning_rate": 1.1346451159522137e-05, "loss": 0.1255, "step": 2694 }, { "epoch": 1.893886156008433, "grad_norm": 0.7172389626502991, "learning_rate": 1.1350667603654252e-05, "loss": 0.1713, "step": 2695 }, { "epoch": 1.8945888966971187, "grad_norm": 1.064613699913025, "learning_rate": 1.1354884047786366e-05, "loss": 0.2409, "step": 2696 }, { "epoch": 1.8952916373858046, "grad_norm": 1.4145694971084595, "learning_rate": 1.1359100491918481e-05, "loss": 0.3685, "step": 2697 }, { "epoch": 1.8959943780744906, "grad_norm": 3.03385591506958, "learning_rate": 1.1363316936050598e-05, "loss": 0.4223, "step": 2698 }, { "epoch": 1.8966971187631763, "grad_norm": 0.4415505528450012, "learning_rate": 1.1367533380182713e-05, "loss": 0.1153, "step": 2699 }, { "epoch": 1.8973998594518622, "grad_norm": 0.2896350026130676, "learning_rate": 1.1371749824314828e-05, "loss": 0.0392, "step": 2700 }, { "epoch": 1.898102600140548, "grad_norm": 0.28578802943229675, "learning_rate": 1.1375966268446945e-05, "loss": 0.0422, "step": 2701 }, { "epoch": 1.898805340829234, "grad_norm": 0.2845664918422699, "learning_rate": 1.138018271257906e-05, "loss": 0.0366, "step": 2702 }, { "epoch": 1.89950808151792, "grad_norm": 0.26700565218925476, "learning_rate": 1.1384399156711174e-05, "loss": 0.0382, "step": 2703 }, { "epoch": 1.9002108222066059, "grad_norm": 0.3448610007762909, "learning_rate": 1.138861560084329e-05, "loss": 0.0434, "step": 2704 }, { "epoch": 1.9009135628952918, "grad_norm": 0.320126473903656, "learning_rate": 1.1392832044975404e-05, "loss": 0.0273, "step": 2705 }, { "epoch": 1.9016163035839775, "grad_norm": 0.21982835233211517, "learning_rate": 1.139704848910752e-05, "loss": 0.0327, "step": 2706 }, { "epoch": 1.9023190442726634, "grad_norm": 0.2549832761287689, "learning_rate": 1.1401264933239636e-05, "loss": 0.0374, "step": 2707 }, { "epoch": 1.903021784961349, "grad_norm": 0.2884438931941986, "learning_rate": 1.1405481377371751e-05, "loss": 0.034, "step": 2708 }, { "epoch": 1.903724525650035, "grad_norm": 0.292632132768631, "learning_rate": 1.1409697821503866e-05, "loss": 0.0432, "step": 2709 }, { "epoch": 1.904427266338721, "grad_norm": 0.3803198039531708, "learning_rate": 1.141391426563598e-05, "loss": 0.0553, "step": 2710 }, { "epoch": 1.9051300070274069, "grad_norm": 0.595048189163208, "learning_rate": 1.1418130709768096e-05, "loss": 0.0578, "step": 2711 }, { "epoch": 1.9058327477160928, "grad_norm": 0.33271029591560364, "learning_rate": 1.142234715390021e-05, "loss": 0.0491, "step": 2712 }, { "epoch": 1.9065354884047787, "grad_norm": 0.34123119711875916, "learning_rate": 1.1426563598032326e-05, "loss": 0.0517, "step": 2713 }, { "epoch": 1.9072382290934646, "grad_norm": 0.3355213403701782, "learning_rate": 1.1430780042164442e-05, "loss": 0.0714, "step": 2714 }, { "epoch": 1.9079409697821503, "grad_norm": 0.38860511779785156, "learning_rate": 1.1434996486296557e-05, "loss": 0.0653, "step": 2715 }, { "epoch": 1.9086437104708363, "grad_norm": 1.059545636177063, "learning_rate": 1.1439212930428672e-05, "loss": 0.0635, "step": 2716 }, { "epoch": 1.9093464511595222, "grad_norm": 0.5601394176483154, "learning_rate": 1.1443429374560787e-05, "loss": 0.0755, "step": 2717 }, { "epoch": 1.9100491918482079, "grad_norm": 0.5049400329589844, "learning_rate": 1.1447645818692902e-05, "loss": 0.0868, "step": 2718 }, { "epoch": 1.9107519325368938, "grad_norm": 0.6097887754440308, "learning_rate": 1.1451862262825017e-05, "loss": 0.1406, "step": 2719 }, { "epoch": 1.9114546732255797, "grad_norm": 0.7579454779624939, "learning_rate": 1.1456078706957132e-05, "loss": 0.1649, "step": 2720 }, { "epoch": 1.9121574139142656, "grad_norm": 1.1805890798568726, "learning_rate": 1.1460295151089249e-05, "loss": 0.2788, "step": 2721 }, { "epoch": 1.9128601546029516, "grad_norm": 1.2014660835266113, "learning_rate": 1.1464511595221364e-05, "loss": 0.3095, "step": 2722 }, { "epoch": 1.9135628952916375, "grad_norm": 1.8143105506896973, "learning_rate": 1.1468728039353478e-05, "loss": 0.4585, "step": 2723 }, { "epoch": 1.9142656359803234, "grad_norm": 0.448321670293808, "learning_rate": 1.1472944483485593e-05, "loss": 0.1104, "step": 2724 }, { "epoch": 1.914968376669009, "grad_norm": 0.31054553389549255, "learning_rate": 1.1477160927617708e-05, "loss": 0.0488, "step": 2725 }, { "epoch": 1.915671117357695, "grad_norm": 0.4049356281757355, "learning_rate": 1.1481377371749823e-05, "loss": 0.0573, "step": 2726 }, { "epoch": 1.916373858046381, "grad_norm": 0.2809562683105469, "learning_rate": 1.148559381588194e-05, "loss": 0.0336, "step": 2727 }, { "epoch": 1.9170765987350666, "grad_norm": 0.2354782372713089, "learning_rate": 1.1489810260014057e-05, "loss": 0.0354, "step": 2728 }, { "epoch": 1.9177793394237526, "grad_norm": 0.22967571020126343, "learning_rate": 1.1494026704146171e-05, "loss": 0.0153, "step": 2729 }, { "epoch": 1.9184820801124385, "grad_norm": 0.25236091017723083, "learning_rate": 1.1498243148278286e-05, "loss": 0.03, "step": 2730 }, { "epoch": 1.9191848208011244, "grad_norm": 0.34307876229286194, "learning_rate": 1.1502459592410401e-05, "loss": 0.0549, "step": 2731 }, { "epoch": 1.9198875614898103, "grad_norm": 0.43649712204933167, "learning_rate": 1.1506676036542516e-05, "loss": 0.0509, "step": 2732 }, { "epoch": 1.9205903021784962, "grad_norm": 0.25351276993751526, "learning_rate": 1.1510892480674631e-05, "loss": 0.0386, "step": 2733 }, { "epoch": 1.9212930428671822, "grad_norm": 0.340570867061615, "learning_rate": 1.1515108924806746e-05, "loss": 0.0453, "step": 2734 }, { "epoch": 1.9219957835558679, "grad_norm": 0.45349904894828796, "learning_rate": 1.1519325368938863e-05, "loss": 0.0356, "step": 2735 }, { "epoch": 1.9226985242445538, "grad_norm": 0.3863717019557953, "learning_rate": 1.1523541813070978e-05, "loss": 0.0508, "step": 2736 }, { "epoch": 1.9234012649332395, "grad_norm": 0.2537796199321747, "learning_rate": 1.1527758257203093e-05, "loss": 0.0313, "step": 2737 }, { "epoch": 1.9241040056219254, "grad_norm": 0.3609955310821533, "learning_rate": 1.1531974701335208e-05, "loss": 0.072, "step": 2738 }, { "epoch": 1.9248067463106113, "grad_norm": 0.29290246963500977, "learning_rate": 1.1536191145467323e-05, "loss": 0.0474, "step": 2739 }, { "epoch": 1.9255094869992972, "grad_norm": 1.0793774127960205, "learning_rate": 1.1540407589599438e-05, "loss": 0.0738, "step": 2740 }, { "epoch": 1.9262122276879832, "grad_norm": 0.3895741105079651, "learning_rate": 1.1544624033731554e-05, "loss": 0.0575, "step": 2741 }, { "epoch": 1.926914968376669, "grad_norm": 0.4739744961261749, "learning_rate": 1.1548840477863669e-05, "loss": 0.0875, "step": 2742 }, { "epoch": 1.927617709065355, "grad_norm": 0.4585334062576294, "learning_rate": 1.1553056921995784e-05, "loss": 0.0908, "step": 2743 }, { "epoch": 1.9283204497540407, "grad_norm": 0.5543898344039917, "learning_rate": 1.1557273366127899e-05, "loss": 0.1312, "step": 2744 }, { "epoch": 1.9290231904427266, "grad_norm": 0.8631575703620911, "learning_rate": 1.1561489810260014e-05, "loss": 0.2421, "step": 2745 }, { "epoch": 1.9297259311314126, "grad_norm": 3.9379825592041016, "learning_rate": 1.1565706254392129e-05, "loss": 0.2287, "step": 2746 }, { "epoch": 1.9304286718200983, "grad_norm": 2.950984477996826, "learning_rate": 1.1569922698524244e-05, "loss": 0.3302, "step": 2747 }, { "epoch": 1.9311314125087842, "grad_norm": 2.253082752227783, "learning_rate": 1.157413914265636e-05, "loss": 0.4774, "step": 2748 }, { "epoch": 1.93183415319747, "grad_norm": 0.37361931800842285, "learning_rate": 1.1578355586788475e-05, "loss": 0.1101, "step": 2749 }, { "epoch": 1.932536893886156, "grad_norm": 0.38015130162239075, "learning_rate": 1.158257203092059e-05, "loss": 0.042, "step": 2750 }, { "epoch": 1.933239634574842, "grad_norm": 0.2510800063610077, "learning_rate": 1.1586788475052705e-05, "loss": 0.0334, "step": 2751 }, { "epoch": 1.9339423752635279, "grad_norm": 0.2893836796283722, "learning_rate": 1.159100491918482e-05, "loss": 0.0567, "step": 2752 }, { "epoch": 1.9346451159522138, "grad_norm": 0.2308386266231537, "learning_rate": 1.1595221363316935e-05, "loss": 0.0357, "step": 2753 }, { "epoch": 1.9353478566408995, "grad_norm": 0.2768094539642334, "learning_rate": 1.159943780744905e-05, "loss": 0.0329, "step": 2754 }, { "epoch": 1.9360505973295854, "grad_norm": 0.4766906797885895, "learning_rate": 1.1603654251581167e-05, "loss": 0.0583, "step": 2755 }, { "epoch": 1.936753338018271, "grad_norm": 0.2268531322479248, "learning_rate": 1.1607870695713282e-05, "loss": 0.0285, "step": 2756 }, { "epoch": 1.937456078706957, "grad_norm": 0.3206336796283722, "learning_rate": 1.1612087139845398e-05, "loss": 0.0656, "step": 2757 }, { "epoch": 1.938158819395643, "grad_norm": 0.23756669461727142, "learning_rate": 1.1616303583977513e-05, "loss": 0.0283, "step": 2758 }, { "epoch": 1.9388615600843289, "grad_norm": 0.2990582585334778, "learning_rate": 1.1620520028109628e-05, "loss": 0.0382, "step": 2759 }, { "epoch": 1.9395643007730148, "grad_norm": 0.29084646701812744, "learning_rate": 1.1624736472241743e-05, "loss": 0.0359, "step": 2760 }, { "epoch": 1.9402670414617007, "grad_norm": 0.6194847226142883, "learning_rate": 1.1628952916373858e-05, "loss": 0.0522, "step": 2761 }, { "epoch": 1.9409697821503866, "grad_norm": 0.22619032859802246, "learning_rate": 1.1633169360505975e-05, "loss": 0.0307, "step": 2762 }, { "epoch": 1.9416725228390725, "grad_norm": 0.2835673391819, "learning_rate": 1.163738580463809e-05, "loss": 0.0414, "step": 2763 }, { "epoch": 1.9423752635277582, "grad_norm": 0.42814886569976807, "learning_rate": 1.1641602248770205e-05, "loss": 0.0439, "step": 2764 }, { "epoch": 1.9430780042164442, "grad_norm": 0.35671260952949524, "learning_rate": 1.164581869290232e-05, "loss": 0.0504, "step": 2765 }, { "epoch": 1.9437807449051299, "grad_norm": 0.2916353642940521, "learning_rate": 1.1650035137034435e-05, "loss": 0.0562, "step": 2766 }, { "epoch": 1.9444834855938158, "grad_norm": 0.4945346415042877, "learning_rate": 1.165425158116655e-05, "loss": 0.0582, "step": 2767 }, { "epoch": 1.9451862262825017, "grad_norm": 0.40402743220329285, "learning_rate": 1.1658468025298664e-05, "loss": 0.0627, "step": 2768 }, { "epoch": 1.9458889669711876, "grad_norm": 0.945106565952301, "learning_rate": 1.1662684469430781e-05, "loss": 0.1468, "step": 2769 }, { "epoch": 1.9465917076598735, "grad_norm": 0.6965932846069336, "learning_rate": 1.1666900913562896e-05, "loss": 0.1596, "step": 2770 }, { "epoch": 1.9472944483485595, "grad_norm": 1.0070009231567383, "learning_rate": 1.1671117357695011e-05, "loss": 0.2702, "step": 2771 }, { "epoch": 1.9479971890372454, "grad_norm": 1.1756887435913086, "learning_rate": 1.1675333801827126e-05, "loss": 0.36, "step": 2772 }, { "epoch": 1.948699929725931, "grad_norm": 2.740370273590088, "learning_rate": 1.1679550245959241e-05, "loss": 0.3983, "step": 2773 }, { "epoch": 1.949402670414617, "grad_norm": 0.5158764719963074, "learning_rate": 1.1683766690091356e-05, "loss": 0.1355, "step": 2774 }, { "epoch": 1.950105411103303, "grad_norm": 0.28075751662254333, "learning_rate": 1.1687983134223472e-05, "loss": 0.0466, "step": 2775 }, { "epoch": 1.9508081517919886, "grad_norm": 0.29191702604293823, "learning_rate": 1.1692199578355587e-05, "loss": 0.0696, "step": 2776 }, { "epoch": 1.9515108924806746, "grad_norm": 0.2342105358839035, "learning_rate": 1.1696416022487702e-05, "loss": 0.0329, "step": 2777 }, { "epoch": 1.9522136331693605, "grad_norm": 0.25120899081230164, "learning_rate": 1.1700632466619817e-05, "loss": 0.038, "step": 2778 }, { "epoch": 1.9529163738580464, "grad_norm": 0.28990647196769714, "learning_rate": 1.1704848910751932e-05, "loss": 0.0348, "step": 2779 }, { "epoch": 1.9536191145467323, "grad_norm": 0.2647773325443268, "learning_rate": 1.1709065354884047e-05, "loss": 0.0348, "step": 2780 }, { "epoch": 1.9543218552354182, "grad_norm": 0.28092798590660095, "learning_rate": 1.1713281799016162e-05, "loss": 0.0315, "step": 2781 }, { "epoch": 1.9550245959241042, "grad_norm": 0.25110742449760437, "learning_rate": 1.1717498243148279e-05, "loss": 0.046, "step": 2782 }, { "epoch": 1.9557273366127899, "grad_norm": 0.36622053384780884, "learning_rate": 1.1721714687280394e-05, "loss": 0.0392, "step": 2783 }, { "epoch": 1.9564300773014758, "grad_norm": 0.3401932716369629, "learning_rate": 1.1725931131412509e-05, "loss": 0.0479, "step": 2784 }, { "epoch": 1.9571328179901615, "grad_norm": 0.3324096202850342, "learning_rate": 1.1730147575544624e-05, "loss": 0.0345, "step": 2785 }, { "epoch": 1.9578355586788474, "grad_norm": 0.3713231086730957, "learning_rate": 1.1734364019676739e-05, "loss": 0.0596, "step": 2786 }, { "epoch": 1.9585382993675333, "grad_norm": 0.393730491399765, "learning_rate": 1.1738580463808855e-05, "loss": 0.0337, "step": 2787 }, { "epoch": 1.9592410400562192, "grad_norm": 0.4405197501182556, "learning_rate": 1.174279690794097e-05, "loss": 0.0826, "step": 2788 }, { "epoch": 1.9599437807449052, "grad_norm": 0.40682414174079895, "learning_rate": 1.1747013352073087e-05, "loss": 0.0621, "step": 2789 }, { "epoch": 1.960646521433591, "grad_norm": 0.3540037274360657, "learning_rate": 1.1751229796205202e-05, "loss": 0.0415, "step": 2790 }, { "epoch": 1.961349262122277, "grad_norm": 0.5013929605484009, "learning_rate": 1.1755446240337317e-05, "loss": 0.0704, "step": 2791 }, { "epoch": 1.9620520028109627, "grad_norm": 0.5110026001930237, "learning_rate": 1.1759662684469432e-05, "loss": 0.0786, "step": 2792 }, { "epoch": 1.9627547434996486, "grad_norm": 0.5240338444709778, "learning_rate": 1.1763879128601547e-05, "loss": 0.112, "step": 2793 }, { "epoch": 1.9634574841883345, "grad_norm": 0.7927791476249695, "learning_rate": 1.1768095572733661e-05, "loss": 0.1072, "step": 2794 }, { "epoch": 1.9641602248770202, "grad_norm": 0.9580129384994507, "learning_rate": 1.1772312016865776e-05, "loss": 0.1815, "step": 2795 }, { "epoch": 1.9648629655657062, "grad_norm": 1.7307252883911133, "learning_rate": 1.1776528460997893e-05, "loss": 0.2408, "step": 2796 }, { "epoch": 1.965565706254392, "grad_norm": 1.2730878591537476, "learning_rate": 1.1780744905130008e-05, "loss": 0.3007, "step": 2797 }, { "epoch": 1.966268446943078, "grad_norm": 2.0633625984191895, "learning_rate": 1.1784961349262123e-05, "loss": 0.3819, "step": 2798 }, { "epoch": 1.966971187631764, "grad_norm": 0.5009836554527283, "learning_rate": 1.1789177793394238e-05, "loss": 0.1277, "step": 2799 }, { "epoch": 1.9676739283204498, "grad_norm": 0.30392658710479736, "learning_rate": 1.1793394237526353e-05, "loss": 0.0531, "step": 2800 }, { "epoch": 1.9683766690091358, "grad_norm": 0.2776542007923126, "learning_rate": 1.1797610681658468e-05, "loss": 0.0464, "step": 2801 }, { "epoch": 1.9690794096978215, "grad_norm": 0.33126100897789, "learning_rate": 1.1801827125790583e-05, "loss": 0.0306, "step": 2802 }, { "epoch": 1.9697821503865074, "grad_norm": 0.3251432180404663, "learning_rate": 1.18060435699227e-05, "loss": 0.0369, "step": 2803 }, { "epoch": 1.970484891075193, "grad_norm": 0.2952777147293091, "learning_rate": 1.1810260014054814e-05, "loss": 0.0404, "step": 2804 }, { "epoch": 1.971187631763879, "grad_norm": 0.3518490195274353, "learning_rate": 1.181447645818693e-05, "loss": 0.0382, "step": 2805 }, { "epoch": 1.971890372452565, "grad_norm": 0.35332536697387695, "learning_rate": 1.1818692902319044e-05, "loss": 0.0397, "step": 2806 }, { "epoch": 1.9725931131412509, "grad_norm": 0.2700692415237427, "learning_rate": 1.1822909346451159e-05, "loss": 0.0417, "step": 2807 }, { "epoch": 1.9732958538299368, "grad_norm": 0.2534126043319702, "learning_rate": 1.1827125790583274e-05, "loss": 0.032, "step": 2808 }, { "epoch": 1.9739985945186227, "grad_norm": 0.28162723779678345, "learning_rate": 1.183134223471539e-05, "loss": 0.0578, "step": 2809 }, { "epoch": 1.9747013352073086, "grad_norm": 0.2786064147949219, "learning_rate": 1.1835558678847506e-05, "loss": 0.0342, "step": 2810 }, { "epoch": 1.9754040758959945, "grad_norm": 0.30585265159606934, "learning_rate": 1.183977512297962e-05, "loss": 0.0406, "step": 2811 }, { "epoch": 1.9761068165846802, "grad_norm": 0.3372497260570526, "learning_rate": 1.1843991567111736e-05, "loss": 0.0445, "step": 2812 }, { "epoch": 1.9768095572733662, "grad_norm": 0.3446290194988251, "learning_rate": 1.184820801124385e-05, "loss": 0.0667, "step": 2813 }, { "epoch": 1.9775122979620519, "grad_norm": 0.26143527030944824, "learning_rate": 1.1852424455375965e-05, "loss": 0.0481, "step": 2814 }, { "epoch": 1.9782150386507378, "grad_norm": 0.43539175391197205, "learning_rate": 1.185664089950808e-05, "loss": 0.0663, "step": 2815 }, { "epoch": 1.9789177793394237, "grad_norm": 0.39635521173477173, "learning_rate": 1.1860857343640197e-05, "loss": 0.0552, "step": 2816 }, { "epoch": 1.9796205200281096, "grad_norm": 0.32904621958732605, "learning_rate": 1.1865073787772314e-05, "loss": 0.0565, "step": 2817 }, { "epoch": 1.9803232607167955, "grad_norm": 0.5834107995033264, "learning_rate": 1.1869290231904429e-05, "loss": 0.1142, "step": 2818 }, { "epoch": 1.9810260014054815, "grad_norm": 0.5940921902656555, "learning_rate": 1.1873506676036544e-05, "loss": 0.1225, "step": 2819 }, { "epoch": 1.9817287420941674, "grad_norm": 2.7760889530181885, "learning_rate": 1.1877723120168658e-05, "loss": 0.222, "step": 2820 }, { "epoch": 1.982431482782853, "grad_norm": 1.1146138906478882, "learning_rate": 1.1881939564300773e-05, "loss": 0.2538, "step": 2821 }, { "epoch": 1.983134223471539, "grad_norm": 1.1787819862365723, "learning_rate": 1.1886156008432888e-05, "loss": 0.2881, "step": 2822 }, { "epoch": 1.983836964160225, "grad_norm": 2.0081374645233154, "learning_rate": 1.1890372452565005e-05, "loss": 0.4194, "step": 2823 }, { "epoch": 1.9845397048489106, "grad_norm": 0.6803329586982727, "learning_rate": 1.189458889669712e-05, "loss": 0.1227, "step": 2824 }, { "epoch": 1.9852424455375965, "grad_norm": 0.3133711516857147, "learning_rate": 1.1898805340829235e-05, "loss": 0.0365, "step": 2825 }, { "epoch": 1.9859451862262825, "grad_norm": 0.31210455298423767, "learning_rate": 1.190302178496135e-05, "loss": 0.0406, "step": 2826 }, { "epoch": 1.9866479269149684, "grad_norm": 0.45400431752204895, "learning_rate": 1.1907238229093465e-05, "loss": 0.0554, "step": 2827 }, { "epoch": 1.9873506676036543, "grad_norm": 0.7217941880226135, "learning_rate": 1.191145467322558e-05, "loss": 0.0253, "step": 2828 }, { "epoch": 1.9880534082923402, "grad_norm": 0.336223304271698, "learning_rate": 1.1915671117357695e-05, "loss": 0.0519, "step": 2829 }, { "epoch": 1.9887561489810262, "grad_norm": 0.23516389727592468, "learning_rate": 1.1919887561489811e-05, "loss": 0.0265, "step": 2830 }, { "epoch": 1.9894588896697118, "grad_norm": 0.30859455466270447, "learning_rate": 1.1924104005621926e-05, "loss": 0.0419, "step": 2831 }, { "epoch": 1.9901616303583978, "grad_norm": 0.27662378549575806, "learning_rate": 1.1928320449754041e-05, "loss": 0.0362, "step": 2832 }, { "epoch": 1.9908643710470835, "grad_norm": 0.3777928948402405, "learning_rate": 1.1932536893886156e-05, "loss": 0.0447, "step": 2833 }, { "epoch": 1.9915671117357694, "grad_norm": 0.2838674783706665, "learning_rate": 1.1936753338018271e-05, "loss": 0.0284, "step": 2834 }, { "epoch": 1.9922698524244553, "grad_norm": 0.32436951994895935, "learning_rate": 1.1940969782150386e-05, "loss": 0.0547, "step": 2835 }, { "epoch": 1.9929725931131412, "grad_norm": 0.6786455512046814, "learning_rate": 1.1945186226282501e-05, "loss": 0.0424, "step": 2836 }, { "epoch": 1.9936753338018272, "grad_norm": 1.2872897386550903, "learning_rate": 1.1949402670414618e-05, "loss": 0.0656, "step": 2837 }, { "epoch": 1.994378074490513, "grad_norm": 0.37479642033576965, "learning_rate": 1.1953619114546733e-05, "loss": 0.0276, "step": 2838 }, { "epoch": 1.995080815179199, "grad_norm": 1.3026833534240723, "learning_rate": 1.1957835558678847e-05, "loss": 0.0705, "step": 2839 }, { "epoch": 1.9957835558678847, "grad_norm": 0.48670580983161926, "learning_rate": 1.1962052002810962e-05, "loss": 0.074, "step": 2840 }, { "epoch": 1.9964862965565706, "grad_norm": 0.6039150953292847, "learning_rate": 1.1966268446943077e-05, "loss": 0.092, "step": 2841 }, { "epoch": 1.9971890372452565, "grad_norm": 0.6628938913345337, "learning_rate": 1.1970484891075192e-05, "loss": 0.1332, "step": 2842 }, { "epoch": 1.9978917779339422, "grad_norm": 0.8993576765060425, "learning_rate": 1.1974701335207309e-05, "loss": 0.2342, "step": 2843 }, { "epoch": 1.9985945186226282, "grad_norm": 1.1457359790802002, "learning_rate": 1.1978917779339424e-05, "loss": 0.2897, "step": 2844 }, { "epoch": 1.999297259311314, "grad_norm": 2.485281467437744, "learning_rate": 1.1983134223471539e-05, "loss": 0.3595, "step": 2845 }, { "epoch": 2.0, "grad_norm": 2.9693121910095215, "learning_rate": 1.1987350667603654e-05, "loss": 0.2286, "step": 2846 }, { "epoch": 2.000702740688686, "grad_norm": 0.4563138782978058, "learning_rate": 1.199156711173577e-05, "loss": 0.1185, "step": 2847 }, { "epoch": 2.001405481377372, "grad_norm": 0.23582391440868378, "learning_rate": 1.1995783555867885e-05, "loss": 0.0452, "step": 2848 }, { "epoch": 2.0021082220660578, "grad_norm": 0.3280401825904846, "learning_rate": 1.2e-05, "loss": 0.0372, "step": 2849 }, { "epoch": 2.0028109627547437, "grad_norm": 0.3775269091129303, "learning_rate": 1.2004216444132117e-05, "loss": 0.0482, "step": 2850 }, { "epoch": 2.003513703443429, "grad_norm": 0.2648159861564636, "learning_rate": 1.2008432888264232e-05, "loss": 0.0279, "step": 2851 }, { "epoch": 2.004216444132115, "grad_norm": 0.2571377158164978, "learning_rate": 1.2012649332396347e-05, "loss": 0.0357, "step": 2852 }, { "epoch": 2.004919184820801, "grad_norm": 0.26678454875946045, "learning_rate": 1.2016865776528462e-05, "loss": 0.0335, "step": 2853 }, { "epoch": 2.005621925509487, "grad_norm": 0.3674272894859314, "learning_rate": 1.2021082220660577e-05, "loss": 0.0628, "step": 2854 }, { "epoch": 2.006324666198173, "grad_norm": 0.2890166640281677, "learning_rate": 1.2025298664792692e-05, "loss": 0.0418, "step": 2855 }, { "epoch": 2.0070274068868588, "grad_norm": 0.2921932637691498, "learning_rate": 1.2029515108924807e-05, "loss": 0.0262, "step": 2856 }, { "epoch": 2.0077301475755447, "grad_norm": 0.2778792381286621, "learning_rate": 1.2033731553056923e-05, "loss": 0.0384, "step": 2857 }, { "epoch": 2.0084328882642306, "grad_norm": 0.25264909863471985, "learning_rate": 1.2037947997189038e-05, "loss": 0.0286, "step": 2858 }, { "epoch": 2.0091356289529165, "grad_norm": 0.37209030985832214, "learning_rate": 1.2042164441321153e-05, "loss": 0.0523, "step": 2859 }, { "epoch": 2.0098383696416025, "grad_norm": 0.26161596179008484, "learning_rate": 1.2046380885453268e-05, "loss": 0.0383, "step": 2860 }, { "epoch": 2.010541110330288, "grad_norm": 0.2997909188270569, "learning_rate": 1.2050597329585383e-05, "loss": 0.0412, "step": 2861 }, { "epoch": 2.011243851018974, "grad_norm": 0.9083963632583618, "learning_rate": 1.2054813773717498e-05, "loss": 0.0704, "step": 2862 }, { "epoch": 2.0119465917076598, "grad_norm": 0.42012226581573486, "learning_rate": 1.2059030217849613e-05, "loss": 0.048, "step": 2863 }, { "epoch": 2.0126493323963457, "grad_norm": 1.0164103507995605, "learning_rate": 1.206324666198173e-05, "loss": 0.071, "step": 2864 }, { "epoch": 2.0133520730850316, "grad_norm": 0.37000778317451477, "learning_rate": 1.2067463106113844e-05, "loss": 0.0591, "step": 2865 }, { "epoch": 2.0140548137737175, "grad_norm": 0.574242889881134, "learning_rate": 1.207167955024596e-05, "loss": 0.0939, "step": 2866 }, { "epoch": 2.0147575544624035, "grad_norm": 0.5159433484077454, "learning_rate": 1.2075895994378074e-05, "loss": 0.1282, "step": 2867 }, { "epoch": 2.0154602951510894, "grad_norm": 0.7215343713760376, "learning_rate": 1.208011243851019e-05, "loss": 0.1799, "step": 2868 }, { "epoch": 2.0161630358397753, "grad_norm": 0.9334829449653625, "learning_rate": 1.2084328882642304e-05, "loss": 0.2177, "step": 2869 }, { "epoch": 2.016865776528461, "grad_norm": 1.8935980796813965, "learning_rate": 1.208854532677442e-05, "loss": 0.3121, "step": 2870 }, { "epoch": 2.0175685172171467, "grad_norm": 3.4539740085601807, "learning_rate": 1.2092761770906536e-05, "loss": 0.3896, "step": 2871 }, { "epoch": 2.0182712579058326, "grad_norm": 0.7639572024345398, "learning_rate": 1.209697821503865e-05, "loss": 0.1068, "step": 2872 }, { "epoch": 2.0189739985945185, "grad_norm": 0.28653401136398315, "learning_rate": 1.2101194659170766e-05, "loss": 0.0478, "step": 2873 }, { "epoch": 2.0196767392832045, "grad_norm": 0.6405016183853149, "learning_rate": 1.210541110330288e-05, "loss": 0.0354, "step": 2874 }, { "epoch": 2.0203794799718904, "grad_norm": 0.28306445479393005, "learning_rate": 1.2109627547434996e-05, "loss": 0.0396, "step": 2875 }, { "epoch": 2.0210822206605763, "grad_norm": 0.23991714417934418, "learning_rate": 1.2113843991567112e-05, "loss": 0.0321, "step": 2876 }, { "epoch": 2.021784961349262, "grad_norm": 0.309123694896698, "learning_rate": 1.2118060435699229e-05, "loss": 0.0228, "step": 2877 }, { "epoch": 2.022487702037948, "grad_norm": 0.3135104179382324, "learning_rate": 1.2122276879831344e-05, "loss": 0.0416, "step": 2878 }, { "epoch": 2.023190442726634, "grad_norm": 0.23606090247631073, "learning_rate": 1.2126493323963459e-05, "loss": 0.0288, "step": 2879 }, { "epoch": 2.0238931834153195, "grad_norm": 0.24160878360271454, "learning_rate": 1.2130709768095574e-05, "loss": 0.0354, "step": 2880 }, { "epoch": 2.0245959241040055, "grad_norm": 0.21764712035655975, "learning_rate": 1.2134926212227689e-05, "loss": 0.0216, "step": 2881 }, { "epoch": 2.0252986647926914, "grad_norm": 0.4048672020435333, "learning_rate": 1.2139142656359804e-05, "loss": 0.0586, "step": 2882 }, { "epoch": 2.0260014054813773, "grad_norm": 0.31101834774017334, "learning_rate": 1.2143359100491919e-05, "loss": 0.0245, "step": 2883 }, { "epoch": 2.0267041461700632, "grad_norm": 0.36289337277412415, "learning_rate": 1.2147575544624035e-05, "loss": 0.0455, "step": 2884 }, { "epoch": 2.027406886858749, "grad_norm": 0.36584821343421936, "learning_rate": 1.215179198875615e-05, "loss": 0.0355, "step": 2885 }, { "epoch": 2.028109627547435, "grad_norm": 0.3486066460609436, "learning_rate": 1.2156008432888265e-05, "loss": 0.0474, "step": 2886 }, { "epoch": 2.028812368236121, "grad_norm": 0.4621623158454895, "learning_rate": 1.216022487702038e-05, "loss": 0.0524, "step": 2887 }, { "epoch": 2.029515108924807, "grad_norm": 0.365517258644104, "learning_rate": 1.2164441321152495e-05, "loss": 0.0466, "step": 2888 }, { "epoch": 2.030217849613493, "grad_norm": 0.40460509061813354, "learning_rate": 1.216865776528461e-05, "loss": 0.0672, "step": 2889 }, { "epoch": 2.0309205903021783, "grad_norm": 1.2685810327529907, "learning_rate": 1.2172874209416725e-05, "loss": 0.0697, "step": 2890 }, { "epoch": 2.0316233309908642, "grad_norm": 0.5714073777198792, "learning_rate": 1.2177090653548841e-05, "loss": 0.11, "step": 2891 }, { "epoch": 2.03232607167955, "grad_norm": 0.6665392518043518, "learning_rate": 1.2181307097680956e-05, "loss": 0.1459, "step": 2892 }, { "epoch": 2.033028812368236, "grad_norm": 0.7599381804466248, "learning_rate": 1.2185523541813071e-05, "loss": 0.1786, "step": 2893 }, { "epoch": 2.033731553056922, "grad_norm": 4.232546806335449, "learning_rate": 1.2189739985945186e-05, "loss": 0.2166, "step": 2894 }, { "epoch": 2.034434293745608, "grad_norm": 1.2311218976974487, "learning_rate": 1.2193956430077301e-05, "loss": 0.3186, "step": 2895 }, { "epoch": 2.035137034434294, "grad_norm": 1.4995874166488647, "learning_rate": 1.2198172874209416e-05, "loss": 0.363, "step": 2896 }, { "epoch": 2.0358397751229798, "grad_norm": 0.6868829727172852, "learning_rate": 1.2202389318341531e-05, "loss": 0.1196, "step": 2897 }, { "epoch": 2.0365425158116657, "grad_norm": 0.3886072635650635, "learning_rate": 1.2206605762473648e-05, "loss": 0.0499, "step": 2898 }, { "epoch": 2.037245256500351, "grad_norm": 0.31103506684303284, "learning_rate": 1.2210822206605763e-05, "loss": 0.0281, "step": 2899 }, { "epoch": 2.037947997189037, "grad_norm": 0.23856762051582336, "learning_rate": 1.2215038650737878e-05, "loss": 0.0383, "step": 2900 }, { "epoch": 2.038650737877723, "grad_norm": 0.2907688319683075, "learning_rate": 1.2219255094869993e-05, "loss": 0.0389, "step": 2901 }, { "epoch": 2.039353478566409, "grad_norm": 0.34626367688179016, "learning_rate": 1.2223471539002108e-05, "loss": 0.0357, "step": 2902 }, { "epoch": 2.040056219255095, "grad_norm": 0.3117671608924866, "learning_rate": 1.2227687983134223e-05, "loss": 0.0349, "step": 2903 }, { "epoch": 2.0407589599437808, "grad_norm": 0.2972533702850342, "learning_rate": 1.2231904427266337e-05, "loss": 0.0386, "step": 2904 }, { "epoch": 2.0414617006324667, "grad_norm": 0.37836503982543945, "learning_rate": 1.2236120871398454e-05, "loss": 0.0469, "step": 2905 }, { "epoch": 2.0421644413211526, "grad_norm": 0.3050505816936493, "learning_rate": 1.224033731553057e-05, "loss": 0.0436, "step": 2906 }, { "epoch": 2.0428671820098385, "grad_norm": 0.33676284551620483, "learning_rate": 1.2244553759662686e-05, "loss": 0.0366, "step": 2907 }, { "epoch": 2.0435699226985244, "grad_norm": 0.24349263310432434, "learning_rate": 1.22487702037948e-05, "loss": 0.0406, "step": 2908 }, { "epoch": 2.04427266338721, "grad_norm": 0.4282669723033905, "learning_rate": 1.2252986647926916e-05, "loss": 0.0641, "step": 2909 }, { "epoch": 2.044975404075896, "grad_norm": 0.37162238359451294, "learning_rate": 1.225720309205903e-05, "loss": 0.0357, "step": 2910 }, { "epoch": 2.0456781447645818, "grad_norm": 0.7500748038291931, "learning_rate": 1.2261419536191147e-05, "loss": 0.0753, "step": 2911 }, { "epoch": 2.0463808854532677, "grad_norm": 0.39147552847862244, "learning_rate": 1.2265635980323262e-05, "loss": 0.0622, "step": 2912 }, { "epoch": 2.0470836261419536, "grad_norm": 0.3086395561695099, "learning_rate": 1.2269852424455377e-05, "loss": 0.0354, "step": 2913 }, { "epoch": 2.0477863668306395, "grad_norm": 0.6542215943336487, "learning_rate": 1.2274068868587492e-05, "loss": 0.0643, "step": 2914 }, { "epoch": 2.0484891075193254, "grad_norm": 0.4524870812892914, "learning_rate": 1.2278285312719607e-05, "loss": 0.063, "step": 2915 }, { "epoch": 2.0491918482080114, "grad_norm": 0.76842200756073, "learning_rate": 1.2282501756851722e-05, "loss": 0.1242, "step": 2916 }, { "epoch": 2.0498945888966973, "grad_norm": 0.6292096972465515, "learning_rate": 1.2286718200983837e-05, "loss": 0.1102, "step": 2917 }, { "epoch": 2.050597329585383, "grad_norm": 0.8783617615699768, "learning_rate": 1.2290934645115953e-05, "loss": 0.1589, "step": 2918 }, { "epoch": 2.0513000702740687, "grad_norm": 1.2400596141815186, "learning_rate": 1.2295151089248068e-05, "loss": 0.2402, "step": 2919 }, { "epoch": 2.0520028109627546, "grad_norm": 1.8703125715255737, "learning_rate": 1.2299367533380183e-05, "loss": 0.3475, "step": 2920 }, { "epoch": 2.0527055516514405, "grad_norm": 2.2813849449157715, "learning_rate": 1.2303583977512298e-05, "loss": 0.3614, "step": 2921 }, { "epoch": 2.0534082923401265, "grad_norm": 0.6248001456260681, "learning_rate": 1.2307800421644413e-05, "loss": 0.1189, "step": 2922 }, { "epoch": 2.0541110330288124, "grad_norm": 0.3556523025035858, "learning_rate": 1.2312016865776528e-05, "loss": 0.0424, "step": 2923 }, { "epoch": 2.0548137737174983, "grad_norm": 0.34135302901268005, "learning_rate": 1.2316233309908643e-05, "loss": 0.0444, "step": 2924 }, { "epoch": 2.055516514406184, "grad_norm": 0.3060636818408966, "learning_rate": 1.232044975404076e-05, "loss": 0.03, "step": 2925 }, { "epoch": 2.05621925509487, "grad_norm": 0.47582587599754333, "learning_rate": 1.2324666198172875e-05, "loss": 0.0425, "step": 2926 }, { "epoch": 2.056921995783556, "grad_norm": 0.2571910619735718, "learning_rate": 1.232888264230499e-05, "loss": 0.0318, "step": 2927 }, { "epoch": 2.0576247364722415, "grad_norm": 0.2563827931880951, "learning_rate": 1.2333099086437105e-05, "loss": 0.0258, "step": 2928 }, { "epoch": 2.0583274771609275, "grad_norm": 0.4007963538169861, "learning_rate": 1.233731553056922e-05, "loss": 0.0398, "step": 2929 }, { "epoch": 2.0590302178496134, "grad_norm": 0.3389331102371216, "learning_rate": 1.2341531974701334e-05, "loss": 0.0481, "step": 2930 }, { "epoch": 2.0597329585382993, "grad_norm": 0.31101182103157043, "learning_rate": 1.234574841883345e-05, "loss": 0.033, "step": 2931 }, { "epoch": 2.060435699226985, "grad_norm": 0.35418251156806946, "learning_rate": 1.2349964862965566e-05, "loss": 0.0302, "step": 2932 }, { "epoch": 2.061138439915671, "grad_norm": 0.39879706501960754, "learning_rate": 1.2354181307097681e-05, "loss": 0.0375, "step": 2933 }, { "epoch": 2.061841180604357, "grad_norm": 0.3623579740524292, "learning_rate": 1.2358397751229796e-05, "loss": 0.0511, "step": 2934 }, { "epoch": 2.062543921293043, "grad_norm": 1.1529754400253296, "learning_rate": 1.2362614195361911e-05, "loss": 0.0421, "step": 2935 }, { "epoch": 2.063246661981729, "grad_norm": 0.3692602217197418, "learning_rate": 1.2366830639494028e-05, "loss": 0.0457, "step": 2936 }, { "epoch": 2.063949402670415, "grad_norm": 0.3376637399196625, "learning_rate": 1.2371047083626142e-05, "loss": 0.0515, "step": 2937 }, { "epoch": 2.0646521433591003, "grad_norm": 0.314151793718338, "learning_rate": 1.2375263527758257e-05, "loss": 0.0331, "step": 2938 }, { "epoch": 2.065354884047786, "grad_norm": 0.42498669028282166, "learning_rate": 1.2379479971890374e-05, "loss": 0.0434, "step": 2939 }, { "epoch": 2.066057624736472, "grad_norm": 0.31491994857788086, "learning_rate": 1.2383696416022489e-05, "loss": 0.052, "step": 2940 }, { "epoch": 2.066760365425158, "grad_norm": 0.6472083926200867, "learning_rate": 1.2387912860154604e-05, "loss": 0.1155, "step": 2941 }, { "epoch": 2.067463106113844, "grad_norm": 0.5078891515731812, "learning_rate": 1.2392129304286719e-05, "loss": 0.1161, "step": 2942 }, { "epoch": 2.06816584680253, "grad_norm": 1.9641292095184326, "learning_rate": 1.2396345748418834e-05, "loss": 0.1716, "step": 2943 }, { "epoch": 2.068868587491216, "grad_norm": 1.037089467048645, "learning_rate": 1.2400562192550949e-05, "loss": 0.2191, "step": 2944 }, { "epoch": 2.0695713281799017, "grad_norm": 1.0027744770050049, "learning_rate": 1.2404778636683065e-05, "loss": 0.2543, "step": 2945 }, { "epoch": 2.0702740688685877, "grad_norm": 1.8291233777999878, "learning_rate": 1.240899508081518e-05, "loss": 0.363, "step": 2946 }, { "epoch": 2.0709768095572736, "grad_norm": 0.5402472019195557, "learning_rate": 1.2413211524947295e-05, "loss": 0.148, "step": 2947 }, { "epoch": 2.071679550245959, "grad_norm": 0.4740919768810272, "learning_rate": 1.241742796907941e-05, "loss": 0.0452, "step": 2948 }, { "epoch": 2.072382290934645, "grad_norm": 0.46599218249320984, "learning_rate": 1.2421644413211525e-05, "loss": 0.0981, "step": 2949 }, { "epoch": 2.073085031623331, "grad_norm": 0.23900499939918518, "learning_rate": 1.242586085734364e-05, "loss": 0.0428, "step": 2950 }, { "epoch": 2.073787772312017, "grad_norm": 0.2577281594276428, "learning_rate": 1.2430077301475755e-05, "loss": 0.0465, "step": 2951 }, { "epoch": 2.0744905130007028, "grad_norm": 0.3151719272136688, "learning_rate": 1.2434293745607872e-05, "loss": 0.0361, "step": 2952 }, { "epoch": 2.0751932536893887, "grad_norm": 0.29078999161720276, "learning_rate": 1.2438510189739987e-05, "loss": 0.032, "step": 2953 }, { "epoch": 2.0758959943780746, "grad_norm": 0.2747107446193695, "learning_rate": 1.2442726633872102e-05, "loss": 0.0364, "step": 2954 }, { "epoch": 2.0765987350667605, "grad_norm": 0.39961159229278564, "learning_rate": 1.2446943078004217e-05, "loss": 0.0352, "step": 2955 }, { "epoch": 2.0773014757554464, "grad_norm": 0.30794844031333923, "learning_rate": 1.2451159522136331e-05, "loss": 0.0259, "step": 2956 }, { "epoch": 2.078004216444132, "grad_norm": 0.4534701406955719, "learning_rate": 1.2455375966268446e-05, "loss": 0.0544, "step": 2957 }, { "epoch": 2.078706957132818, "grad_norm": 0.30754607915878296, "learning_rate": 1.2459592410400561e-05, "loss": 0.0382, "step": 2958 }, { "epoch": 2.0794096978215038, "grad_norm": 0.5031552910804749, "learning_rate": 1.2463808854532678e-05, "loss": 0.0558, "step": 2959 }, { "epoch": 2.0801124385101897, "grad_norm": 0.9901471734046936, "learning_rate": 1.2468025298664793e-05, "loss": 0.0282, "step": 2960 }, { "epoch": 2.0808151791988756, "grad_norm": 0.43881064653396606, "learning_rate": 1.2472241742796908e-05, "loss": 0.0701, "step": 2961 }, { "epoch": 2.0815179198875615, "grad_norm": 0.5653929114341736, "learning_rate": 1.2476458186929023e-05, "loss": 0.0596, "step": 2962 }, { "epoch": 2.0822206605762474, "grad_norm": 0.30258908867836, "learning_rate": 1.2480674631061138e-05, "loss": 0.0335, "step": 2963 }, { "epoch": 2.0829234012649334, "grad_norm": 0.38461220264434814, "learning_rate": 1.2484891075193253e-05, "loss": 0.0521, "step": 2964 }, { "epoch": 2.0836261419536193, "grad_norm": 0.5833608508110046, "learning_rate": 1.2489107519325368e-05, "loss": 0.0861, "step": 2965 }, { "epoch": 2.084328882642305, "grad_norm": 0.5585396885871887, "learning_rate": 1.2493323963457486e-05, "loss": 0.0848, "step": 2966 }, { "epoch": 2.0850316233309907, "grad_norm": 0.7006297707557678, "learning_rate": 1.2497540407589601e-05, "loss": 0.1608, "step": 2967 }, { "epoch": 2.0857343640196766, "grad_norm": 1.414500117301941, "learning_rate": 1.2501756851721716e-05, "loss": 0.1726, "step": 2968 }, { "epoch": 2.0864371047083625, "grad_norm": 1.8166933059692383, "learning_rate": 1.250597329585383e-05, "loss": 0.2116, "step": 2969 }, { "epoch": 2.0871398453970484, "grad_norm": 1.1778051853179932, "learning_rate": 1.2510189739985946e-05, "loss": 0.2784, "step": 2970 }, { "epoch": 2.0878425860857344, "grad_norm": 4.777845859527588, "learning_rate": 1.251440618411806e-05, "loss": 0.3808, "step": 2971 }, { "epoch": 2.0885453267744203, "grad_norm": 0.4349726736545563, "learning_rate": 1.2518622628250176e-05, "loss": 0.1173, "step": 2972 }, { "epoch": 2.089248067463106, "grad_norm": 0.26682305335998535, "learning_rate": 1.2522839072382292e-05, "loss": 0.0409, "step": 2973 }, { "epoch": 2.089950808151792, "grad_norm": 0.4432196617126465, "learning_rate": 1.2527055516514407e-05, "loss": 0.0343, "step": 2974 }, { "epoch": 2.090653548840478, "grad_norm": 0.2685673236846924, "learning_rate": 1.2531271960646522e-05, "loss": 0.0435, "step": 2975 }, { "epoch": 2.0913562895291635, "grad_norm": 0.28980180621147156, "learning_rate": 1.2535488404778637e-05, "loss": 0.0392, "step": 2976 }, { "epoch": 2.0920590302178494, "grad_norm": 0.24780523777008057, "learning_rate": 1.2539704848910752e-05, "loss": 0.0253, "step": 2977 }, { "epoch": 2.0927617709065354, "grad_norm": 0.31899112462997437, "learning_rate": 1.2543921293042867e-05, "loss": 0.0342, "step": 2978 }, { "epoch": 2.0934645115952213, "grad_norm": 0.5176441073417664, "learning_rate": 1.2548137737174984e-05, "loss": 0.0241, "step": 2979 }, { "epoch": 2.094167252283907, "grad_norm": 0.32075491547584534, "learning_rate": 1.2552354181307099e-05, "loss": 0.0376, "step": 2980 }, { "epoch": 2.094869992972593, "grad_norm": 0.24556919932365417, "learning_rate": 1.2556570625439214e-05, "loss": 0.0233, "step": 2981 }, { "epoch": 2.095572733661279, "grad_norm": 0.4250466823577881, "learning_rate": 1.2560787069571328e-05, "loss": 0.0546, "step": 2982 }, { "epoch": 2.096275474349965, "grad_norm": 0.7153691649436951, "learning_rate": 1.2565003513703443e-05, "loss": 0.0364, "step": 2983 }, { "epoch": 2.096978215038651, "grad_norm": 0.32982057332992554, "learning_rate": 1.2569219957835558e-05, "loss": 0.0484, "step": 2984 }, { "epoch": 2.097680955727337, "grad_norm": 0.4787393808364868, "learning_rate": 1.2573436401967673e-05, "loss": 0.0371, "step": 2985 }, { "epoch": 2.0983836964160223, "grad_norm": 0.7167603373527527, "learning_rate": 1.257765284609979e-05, "loss": 0.0775, "step": 2986 }, { "epoch": 2.099086437104708, "grad_norm": 0.3367031514644623, "learning_rate": 1.2581869290231905e-05, "loss": 0.0553, "step": 2987 }, { "epoch": 2.099789177793394, "grad_norm": 0.6352349519729614, "learning_rate": 1.258608573436402e-05, "loss": 0.0573, "step": 2988 }, { "epoch": 2.10049191848208, "grad_norm": 0.32246488332748413, "learning_rate": 1.2590302178496135e-05, "loss": 0.0541, "step": 2989 }, { "epoch": 2.101194659170766, "grad_norm": 0.43373098969459534, "learning_rate": 1.259451862262825e-05, "loss": 0.0926, "step": 2990 }, { "epoch": 2.101897399859452, "grad_norm": 0.42193248867988586, "learning_rate": 1.2598735066760365e-05, "loss": 0.0745, "step": 2991 }, { "epoch": 2.102600140548138, "grad_norm": 0.8140786290168762, "learning_rate": 1.260295151089248e-05, "loss": 0.1211, "step": 2992 }, { "epoch": 2.1033028812368237, "grad_norm": 0.8078717589378357, "learning_rate": 1.2607167955024596e-05, "loss": 0.1637, "step": 2993 }, { "epoch": 2.1040056219255097, "grad_norm": 1.4868522882461548, "learning_rate": 1.2611384399156711e-05, "loss": 0.2643, "step": 2994 }, { "epoch": 2.104708362614195, "grad_norm": 1.482232689857483, "learning_rate": 1.2615600843288828e-05, "loss": 0.3177, "step": 2995 }, { "epoch": 2.105411103302881, "grad_norm": 2.122424602508545, "learning_rate": 1.2619817287420943e-05, "loss": 0.4279, "step": 2996 }, { "epoch": 2.106113843991567, "grad_norm": 0.3560352623462677, "learning_rate": 1.2624033731553058e-05, "loss": 0.092, "step": 2997 }, { "epoch": 2.106816584680253, "grad_norm": 0.22723506391048431, "learning_rate": 1.2628250175685173e-05, "loss": 0.047, "step": 2998 }, { "epoch": 2.107519325368939, "grad_norm": 0.4495648145675659, "learning_rate": 1.2632466619817288e-05, "loss": 0.0396, "step": 2999 }, { "epoch": 2.1082220660576247, "grad_norm": 0.31141534447669983, "learning_rate": 1.2636683063949404e-05, "loss": 0.0369, "step": 3000 }, { "epoch": 2.1082220660576247, "eval_cer": 0.20234471321178793, "eval_loss": 0.3288933038711548, "eval_runtime": 18.0788, "eval_samples_per_second": 251.012, "eval_steps_per_second": 0.83, "eval_wer": 0.37445881399960895, "step": 3000 }, { "epoch": 2.1089248067463107, "grad_norm": 0.2855953574180603, "learning_rate": 1.264089950808152e-05, "loss": 0.0422, "step": 3001 }, { "epoch": 2.1096275474349966, "grad_norm": 0.2087012380361557, "learning_rate": 1.2645115952213634e-05, "loss": 0.0253, "step": 3002 }, { "epoch": 2.1103302881236825, "grad_norm": 0.2310590296983719, "learning_rate": 1.2649332396345749e-05, "loss": 0.0293, "step": 3003 }, { "epoch": 2.1110330288123684, "grad_norm": 0.28470245003700256, "learning_rate": 1.2653548840477864e-05, "loss": 0.0351, "step": 3004 }, { "epoch": 2.111735769501054, "grad_norm": 0.2857593595981598, "learning_rate": 1.2657765284609979e-05, "loss": 0.047, "step": 3005 }, { "epoch": 2.11243851018974, "grad_norm": 0.3282656669616699, "learning_rate": 1.2661981728742094e-05, "loss": 0.0352, "step": 3006 }, { "epoch": 2.1131412508784257, "grad_norm": 12.11142635345459, "learning_rate": 1.266619817287421e-05, "loss": 0.0321, "step": 3007 }, { "epoch": 2.1138439915671117, "grad_norm": 0.23054993152618408, "learning_rate": 1.2670414617006325e-05, "loss": 0.0198, "step": 3008 }, { "epoch": 2.1145467322557976, "grad_norm": 0.3110561966896057, "learning_rate": 1.267463106113844e-05, "loss": 0.047, "step": 3009 }, { "epoch": 2.1152494729444835, "grad_norm": 0.4849814772605896, "learning_rate": 1.2678847505270555e-05, "loss": 0.0402, "step": 3010 }, { "epoch": 2.1159522136331694, "grad_norm": 0.7151211500167847, "learning_rate": 1.268306394940267e-05, "loss": 0.0439, "step": 3011 }, { "epoch": 2.1166549543218554, "grad_norm": 1.0706678628921509, "learning_rate": 1.2687280393534785e-05, "loss": 0.0455, "step": 3012 }, { "epoch": 2.1173576950105413, "grad_norm": 0.3770522177219391, "learning_rate": 1.2691496837666902e-05, "loss": 0.0561, "step": 3013 }, { "epoch": 2.118060435699227, "grad_norm": 0.4126117527484894, "learning_rate": 1.2695713281799017e-05, "loss": 0.0409, "step": 3014 }, { "epoch": 2.1187631763879127, "grad_norm": 2.052386522293091, "learning_rate": 1.2699929725931132e-05, "loss": 0.0854, "step": 3015 }, { "epoch": 2.1194659170765986, "grad_norm": 0.5254161953926086, "learning_rate": 1.2704146170063247e-05, "loss": 0.0609, "step": 3016 }, { "epoch": 2.1201686577652845, "grad_norm": 0.9241958260536194, "learning_rate": 1.2708362614195362e-05, "loss": 0.1455, "step": 3017 }, { "epoch": 2.1208713984539704, "grad_norm": 1.0538339614868164, "learning_rate": 1.2712579058327477e-05, "loss": 0.1668, "step": 3018 }, { "epoch": 2.1215741391426564, "grad_norm": 1.3994110822677612, "learning_rate": 1.2716795502459592e-05, "loss": 0.2712, "step": 3019 }, { "epoch": 2.1222768798313423, "grad_norm": 1.57978093624115, "learning_rate": 1.2721011946591708e-05, "loss": 0.3226, "step": 3020 }, { "epoch": 2.122979620520028, "grad_norm": 2.44092059135437, "learning_rate": 1.2725228390723823e-05, "loss": 0.3656, "step": 3021 }, { "epoch": 2.123682361208714, "grad_norm": 0.5525580048561096, "learning_rate": 1.2729444834855938e-05, "loss": 0.1138, "step": 3022 }, { "epoch": 2.1243851018974, "grad_norm": 0.31011345982551575, "learning_rate": 1.2733661278988053e-05, "loss": 0.0412, "step": 3023 }, { "epoch": 2.125087842586086, "grad_norm": 0.2504320442676544, "learning_rate": 1.2737877723120168e-05, "loss": 0.0309, "step": 3024 }, { "epoch": 2.1257905832747714, "grad_norm": 0.6903676390647888, "learning_rate": 1.2742094167252285e-05, "loss": 0.0519, "step": 3025 }, { "epoch": 2.1264933239634574, "grad_norm": 0.2780322730541229, "learning_rate": 1.27463106113844e-05, "loss": 0.052, "step": 3026 }, { "epoch": 2.1271960646521433, "grad_norm": 0.34662434458732605, "learning_rate": 1.2750527055516516e-05, "loss": 0.0324, "step": 3027 }, { "epoch": 2.127898805340829, "grad_norm": 0.28032705187797546, "learning_rate": 1.2754743499648631e-05, "loss": 0.0317, "step": 3028 }, { "epoch": 2.128601546029515, "grad_norm": 0.33138400316238403, "learning_rate": 1.2758959943780746e-05, "loss": 0.0409, "step": 3029 }, { "epoch": 2.129304286718201, "grad_norm": 0.3515057861804962, "learning_rate": 1.2763176387912861e-05, "loss": 0.0454, "step": 3030 }, { "epoch": 2.130007027406887, "grad_norm": 0.35484778881073, "learning_rate": 1.2767392832044976e-05, "loss": 0.0345, "step": 3031 }, { "epoch": 2.130709768095573, "grad_norm": 0.40305837988853455, "learning_rate": 1.2771609276177091e-05, "loss": 0.0592, "step": 3032 }, { "epoch": 2.131412508784259, "grad_norm": 0.25195011496543884, "learning_rate": 1.2775825720309206e-05, "loss": 0.0317, "step": 3033 }, { "epoch": 2.1321152494729443, "grad_norm": 0.5019016265869141, "learning_rate": 1.2780042164441322e-05, "loss": 0.0579, "step": 3034 }, { "epoch": 2.13281799016163, "grad_norm": 0.3073757290840149, "learning_rate": 1.2784258608573437e-05, "loss": 0.0263, "step": 3035 }, { "epoch": 2.133520730850316, "grad_norm": 0.3563641905784607, "learning_rate": 1.2788475052705552e-05, "loss": 0.0615, "step": 3036 }, { "epoch": 2.134223471539002, "grad_norm": 0.34574973583221436, "learning_rate": 1.2792691496837667e-05, "loss": 0.0546, "step": 3037 }, { "epoch": 2.134926212227688, "grad_norm": 0.26645392179489136, "learning_rate": 1.2796907940969782e-05, "loss": 0.0306, "step": 3038 }, { "epoch": 2.135628952916374, "grad_norm": 0.4935447573661804, "learning_rate": 1.2801124385101897e-05, "loss": 0.0679, "step": 3039 }, { "epoch": 2.13633169360506, "grad_norm": 0.4090310037136078, "learning_rate": 1.2805340829234012e-05, "loss": 0.0555, "step": 3040 }, { "epoch": 2.1370344342937457, "grad_norm": 0.429732084274292, "learning_rate": 1.2809557273366129e-05, "loss": 0.0961, "step": 3041 }, { "epoch": 2.1377371749824317, "grad_norm": 1.1195778846740723, "learning_rate": 1.2813773717498244e-05, "loss": 0.1064, "step": 3042 }, { "epoch": 2.1384399156711176, "grad_norm": 0.7463206648826599, "learning_rate": 1.2817990161630359e-05, "loss": 0.1973, "step": 3043 }, { "epoch": 2.139142656359803, "grad_norm": 1.1379718780517578, "learning_rate": 1.2822206605762474e-05, "loss": 0.248, "step": 3044 }, { "epoch": 2.139845397048489, "grad_norm": 1.156160593032837, "learning_rate": 1.2826423049894589e-05, "loss": 0.2759, "step": 3045 }, { "epoch": 2.140548137737175, "grad_norm": 1.9348074197769165, "learning_rate": 1.2830639494026703e-05, "loss": 0.3601, "step": 3046 }, { "epoch": 2.141250878425861, "grad_norm": 0.6660420894622803, "learning_rate": 1.283485593815882e-05, "loss": 0.1035, "step": 3047 }, { "epoch": 2.1419536191145467, "grad_norm": 0.23457416892051697, "learning_rate": 1.2839072382290935e-05, "loss": 0.0434, "step": 3048 }, { "epoch": 2.1426563598032327, "grad_norm": 0.283096581697464, "learning_rate": 1.284328882642305e-05, "loss": 0.061, "step": 3049 }, { "epoch": 2.1433591004919186, "grad_norm": 0.2589288055896759, "learning_rate": 1.2847505270555165e-05, "loss": 0.0483, "step": 3050 }, { "epoch": 2.1440618411806045, "grad_norm": 0.21459190547466278, "learning_rate": 1.285172171468728e-05, "loss": 0.0364, "step": 3051 }, { "epoch": 2.1447645818692904, "grad_norm": 0.2350425273180008, "learning_rate": 1.2855938158819395e-05, "loss": 0.0281, "step": 3052 }, { "epoch": 2.145467322557976, "grad_norm": 0.31857752799987793, "learning_rate": 1.286015460295151e-05, "loss": 0.034, "step": 3053 }, { "epoch": 2.146170063246662, "grad_norm": 0.28287264704704285, "learning_rate": 1.2864371047083626e-05, "loss": 0.0358, "step": 3054 }, { "epoch": 2.1468728039353477, "grad_norm": 0.2543136179447174, "learning_rate": 1.2868587491215743e-05, "loss": 0.0305, "step": 3055 }, { "epoch": 2.1475755446240337, "grad_norm": 0.18682125210762024, "learning_rate": 1.2872803935347858e-05, "loss": 0.0205, "step": 3056 }, { "epoch": 2.1482782853127196, "grad_norm": 0.30859819054603577, "learning_rate": 1.2877020379479973e-05, "loss": 0.0327, "step": 3057 }, { "epoch": 2.1489810260014055, "grad_norm": 0.280215859413147, "learning_rate": 1.2881236823612088e-05, "loss": 0.0341, "step": 3058 }, { "epoch": 2.1496837666900914, "grad_norm": 0.27274224162101746, "learning_rate": 1.2885453267744203e-05, "loss": 0.0419, "step": 3059 }, { "epoch": 2.1503865073787773, "grad_norm": 0.3223792016506195, "learning_rate": 1.2889669711876318e-05, "loss": 0.0426, "step": 3060 }, { "epoch": 2.1510892480674633, "grad_norm": 0.29320210218429565, "learning_rate": 1.2893886156008434e-05, "loss": 0.0487, "step": 3061 }, { "epoch": 2.151791988756149, "grad_norm": 0.24837614595890045, "learning_rate": 1.289810260014055e-05, "loss": 0.0445, "step": 3062 }, { "epoch": 2.1524947294448347, "grad_norm": 0.5331799983978271, "learning_rate": 1.2902319044272664e-05, "loss": 0.0559, "step": 3063 }, { "epoch": 2.1531974701335206, "grad_norm": 0.3422524034976959, "learning_rate": 1.290653548840478e-05, "loss": 0.048, "step": 3064 }, { "epoch": 2.1539002108222065, "grad_norm": 0.45376214385032654, "learning_rate": 1.2910751932536894e-05, "loss": 0.0906, "step": 3065 }, { "epoch": 2.1546029515108924, "grad_norm": 0.4387713670730591, "learning_rate": 1.2914968376669009e-05, "loss": 0.0896, "step": 3066 }, { "epoch": 2.1553056921995783, "grad_norm": 0.4942369759082794, "learning_rate": 1.2919184820801124e-05, "loss": 0.1041, "step": 3067 }, { "epoch": 2.1560084328882643, "grad_norm": 1.3524384498596191, "learning_rate": 1.292340126493324e-05, "loss": 0.1672, "step": 3068 }, { "epoch": 2.15671117357695, "grad_norm": 0.9710716605186462, "learning_rate": 1.2927617709065356e-05, "loss": 0.2282, "step": 3069 }, { "epoch": 2.157413914265636, "grad_norm": 1.367226243019104, "learning_rate": 1.293183415319747e-05, "loss": 0.277, "step": 3070 }, { "epoch": 2.158116654954322, "grad_norm": 2.1601908206939697, "learning_rate": 1.2936050597329586e-05, "loss": 0.3513, "step": 3071 }, { "epoch": 2.1588193956430075, "grad_norm": 0.3873579800128937, "learning_rate": 1.29402670414617e-05, "loss": 0.1103, "step": 3072 }, { "epoch": 2.1595221363316934, "grad_norm": 0.3076515793800354, "learning_rate": 1.2944483485593815e-05, "loss": 0.046, "step": 3073 }, { "epoch": 2.1602248770203794, "grad_norm": 0.20929107069969177, "learning_rate": 1.294869992972593e-05, "loss": 0.0331, "step": 3074 }, { "epoch": 2.1609276177090653, "grad_norm": 0.2781367301940918, "learning_rate": 1.2952916373858047e-05, "loss": 0.0367, "step": 3075 }, { "epoch": 2.161630358397751, "grad_norm": 0.33961671590805054, "learning_rate": 1.2957132817990162e-05, "loss": 0.0411, "step": 3076 }, { "epoch": 2.162333099086437, "grad_norm": 0.2485889345407486, "learning_rate": 1.2961349262122277e-05, "loss": 0.0403, "step": 3077 }, { "epoch": 2.163035839775123, "grad_norm": 0.32732972502708435, "learning_rate": 1.2965565706254392e-05, "loss": 0.0452, "step": 3078 }, { "epoch": 2.163738580463809, "grad_norm": 0.27798500657081604, "learning_rate": 1.2969782150386507e-05, "loss": 0.0328, "step": 3079 }, { "epoch": 2.164441321152495, "grad_norm": 0.37998178601264954, "learning_rate": 1.2973998594518622e-05, "loss": 0.044, "step": 3080 }, { "epoch": 2.165144061841181, "grad_norm": 0.20000596344470978, "learning_rate": 1.2978215038650738e-05, "loss": 0.0205, "step": 3081 }, { "epoch": 2.1658468025298663, "grad_norm": 0.3552049696445465, "learning_rate": 1.2982431482782853e-05, "loss": 0.0409, "step": 3082 }, { "epoch": 2.166549543218552, "grad_norm": 0.3400084972381592, "learning_rate": 1.2986647926914968e-05, "loss": 0.0294, "step": 3083 }, { "epoch": 2.167252283907238, "grad_norm": 0.4120009243488312, "learning_rate": 1.2990864371047083e-05, "loss": 0.0513, "step": 3084 }, { "epoch": 2.167955024595924, "grad_norm": 0.30695006251335144, "learning_rate": 1.29950808151792e-05, "loss": 0.0323, "step": 3085 }, { "epoch": 2.16865776528461, "grad_norm": 0.3618544936180115, "learning_rate": 1.2999297259311315e-05, "loss": 0.0525, "step": 3086 }, { "epoch": 2.169360505973296, "grad_norm": 0.36396515369415283, "learning_rate": 1.300351370344343e-05, "loss": 0.0384, "step": 3087 }, { "epoch": 2.170063246661982, "grad_norm": 0.35046884417533875, "learning_rate": 1.3007730147575546e-05, "loss": 0.0355, "step": 3088 }, { "epoch": 2.1707659873506677, "grad_norm": 0.46324628591537476, "learning_rate": 1.3011946591707661e-05, "loss": 0.0989, "step": 3089 }, { "epoch": 2.1714687280393536, "grad_norm": 0.4274302124977112, "learning_rate": 1.3016163035839776e-05, "loss": 0.0811, "step": 3090 }, { "epoch": 2.172171468728039, "grad_norm": 0.4428364932537079, "learning_rate": 1.3020379479971891e-05, "loss": 0.0864, "step": 3091 }, { "epoch": 2.172874209416725, "grad_norm": 0.6100673079490662, "learning_rate": 1.3024595924104006e-05, "loss": 0.1048, "step": 3092 }, { "epoch": 2.173576950105411, "grad_norm": 0.8233513832092285, "learning_rate": 1.3028812368236121e-05, "loss": 0.173, "step": 3093 }, { "epoch": 2.174279690794097, "grad_norm": 1.9886720180511475, "learning_rate": 1.3033028812368236e-05, "loss": 0.2451, "step": 3094 }, { "epoch": 2.174982431482783, "grad_norm": 1.3362778425216675, "learning_rate": 1.3037245256500353e-05, "loss": 0.3098, "step": 3095 }, { "epoch": 2.1756851721714687, "grad_norm": 2.290135383605957, "learning_rate": 1.3041461700632468e-05, "loss": 0.365, "step": 3096 }, { "epoch": 2.1763879128601546, "grad_norm": 0.4353142976760864, "learning_rate": 1.3045678144764583e-05, "loss": 0.11, "step": 3097 }, { "epoch": 2.1770906535488406, "grad_norm": 0.27298256754875183, "learning_rate": 1.3049894588896697e-05, "loss": 0.0474, "step": 3098 }, { "epoch": 2.1777933942375265, "grad_norm": 0.25703513622283936, "learning_rate": 1.3054111033028812e-05, "loss": 0.0488, "step": 3099 }, { "epoch": 2.1784961349262124, "grad_norm": 0.26303088665008545, "learning_rate": 1.3058327477160927e-05, "loss": 0.0282, "step": 3100 }, { "epoch": 2.1791988756148983, "grad_norm": 0.22789409756660461, "learning_rate": 1.3062543921293042e-05, "loss": 0.029, "step": 3101 }, { "epoch": 2.179901616303584, "grad_norm": 0.21372488141059875, "learning_rate": 1.3066760365425159e-05, "loss": 0.0315, "step": 3102 }, { "epoch": 2.1806043569922697, "grad_norm": 0.2928427755832672, "learning_rate": 1.3070976809557274e-05, "loss": 0.039, "step": 3103 }, { "epoch": 2.1813070976809557, "grad_norm": 0.790154755115509, "learning_rate": 1.3075193253689389e-05, "loss": 0.0376, "step": 3104 }, { "epoch": 2.1820098383696416, "grad_norm": 0.30888432264328003, "learning_rate": 1.3079409697821504e-05, "loss": 0.0494, "step": 3105 }, { "epoch": 2.1827125790583275, "grad_norm": 0.2653578817844391, "learning_rate": 1.3083626141953619e-05, "loss": 0.0195, "step": 3106 }, { "epoch": 2.1834153197470134, "grad_norm": 0.3262263834476471, "learning_rate": 1.3087842586085734e-05, "loss": 0.0473, "step": 3107 }, { "epoch": 2.1841180604356993, "grad_norm": 0.2975948452949524, "learning_rate": 1.3092059030217849e-05, "loss": 0.0326, "step": 3108 }, { "epoch": 2.1848208011243853, "grad_norm": 0.39753293991088867, "learning_rate": 1.3096275474349965e-05, "loss": 0.0504, "step": 3109 }, { "epoch": 2.185523541813071, "grad_norm": 0.23435404896736145, "learning_rate": 1.310049191848208e-05, "loss": 0.0337, "step": 3110 }, { "epoch": 2.1862262825017567, "grad_norm": 0.33791452646255493, "learning_rate": 1.3104708362614195e-05, "loss": 0.0517, "step": 3111 }, { "epoch": 2.1869290231904426, "grad_norm": 0.45352768898010254, "learning_rate": 1.310892480674631e-05, "loss": 0.0528, "step": 3112 }, { "epoch": 2.1876317638791285, "grad_norm": 0.3227742910385132, "learning_rate": 1.3113141250878425e-05, "loss": 0.0458, "step": 3113 }, { "epoch": 2.1883345045678144, "grad_norm": 0.3919812738895416, "learning_rate": 1.3117357695010542e-05, "loss": 0.0579, "step": 3114 }, { "epoch": 2.1890372452565003, "grad_norm": 0.371197372674942, "learning_rate": 1.3121574139142657e-05, "loss": 0.0836, "step": 3115 }, { "epoch": 2.1897399859451863, "grad_norm": 0.571222722530365, "learning_rate": 1.3125790583274773e-05, "loss": 0.0968, "step": 3116 }, { "epoch": 2.190442726633872, "grad_norm": 0.47646835446357727, "learning_rate": 1.3130007027406888e-05, "loss": 0.0996, "step": 3117 }, { "epoch": 2.191145467322558, "grad_norm": 0.8091238141059875, "learning_rate": 1.3134223471539003e-05, "loss": 0.1991, "step": 3118 }, { "epoch": 2.191848208011244, "grad_norm": 0.8845686316490173, "learning_rate": 1.3138439915671118e-05, "loss": 0.262, "step": 3119 }, { "epoch": 2.19255094869993, "grad_norm": 1.1125710010528564, "learning_rate": 1.3142656359803233e-05, "loss": 0.3232, "step": 3120 }, { "epoch": 2.1932536893886154, "grad_norm": 5.924344062805176, "learning_rate": 1.3146872803935348e-05, "loss": 0.4186, "step": 3121 }, { "epoch": 2.1939564300773013, "grad_norm": 0.36835604906082153, "learning_rate": 1.3151089248067465e-05, "loss": 0.0986, "step": 3122 }, { "epoch": 2.1946591707659873, "grad_norm": 0.4904063045978546, "learning_rate": 1.315530569219958e-05, "loss": 0.0477, "step": 3123 }, { "epoch": 2.195361911454673, "grad_norm": 0.28600019216537476, "learning_rate": 1.3159522136331695e-05, "loss": 0.0496, "step": 3124 }, { "epoch": 2.196064652143359, "grad_norm": 0.23607724905014038, "learning_rate": 1.316373858046381e-05, "loss": 0.0396, "step": 3125 }, { "epoch": 2.196767392832045, "grad_norm": 0.23304864764213562, "learning_rate": 1.3167955024595924e-05, "loss": 0.0349, "step": 3126 }, { "epoch": 2.197470133520731, "grad_norm": 0.42506828904151917, "learning_rate": 1.317217146872804e-05, "loss": 0.0364, "step": 3127 }, { "epoch": 2.198172874209417, "grad_norm": 0.2832190692424774, "learning_rate": 1.3176387912860154e-05, "loss": 0.0422, "step": 3128 }, { "epoch": 2.198875614898103, "grad_norm": 1.0195327997207642, "learning_rate": 1.3180604356992271e-05, "loss": 0.0328, "step": 3129 }, { "epoch": 2.1995783555867883, "grad_norm": 0.2617390751838684, "learning_rate": 1.3184820801124386e-05, "loss": 0.0441, "step": 3130 }, { "epoch": 2.200281096275474, "grad_norm": 0.21353855729103088, "learning_rate": 1.31890372452565e-05, "loss": 0.0262, "step": 3131 }, { "epoch": 2.20098383696416, "grad_norm": 0.4376658797264099, "learning_rate": 1.3193253689388616e-05, "loss": 0.0408, "step": 3132 }, { "epoch": 2.201686577652846, "grad_norm": 0.25406768918037415, "learning_rate": 1.319747013352073e-05, "loss": 0.0249, "step": 3133 }, { "epoch": 2.202389318341532, "grad_norm": 0.24452945590019226, "learning_rate": 1.3201686577652846e-05, "loss": 0.0477, "step": 3134 }, { "epoch": 2.203092059030218, "grad_norm": 0.6502084136009216, "learning_rate": 1.320590302178496e-05, "loss": 0.0291, "step": 3135 }, { "epoch": 2.203794799718904, "grad_norm": 0.4697689414024353, "learning_rate": 1.3210119465917077e-05, "loss": 0.0662, "step": 3136 }, { "epoch": 2.2044975404075897, "grad_norm": 0.31303325295448303, "learning_rate": 1.3214335910049192e-05, "loss": 0.0563, "step": 3137 }, { "epoch": 2.2052002810962756, "grad_norm": 0.2934325933456421, "learning_rate": 1.3218552354181307e-05, "loss": 0.0484, "step": 3138 }, { "epoch": 2.2059030217849616, "grad_norm": 0.4574054181575775, "learning_rate": 1.3222768798313422e-05, "loss": 0.0567, "step": 3139 }, { "epoch": 2.206605762473647, "grad_norm": 0.5277433395385742, "learning_rate": 1.3226985242445537e-05, "loss": 0.085, "step": 3140 }, { "epoch": 2.207308503162333, "grad_norm": 1.2316148281097412, "learning_rate": 1.3231201686577652e-05, "loss": 0.0842, "step": 3141 }, { "epoch": 2.208011243851019, "grad_norm": 1.43088698387146, "learning_rate": 1.3235418130709767e-05, "loss": 0.1261, "step": 3142 }, { "epoch": 2.208713984539705, "grad_norm": 0.661087691783905, "learning_rate": 1.3239634574841884e-05, "loss": 0.1768, "step": 3143 }, { "epoch": 2.2094167252283907, "grad_norm": 1.2505240440368652, "learning_rate": 1.3243851018974e-05, "loss": 0.2443, "step": 3144 }, { "epoch": 2.2101194659170766, "grad_norm": 1.3945940732955933, "learning_rate": 1.3248067463106115e-05, "loss": 0.299, "step": 3145 }, { "epoch": 2.2108222066057626, "grad_norm": 1.759690523147583, "learning_rate": 1.325228390723823e-05, "loss": 0.3821, "step": 3146 }, { "epoch": 2.2115249472944485, "grad_norm": 0.3721087574958801, "learning_rate": 1.3256500351370345e-05, "loss": 0.0958, "step": 3147 }, { "epoch": 2.2122276879831344, "grad_norm": 0.24932003021240234, "learning_rate": 1.326071679550246e-05, "loss": 0.0523, "step": 3148 }, { "epoch": 2.21293042867182, "grad_norm": 0.2168106883764267, "learning_rate": 1.3264933239634575e-05, "loss": 0.031, "step": 3149 }, { "epoch": 2.213633169360506, "grad_norm": 0.19056878983974457, "learning_rate": 1.3269149683766692e-05, "loss": 0.0281, "step": 3150 }, { "epoch": 2.2143359100491917, "grad_norm": 0.20576216280460358, "learning_rate": 1.3273366127898806e-05, "loss": 0.0345, "step": 3151 }, { "epoch": 2.2150386507378776, "grad_norm": 0.2165052890777588, "learning_rate": 1.3277582572030921e-05, "loss": 0.0326, "step": 3152 }, { "epoch": 2.2157413914265636, "grad_norm": 0.2670172452926636, "learning_rate": 1.3281799016163036e-05, "loss": 0.0368, "step": 3153 }, { "epoch": 2.2164441321152495, "grad_norm": 0.22803378105163574, "learning_rate": 1.3286015460295151e-05, "loss": 0.0379, "step": 3154 }, { "epoch": 2.2171468728039354, "grad_norm": 0.30932697653770447, "learning_rate": 1.3290231904427266e-05, "loss": 0.0439, "step": 3155 }, { "epoch": 2.2178496134926213, "grad_norm": 0.26504698395729065, "learning_rate": 1.3294448348559383e-05, "loss": 0.0255, "step": 3156 }, { "epoch": 2.2185523541813073, "grad_norm": 1.7388235330581665, "learning_rate": 1.3298664792691498e-05, "loss": 0.0896, "step": 3157 }, { "epoch": 2.219255094869993, "grad_norm": 0.4434608519077301, "learning_rate": 1.3302881236823613e-05, "loss": 0.0344, "step": 3158 }, { "epoch": 2.2199578355586786, "grad_norm": 0.3595481812953949, "learning_rate": 1.3307097680955728e-05, "loss": 0.0656, "step": 3159 }, { "epoch": 2.2206605762473646, "grad_norm": 0.2652469277381897, "learning_rate": 1.3311314125087843e-05, "loss": 0.0314, "step": 3160 }, { "epoch": 2.2213633169360505, "grad_norm": 0.3041810691356659, "learning_rate": 1.3315530569219958e-05, "loss": 0.0537, "step": 3161 }, { "epoch": 2.2220660576247364, "grad_norm": 0.409078985452652, "learning_rate": 1.3319747013352073e-05, "loss": 0.0482, "step": 3162 }, { "epoch": 2.2227687983134223, "grad_norm": 0.43749186396598816, "learning_rate": 1.332396345748419e-05, "loss": 0.0604, "step": 3163 }, { "epoch": 2.2234715390021083, "grad_norm": 0.37567707896232605, "learning_rate": 1.3328179901616304e-05, "loss": 0.0592, "step": 3164 }, { "epoch": 2.224174279690794, "grad_norm": 0.44984495639801025, "learning_rate": 1.3332396345748419e-05, "loss": 0.0855, "step": 3165 }, { "epoch": 2.22487702037948, "grad_norm": 0.5357404351234436, "learning_rate": 1.3336612789880534e-05, "loss": 0.0862, "step": 3166 }, { "epoch": 2.225579761068166, "grad_norm": 0.5499841570854187, "learning_rate": 1.3340829234012649e-05, "loss": 0.1216, "step": 3167 }, { "epoch": 2.2262825017568515, "grad_norm": 0.7411800026893616, "learning_rate": 1.3345045678144764e-05, "loss": 0.1533, "step": 3168 }, { "epoch": 2.2269852424455374, "grad_norm": 2.7273316383361816, "learning_rate": 1.3349262122276879e-05, "loss": 0.2451, "step": 3169 }, { "epoch": 2.2276879831342233, "grad_norm": 1.40744149684906, "learning_rate": 1.3353478566408995e-05, "loss": 0.3311, "step": 3170 }, { "epoch": 2.2283907238229093, "grad_norm": 2.2251124382019043, "learning_rate": 1.335769501054111e-05, "loss": 0.441, "step": 3171 }, { "epoch": 2.229093464511595, "grad_norm": 0.3625803589820862, "learning_rate": 1.3361911454673225e-05, "loss": 0.0968, "step": 3172 }, { "epoch": 2.229796205200281, "grad_norm": 0.22989565134048462, "learning_rate": 1.336612789880534e-05, "loss": 0.0396, "step": 3173 }, { "epoch": 2.230498945888967, "grad_norm": 0.2599412202835083, "learning_rate": 1.3370344342937457e-05, "loss": 0.0379, "step": 3174 }, { "epoch": 2.231201686577653, "grad_norm": 0.25295916199684143, "learning_rate": 1.3374560787069572e-05, "loss": 0.0383, "step": 3175 }, { "epoch": 2.231904427266339, "grad_norm": 0.19145286083221436, "learning_rate": 1.3378777231201687e-05, "loss": 0.0266, "step": 3176 }, { "epoch": 2.232607167955025, "grad_norm": 0.25801771879196167, "learning_rate": 1.3382993675333803e-05, "loss": 0.0414, "step": 3177 }, { "epoch": 2.2333099086437107, "grad_norm": 0.24709250032901764, "learning_rate": 1.3387210119465918e-05, "loss": 0.0375, "step": 3178 }, { "epoch": 2.234012649332396, "grad_norm": 0.3932255506515503, "learning_rate": 1.3391426563598033e-05, "loss": 0.047, "step": 3179 }, { "epoch": 2.234715390021082, "grad_norm": 0.3239891529083252, "learning_rate": 1.3395643007730148e-05, "loss": 0.0511, "step": 3180 }, { "epoch": 2.235418130709768, "grad_norm": 0.24955859780311584, "learning_rate": 1.3399859451862263e-05, "loss": 0.0284, "step": 3181 }, { "epoch": 2.236120871398454, "grad_norm": 0.24324165284633636, "learning_rate": 1.3404075895994378e-05, "loss": 0.0402, "step": 3182 }, { "epoch": 2.23682361208714, "grad_norm": 0.2951580584049225, "learning_rate": 1.3408292340126493e-05, "loss": 0.0395, "step": 3183 }, { "epoch": 2.237526352775826, "grad_norm": 0.23842023313045502, "learning_rate": 1.341250878425861e-05, "loss": 0.0402, "step": 3184 }, { "epoch": 2.2382290934645117, "grad_norm": 0.3254795968532562, "learning_rate": 1.3416725228390725e-05, "loss": 0.0217, "step": 3185 }, { "epoch": 2.2389318341531976, "grad_norm": 0.48053258657455444, "learning_rate": 1.342094167252284e-05, "loss": 0.0561, "step": 3186 }, { "epoch": 2.2396345748418836, "grad_norm": 0.3371266722679138, "learning_rate": 1.3425158116654955e-05, "loss": 0.0571, "step": 3187 }, { "epoch": 2.240337315530569, "grad_norm": 0.25123727321624756, "learning_rate": 1.342937456078707e-05, "loss": 0.0403, "step": 3188 }, { "epoch": 2.241040056219255, "grad_norm": 0.5430963039398193, "learning_rate": 1.3433591004919184e-05, "loss": 0.0644, "step": 3189 }, { "epoch": 2.241742796907941, "grad_norm": 0.4922007620334625, "learning_rate": 1.3437807449051301e-05, "loss": 0.0743, "step": 3190 }, { "epoch": 2.242445537596627, "grad_norm": 0.5128400325775146, "learning_rate": 1.3442023893183416e-05, "loss": 0.1001, "step": 3191 }, { "epoch": 2.2431482782853127, "grad_norm": 0.529565155506134, "learning_rate": 1.3446240337315531e-05, "loss": 0.124, "step": 3192 }, { "epoch": 2.2438510189739986, "grad_norm": 1.1296465396881104, "learning_rate": 1.3450456781447646e-05, "loss": 0.1675, "step": 3193 }, { "epoch": 2.2445537596626846, "grad_norm": 0.8408706188201904, "learning_rate": 1.3454673225579761e-05, "loss": 0.2161, "step": 3194 }, { "epoch": 2.2452565003513705, "grad_norm": 1.2847223281860352, "learning_rate": 1.3458889669711876e-05, "loss": 0.2833, "step": 3195 }, { "epoch": 2.2459592410400564, "grad_norm": 2.5208239555358887, "learning_rate": 1.346310611384399e-05, "loss": 0.3981, "step": 3196 }, { "epoch": 2.2466619817287423, "grad_norm": 0.39608240127563477, "learning_rate": 1.3467322557976107e-05, "loss": 0.1115, "step": 3197 }, { "epoch": 2.247364722417428, "grad_norm": 0.4443519711494446, "learning_rate": 1.3471539002108222e-05, "loss": 0.0528, "step": 3198 }, { "epoch": 2.2480674631061137, "grad_norm": 0.6669421792030334, "learning_rate": 1.3475755446240337e-05, "loss": 0.0555, "step": 3199 }, { "epoch": 2.2487702037947996, "grad_norm": 0.18214555084705353, "learning_rate": 1.3479971890372452e-05, "loss": 0.0271, "step": 3200 }, { "epoch": 2.2494729444834856, "grad_norm": 0.24413613975048065, "learning_rate": 1.3484188334504567e-05, "loss": 0.0256, "step": 3201 }, { "epoch": 2.2501756851721715, "grad_norm": 0.2596517503261566, "learning_rate": 1.3488404778636682e-05, "loss": 0.0362, "step": 3202 }, { "epoch": 2.2508784258608574, "grad_norm": 0.2140227109193802, "learning_rate": 1.3492621222768797e-05, "loss": 0.0282, "step": 3203 }, { "epoch": 2.2515811665495433, "grad_norm": 0.25029003620147705, "learning_rate": 1.3496837666900915e-05, "loss": 0.0393, "step": 3204 }, { "epoch": 2.2522839072382292, "grad_norm": 0.2788681983947754, "learning_rate": 1.350105411103303e-05, "loss": 0.0464, "step": 3205 }, { "epoch": 2.2529866479269147, "grad_norm": 0.3422311544418335, "learning_rate": 1.3505270555165145e-05, "loss": 0.043, "step": 3206 }, { "epoch": 2.2536893886156006, "grad_norm": 0.3071677088737488, "learning_rate": 1.350948699929726e-05, "loss": 0.0521, "step": 3207 }, { "epoch": 2.2543921293042866, "grad_norm": 0.27246955037117004, "learning_rate": 1.3513703443429375e-05, "loss": 0.0322, "step": 3208 }, { "epoch": 2.2550948699929725, "grad_norm": 0.35023507475852966, "learning_rate": 1.351791988756149e-05, "loss": 0.0547, "step": 3209 }, { "epoch": 2.2557976106816584, "grad_norm": 0.31625455617904663, "learning_rate": 1.3522136331693605e-05, "loss": 0.0294, "step": 3210 }, { "epoch": 2.2565003513703443, "grad_norm": 0.3264665901660919, "learning_rate": 1.3526352775825722e-05, "loss": 0.0468, "step": 3211 }, { "epoch": 2.2572030920590302, "grad_norm": 0.43031784892082214, "learning_rate": 1.3530569219957837e-05, "loss": 0.0582, "step": 3212 }, { "epoch": 2.257905832747716, "grad_norm": 0.305205762386322, "learning_rate": 1.3534785664089952e-05, "loss": 0.0389, "step": 3213 }, { "epoch": 2.258608573436402, "grad_norm": 0.5539513230323792, "learning_rate": 1.3539002108222067e-05, "loss": 0.0716, "step": 3214 }, { "epoch": 2.259311314125088, "grad_norm": 0.38942474126815796, "learning_rate": 1.3543218552354181e-05, "loss": 0.0591, "step": 3215 }, { "epoch": 2.260014054813774, "grad_norm": 0.6809936165809631, "learning_rate": 1.3547434996486296e-05, "loss": 0.0896, "step": 3216 }, { "epoch": 2.2607167955024594, "grad_norm": 0.4015025496482849, "learning_rate": 1.3551651440618413e-05, "loss": 0.0996, "step": 3217 }, { "epoch": 2.2614195361911453, "grad_norm": 0.7304894924163818, "learning_rate": 1.3555867884750528e-05, "loss": 0.1581, "step": 3218 }, { "epoch": 2.2621222768798313, "grad_norm": 1.0164284706115723, "learning_rate": 1.3560084328882643e-05, "loss": 0.2236, "step": 3219 }, { "epoch": 2.262825017568517, "grad_norm": 1.018570065498352, "learning_rate": 1.3564300773014758e-05, "loss": 0.2609, "step": 3220 }, { "epoch": 2.263527758257203, "grad_norm": 1.6784039735794067, "learning_rate": 1.3568517217146873e-05, "loss": 0.385, "step": 3221 }, { "epoch": 2.264230498945889, "grad_norm": 0.25071272253990173, "learning_rate": 1.3572733661278988e-05, "loss": 0.069, "step": 3222 }, { "epoch": 2.264933239634575, "grad_norm": 0.24364469945430756, "learning_rate": 1.3576950105411103e-05, "loss": 0.0433, "step": 3223 }, { "epoch": 2.265635980323261, "grad_norm": 0.40623876452445984, "learning_rate": 1.358116654954322e-05, "loss": 0.0511, "step": 3224 }, { "epoch": 2.2663387210119468, "grad_norm": 0.249302476644516, "learning_rate": 1.3585382993675334e-05, "loss": 0.0331, "step": 3225 }, { "epoch": 2.2670414617006323, "grad_norm": 0.3460821807384491, "learning_rate": 1.358959943780745e-05, "loss": 0.0515, "step": 3226 }, { "epoch": 2.267744202389318, "grad_norm": 0.23608440160751343, "learning_rate": 1.3593815881939564e-05, "loss": 0.0202, "step": 3227 }, { "epoch": 2.268446943078004, "grad_norm": 0.24020685255527496, "learning_rate": 1.3598032326071679e-05, "loss": 0.0314, "step": 3228 }, { "epoch": 2.26914968376669, "grad_norm": 0.5192415118217468, "learning_rate": 1.3602248770203794e-05, "loss": 0.0504, "step": 3229 }, { "epoch": 2.269852424455376, "grad_norm": 0.312498539686203, "learning_rate": 1.3606465214335909e-05, "loss": 0.0444, "step": 3230 }, { "epoch": 2.270555165144062, "grad_norm": 0.2635650634765625, "learning_rate": 1.3610681658468026e-05, "loss": 0.0399, "step": 3231 }, { "epoch": 2.271257905832748, "grad_norm": 0.31939613819122314, "learning_rate": 1.361489810260014e-05, "loss": 0.0309, "step": 3232 }, { "epoch": 2.2719606465214337, "grad_norm": 0.23442929983139038, "learning_rate": 1.3619114546732256e-05, "loss": 0.0245, "step": 3233 }, { "epoch": 2.2726633872101196, "grad_norm": 0.30677178502082825, "learning_rate": 1.3623330990864372e-05, "loss": 0.0505, "step": 3234 }, { "epoch": 2.2733661278988055, "grad_norm": 0.2568587064743042, "learning_rate": 1.3627547434996487e-05, "loss": 0.0339, "step": 3235 }, { "epoch": 2.2740688685874915, "grad_norm": 0.35176411271095276, "learning_rate": 1.3631763879128602e-05, "loss": 0.0539, "step": 3236 }, { "epoch": 2.274771609276177, "grad_norm": 0.42089924216270447, "learning_rate": 1.3635980323260717e-05, "loss": 0.0498, "step": 3237 }, { "epoch": 2.275474349964863, "grad_norm": 0.2862355411052704, "learning_rate": 1.3640196767392834e-05, "loss": 0.0289, "step": 3238 }, { "epoch": 2.276177090653549, "grad_norm": 0.9567038416862488, "learning_rate": 1.3644413211524949e-05, "loss": 0.0683, "step": 3239 }, { "epoch": 2.2768798313422347, "grad_norm": 0.28430479764938354, "learning_rate": 1.3648629655657064e-05, "loss": 0.049, "step": 3240 }, { "epoch": 2.2775825720309206, "grad_norm": 0.6331989169120789, "learning_rate": 1.3652846099789178e-05, "loss": 0.0829, "step": 3241 }, { "epoch": 2.2782853127196065, "grad_norm": 0.5228856205940247, "learning_rate": 1.3657062543921293e-05, "loss": 0.1438, "step": 3242 }, { "epoch": 2.2789880534082925, "grad_norm": 0.6086238026618958, "learning_rate": 1.3661278988053408e-05, "loss": 0.1453, "step": 3243 }, { "epoch": 2.2796907940969784, "grad_norm": 1.1045573949813843, "learning_rate": 1.3665495432185523e-05, "loss": 0.2171, "step": 3244 }, { "epoch": 2.280393534785664, "grad_norm": 2.4791102409362793, "learning_rate": 1.366971187631764e-05, "loss": 0.3177, "step": 3245 }, { "epoch": 2.28109627547435, "grad_norm": 2.3178036212921143, "learning_rate": 1.3673928320449755e-05, "loss": 0.3489, "step": 3246 }, { "epoch": 2.2817990161630357, "grad_norm": 0.7250840663909912, "learning_rate": 1.367814476458187e-05, "loss": 0.1587, "step": 3247 }, { "epoch": 2.2825017568517216, "grad_norm": 0.29029154777526855, "learning_rate": 1.3682361208713985e-05, "loss": 0.0489, "step": 3248 }, { "epoch": 2.2832044975404076, "grad_norm": 0.37022534012794495, "learning_rate": 1.36865776528461e-05, "loss": 0.0567, "step": 3249 }, { "epoch": 2.2839072382290935, "grad_norm": 0.25606873631477356, "learning_rate": 1.3690794096978215e-05, "loss": 0.0309, "step": 3250 }, { "epoch": 2.2846099789177794, "grad_norm": 0.31609731912612915, "learning_rate": 1.3695010541110331e-05, "loss": 0.0424, "step": 3251 }, { "epoch": 2.2853127196064653, "grad_norm": 0.23898446559906006, "learning_rate": 1.3699226985242446e-05, "loss": 0.0348, "step": 3252 }, { "epoch": 2.2860154602951512, "grad_norm": 0.2779499888420105, "learning_rate": 1.3703443429374561e-05, "loss": 0.0263, "step": 3253 }, { "epoch": 2.286718200983837, "grad_norm": 0.2221638560295105, "learning_rate": 1.3707659873506676e-05, "loss": 0.0341, "step": 3254 }, { "epoch": 2.287420941672523, "grad_norm": 0.39680030941963196, "learning_rate": 1.3711876317638791e-05, "loss": 0.0437, "step": 3255 }, { "epoch": 2.2881236823612086, "grad_norm": 0.274086594581604, "learning_rate": 1.3716092761770906e-05, "loss": 0.03, "step": 3256 }, { "epoch": 2.2888264230498945, "grad_norm": 0.602416455745697, "learning_rate": 1.3720309205903021e-05, "loss": 0.0389, "step": 3257 }, { "epoch": 2.2895291637385804, "grad_norm": 0.6080679893493652, "learning_rate": 1.3724525650035138e-05, "loss": 0.0285, "step": 3258 }, { "epoch": 2.2902319044272663, "grad_norm": 0.3270157277584076, "learning_rate": 1.3728742094167253e-05, "loss": 0.043, "step": 3259 }, { "epoch": 2.2909346451159522, "grad_norm": 0.2817038893699646, "learning_rate": 1.3732958538299367e-05, "loss": 0.0504, "step": 3260 }, { "epoch": 2.291637385804638, "grad_norm": 0.4406129717826843, "learning_rate": 1.3737174982431482e-05, "loss": 0.0772, "step": 3261 }, { "epoch": 2.292340126493324, "grad_norm": 0.47057539224624634, "learning_rate": 1.3741391426563597e-05, "loss": 0.0793, "step": 3262 }, { "epoch": 2.29304286718201, "grad_norm": 0.3296564221382141, "learning_rate": 1.3745607870695714e-05, "loss": 0.0538, "step": 3263 }, { "epoch": 2.2937456078706955, "grad_norm": 0.5843701362609863, "learning_rate": 1.3749824314827829e-05, "loss": 0.0533, "step": 3264 }, { "epoch": 2.2944483485593814, "grad_norm": 0.3238708972930908, "learning_rate": 1.3754040758959946e-05, "loss": 0.0612, "step": 3265 }, { "epoch": 2.2951510892480673, "grad_norm": 0.46961793303489685, "learning_rate": 1.375825720309206e-05, "loss": 0.1247, "step": 3266 }, { "epoch": 2.2958538299367532, "grad_norm": 0.5814113616943359, "learning_rate": 1.3762473647224175e-05, "loss": 0.1173, "step": 3267 }, { "epoch": 2.296556570625439, "grad_norm": 0.7028322219848633, "learning_rate": 1.376669009135629e-05, "loss": 0.1712, "step": 3268 }, { "epoch": 2.297259311314125, "grad_norm": 0.9265245199203491, "learning_rate": 1.3770906535488405e-05, "loss": 0.2258, "step": 3269 }, { "epoch": 2.297962052002811, "grad_norm": 1.002249002456665, "learning_rate": 1.377512297962052e-05, "loss": 0.2493, "step": 3270 }, { "epoch": 2.298664792691497, "grad_norm": 1.7470648288726807, "learning_rate": 1.3779339423752635e-05, "loss": 0.33, "step": 3271 }, { "epoch": 2.299367533380183, "grad_norm": 0.7685773968696594, "learning_rate": 1.3783555867884752e-05, "loss": 0.1115, "step": 3272 }, { "epoch": 2.3000702740688688, "grad_norm": 0.26242631673812866, "learning_rate": 1.3787772312016867e-05, "loss": 0.0538, "step": 3273 }, { "epoch": 2.3007730147575547, "grad_norm": 0.26708871126174927, "learning_rate": 1.3791988756148982e-05, "loss": 0.0475, "step": 3274 }, { "epoch": 2.30147575544624, "grad_norm": 0.17911364138126373, "learning_rate": 1.3796205200281097e-05, "loss": 0.0304, "step": 3275 }, { "epoch": 2.302178496134926, "grad_norm": 0.20092913508415222, "learning_rate": 1.3800421644413212e-05, "loss": 0.0343, "step": 3276 }, { "epoch": 2.302881236823612, "grad_norm": 0.5377036929130554, "learning_rate": 1.3804638088545327e-05, "loss": 0.0412, "step": 3277 }, { "epoch": 2.303583977512298, "grad_norm": 0.280720591545105, "learning_rate": 1.3808854532677442e-05, "loss": 0.0301, "step": 3278 }, { "epoch": 2.304286718200984, "grad_norm": 0.21405811607837677, "learning_rate": 1.3813070976809558e-05, "loss": 0.0302, "step": 3279 }, { "epoch": 2.3049894588896698, "grad_norm": 0.26902270317077637, "learning_rate": 1.3817287420941673e-05, "loss": 0.0344, "step": 3280 }, { "epoch": 2.3056921995783557, "grad_norm": 0.287384033203125, "learning_rate": 1.3821503865073788e-05, "loss": 0.0371, "step": 3281 }, { "epoch": 2.3063949402670416, "grad_norm": 0.33529987931251526, "learning_rate": 1.3825720309205903e-05, "loss": 0.037, "step": 3282 }, { "epoch": 2.307097680955727, "grad_norm": 0.22246865928173065, "learning_rate": 1.3829936753338018e-05, "loss": 0.0324, "step": 3283 }, { "epoch": 2.307800421644413, "grad_norm": 0.25323474407196045, "learning_rate": 1.3834153197470133e-05, "loss": 0.034, "step": 3284 }, { "epoch": 2.308503162333099, "grad_norm": 0.27454984188079834, "learning_rate": 1.383836964160225e-05, "loss": 0.0401, "step": 3285 }, { "epoch": 2.309205903021785, "grad_norm": 0.3514236509799957, "learning_rate": 1.3842586085734364e-05, "loss": 0.0438, "step": 3286 }, { "epoch": 2.3099086437104708, "grad_norm": 0.48942214250564575, "learning_rate": 1.384680252986648e-05, "loss": 0.1071, "step": 3287 }, { "epoch": 2.3106113843991567, "grad_norm": 0.29276418685913086, "learning_rate": 1.3851018973998594e-05, "loss": 0.05, "step": 3288 }, { "epoch": 2.3113141250878426, "grad_norm": 0.7103308439254761, "learning_rate": 1.385523541813071e-05, "loss": 0.0603, "step": 3289 }, { "epoch": 2.3120168657765285, "grad_norm": 0.4448978006839752, "learning_rate": 1.3859451862262824e-05, "loss": 0.0708, "step": 3290 }, { "epoch": 2.3127196064652145, "grad_norm": 0.5096606612205505, "learning_rate": 1.386366830639494e-05, "loss": 0.0883, "step": 3291 }, { "epoch": 2.3134223471539004, "grad_norm": 0.7761560082435608, "learning_rate": 1.3867884750527056e-05, "loss": 0.1079, "step": 3292 }, { "epoch": 2.3141250878425863, "grad_norm": 0.6792739629745483, "learning_rate": 1.3872101194659172e-05, "loss": 0.1595, "step": 3293 }, { "epoch": 2.314827828531272, "grad_norm": 0.7794402241706848, "learning_rate": 1.3876317638791287e-05, "loss": 0.2052, "step": 3294 }, { "epoch": 2.3155305692199577, "grad_norm": 2.8527026176452637, "learning_rate": 1.3880534082923402e-05, "loss": 0.3025, "step": 3295 }, { "epoch": 2.3162333099086436, "grad_norm": 1.9183038473129272, "learning_rate": 1.3884750527055517e-05, "loss": 0.3517, "step": 3296 }, { "epoch": 2.3169360505973295, "grad_norm": 0.44932183623313904, "learning_rate": 1.3888966971187632e-05, "loss": 0.0966, "step": 3297 }, { "epoch": 2.3176387912860155, "grad_norm": 0.3786161541938782, "learning_rate": 1.3893183415319747e-05, "loss": 0.0422, "step": 3298 }, { "epoch": 2.3183415319747014, "grad_norm": 0.32649627327919006, "learning_rate": 1.3897399859451864e-05, "loss": 0.0487, "step": 3299 }, { "epoch": 2.3190442726633873, "grad_norm": 0.3434774577617645, "learning_rate": 1.3901616303583979e-05, "loss": 0.0484, "step": 3300 }, { "epoch": 2.3197470133520732, "grad_norm": 0.25850677490234375, "learning_rate": 1.3905832747716094e-05, "loss": 0.044, "step": 3301 }, { "epoch": 2.320449754040759, "grad_norm": 0.38014161586761475, "learning_rate": 1.3910049191848209e-05, "loss": 0.0323, "step": 3302 }, { "epoch": 2.3211524947294446, "grad_norm": 0.25905802845954895, "learning_rate": 1.3914265635980324e-05, "loss": 0.0289, "step": 3303 }, { "epoch": 2.3218552354181305, "grad_norm": 0.286445289850235, "learning_rate": 1.3918482080112439e-05, "loss": 0.0508, "step": 3304 }, { "epoch": 2.3225579761068165, "grad_norm": 0.290143758058548, "learning_rate": 1.3922698524244554e-05, "loss": 0.0506, "step": 3305 }, { "epoch": 2.3232607167955024, "grad_norm": 0.2093837708234787, "learning_rate": 1.392691496837667e-05, "loss": 0.0319, "step": 3306 }, { "epoch": 2.3239634574841883, "grad_norm": 0.28235024213790894, "learning_rate": 1.3931131412508785e-05, "loss": 0.0401, "step": 3307 }, { "epoch": 2.3246661981728742, "grad_norm": 0.22320349514484406, "learning_rate": 1.39353478566409e-05, "loss": 0.0293, "step": 3308 }, { "epoch": 2.32536893886156, "grad_norm": 0.28252044320106506, "learning_rate": 1.3939564300773015e-05, "loss": 0.0364, "step": 3309 }, { "epoch": 2.326071679550246, "grad_norm": 0.24402828514575958, "learning_rate": 1.394378074490513e-05, "loss": 0.0286, "step": 3310 }, { "epoch": 2.326774420238932, "grad_norm": 0.5991197228431702, "learning_rate": 1.3947997189037245e-05, "loss": 0.052, "step": 3311 }, { "epoch": 2.327477160927618, "grad_norm": 0.5048375725746155, "learning_rate": 1.395221363316936e-05, "loss": 0.0762, "step": 3312 }, { "epoch": 2.3281799016163034, "grad_norm": 0.5505686402320862, "learning_rate": 1.3956430077301476e-05, "loss": 0.0582, "step": 3313 }, { "epoch": 2.3288826423049893, "grad_norm": 0.45022690296173096, "learning_rate": 1.3960646521433591e-05, "loss": 0.057, "step": 3314 }, { "epoch": 2.3295853829936752, "grad_norm": 0.5352135896682739, "learning_rate": 1.3964862965565706e-05, "loss": 0.1115, "step": 3315 }, { "epoch": 2.330288123682361, "grad_norm": 0.5348096489906311, "learning_rate": 1.3969079409697821e-05, "loss": 0.0576, "step": 3316 }, { "epoch": 2.330990864371047, "grad_norm": 0.6303279399871826, "learning_rate": 1.3973295853829936e-05, "loss": 0.1374, "step": 3317 }, { "epoch": 2.331693605059733, "grad_norm": 0.7523888349533081, "learning_rate": 1.3977512297962051e-05, "loss": 0.2228, "step": 3318 }, { "epoch": 2.332396345748419, "grad_norm": 1.0385938882827759, "learning_rate": 1.3981728742094168e-05, "loss": 0.2239, "step": 3319 }, { "epoch": 2.333099086437105, "grad_norm": 1.5546519756317139, "learning_rate": 1.3985945186226283e-05, "loss": 0.3053, "step": 3320 }, { "epoch": 2.3338018271257908, "grad_norm": 2.4560482501983643, "learning_rate": 1.3990161630358398e-05, "loss": 0.4143, "step": 3321 }, { "epoch": 2.3345045678144762, "grad_norm": 0.495096355676651, "learning_rate": 1.3994378074490513e-05, "loss": 0.1307, "step": 3322 }, { "epoch": 2.335207308503162, "grad_norm": 0.3405030071735382, "learning_rate": 1.399859451862263e-05, "loss": 0.0537, "step": 3323 }, { "epoch": 2.335910049191848, "grad_norm": 0.19094496965408325, "learning_rate": 1.4002810962754744e-05, "loss": 0.0376, "step": 3324 }, { "epoch": 2.336612789880534, "grad_norm": 0.2641547620296478, "learning_rate": 1.400702740688686e-05, "loss": 0.0262, "step": 3325 }, { "epoch": 2.33731553056922, "grad_norm": 0.28343817591667175, "learning_rate": 1.4011243851018976e-05, "loss": 0.0276, "step": 3326 }, { "epoch": 2.338018271257906, "grad_norm": 0.24927422404289246, "learning_rate": 1.401546029515109e-05, "loss": 0.0299, "step": 3327 }, { "epoch": 2.3387210119465918, "grad_norm": 0.2230633795261383, "learning_rate": 1.4019676739283206e-05, "loss": 0.0236, "step": 3328 }, { "epoch": 2.3394237526352777, "grad_norm": 0.2494703084230423, "learning_rate": 1.402389318341532e-05, "loss": 0.0317, "step": 3329 }, { "epoch": 2.3401264933239636, "grad_norm": 0.2562674283981323, "learning_rate": 1.4028109627547436e-05, "loss": 0.0287, "step": 3330 }, { "epoch": 2.3408292340126495, "grad_norm": 0.21009135246276855, "learning_rate": 1.403232607167955e-05, "loss": 0.0218, "step": 3331 }, { "epoch": 2.3415319747013355, "grad_norm": 0.6608794331550598, "learning_rate": 1.4036542515811665e-05, "loss": 0.0411, "step": 3332 }, { "epoch": 2.342234715390021, "grad_norm": 0.23723439872264862, "learning_rate": 1.4040758959943782e-05, "loss": 0.0341, "step": 3333 }, { "epoch": 2.342937456078707, "grad_norm": 0.45123326778411865, "learning_rate": 1.4044975404075897e-05, "loss": 0.0611, "step": 3334 }, { "epoch": 2.3436401967673928, "grad_norm": 0.44394010305404663, "learning_rate": 1.4049191848208012e-05, "loss": 0.0374, "step": 3335 }, { "epoch": 2.3443429374560787, "grad_norm": 0.37015628814697266, "learning_rate": 1.4053408292340127e-05, "loss": 0.0554, "step": 3336 }, { "epoch": 2.3450456781447646, "grad_norm": 0.31616586446762085, "learning_rate": 1.4057624736472242e-05, "loss": 0.077, "step": 3337 }, { "epoch": 2.3457484188334505, "grad_norm": 0.32585352659225464, "learning_rate": 1.4061841180604357e-05, "loss": 0.0457, "step": 3338 }, { "epoch": 2.3464511595221365, "grad_norm": 0.31724005937576294, "learning_rate": 1.4066057624736472e-05, "loss": 0.0795, "step": 3339 }, { "epoch": 2.3471539002108224, "grad_norm": 0.363290011882782, "learning_rate": 1.4070274068868588e-05, "loss": 0.0581, "step": 3340 }, { "epoch": 2.347856640899508, "grad_norm": 0.4224042296409607, "learning_rate": 1.4074490513000703e-05, "loss": 0.0704, "step": 3341 }, { "epoch": 2.3485593815881938, "grad_norm": 0.569266140460968, "learning_rate": 1.4078706957132818e-05, "loss": 0.1202, "step": 3342 }, { "epoch": 2.3492621222768797, "grad_norm": 0.6829925775527954, "learning_rate": 1.4082923401264933e-05, "loss": 0.1636, "step": 3343 }, { "epoch": 2.3499648629655656, "grad_norm": 1.1720163822174072, "learning_rate": 1.4087139845397048e-05, "loss": 0.2545, "step": 3344 }, { "epoch": 2.3506676036542515, "grad_norm": 1.2502496242523193, "learning_rate": 1.4091356289529163e-05, "loss": 0.2895, "step": 3345 }, { "epoch": 2.3513703443429375, "grad_norm": 1.7600681781768799, "learning_rate": 1.4095572733661278e-05, "loss": 0.4034, "step": 3346 }, { "epoch": 2.3520730850316234, "grad_norm": 0.5339741110801697, "learning_rate": 1.4099789177793395e-05, "loss": 0.0938, "step": 3347 }, { "epoch": 2.3527758257203093, "grad_norm": 0.2705543041229248, "learning_rate": 1.410400562192551e-05, "loss": 0.0484, "step": 3348 }, { "epoch": 2.353478566408995, "grad_norm": 0.24811282753944397, "learning_rate": 1.4108222066057625e-05, "loss": 0.0389, "step": 3349 }, { "epoch": 2.354181307097681, "grad_norm": 0.26473328471183777, "learning_rate": 1.411243851018974e-05, "loss": 0.0316, "step": 3350 }, { "epoch": 2.354884047786367, "grad_norm": 0.2813819646835327, "learning_rate": 1.4116654954321854e-05, "loss": 0.0454, "step": 3351 }, { "epoch": 2.3555867884750525, "grad_norm": 0.19483429193496704, "learning_rate": 1.412087139845397e-05, "loss": 0.0322, "step": 3352 }, { "epoch": 2.3562895291637385, "grad_norm": 0.2645803391933441, "learning_rate": 1.4125087842586086e-05, "loss": 0.0288, "step": 3353 }, { "epoch": 2.3569922698524244, "grad_norm": 0.2739138901233673, "learning_rate": 1.4129304286718203e-05, "loss": 0.0349, "step": 3354 }, { "epoch": 2.3576950105411103, "grad_norm": 0.30203697085380554, "learning_rate": 1.4133520730850318e-05, "loss": 0.0416, "step": 3355 }, { "epoch": 2.3583977512297962, "grad_norm": 0.18639631569385529, "learning_rate": 1.4137737174982433e-05, "loss": 0.0304, "step": 3356 }, { "epoch": 2.359100491918482, "grad_norm": 0.24256548285484314, "learning_rate": 1.4141953619114548e-05, "loss": 0.032, "step": 3357 }, { "epoch": 2.359803232607168, "grad_norm": 0.29372134804725647, "learning_rate": 1.4146170063246662e-05, "loss": 0.0369, "step": 3358 }, { "epoch": 2.360505973295854, "grad_norm": 0.3121417164802551, "learning_rate": 1.4150386507378777e-05, "loss": 0.0489, "step": 3359 }, { "epoch": 2.3612087139845395, "grad_norm": 0.3481769561767578, "learning_rate": 1.4154602951510894e-05, "loss": 0.0463, "step": 3360 }, { "epoch": 2.3619114546732254, "grad_norm": 0.3122371733188629, "learning_rate": 1.4158819395643009e-05, "loss": 0.0526, "step": 3361 }, { "epoch": 2.3626141953619113, "grad_norm": 0.2837856709957123, "learning_rate": 1.4163035839775124e-05, "loss": 0.0514, "step": 3362 }, { "epoch": 2.3633169360505972, "grad_norm": 0.3209629952907562, "learning_rate": 1.4167252283907239e-05, "loss": 0.0448, "step": 3363 }, { "epoch": 2.364019676739283, "grad_norm": 0.3428134322166443, "learning_rate": 1.4171468728039354e-05, "loss": 0.0576, "step": 3364 }, { "epoch": 2.364722417427969, "grad_norm": 0.36037692427635193, "learning_rate": 1.4175685172171469e-05, "loss": 0.0513, "step": 3365 }, { "epoch": 2.365425158116655, "grad_norm": 0.3831271827220917, "learning_rate": 1.4179901616303584e-05, "loss": 0.0704, "step": 3366 }, { "epoch": 2.366127898805341, "grad_norm": 0.6167788505554199, "learning_rate": 1.41841180604357e-05, "loss": 0.0849, "step": 3367 }, { "epoch": 2.366830639494027, "grad_norm": 0.7320848107337952, "learning_rate": 1.4188334504567815e-05, "loss": 0.1841, "step": 3368 }, { "epoch": 2.3675333801827128, "grad_norm": 0.8266273736953735, "learning_rate": 1.419255094869993e-05, "loss": 0.2292, "step": 3369 }, { "epoch": 2.3682361208713987, "grad_norm": 1.0404242277145386, "learning_rate": 1.4196767392832045e-05, "loss": 0.2735, "step": 3370 }, { "epoch": 2.368938861560084, "grad_norm": 1.7094444036483765, "learning_rate": 1.420098383696416e-05, "loss": 0.3698, "step": 3371 }, { "epoch": 2.36964160224877, "grad_norm": 0.4014900028705597, "learning_rate": 1.4205200281096275e-05, "loss": 0.1046, "step": 3372 }, { "epoch": 2.370344342937456, "grad_norm": 0.23923063278198242, "learning_rate": 1.420941672522839e-05, "loss": 0.0463, "step": 3373 }, { "epoch": 2.371047083626142, "grad_norm": 0.38480404019355774, "learning_rate": 1.4213633169360507e-05, "loss": 0.0471, "step": 3374 }, { "epoch": 2.371749824314828, "grad_norm": 0.19407792389392853, "learning_rate": 1.4217849613492622e-05, "loss": 0.0284, "step": 3375 }, { "epoch": 2.3724525650035138, "grad_norm": 0.2562459707260132, "learning_rate": 1.4222066057624737e-05, "loss": 0.0363, "step": 3376 }, { "epoch": 2.3731553056921997, "grad_norm": 0.23097635805606842, "learning_rate": 1.4226282501756851e-05, "loss": 0.0326, "step": 3377 }, { "epoch": 2.3738580463808856, "grad_norm": 0.27697283029556274, "learning_rate": 1.4230498945888966e-05, "loss": 0.0392, "step": 3378 }, { "epoch": 2.3745607870695715, "grad_norm": 0.2775770425796509, "learning_rate": 1.4234715390021081e-05, "loss": 0.0348, "step": 3379 }, { "epoch": 2.375263527758257, "grad_norm": 0.25134024024009705, "learning_rate": 1.4238931834153196e-05, "loss": 0.043, "step": 3380 }, { "epoch": 2.375966268446943, "grad_norm": 0.2717861235141754, "learning_rate": 1.4243148278285313e-05, "loss": 0.0233, "step": 3381 }, { "epoch": 2.376669009135629, "grad_norm": 0.362402081489563, "learning_rate": 1.424736472241743e-05, "loss": 0.0616, "step": 3382 }, { "epoch": 2.3773717498243148, "grad_norm": 0.27408692240715027, "learning_rate": 1.4251581166549545e-05, "loss": 0.0383, "step": 3383 }, { "epoch": 2.3780744905130007, "grad_norm": 0.7401502132415771, "learning_rate": 1.425579761068166e-05, "loss": 0.0393, "step": 3384 }, { "epoch": 2.3787772312016866, "grad_norm": 0.36721646785736084, "learning_rate": 1.4260014054813774e-05, "loss": 0.0362, "step": 3385 }, { "epoch": 2.3794799718903725, "grad_norm": 0.5937895774841309, "learning_rate": 1.426423049894589e-05, "loss": 0.058, "step": 3386 }, { "epoch": 2.3801827125790584, "grad_norm": 0.34069305658340454, "learning_rate": 1.4268446943078004e-05, "loss": 0.0586, "step": 3387 }, { "epoch": 2.3808854532677444, "grad_norm": 0.33118873834609985, "learning_rate": 1.4272663387210121e-05, "loss": 0.0293, "step": 3388 }, { "epoch": 2.3815881939564303, "grad_norm": 0.3828001320362091, "learning_rate": 1.4276879831342236e-05, "loss": 0.0566, "step": 3389 }, { "epoch": 2.3822909346451158, "grad_norm": 0.3801080882549286, "learning_rate": 1.428109627547435e-05, "loss": 0.0693, "step": 3390 }, { "epoch": 2.3829936753338017, "grad_norm": 0.7566280961036682, "learning_rate": 1.4285312719606466e-05, "loss": 0.0871, "step": 3391 }, { "epoch": 2.3836964160224876, "grad_norm": 0.48874935507774353, "learning_rate": 1.428952916373858e-05, "loss": 0.1018, "step": 3392 }, { "epoch": 2.3843991567111735, "grad_norm": 0.7577298283576965, "learning_rate": 1.4293745607870696e-05, "loss": 0.1632, "step": 3393 }, { "epoch": 2.3851018973998594, "grad_norm": 1.1178438663482666, "learning_rate": 1.4297962052002812e-05, "loss": 0.2046, "step": 3394 }, { "epoch": 2.3858046380885454, "grad_norm": 1.5059897899627686, "learning_rate": 1.4302178496134927e-05, "loss": 0.3056, "step": 3395 }, { "epoch": 2.3865073787772313, "grad_norm": 2.5121805667877197, "learning_rate": 1.4306394940267042e-05, "loss": 0.3264, "step": 3396 }, { "epoch": 2.387210119465917, "grad_norm": 0.5984882712364197, "learning_rate": 1.4310611384399157e-05, "loss": 0.0888, "step": 3397 }, { "epoch": 2.387912860154603, "grad_norm": 0.2384800910949707, "learning_rate": 1.4314827828531272e-05, "loss": 0.0476, "step": 3398 }, { "epoch": 2.3886156008432886, "grad_norm": 0.242109015583992, "learning_rate": 1.4319044272663387e-05, "loss": 0.0429, "step": 3399 }, { "epoch": 2.3893183415319745, "grad_norm": 0.31154364347457886, "learning_rate": 1.4323260716795502e-05, "loss": 0.0481, "step": 3400 }, { "epoch": 2.3900210822206605, "grad_norm": 0.2217644453048706, "learning_rate": 1.4327477160927619e-05, "loss": 0.0373, "step": 3401 }, { "epoch": 2.3907238229093464, "grad_norm": 0.25375810265541077, "learning_rate": 1.4331693605059734e-05, "loss": 0.0368, "step": 3402 }, { "epoch": 2.3914265635980323, "grad_norm": 0.30135929584503174, "learning_rate": 1.4335910049191848e-05, "loss": 0.0394, "step": 3403 }, { "epoch": 2.392129304286718, "grad_norm": 0.290024995803833, "learning_rate": 1.4340126493323963e-05, "loss": 0.0426, "step": 3404 }, { "epoch": 2.392832044975404, "grad_norm": 0.2133394330739975, "learning_rate": 1.4344342937456078e-05, "loss": 0.0465, "step": 3405 }, { "epoch": 2.39353478566409, "grad_norm": 0.20148053765296936, "learning_rate": 1.4348559381588193e-05, "loss": 0.0231, "step": 3406 }, { "epoch": 2.394237526352776, "grad_norm": 0.3472469747066498, "learning_rate": 1.4352775825720308e-05, "loss": 0.0561, "step": 3407 }, { "epoch": 2.394940267041462, "grad_norm": 0.35633739829063416, "learning_rate": 1.4356992269852425e-05, "loss": 0.0364, "step": 3408 }, { "epoch": 2.395643007730148, "grad_norm": 0.3205013871192932, "learning_rate": 1.436120871398454e-05, "loss": 0.0478, "step": 3409 }, { "epoch": 2.3963457484188333, "grad_norm": 0.25371846556663513, "learning_rate": 1.4365425158116655e-05, "loss": 0.0283, "step": 3410 }, { "epoch": 2.397048489107519, "grad_norm": 0.31015413999557495, "learning_rate": 1.436964160224877e-05, "loss": 0.0424, "step": 3411 }, { "epoch": 2.397751229796205, "grad_norm": 0.33794960379600525, "learning_rate": 1.4373858046380886e-05, "loss": 0.0658, "step": 3412 }, { "epoch": 2.398453970484891, "grad_norm": 0.25703826546669006, "learning_rate": 1.4378074490513001e-05, "loss": 0.0288, "step": 3413 }, { "epoch": 2.399156711173577, "grad_norm": 0.4641595184803009, "learning_rate": 1.4382290934645116e-05, "loss": 0.0732, "step": 3414 }, { "epoch": 2.399859451862263, "grad_norm": 0.37456315755844116, "learning_rate": 1.4386507378777233e-05, "loss": 0.0694, "step": 3415 }, { "epoch": 2.400562192550949, "grad_norm": 0.38197964429855347, "learning_rate": 1.4390723822909348e-05, "loss": 0.0838, "step": 3416 }, { "epoch": 2.4012649332396347, "grad_norm": 0.5637369155883789, "learning_rate": 1.4394940267041463e-05, "loss": 0.0943, "step": 3417 }, { "epoch": 2.4019676739283202, "grad_norm": 2.7926406860351562, "learning_rate": 1.4399156711173578e-05, "loss": 0.1834, "step": 3418 }, { "epoch": 2.402670414617006, "grad_norm": 0.8567859530448914, "learning_rate": 1.4403373155305693e-05, "loss": 0.2306, "step": 3419 }, { "epoch": 2.403373155305692, "grad_norm": 0.991273045539856, "learning_rate": 1.4407589599437808e-05, "loss": 0.2781, "step": 3420 }, { "epoch": 2.404075895994378, "grad_norm": 1.5369490385055542, "learning_rate": 1.4411806043569923e-05, "loss": 0.3726, "step": 3421 }, { "epoch": 2.404778636683064, "grad_norm": 0.571907639503479, "learning_rate": 1.441602248770204e-05, "loss": 0.0917, "step": 3422 }, { "epoch": 2.40548137737175, "grad_norm": 0.22036424279212952, "learning_rate": 1.4420238931834154e-05, "loss": 0.0335, "step": 3423 }, { "epoch": 2.4061841180604358, "grad_norm": 1.5221922397613525, "learning_rate": 1.4424455375966269e-05, "loss": 0.0362, "step": 3424 }, { "epoch": 2.4068868587491217, "grad_norm": 0.18007510900497437, "learning_rate": 1.4428671820098384e-05, "loss": 0.0251, "step": 3425 }, { "epoch": 2.4075895994378076, "grad_norm": 0.25896477699279785, "learning_rate": 1.4432888264230499e-05, "loss": 0.0351, "step": 3426 }, { "epoch": 2.4082923401264935, "grad_norm": 0.25360971689224243, "learning_rate": 1.4437104708362614e-05, "loss": 0.0235, "step": 3427 }, { "epoch": 2.4089950808151794, "grad_norm": 0.31867870688438416, "learning_rate": 1.444132115249473e-05, "loss": 0.0387, "step": 3428 }, { "epoch": 2.409697821503865, "grad_norm": 0.24785318970680237, "learning_rate": 1.4445537596626845e-05, "loss": 0.0281, "step": 3429 }, { "epoch": 2.410400562192551, "grad_norm": 0.2819875478744507, "learning_rate": 1.444975404075896e-05, "loss": 0.0357, "step": 3430 }, { "epoch": 2.4111033028812368, "grad_norm": 0.23040196299552917, "learning_rate": 1.4453970484891075e-05, "loss": 0.0245, "step": 3431 }, { "epoch": 2.4118060435699227, "grad_norm": 0.26160141825675964, "learning_rate": 1.445818692902319e-05, "loss": 0.0435, "step": 3432 }, { "epoch": 2.4125087842586086, "grad_norm": 0.275932639837265, "learning_rate": 1.4462403373155305e-05, "loss": 0.0288, "step": 3433 }, { "epoch": 2.4132115249472945, "grad_norm": 0.3517106771469116, "learning_rate": 1.446661981728742e-05, "loss": 0.0377, "step": 3434 }, { "epoch": 2.4139142656359804, "grad_norm": 0.27155548334121704, "learning_rate": 1.4470836261419537e-05, "loss": 0.0522, "step": 3435 }, { "epoch": 2.4146170063246664, "grad_norm": 0.36793267726898193, "learning_rate": 1.4475052705551652e-05, "loss": 0.0312, "step": 3436 }, { "epoch": 2.415319747013352, "grad_norm": 0.39352932572364807, "learning_rate": 1.4479269149683767e-05, "loss": 0.0739, "step": 3437 }, { "epoch": 2.4160224877020378, "grad_norm": 0.21523676812648773, "learning_rate": 1.4483485593815882e-05, "loss": 0.0299, "step": 3438 }, { "epoch": 2.4167252283907237, "grad_norm": 0.3661257326602936, "learning_rate": 1.4487702037947997e-05, "loss": 0.052, "step": 3439 }, { "epoch": 2.4174279690794096, "grad_norm": 0.5253397822380066, "learning_rate": 1.4491918482080112e-05, "loss": 0.0572, "step": 3440 }, { "epoch": 2.4181307097680955, "grad_norm": 0.44567060470581055, "learning_rate": 1.4496134926212227e-05, "loss": 0.0583, "step": 3441 }, { "epoch": 2.4188334504567814, "grad_norm": 0.5172283053398132, "learning_rate": 1.4500351370344345e-05, "loss": 0.1338, "step": 3442 }, { "epoch": 2.4195361911454674, "grad_norm": 1.1310644149780273, "learning_rate": 1.450456781447646e-05, "loss": 0.161, "step": 3443 }, { "epoch": 2.4202389318341533, "grad_norm": 0.8058957457542419, "learning_rate": 1.4508784258608575e-05, "loss": 0.2467, "step": 3444 }, { "epoch": 2.420941672522839, "grad_norm": 1.1903133392333984, "learning_rate": 1.451300070274069e-05, "loss": 0.2804, "step": 3445 }, { "epoch": 2.421644413211525, "grad_norm": 2.0496950149536133, "learning_rate": 1.4517217146872805e-05, "loss": 0.3068, "step": 3446 }, { "epoch": 2.422347153900211, "grad_norm": 0.4151095747947693, "learning_rate": 1.452143359100492e-05, "loss": 0.0926, "step": 3447 }, { "epoch": 2.4230498945888965, "grad_norm": 0.26000019907951355, "learning_rate": 1.4525650035137034e-05, "loss": 0.0525, "step": 3448 }, { "epoch": 2.4237526352775824, "grad_norm": 0.2552748918533325, "learning_rate": 1.4529866479269151e-05, "loss": 0.0377, "step": 3449 }, { "epoch": 2.4244553759662684, "grad_norm": 0.28284838795661926, "learning_rate": 1.4534082923401266e-05, "loss": 0.0465, "step": 3450 }, { "epoch": 2.4251581166549543, "grad_norm": 0.26874974370002747, "learning_rate": 1.4538299367533381e-05, "loss": 0.0376, "step": 3451 }, { "epoch": 2.42586085734364, "grad_norm": 0.28883370757102966, "learning_rate": 1.4542515811665496e-05, "loss": 0.0292, "step": 3452 }, { "epoch": 2.426563598032326, "grad_norm": 0.21107397973537445, "learning_rate": 1.4546732255797611e-05, "loss": 0.0236, "step": 3453 }, { "epoch": 2.427266338721012, "grad_norm": 0.20843517780303955, "learning_rate": 1.4550948699929726e-05, "loss": 0.0271, "step": 3454 }, { "epoch": 2.427969079409698, "grad_norm": 0.2805272042751312, "learning_rate": 1.455516514406184e-05, "loss": 0.0424, "step": 3455 }, { "epoch": 2.4286718200983834, "grad_norm": 0.2217632234096527, "learning_rate": 1.4559381588193957e-05, "loss": 0.0199, "step": 3456 }, { "epoch": 2.4293745607870694, "grad_norm": 2.596923828125, "learning_rate": 1.4563598032326072e-05, "loss": 0.0276, "step": 3457 }, { "epoch": 2.4300773014757553, "grad_norm": 0.295259952545166, "learning_rate": 1.4567814476458187e-05, "loss": 0.0251, "step": 3458 }, { "epoch": 2.430780042164441, "grad_norm": 0.3011491894721985, "learning_rate": 1.4572030920590302e-05, "loss": 0.057, "step": 3459 }, { "epoch": 2.431482782853127, "grad_norm": 0.24999086558818817, "learning_rate": 1.4576247364722417e-05, "loss": 0.0391, "step": 3460 }, { "epoch": 2.432185523541813, "grad_norm": 1.6255470514297485, "learning_rate": 1.4580463808854532e-05, "loss": 0.0565, "step": 3461 }, { "epoch": 2.432888264230499, "grad_norm": 0.28864169120788574, "learning_rate": 1.4584680252986649e-05, "loss": 0.0474, "step": 3462 }, { "epoch": 2.433591004919185, "grad_norm": 0.2940654456615448, "learning_rate": 1.4588896697118764e-05, "loss": 0.0417, "step": 3463 }, { "epoch": 2.434293745607871, "grad_norm": 0.4866448640823364, "learning_rate": 1.4593113141250879e-05, "loss": 0.0592, "step": 3464 }, { "epoch": 2.4349964862965567, "grad_norm": 0.29375940561294556, "learning_rate": 1.4597329585382994e-05, "loss": 0.0546, "step": 3465 }, { "epoch": 2.4356992269852427, "grad_norm": 0.5760843753814697, "learning_rate": 1.4601546029515109e-05, "loss": 0.0871, "step": 3466 }, { "epoch": 2.436401967673928, "grad_norm": 0.5966302156448364, "learning_rate": 1.4605762473647224e-05, "loss": 0.1239, "step": 3467 }, { "epoch": 2.437104708362614, "grad_norm": 1.1092634201049805, "learning_rate": 1.4609978917779338e-05, "loss": 0.1794, "step": 3468 }, { "epoch": 2.4378074490513, "grad_norm": 2.497875928878784, "learning_rate": 1.4614195361911455e-05, "loss": 0.2477, "step": 3469 }, { "epoch": 2.438510189739986, "grad_norm": 1.4512072801589966, "learning_rate": 1.461841180604357e-05, "loss": 0.2983, "step": 3470 }, { "epoch": 2.439212930428672, "grad_norm": 1.5311909914016724, "learning_rate": 1.4622628250175685e-05, "loss": 0.4032, "step": 3471 }, { "epoch": 2.4399156711173577, "grad_norm": 0.40472471714019775, "learning_rate": 1.4626844694307802e-05, "loss": 0.1219, "step": 3472 }, { "epoch": 2.4406184118060437, "grad_norm": 0.3257664442062378, "learning_rate": 1.4631061138439917e-05, "loss": 0.0572, "step": 3473 }, { "epoch": 2.4413211524947296, "grad_norm": 0.26582589745521545, "learning_rate": 1.4635277582572031e-05, "loss": 0.0374, "step": 3474 }, { "epoch": 2.4420238931834155, "grad_norm": 0.21833117306232452, "learning_rate": 1.4639494026704146e-05, "loss": 0.0348, "step": 3475 }, { "epoch": 2.442726633872101, "grad_norm": 0.25768977403640747, "learning_rate": 1.4643710470836263e-05, "loss": 0.0413, "step": 3476 }, { "epoch": 2.443429374560787, "grad_norm": 0.269132524728775, "learning_rate": 1.4647926914968378e-05, "loss": 0.0227, "step": 3477 }, { "epoch": 2.444132115249473, "grad_norm": 0.23749828338623047, "learning_rate": 1.4652143359100493e-05, "loss": 0.0353, "step": 3478 }, { "epoch": 2.4448348559381587, "grad_norm": 0.4441564381122589, "learning_rate": 1.4656359803232608e-05, "loss": 0.046, "step": 3479 }, { "epoch": 2.4455375966268447, "grad_norm": 0.27728942036628723, "learning_rate": 1.4660576247364723e-05, "loss": 0.0321, "step": 3480 }, { "epoch": 2.4462403373155306, "grad_norm": 0.22997403144836426, "learning_rate": 1.4664792691496838e-05, "loss": 0.0183, "step": 3481 }, { "epoch": 2.4469430780042165, "grad_norm": 0.3341456651687622, "learning_rate": 1.4669009135628953e-05, "loss": 0.0381, "step": 3482 }, { "epoch": 2.4476458186929024, "grad_norm": 0.2706676125526428, "learning_rate": 1.467322557976107e-05, "loss": 0.0346, "step": 3483 }, { "epoch": 2.4483485593815884, "grad_norm": 0.2582002878189087, "learning_rate": 1.4677442023893184e-05, "loss": 0.0361, "step": 3484 }, { "epoch": 2.4490513000702743, "grad_norm": 0.20575353503227234, "learning_rate": 1.46816584680253e-05, "loss": 0.026, "step": 3485 }, { "epoch": 2.4497540407589597, "grad_norm": 0.23538632690906525, "learning_rate": 1.4685874912157414e-05, "loss": 0.0393, "step": 3486 }, { "epoch": 2.4504567814476457, "grad_norm": 0.49825412034988403, "learning_rate": 1.469009135628953e-05, "loss": 0.083, "step": 3487 }, { "epoch": 2.4511595221363316, "grad_norm": 0.4539106488227844, "learning_rate": 1.4694307800421644e-05, "loss": 0.0744, "step": 3488 }, { "epoch": 2.4518622628250175, "grad_norm": 0.401479572057724, "learning_rate": 1.4698524244553759e-05, "loss": 0.0697, "step": 3489 }, { "epoch": 2.4525650035137034, "grad_norm": 0.46213358640670776, "learning_rate": 1.4702740688685876e-05, "loss": 0.072, "step": 3490 }, { "epoch": 2.4532677442023894, "grad_norm": 0.8771588802337646, "learning_rate": 1.470695713281799e-05, "loss": 0.1068, "step": 3491 }, { "epoch": 2.4539704848910753, "grad_norm": 0.45374369621276855, "learning_rate": 1.4711173576950106e-05, "loss": 0.1044, "step": 3492 }, { "epoch": 2.454673225579761, "grad_norm": 0.6995745897293091, "learning_rate": 1.471539002108222e-05, "loss": 0.1478, "step": 3493 }, { "epoch": 2.455375966268447, "grad_norm": 1.4991692304611206, "learning_rate": 1.4719606465214335e-05, "loss": 0.2186, "step": 3494 }, { "epoch": 2.4560787069571326, "grad_norm": 1.3864758014678955, "learning_rate": 1.472382290934645e-05, "loss": 0.299, "step": 3495 }, { "epoch": 2.4567814476458185, "grad_norm": 3.469616651535034, "learning_rate": 1.4728039353478567e-05, "loss": 0.3863, "step": 3496 }, { "epoch": 2.4574841883345044, "grad_norm": 0.8269253373146057, "learning_rate": 1.4732255797610682e-05, "loss": 0.0879, "step": 3497 }, { "epoch": 2.4581869290231904, "grad_norm": 0.34332671761512756, "learning_rate": 1.4736472241742797e-05, "loss": 0.0568, "step": 3498 }, { "epoch": 2.4588896697118763, "grad_norm": 0.19618068635463715, "learning_rate": 1.4740688685874912e-05, "loss": 0.0339, "step": 3499 }, { "epoch": 2.459592410400562, "grad_norm": 0.34279629588127136, "learning_rate": 1.4744905130007027e-05, "loss": 0.037, "step": 3500 }, { "epoch": 2.460295151089248, "grad_norm": 0.22796811163425446, "learning_rate": 1.4749121574139143e-05, "loss": 0.041, "step": 3501 }, { "epoch": 2.460997891777934, "grad_norm": 0.22545884549617767, "learning_rate": 1.4753338018271258e-05, "loss": 0.0304, "step": 3502 }, { "epoch": 2.46170063246662, "grad_norm": 0.1780412495136261, "learning_rate": 1.4757554462403375e-05, "loss": 0.017, "step": 3503 }, { "epoch": 2.462403373155306, "grad_norm": 0.3543473482131958, "learning_rate": 1.476177090653549e-05, "loss": 0.0344, "step": 3504 }, { "epoch": 2.463106113843992, "grad_norm": 0.21112868189811707, "learning_rate": 1.4765987350667605e-05, "loss": 0.0275, "step": 3505 }, { "epoch": 2.4638088545326773, "grad_norm": 0.20647741854190826, "learning_rate": 1.477020379479972e-05, "loss": 0.0271, "step": 3506 }, { "epoch": 2.464511595221363, "grad_norm": 0.23676608502864838, "learning_rate": 1.4774420238931835e-05, "loss": 0.0365, "step": 3507 }, { "epoch": 2.465214335910049, "grad_norm": 0.247759610414505, "learning_rate": 1.477863668306395e-05, "loss": 0.0306, "step": 3508 }, { "epoch": 2.465917076598735, "grad_norm": 0.34531646966934204, "learning_rate": 1.4782853127196065e-05, "loss": 0.0411, "step": 3509 }, { "epoch": 2.466619817287421, "grad_norm": 0.2840847969055176, "learning_rate": 1.4787069571328181e-05, "loss": 0.0334, "step": 3510 }, { "epoch": 2.467322557976107, "grad_norm": 0.3829038739204407, "learning_rate": 1.4791286015460296e-05, "loss": 0.0588, "step": 3511 }, { "epoch": 2.468025298664793, "grad_norm": 0.3352420926094055, "learning_rate": 1.4795502459592411e-05, "loss": 0.0541, "step": 3512 }, { "epoch": 2.4687280393534787, "grad_norm": 0.29344433546066284, "learning_rate": 1.4799718903724526e-05, "loss": 0.0489, "step": 3513 }, { "epoch": 2.469430780042164, "grad_norm": 0.38985633850097656, "learning_rate": 1.4803935347856641e-05, "loss": 0.0851, "step": 3514 }, { "epoch": 2.47013352073085, "grad_norm": 0.3517182171344757, "learning_rate": 1.4808151791988756e-05, "loss": 0.0614, "step": 3515 }, { "epoch": 2.470836261419536, "grad_norm": 0.4760952293872833, "learning_rate": 1.4812368236120871e-05, "loss": 0.0633, "step": 3516 }, { "epoch": 2.471539002108222, "grad_norm": 1.0437012910842896, "learning_rate": 1.4816584680252988e-05, "loss": 0.1046, "step": 3517 }, { "epoch": 2.472241742796908, "grad_norm": 0.7384281158447266, "learning_rate": 1.4820801124385103e-05, "loss": 0.1637, "step": 3518 }, { "epoch": 2.472944483485594, "grad_norm": 0.7662146687507629, "learning_rate": 1.4825017568517218e-05, "loss": 0.2204, "step": 3519 }, { "epoch": 2.4736472241742797, "grad_norm": 1.2697933912277222, "learning_rate": 1.4829234012649332e-05, "loss": 0.3182, "step": 3520 }, { "epoch": 2.4743499648629657, "grad_norm": 1.7271876335144043, "learning_rate": 1.4833450456781447e-05, "loss": 0.349, "step": 3521 }, { "epoch": 2.4750527055516516, "grad_norm": 0.304651141166687, "learning_rate": 1.4837666900913562e-05, "loss": 0.099, "step": 3522 }, { "epoch": 2.4757554462403375, "grad_norm": 0.20063982903957367, "learning_rate": 1.4841883345045677e-05, "loss": 0.035, "step": 3523 }, { "epoch": 2.4764581869290234, "grad_norm": 0.24500903487205505, "learning_rate": 1.4846099789177794e-05, "loss": 0.0307, "step": 3524 }, { "epoch": 2.477160927617709, "grad_norm": 0.2928718030452728, "learning_rate": 1.4850316233309909e-05, "loss": 0.0364, "step": 3525 }, { "epoch": 2.477863668306395, "grad_norm": 0.2751784324645996, "learning_rate": 1.4854532677442024e-05, "loss": 0.0312, "step": 3526 }, { "epoch": 2.4785664089950807, "grad_norm": 0.28395524621009827, "learning_rate": 1.4858749121574139e-05, "loss": 0.0292, "step": 3527 }, { "epoch": 2.4792691496837667, "grad_norm": 0.1729377955198288, "learning_rate": 1.4862965565706254e-05, "loss": 0.0201, "step": 3528 }, { "epoch": 2.4799718903724526, "grad_norm": 0.155472993850708, "learning_rate": 1.4867182009838369e-05, "loss": 0.026, "step": 3529 }, { "epoch": 2.4806746310611385, "grad_norm": 0.23645180463790894, "learning_rate": 1.4871398453970485e-05, "loss": 0.0508, "step": 3530 }, { "epoch": 2.4813773717498244, "grad_norm": 0.19468393921852112, "learning_rate": 1.4875614898102602e-05, "loss": 0.0273, "step": 3531 }, { "epoch": 2.4820801124385103, "grad_norm": 0.24426311254501343, "learning_rate": 1.4879831342234717e-05, "loss": 0.0431, "step": 3532 }, { "epoch": 2.482782853127196, "grad_norm": 0.263327419757843, "learning_rate": 1.4884047786366832e-05, "loss": 0.0373, "step": 3533 }, { "epoch": 2.4834855938158817, "grad_norm": 0.3785524070262909, "learning_rate": 1.4888264230498947e-05, "loss": 0.0414, "step": 3534 }, { "epoch": 2.4841883345045677, "grad_norm": 0.2946217358112335, "learning_rate": 1.4892480674631062e-05, "loss": 0.0305, "step": 3535 }, { "epoch": 2.4848910751932536, "grad_norm": 0.567198634147644, "learning_rate": 1.4896697118763177e-05, "loss": 0.034, "step": 3536 }, { "epoch": 2.4855938158819395, "grad_norm": 0.24756993353366852, "learning_rate": 1.4900913562895293e-05, "loss": 0.0432, "step": 3537 }, { "epoch": 2.4862965565706254, "grad_norm": 0.3277271091938019, "learning_rate": 1.4905130007027408e-05, "loss": 0.0506, "step": 3538 }, { "epoch": 2.4869992972593113, "grad_norm": 0.43649616837501526, "learning_rate": 1.4909346451159523e-05, "loss": 0.0375, "step": 3539 }, { "epoch": 2.4877020379479973, "grad_norm": 0.4448278248310089, "learning_rate": 1.4913562895291638e-05, "loss": 0.0959, "step": 3540 }, { "epoch": 2.488404778636683, "grad_norm": 0.3350195288658142, "learning_rate": 1.4917779339423753e-05, "loss": 0.0725, "step": 3541 }, { "epoch": 2.489107519325369, "grad_norm": 0.5876370668411255, "learning_rate": 1.4921995783555868e-05, "loss": 0.1358, "step": 3542 }, { "epoch": 2.489810260014055, "grad_norm": 0.8179095983505249, "learning_rate": 1.4926212227687983e-05, "loss": 0.2014, "step": 3543 }, { "epoch": 2.4905130007027405, "grad_norm": 1.205495834350586, "learning_rate": 1.49304286718201e-05, "loss": 0.2204, "step": 3544 }, { "epoch": 2.4912157413914264, "grad_norm": 1.5387852191925049, "learning_rate": 1.4934645115952215e-05, "loss": 0.316, "step": 3545 }, { "epoch": 2.4919184820801124, "grad_norm": 1.393608808517456, "learning_rate": 1.493886156008433e-05, "loss": 0.3732, "step": 3546 }, { "epoch": 2.4926212227687983, "grad_norm": 0.4779973328113556, "learning_rate": 1.4943078004216444e-05, "loss": 0.0811, "step": 3547 }, { "epoch": 2.493323963457484, "grad_norm": 0.2650868892669678, "learning_rate": 1.494729444834856e-05, "loss": 0.0732, "step": 3548 }, { "epoch": 2.49402670414617, "grad_norm": 0.24209852516651154, "learning_rate": 1.4951510892480674e-05, "loss": 0.0434, "step": 3549 }, { "epoch": 2.494729444834856, "grad_norm": 0.21831677854061127, "learning_rate": 1.495572733661279e-05, "loss": 0.0205, "step": 3550 }, { "epoch": 2.495432185523542, "grad_norm": 0.21376509964466095, "learning_rate": 1.4959943780744906e-05, "loss": 0.0194, "step": 3551 }, { "epoch": 2.496134926212228, "grad_norm": 0.22911398112773895, "learning_rate": 1.496416022487702e-05, "loss": 0.0339, "step": 3552 }, { "epoch": 2.4968376669009134, "grad_norm": 0.2768234312534332, "learning_rate": 1.4968376669009136e-05, "loss": 0.0303, "step": 3553 }, { "epoch": 2.4975404075895993, "grad_norm": 0.3868647515773773, "learning_rate": 1.497259311314125e-05, "loss": 0.0492, "step": 3554 }, { "epoch": 2.498243148278285, "grad_norm": 0.26258692145347595, "learning_rate": 1.4976809557273366e-05, "loss": 0.027, "step": 3555 }, { "epoch": 2.498945888966971, "grad_norm": 0.26301801204681396, "learning_rate": 1.498102600140548e-05, "loss": 0.0598, "step": 3556 }, { "epoch": 2.499648629655657, "grad_norm": 0.2520259618759155, "learning_rate": 1.4985242445537596e-05, "loss": 0.0469, "step": 3557 }, { "epoch": 2.500351370344343, "grad_norm": 0.45003634691238403, "learning_rate": 1.4989458889669712e-05, "loss": 0.0387, "step": 3558 }, { "epoch": 2.501054111033029, "grad_norm": 0.2912486791610718, "learning_rate": 1.4993675333801827e-05, "loss": 0.0418, "step": 3559 }, { "epoch": 2.501756851721715, "grad_norm": 0.2972334623336792, "learning_rate": 1.4997891777933942e-05, "loss": 0.032, "step": 3560 }, { "epoch": 2.5024595924104007, "grad_norm": 0.28102850914001465, "learning_rate": 1.5002108222066059e-05, "loss": 0.0428, "step": 3561 }, { "epoch": 2.5031623330990866, "grad_norm": 0.3731078803539276, "learning_rate": 1.5006324666198172e-05, "loss": 0.0503, "step": 3562 }, { "epoch": 2.5038650737877726, "grad_norm": 0.2454453706741333, "learning_rate": 1.5010541110330289e-05, "loss": 0.032, "step": 3563 }, { "epoch": 2.504567814476458, "grad_norm": 0.4568895697593689, "learning_rate": 1.5014757554462404e-05, "loss": 0.0578, "step": 3564 }, { "epoch": 2.505270555165144, "grad_norm": 0.45049214363098145, "learning_rate": 1.501897399859452e-05, "loss": 0.0755, "step": 3565 }, { "epoch": 2.50597329585383, "grad_norm": 0.5229963660240173, "learning_rate": 1.5023190442726633e-05, "loss": 0.1121, "step": 3566 }, { "epoch": 2.506676036542516, "grad_norm": 0.6353605389595032, "learning_rate": 1.502740688685875e-05, "loss": 0.0925, "step": 3567 }, { "epoch": 2.5073787772312017, "grad_norm": 0.6401220560073853, "learning_rate": 1.5031623330990863e-05, "loss": 0.175, "step": 3568 }, { "epoch": 2.5080815179198876, "grad_norm": 1.0391932725906372, "learning_rate": 1.503583977512298e-05, "loss": 0.229, "step": 3569 }, { "epoch": 2.5087842586085736, "grad_norm": 1.1678333282470703, "learning_rate": 1.5040056219255093e-05, "loss": 0.2521, "step": 3570 }, { "epoch": 2.509486999297259, "grad_norm": 1.7500545978546143, "learning_rate": 1.5044272663387212e-05, "loss": 0.4197, "step": 3571 }, { "epoch": 2.510189739985945, "grad_norm": 0.4356282353401184, "learning_rate": 1.5048489107519325e-05, "loss": 0.0911, "step": 3572 }, { "epoch": 2.510892480674631, "grad_norm": 0.34436705708503723, "learning_rate": 1.5052705551651441e-05, "loss": 0.0473, "step": 3573 }, { "epoch": 2.511595221363317, "grad_norm": 0.2870476543903351, "learning_rate": 1.5056921995783555e-05, "loss": 0.039, "step": 3574 }, { "epoch": 2.5122979620520027, "grad_norm": 0.2562531530857086, "learning_rate": 1.5061138439915671e-05, "loss": 0.0443, "step": 3575 }, { "epoch": 2.5130007027406887, "grad_norm": 0.1846725344657898, "learning_rate": 1.5065354884047788e-05, "loss": 0.0198, "step": 3576 }, { "epoch": 2.5137034434293746, "grad_norm": 0.3723224103450775, "learning_rate": 1.5069571328179901e-05, "loss": 0.0291, "step": 3577 }, { "epoch": 2.5144061841180605, "grad_norm": 0.269846111536026, "learning_rate": 1.507378777231202e-05, "loss": 0.0443, "step": 3578 }, { "epoch": 2.5151089248067464, "grad_norm": 0.3093031346797943, "learning_rate": 1.5078004216444133e-05, "loss": 0.0385, "step": 3579 }, { "epoch": 2.5158116654954323, "grad_norm": 0.41703879833221436, "learning_rate": 1.508222066057625e-05, "loss": 0.048, "step": 3580 }, { "epoch": 2.5165144061841183, "grad_norm": 0.1947413682937622, "learning_rate": 1.5086437104708363e-05, "loss": 0.0194, "step": 3581 }, { "epoch": 2.517217146872804, "grad_norm": 0.22589461505413055, "learning_rate": 1.509065354884048e-05, "loss": 0.0406, "step": 3582 }, { "epoch": 2.5179198875614897, "grad_norm": 0.29916486144065857, "learning_rate": 1.5094869992972593e-05, "loss": 0.0348, "step": 3583 }, { "epoch": 2.5186226282501756, "grad_norm": 0.2609555721282959, "learning_rate": 1.509908643710471e-05, "loss": 0.0447, "step": 3584 }, { "epoch": 2.5193253689388615, "grad_norm": 0.28455573320388794, "learning_rate": 1.5103302881236824e-05, "loss": 0.0384, "step": 3585 }, { "epoch": 2.5200281096275474, "grad_norm": 0.40486961603164673, "learning_rate": 1.510751932536894e-05, "loss": 0.0623, "step": 3586 }, { "epoch": 2.5207308503162333, "grad_norm": 0.3347594738006592, "learning_rate": 1.5111735769501054e-05, "loss": 0.059, "step": 3587 }, { "epoch": 2.5214335910049193, "grad_norm": 0.30399376153945923, "learning_rate": 1.511595221363317e-05, "loss": 0.0323, "step": 3588 }, { "epoch": 2.522136331693605, "grad_norm": 0.3850167393684387, "learning_rate": 1.5120168657765284e-05, "loss": 0.0792, "step": 3589 }, { "epoch": 2.5228390723822907, "grad_norm": 0.3907259404659271, "learning_rate": 1.51243851018974e-05, "loss": 0.0594, "step": 3590 }, { "epoch": 2.5235418130709766, "grad_norm": 0.37316450476646423, "learning_rate": 1.5128601546029514e-05, "loss": 0.0793, "step": 3591 }, { "epoch": 2.5242445537596625, "grad_norm": 0.6348283886909485, "learning_rate": 1.5132817990161632e-05, "loss": 0.1521, "step": 3592 }, { "epoch": 2.5249472944483484, "grad_norm": 0.6204752922058105, "learning_rate": 1.5137034434293745e-05, "loss": 0.1457, "step": 3593 }, { "epoch": 2.5256500351370343, "grad_norm": 1.1051058769226074, "learning_rate": 1.5141250878425862e-05, "loss": 0.225, "step": 3594 }, { "epoch": 2.5263527758257203, "grad_norm": 1.348564624786377, "learning_rate": 1.5145467322557975e-05, "loss": 0.2691, "step": 3595 }, { "epoch": 2.527055516514406, "grad_norm": 1.9712861776351929, "learning_rate": 1.5149683766690092e-05, "loss": 0.4167, "step": 3596 }, { "epoch": 2.527758257203092, "grad_norm": 0.39585310220718384, "learning_rate": 1.5153900210822205e-05, "loss": 0.0796, "step": 3597 }, { "epoch": 2.528460997891778, "grad_norm": 0.16308477520942688, "learning_rate": 1.5158116654954323e-05, "loss": 0.0285, "step": 3598 }, { "epoch": 2.529163738580464, "grad_norm": 0.4515666365623474, "learning_rate": 1.5162333099086437e-05, "loss": 0.0477, "step": 3599 }, { "epoch": 2.52986647926915, "grad_norm": 0.22030538320541382, "learning_rate": 1.5166549543218553e-05, "loss": 0.0378, "step": 3600 }, { "epoch": 2.530569219957836, "grad_norm": 0.23477651178836823, "learning_rate": 1.5170765987350667e-05, "loss": 0.0294, "step": 3601 }, { "epoch": 2.5312719606465213, "grad_norm": 0.20514491200447083, "learning_rate": 1.5174982431482783e-05, "loss": 0.0206, "step": 3602 }, { "epoch": 2.531974701335207, "grad_norm": 0.2148810476064682, "learning_rate": 1.5179198875614896e-05, "loss": 0.0232, "step": 3603 }, { "epoch": 2.532677442023893, "grad_norm": 0.2895732522010803, "learning_rate": 1.5183415319747013e-05, "loss": 0.0332, "step": 3604 }, { "epoch": 2.533380182712579, "grad_norm": 0.23781897127628326, "learning_rate": 1.5187631763879128e-05, "loss": 0.0338, "step": 3605 }, { "epoch": 2.534082923401265, "grad_norm": 0.28562822937965393, "learning_rate": 1.5191848208011245e-05, "loss": 0.0322, "step": 3606 }, { "epoch": 2.534785664089951, "grad_norm": 0.29044830799102783, "learning_rate": 1.5196064652143361e-05, "loss": 0.0204, "step": 3607 }, { "epoch": 2.535488404778637, "grad_norm": 0.3088037669658661, "learning_rate": 1.5200281096275475e-05, "loss": 0.0497, "step": 3608 }, { "epoch": 2.5361911454673227, "grad_norm": 0.30308613181114197, "learning_rate": 1.5204497540407591e-05, "loss": 0.0407, "step": 3609 }, { "epoch": 2.536893886156008, "grad_norm": 0.29545125365257263, "learning_rate": 1.5208713984539704e-05, "loss": 0.0356, "step": 3610 }, { "epoch": 2.537596626844694, "grad_norm": 0.3444337248802185, "learning_rate": 1.5212930428671821e-05, "loss": 0.0375, "step": 3611 }, { "epoch": 2.53829936753338, "grad_norm": 0.27380356192588806, "learning_rate": 1.5217146872803936e-05, "loss": 0.04, "step": 3612 }, { "epoch": 2.539002108222066, "grad_norm": 0.5498489737510681, "learning_rate": 1.5221363316936053e-05, "loss": 0.1073, "step": 3613 }, { "epoch": 2.539704848910752, "grad_norm": 0.28304973244667053, "learning_rate": 1.5225579761068166e-05, "loss": 0.0491, "step": 3614 }, { "epoch": 2.540407589599438, "grad_norm": 0.5313096046447754, "learning_rate": 1.5229796205200283e-05, "loss": 0.095, "step": 3615 }, { "epoch": 2.5411103302881237, "grad_norm": 0.4266931414604187, "learning_rate": 1.5234012649332396e-05, "loss": 0.0897, "step": 3616 }, { "epoch": 2.5418130709768096, "grad_norm": 1.1295496225357056, "learning_rate": 1.5238229093464512e-05, "loss": 0.1028, "step": 3617 }, { "epoch": 2.5425158116654956, "grad_norm": 0.771978497505188, "learning_rate": 1.5242445537596626e-05, "loss": 0.1995, "step": 3618 }, { "epoch": 2.5432185523541815, "grad_norm": 1.4318395853042603, "learning_rate": 1.5246661981728744e-05, "loss": 0.2789, "step": 3619 }, { "epoch": 2.5439212930428674, "grad_norm": 1.2611746788024902, "learning_rate": 1.5250878425860857e-05, "loss": 0.3132, "step": 3620 }, { "epoch": 2.5446240337315533, "grad_norm": 1.4736441373825073, "learning_rate": 1.5255094869992974e-05, "loss": 0.321, "step": 3621 }, { "epoch": 2.545326774420239, "grad_norm": 0.4284851551055908, "learning_rate": 1.5259311314125087e-05, "loss": 0.1176, "step": 3622 }, { "epoch": 2.5460295151089247, "grad_norm": 0.22899365425109863, "learning_rate": 1.5263527758257204e-05, "loss": 0.0426, "step": 3623 }, { "epoch": 2.5467322557976106, "grad_norm": 0.25098076462745667, "learning_rate": 1.5267744202389317e-05, "loss": 0.0436, "step": 3624 }, { "epoch": 2.5474349964862966, "grad_norm": 0.2969900071620941, "learning_rate": 1.5271960646521434e-05, "loss": 0.0357, "step": 3625 }, { "epoch": 2.5481377371749825, "grad_norm": 0.268356055021286, "learning_rate": 1.5276177090653547e-05, "loss": 0.0335, "step": 3626 }, { "epoch": 2.5488404778636684, "grad_norm": 0.18480269610881805, "learning_rate": 1.5280393534785664e-05, "loss": 0.0152, "step": 3627 }, { "epoch": 2.5495432185523543, "grad_norm": 0.22141677141189575, "learning_rate": 1.5284609978917777e-05, "loss": 0.0329, "step": 3628 }, { "epoch": 2.55024595924104, "grad_norm": 0.43298134207725525, "learning_rate": 1.5288826423049897e-05, "loss": 0.0464, "step": 3629 }, { "epoch": 2.5509486999297257, "grad_norm": 0.2383909970521927, "learning_rate": 1.529304286718201e-05, "loss": 0.0294, "step": 3630 }, { "epoch": 2.5516514406184116, "grad_norm": 0.22575202584266663, "learning_rate": 1.5297259311314127e-05, "loss": 0.0306, "step": 3631 }, { "epoch": 2.5523541813070976, "grad_norm": 0.2300325334072113, "learning_rate": 1.530147575544624e-05, "loss": 0.041, "step": 3632 }, { "epoch": 2.5530569219957835, "grad_norm": 0.31380710005760193, "learning_rate": 1.5305692199578357e-05, "loss": 0.0268, "step": 3633 }, { "epoch": 2.5537596626844694, "grad_norm": 0.2669166922569275, "learning_rate": 1.530990864371047e-05, "loss": 0.0441, "step": 3634 }, { "epoch": 2.5544624033731553, "grad_norm": 0.6494589447975159, "learning_rate": 1.5314125087842587e-05, "loss": 0.0337, "step": 3635 }, { "epoch": 2.5551651440618413, "grad_norm": 0.29358068108558655, "learning_rate": 1.5318341531974703e-05, "loss": 0.0493, "step": 3636 }, { "epoch": 2.555867884750527, "grad_norm": 0.3224460780620575, "learning_rate": 1.5322557976106816e-05, "loss": 0.0537, "step": 3637 }, { "epoch": 2.556570625439213, "grad_norm": 0.4082740545272827, "learning_rate": 1.5326774420238933e-05, "loss": 0.0357, "step": 3638 }, { "epoch": 2.557273366127899, "grad_norm": 0.32825204730033875, "learning_rate": 1.5330990864371046e-05, "loss": 0.0461, "step": 3639 }, { "epoch": 2.557976106816585, "grad_norm": 0.3272165358066559, "learning_rate": 1.5335207308503163e-05, "loss": 0.0516, "step": 3640 }, { "epoch": 2.5586788475052704, "grad_norm": 0.48853978514671326, "learning_rate": 1.5339423752635276e-05, "loss": 0.0813, "step": 3641 }, { "epoch": 2.5593815881939563, "grad_norm": 0.4805428385734558, "learning_rate": 1.5343640196767393e-05, "loss": 0.1212, "step": 3642 }, { "epoch": 2.5600843288826423, "grad_norm": 0.6583701372146606, "learning_rate": 1.534785664089951e-05, "loss": 0.1972, "step": 3643 }, { "epoch": 2.560787069571328, "grad_norm": 0.9649458527565002, "learning_rate": 1.5352073085031626e-05, "loss": 0.2514, "step": 3644 }, { "epoch": 2.561489810260014, "grad_norm": 1.4467853307724, "learning_rate": 1.535628952916374e-05, "loss": 0.315, "step": 3645 }, { "epoch": 2.5621925509487, "grad_norm": 8.076333999633789, "learning_rate": 1.5360505973295856e-05, "loss": 0.3717, "step": 3646 }, { "epoch": 2.562895291637386, "grad_norm": 0.3209485113620758, "learning_rate": 1.536472241742797e-05, "loss": 0.0858, "step": 3647 }, { "epoch": 2.5635980323260714, "grad_norm": 0.38228318095207214, "learning_rate": 1.5368938861560086e-05, "loss": 0.0518, "step": 3648 }, { "epoch": 2.5643007730147573, "grad_norm": 0.27336907386779785, "learning_rate": 1.53731553056922e-05, "loss": 0.0414, "step": 3649 }, { "epoch": 2.5650035137034433, "grad_norm": 0.2778533101081848, "learning_rate": 1.5377371749824316e-05, "loss": 0.0405, "step": 3650 }, { "epoch": 2.565706254392129, "grad_norm": 0.1965675801038742, "learning_rate": 1.538158819395643e-05, "loss": 0.031, "step": 3651 }, { "epoch": 2.566408995080815, "grad_norm": 0.3299618065357208, "learning_rate": 1.5385804638088546e-05, "loss": 0.0208, "step": 3652 }, { "epoch": 2.567111735769501, "grad_norm": 0.3993504047393799, "learning_rate": 1.539002108222066e-05, "loss": 0.0541, "step": 3653 }, { "epoch": 2.567814476458187, "grad_norm": 0.3696022629737854, "learning_rate": 1.5394237526352776e-05, "loss": 0.0605, "step": 3654 }, { "epoch": 2.568517217146873, "grad_norm": 0.17544826865196228, "learning_rate": 1.539845397048489e-05, "loss": 0.0271, "step": 3655 }, { "epoch": 2.569219957835559, "grad_norm": 0.20886799693107605, "learning_rate": 1.540267041461701e-05, "loss": 0.0214, "step": 3656 }, { "epoch": 2.5699226985242447, "grad_norm": 0.2577146291732788, "learning_rate": 1.5406886858749122e-05, "loss": 0.0342, "step": 3657 }, { "epoch": 2.5706254392129306, "grad_norm": 0.2209271341562271, "learning_rate": 1.541110330288124e-05, "loss": 0.0309, "step": 3658 }, { "epoch": 2.5713281799016166, "grad_norm": 0.36908406019210815, "learning_rate": 1.5415319747013352e-05, "loss": 0.0592, "step": 3659 }, { "epoch": 2.572030920590302, "grad_norm": 0.19347256422042847, "learning_rate": 1.541953619114547e-05, "loss": 0.0161, "step": 3660 }, { "epoch": 2.572733661278988, "grad_norm": 0.3416833281517029, "learning_rate": 1.5423752635277582e-05, "loss": 0.059, "step": 3661 }, { "epoch": 2.573436401967674, "grad_norm": 0.26314467191696167, "learning_rate": 1.54279690794097e-05, "loss": 0.0477, "step": 3662 }, { "epoch": 2.57413914265636, "grad_norm": 0.30452045798301697, "learning_rate": 1.5432185523541812e-05, "loss": 0.0289, "step": 3663 }, { "epoch": 2.5748418833450457, "grad_norm": 0.4550258219242096, "learning_rate": 1.543640196767393e-05, "loss": 0.0467, "step": 3664 }, { "epoch": 2.5755446240337316, "grad_norm": 0.39280372858047485, "learning_rate": 1.5440618411806045e-05, "loss": 0.0774, "step": 3665 }, { "epoch": 2.5762473647224176, "grad_norm": 0.43763622641563416, "learning_rate": 1.5444834855938158e-05, "loss": 0.0902, "step": 3666 }, { "epoch": 2.576950105411103, "grad_norm": 0.7629660367965698, "learning_rate": 1.5449051300070275e-05, "loss": 0.1445, "step": 3667 }, { "epoch": 2.577652846099789, "grad_norm": 0.5787600874900818, "learning_rate": 1.5453267744202388e-05, "loss": 0.1745, "step": 3668 }, { "epoch": 2.578355586788475, "grad_norm": 0.9743255376815796, "learning_rate": 1.5457484188334505e-05, "loss": 0.2161, "step": 3669 }, { "epoch": 2.579058327477161, "grad_norm": 1.1665128469467163, "learning_rate": 1.546170063246662e-05, "loss": 0.2998, "step": 3670 }, { "epoch": 2.5797610681658467, "grad_norm": 2.9765102863311768, "learning_rate": 1.5465917076598738e-05, "loss": 0.4017, "step": 3671 }, { "epoch": 2.5804638088545326, "grad_norm": 0.3018941581249237, "learning_rate": 1.547013352073085e-05, "loss": 0.0905, "step": 3672 }, { "epoch": 2.5811665495432186, "grad_norm": 0.21499855816364288, "learning_rate": 1.5474349964862968e-05, "loss": 0.0439, "step": 3673 }, { "epoch": 2.5818692902319045, "grad_norm": 0.18719840049743652, "learning_rate": 1.547856640899508e-05, "loss": 0.0299, "step": 3674 }, { "epoch": 2.5825720309205904, "grad_norm": 0.2503771185874939, "learning_rate": 1.5482782853127198e-05, "loss": 0.0298, "step": 3675 }, { "epoch": 2.5832747716092763, "grad_norm": 0.31099992990493774, "learning_rate": 1.548699929725931e-05, "loss": 0.0441, "step": 3676 }, { "epoch": 2.5839775122979622, "grad_norm": 0.2970248758792877, "learning_rate": 1.5491215741391428e-05, "loss": 0.0283, "step": 3677 }, { "epoch": 2.584680252986648, "grad_norm": 0.16958506405353546, "learning_rate": 1.549543218552354e-05, "loss": 0.0206, "step": 3678 }, { "epoch": 2.5853829936753336, "grad_norm": 0.22811810672283173, "learning_rate": 1.5499648629655658e-05, "loss": 0.0299, "step": 3679 }, { "epoch": 2.5860857343640196, "grad_norm": 0.24164561927318573, "learning_rate": 1.550386507378777e-05, "loss": 0.0296, "step": 3680 }, { "epoch": 2.5867884750527055, "grad_norm": 0.2164926379919052, "learning_rate": 1.5508081517919888e-05, "loss": 0.0238, "step": 3681 }, { "epoch": 2.5874912157413914, "grad_norm": 0.32610517740249634, "learning_rate": 1.5512297962052e-05, "loss": 0.0452, "step": 3682 }, { "epoch": 2.5881939564300773, "grad_norm": 0.2239130437374115, "learning_rate": 1.551651440618412e-05, "loss": 0.0306, "step": 3683 }, { "epoch": 2.5888966971187632, "grad_norm": 0.4034746289253235, "learning_rate": 1.5520730850316234e-05, "loss": 0.0554, "step": 3684 }, { "epoch": 2.589599437807449, "grad_norm": 0.18870528042316437, "learning_rate": 1.552494729444835e-05, "loss": 0.0248, "step": 3685 }, { "epoch": 2.590302178496135, "grad_norm": 0.27149680256843567, "learning_rate": 1.5529163738580464e-05, "loss": 0.0474, "step": 3686 }, { "epoch": 2.5910049191848206, "grad_norm": 0.29522550106048584, "learning_rate": 1.553338018271258e-05, "loss": 0.0559, "step": 3687 }, { "epoch": 2.5917076598735065, "grad_norm": 0.46424856781959534, "learning_rate": 1.5537596626844694e-05, "loss": 0.0611, "step": 3688 }, { "epoch": 2.5924104005621924, "grad_norm": 0.5930078625679016, "learning_rate": 1.554181307097681e-05, "loss": 0.0372, "step": 3689 }, { "epoch": 2.5931131412508783, "grad_norm": 0.5229631066322327, "learning_rate": 1.5546029515108924e-05, "loss": 0.0876, "step": 3690 }, { "epoch": 2.5938158819395642, "grad_norm": 0.8575285077095032, "learning_rate": 1.555024595924104e-05, "loss": 0.0704, "step": 3691 }, { "epoch": 2.59451862262825, "grad_norm": 1.3452479839324951, "learning_rate": 1.5554462403373154e-05, "loss": 0.1708, "step": 3692 }, { "epoch": 2.595221363316936, "grad_norm": 0.7665390968322754, "learning_rate": 1.555867884750527e-05, "loss": 0.1743, "step": 3693 }, { "epoch": 2.595924104005622, "grad_norm": 1.0518441200256348, "learning_rate": 1.5562895291637383e-05, "loss": 0.2035, "step": 3694 }, { "epoch": 2.596626844694308, "grad_norm": 1.0726568698883057, "learning_rate": 1.55671117357695e-05, "loss": 0.2512, "step": 3695 }, { "epoch": 2.597329585382994, "grad_norm": 1.5932284593582153, "learning_rate": 1.5571328179901617e-05, "loss": 0.3815, "step": 3696 }, { "epoch": 2.5980323260716798, "grad_norm": 0.3073001801967621, "learning_rate": 1.5575544624033733e-05, "loss": 0.0935, "step": 3697 }, { "epoch": 2.5987350667603657, "grad_norm": 0.2380916327238083, "learning_rate": 1.557976106816585e-05, "loss": 0.049, "step": 3698 }, { "epoch": 2.599437807449051, "grad_norm": 0.2637935280799866, "learning_rate": 1.5583977512297963e-05, "loss": 0.0466, "step": 3699 }, { "epoch": 2.600140548137737, "grad_norm": 0.21188563108444214, "learning_rate": 1.558819395643008e-05, "loss": 0.0355, "step": 3700 }, { "epoch": 2.600843288826423, "grad_norm": 0.22236143052577972, "learning_rate": 1.5592410400562193e-05, "loss": 0.025, "step": 3701 }, { "epoch": 2.601546029515109, "grad_norm": 0.23494909703731537, "learning_rate": 1.559662684469431e-05, "loss": 0.0337, "step": 3702 }, { "epoch": 2.602248770203795, "grad_norm": 0.22976577281951904, "learning_rate": 1.5600843288826423e-05, "loss": 0.0328, "step": 3703 }, { "epoch": 2.602951510892481, "grad_norm": 0.22649627923965454, "learning_rate": 1.560505973295854e-05, "loss": 0.0414, "step": 3704 }, { "epoch": 2.6036542515811667, "grad_norm": 0.2196725606918335, "learning_rate": 1.5609276177090653e-05, "loss": 0.0345, "step": 3705 }, { "epoch": 2.604356992269852, "grad_norm": 0.43055298924446106, "learning_rate": 1.561349262122277e-05, "loss": 0.061, "step": 3706 }, { "epoch": 2.605059732958538, "grad_norm": 0.4310080111026764, "learning_rate": 1.5617709065354883e-05, "loss": 0.0497, "step": 3707 }, { "epoch": 2.605762473647224, "grad_norm": 0.2884175479412079, "learning_rate": 1.5621925509487e-05, "loss": 0.0295, "step": 3708 }, { "epoch": 2.60646521433591, "grad_norm": 0.21603800356388092, "learning_rate": 1.5626141953619113e-05, "loss": 0.0227, "step": 3709 }, { "epoch": 2.607167955024596, "grad_norm": 0.2683541476726532, "learning_rate": 1.563035839775123e-05, "loss": 0.0583, "step": 3710 }, { "epoch": 2.607870695713282, "grad_norm": 0.2586519420146942, "learning_rate": 1.5634574841883346e-05, "loss": 0.0494, "step": 3711 }, { "epoch": 2.6085734364019677, "grad_norm": 0.3975968062877655, "learning_rate": 1.5638791286015463e-05, "loss": 0.0594, "step": 3712 }, { "epoch": 2.6092761770906536, "grad_norm": 0.2718150317668915, "learning_rate": 1.5643007730147576e-05, "loss": 0.0358, "step": 3713 }, { "epoch": 2.6099789177793395, "grad_norm": 1.6978464126586914, "learning_rate": 1.5647224174279693e-05, "loss": 0.0559, "step": 3714 }, { "epoch": 2.6106816584680255, "grad_norm": 0.4534337818622589, "learning_rate": 1.5651440618411806e-05, "loss": 0.0539, "step": 3715 }, { "epoch": 2.6113843991567114, "grad_norm": 0.566103994846344, "learning_rate": 1.5655657062543922e-05, "loss": 0.1114, "step": 3716 }, { "epoch": 2.6120871398453973, "grad_norm": 0.7450999617576599, "learning_rate": 1.5659873506676036e-05, "loss": 0.1126, "step": 3717 }, { "epoch": 2.612789880534083, "grad_norm": 0.8763279318809509, "learning_rate": 1.5664089950808152e-05, "loss": 0.1794, "step": 3718 }, { "epoch": 2.6134926212227687, "grad_norm": 1.1569647789001465, "learning_rate": 1.5668306394940266e-05, "loss": 0.2421, "step": 3719 }, { "epoch": 2.6141953619114546, "grad_norm": 1.5058104991912842, "learning_rate": 1.5672522839072382e-05, "loss": 0.3039, "step": 3720 }, { "epoch": 2.6148981026001406, "grad_norm": 1.5279064178466797, "learning_rate": 1.5676739283204495e-05, "loss": 0.3253, "step": 3721 }, { "epoch": 2.6156008432888265, "grad_norm": 0.3896099925041199, "learning_rate": 1.5680955727336612e-05, "loss": 0.1056, "step": 3722 }, { "epoch": 2.6163035839775124, "grad_norm": 0.24343369901180267, "learning_rate": 1.5685172171468725e-05, "loss": 0.0369, "step": 3723 }, { "epoch": 2.6170063246661983, "grad_norm": 0.2620580792427063, "learning_rate": 1.5689388615600845e-05, "loss": 0.0395, "step": 3724 }, { "epoch": 2.617709065354884, "grad_norm": 0.26822054386138916, "learning_rate": 1.5693605059732962e-05, "loss": 0.0331, "step": 3725 }, { "epoch": 2.6184118060435697, "grad_norm": 0.2480001151561737, "learning_rate": 1.5697821503865075e-05, "loss": 0.032, "step": 3726 }, { "epoch": 2.6191145467322556, "grad_norm": 0.26679518818855286, "learning_rate": 1.5702037947997192e-05, "loss": 0.0306, "step": 3727 }, { "epoch": 2.6198172874209416, "grad_norm": 0.4468483030796051, "learning_rate": 1.5706254392129305e-05, "loss": 0.032, "step": 3728 }, { "epoch": 2.6205200281096275, "grad_norm": 0.2474292516708374, "learning_rate": 1.5710470836261422e-05, "loss": 0.0325, "step": 3729 }, { "epoch": 2.6212227687983134, "grad_norm": 0.3028281629085541, "learning_rate": 1.5714687280393535e-05, "loss": 0.0493, "step": 3730 }, { "epoch": 2.6219255094869993, "grad_norm": 0.24244150519371033, "learning_rate": 1.571890372452565e-05, "loss": 0.0278, "step": 3731 }, { "epoch": 2.6226282501756852, "grad_norm": 0.323403000831604, "learning_rate": 1.5723120168657765e-05, "loss": 0.0424, "step": 3732 }, { "epoch": 2.623330990864371, "grad_norm": 0.3207201659679413, "learning_rate": 1.572733661278988e-05, "loss": 0.0436, "step": 3733 }, { "epoch": 2.624033731553057, "grad_norm": 0.28457197546958923, "learning_rate": 1.5731553056921995e-05, "loss": 0.0336, "step": 3734 }, { "epoch": 2.624736472241743, "grad_norm": 0.3100004196166992, "learning_rate": 1.573576950105411e-05, "loss": 0.0251, "step": 3735 }, { "epoch": 2.625439212930429, "grad_norm": 0.40509873628616333, "learning_rate": 1.5739985945186225e-05, "loss": 0.065, "step": 3736 }, { "epoch": 2.6261419536191144, "grad_norm": 0.33840277791023254, "learning_rate": 1.574420238931834e-05, "loss": 0.0562, "step": 3737 }, { "epoch": 2.6268446943078003, "grad_norm": 0.3312283754348755, "learning_rate": 1.5748418833450458e-05, "loss": 0.0314, "step": 3738 }, { "epoch": 2.6275474349964862, "grad_norm": 0.2828814387321472, "learning_rate": 1.5752635277582575e-05, "loss": 0.0548, "step": 3739 }, { "epoch": 2.628250175685172, "grad_norm": 0.461818665266037, "learning_rate": 1.5756851721714688e-05, "loss": 0.0895, "step": 3740 }, { "epoch": 2.628952916373858, "grad_norm": 0.4964267909526825, "learning_rate": 1.5761068165846804e-05, "loss": 0.1007, "step": 3741 }, { "epoch": 2.629655657062544, "grad_norm": 0.5678427219390869, "learning_rate": 1.5765284609978918e-05, "loss": 0.1058, "step": 3742 }, { "epoch": 2.63035839775123, "grad_norm": 0.5454670190811157, "learning_rate": 1.5769501054111034e-05, "loss": 0.1577, "step": 3743 }, { "epoch": 2.6310611384399154, "grad_norm": 1.1268045902252197, "learning_rate": 1.5773717498243148e-05, "loss": 0.2571, "step": 3744 }, { "epoch": 2.6317638791286013, "grad_norm": 1.287457823753357, "learning_rate": 1.5777933942375264e-05, "loss": 0.3046, "step": 3745 }, { "epoch": 2.6324666198172872, "grad_norm": 5.686559677124023, "learning_rate": 1.5782150386507377e-05, "loss": 0.4203, "step": 3746 }, { "epoch": 2.633169360505973, "grad_norm": 0.39886075258255005, "learning_rate": 1.5786366830639494e-05, "loss": 0.0932, "step": 3747 }, { "epoch": 2.633872101194659, "grad_norm": 0.33239901065826416, "learning_rate": 1.5790583274771607e-05, "loss": 0.0419, "step": 3748 }, { "epoch": 2.634574841883345, "grad_norm": 0.17040538787841797, "learning_rate": 1.5794799718903724e-05, "loss": 0.0296, "step": 3749 }, { "epoch": 2.635277582572031, "grad_norm": 0.2989369034767151, "learning_rate": 1.5799016163035837e-05, "loss": 0.0426, "step": 3750 }, { "epoch": 2.635980323260717, "grad_norm": 0.1890328973531723, "learning_rate": 1.5803232607167957e-05, "loss": 0.0256, "step": 3751 }, { "epoch": 2.6366830639494028, "grad_norm": 0.2699463665485382, "learning_rate": 1.580744905130007e-05, "loss": 0.0226, "step": 3752 }, { "epoch": 2.6373858046380887, "grad_norm": 0.30285149812698364, "learning_rate": 1.5811665495432187e-05, "loss": 0.0359, "step": 3753 }, { "epoch": 2.6380885453267746, "grad_norm": 0.23971204459667206, "learning_rate": 1.58158819395643e-05, "loss": 0.0288, "step": 3754 }, { "epoch": 2.6387912860154605, "grad_norm": 0.29329657554626465, "learning_rate": 1.5820098383696417e-05, "loss": 0.0302, "step": 3755 }, { "epoch": 2.639494026704146, "grad_norm": 0.4361337423324585, "learning_rate": 1.5824314827828534e-05, "loss": 0.0315, "step": 3756 }, { "epoch": 2.640196767392832, "grad_norm": 0.2769629657268524, "learning_rate": 1.5828531271960647e-05, "loss": 0.0325, "step": 3757 }, { "epoch": 2.640899508081518, "grad_norm": 0.22709588706493378, "learning_rate": 1.5832747716092764e-05, "loss": 0.0268, "step": 3758 }, { "epoch": 2.6416022487702038, "grad_norm": 0.29211926460266113, "learning_rate": 1.5836964160224877e-05, "loss": 0.0392, "step": 3759 }, { "epoch": 2.6423049894588897, "grad_norm": 0.327166348695755, "learning_rate": 1.5841180604356993e-05, "loss": 0.0468, "step": 3760 }, { "epoch": 2.6430077301475756, "grad_norm": 0.25973719358444214, "learning_rate": 1.5845397048489107e-05, "loss": 0.0369, "step": 3761 }, { "epoch": 2.6437104708362615, "grad_norm": 0.29236775636672974, "learning_rate": 1.5849613492621223e-05, "loss": 0.0558, "step": 3762 }, { "epoch": 2.6444132115249475, "grad_norm": 0.24844273924827576, "learning_rate": 1.5853829936753337e-05, "loss": 0.0308, "step": 3763 }, { "epoch": 2.645115952213633, "grad_norm": 0.32577478885650635, "learning_rate": 1.5858046380885453e-05, "loss": 0.0498, "step": 3764 }, { "epoch": 2.645818692902319, "grad_norm": 0.3039495050907135, "learning_rate": 1.586226282501757e-05, "loss": 0.0458, "step": 3765 }, { "epoch": 2.646521433591005, "grad_norm": 0.4187602996826172, "learning_rate": 1.5866479269149687e-05, "loss": 0.0525, "step": 3766 }, { "epoch": 2.6472241742796907, "grad_norm": 0.5137066841125488, "learning_rate": 1.58706957132818e-05, "loss": 0.132, "step": 3767 }, { "epoch": 2.6479269149683766, "grad_norm": 0.6086037755012512, "learning_rate": 1.5874912157413916e-05, "loss": 0.136, "step": 3768 }, { "epoch": 2.6486296556570625, "grad_norm": 1.9668315649032593, "learning_rate": 1.587912860154603e-05, "loss": 0.2567, "step": 3769 }, { "epoch": 2.6493323963457485, "grad_norm": 1.2775843143463135, "learning_rate": 1.5883345045678146e-05, "loss": 0.2779, "step": 3770 }, { "epoch": 2.6500351370344344, "grad_norm": 1.4547398090362549, "learning_rate": 1.588756148981026e-05, "loss": 0.3281, "step": 3771 }, { "epoch": 2.6507378777231203, "grad_norm": 0.33852872252464294, "learning_rate": 1.5891777933942376e-05, "loss": 0.1031, "step": 3772 }, { "epoch": 2.6514406184118062, "grad_norm": 0.5528796315193176, "learning_rate": 1.589599437807449e-05, "loss": 0.0251, "step": 3773 }, { "epoch": 2.652143359100492, "grad_norm": 0.26536664366722107, "learning_rate": 1.5900210822206606e-05, "loss": 0.0518, "step": 3774 }, { "epoch": 2.652846099789178, "grad_norm": 0.2182135134935379, "learning_rate": 1.590442726633872e-05, "loss": 0.0249, "step": 3775 }, { "epoch": 2.6535488404778635, "grad_norm": 0.2096257507801056, "learning_rate": 1.5908643710470836e-05, "loss": 0.0355, "step": 3776 }, { "epoch": 2.6542515811665495, "grad_norm": 0.2695680558681488, "learning_rate": 1.591286015460295e-05, "loss": 0.0399, "step": 3777 }, { "epoch": 2.6549543218552354, "grad_norm": 0.19727708399295807, "learning_rate": 1.5917076598735066e-05, "loss": 0.0204, "step": 3778 }, { "epoch": 2.6556570625439213, "grad_norm": 0.20925076305866241, "learning_rate": 1.5921293042867182e-05, "loss": 0.0293, "step": 3779 }, { "epoch": 2.6563598032326072, "grad_norm": 0.24785757064819336, "learning_rate": 1.59255094869993e-05, "loss": 0.0383, "step": 3780 }, { "epoch": 2.657062543921293, "grad_norm": 0.3159973919391632, "learning_rate": 1.5929725931131412e-05, "loss": 0.0322, "step": 3781 }, { "epoch": 2.657765284609979, "grad_norm": 0.31252431869506836, "learning_rate": 1.593394237526353e-05, "loss": 0.0325, "step": 3782 }, { "epoch": 2.6584680252986645, "grad_norm": 0.2835978865623474, "learning_rate": 1.5938158819395642e-05, "loss": 0.041, "step": 3783 }, { "epoch": 2.6591707659873505, "grad_norm": 0.35704395174980164, "learning_rate": 1.594237526352776e-05, "loss": 0.036, "step": 3784 }, { "epoch": 2.6598735066760364, "grad_norm": 1.6090619564056396, "learning_rate": 1.5946591707659876e-05, "loss": 0.0531, "step": 3785 }, { "epoch": 2.6605762473647223, "grad_norm": 0.29859301447868347, "learning_rate": 1.595080815179199e-05, "loss": 0.0469, "step": 3786 }, { "epoch": 2.6612789880534082, "grad_norm": 0.4124208986759186, "learning_rate": 1.5955024595924105e-05, "loss": 0.0594, "step": 3787 }, { "epoch": 2.661981728742094, "grad_norm": 0.3127163350582123, "learning_rate": 1.595924104005622e-05, "loss": 0.0428, "step": 3788 }, { "epoch": 2.66268446943078, "grad_norm": 0.3087593615055084, "learning_rate": 1.5963457484188335e-05, "loss": 0.0458, "step": 3789 }, { "epoch": 2.663387210119466, "grad_norm": 0.4978085160255432, "learning_rate": 1.596767392832045e-05, "loss": 0.0704, "step": 3790 }, { "epoch": 2.664089950808152, "grad_norm": 0.5215127468109131, "learning_rate": 1.5971890372452565e-05, "loss": 0.1273, "step": 3791 }, { "epoch": 2.664792691496838, "grad_norm": 0.6018799543380737, "learning_rate": 1.5976106816584682e-05, "loss": 0.1402, "step": 3792 }, { "epoch": 2.6654954321855238, "grad_norm": 0.7769544124603271, "learning_rate": 1.59803232607168e-05, "loss": 0.1606, "step": 3793 }, { "epoch": 2.6661981728742097, "grad_norm": 1.1286091804504395, "learning_rate": 1.5984539704848912e-05, "loss": 0.2449, "step": 3794 }, { "epoch": 2.666900913562895, "grad_norm": 1.558337926864624, "learning_rate": 1.598875614898103e-05, "loss": 0.2681, "step": 3795 }, { "epoch": 2.667603654251581, "grad_norm": 1.7626891136169434, "learning_rate": 1.599297259311314e-05, "loss": 0.3824, "step": 3796 }, { "epoch": 2.668306394940267, "grad_norm": 0.5080711841583252, "learning_rate": 1.5997189037245258e-05, "loss": 0.1025, "step": 3797 }, { "epoch": 2.669009135628953, "grad_norm": 0.2783711552619934, "learning_rate": 1.600140548137737e-05, "loss": 0.0493, "step": 3798 }, { "epoch": 2.669711876317639, "grad_norm": 0.2004442811012268, "learning_rate": 1.6005621925509488e-05, "loss": 0.024, "step": 3799 }, { "epoch": 2.6704146170063248, "grad_norm": 0.21838867664337158, "learning_rate": 1.60098383696416e-05, "loss": 0.0269, "step": 3800 }, { "epoch": 2.6711173576950107, "grad_norm": 0.21632051467895508, "learning_rate": 1.6014054813773718e-05, "loss": 0.0332, "step": 3801 }, { "epoch": 2.671820098383696, "grad_norm": 0.20058076083660126, "learning_rate": 1.601827125790583e-05, "loss": 0.0229, "step": 3802 }, { "epoch": 2.672522839072382, "grad_norm": 0.2719428539276123, "learning_rate": 1.6022487702037948e-05, "loss": 0.0479, "step": 3803 }, { "epoch": 2.673225579761068, "grad_norm": 0.20237508416175842, "learning_rate": 1.602670414617006e-05, "loss": 0.0205, "step": 3804 }, { "epoch": 2.673928320449754, "grad_norm": 0.2407163381576538, "learning_rate": 1.6030920590302178e-05, "loss": 0.0342, "step": 3805 }, { "epoch": 2.67463106113844, "grad_norm": 0.19568048417568207, "learning_rate": 1.6035137034434294e-05, "loss": 0.0269, "step": 3806 }, { "epoch": 2.6753338018271258, "grad_norm": 0.3127380907535553, "learning_rate": 1.603935347856641e-05, "loss": 0.0612, "step": 3807 }, { "epoch": 2.6760365425158117, "grad_norm": 0.34178829193115234, "learning_rate": 1.6043569922698524e-05, "loss": 0.0275, "step": 3808 }, { "epoch": 2.6767392832044976, "grad_norm": 0.3097424805164337, "learning_rate": 1.604778636683064e-05, "loss": 0.0467, "step": 3809 }, { "epoch": 2.6774420238931835, "grad_norm": 0.21488171815872192, "learning_rate": 1.6052002810962754e-05, "loss": 0.0282, "step": 3810 }, { "epoch": 2.6781447645818695, "grad_norm": 0.32982543110847473, "learning_rate": 1.605621925509487e-05, "loss": 0.0393, "step": 3811 }, { "epoch": 2.6788475052705554, "grad_norm": 0.338469535112381, "learning_rate": 1.6060435699226984e-05, "loss": 0.0526, "step": 3812 }, { "epoch": 2.6795502459592413, "grad_norm": 0.23561248183250427, "learning_rate": 1.60646521433591e-05, "loss": 0.0336, "step": 3813 }, { "epoch": 2.6802529866479268, "grad_norm": 0.45130717754364014, "learning_rate": 1.6068868587491217e-05, "loss": 0.0557, "step": 3814 }, { "epoch": 2.6809557273366127, "grad_norm": 0.3590386211872101, "learning_rate": 1.607308503162333e-05, "loss": 0.0448, "step": 3815 }, { "epoch": 2.6816584680252986, "grad_norm": 1.3943983316421509, "learning_rate": 1.6077301475755447e-05, "loss": 0.0717, "step": 3816 }, { "epoch": 2.6823612087139845, "grad_norm": 0.5972651243209839, "learning_rate": 1.608151791988756e-05, "loss": 0.1016, "step": 3817 }, { "epoch": 2.6830639494026705, "grad_norm": 0.7811306118965149, "learning_rate": 1.6085734364019677e-05, "loss": 0.1489, "step": 3818 }, { "epoch": 2.6837666900913564, "grad_norm": 0.7939417362213135, "learning_rate": 1.6089950808151794e-05, "loss": 0.2216, "step": 3819 }, { "epoch": 2.6844694307800423, "grad_norm": 1.655401349067688, "learning_rate": 1.609416725228391e-05, "loss": 0.2788, "step": 3820 }, { "epoch": 2.6851721714687278, "grad_norm": 1.680849313735962, "learning_rate": 1.6098383696416024e-05, "loss": 0.3017, "step": 3821 }, { "epoch": 2.6858749121574137, "grad_norm": 0.3202197253704071, "learning_rate": 1.610260014054814e-05, "loss": 0.0925, "step": 3822 }, { "epoch": 2.6865776528460996, "grad_norm": 0.30944177508354187, "learning_rate": 1.6106816584680254e-05, "loss": 0.0414, "step": 3823 }, { "epoch": 2.6872803935347855, "grad_norm": 0.29778051376342773, "learning_rate": 1.611103302881237e-05, "loss": 0.0402, "step": 3824 }, { "epoch": 2.6879831342234715, "grad_norm": 0.2968692183494568, "learning_rate": 1.6115249472944483e-05, "loss": 0.0553, "step": 3825 }, { "epoch": 2.6886858749121574, "grad_norm": 0.19530241191387177, "learning_rate": 1.61194659170766e-05, "loss": 0.0268, "step": 3826 }, { "epoch": 2.6893886156008433, "grad_norm": 0.23175203800201416, "learning_rate": 1.6123682361208713e-05, "loss": 0.0148, "step": 3827 }, { "epoch": 2.6900913562895292, "grad_norm": 0.20814940333366394, "learning_rate": 1.612789880534083e-05, "loss": 0.0287, "step": 3828 }, { "epoch": 2.690794096978215, "grad_norm": 0.23978962004184723, "learning_rate": 1.6132115249472943e-05, "loss": 0.0352, "step": 3829 }, { "epoch": 2.691496837666901, "grad_norm": 0.29808375239372253, "learning_rate": 1.613633169360506e-05, "loss": 0.0381, "step": 3830 }, { "epoch": 2.692199578355587, "grad_norm": 0.21954356133937836, "learning_rate": 1.6140548137737173e-05, "loss": 0.0206, "step": 3831 }, { "epoch": 2.692902319044273, "grad_norm": 0.25609326362609863, "learning_rate": 1.614476458186929e-05, "loss": 0.0379, "step": 3832 }, { "epoch": 2.6936050597329584, "grad_norm": 0.2777397632598877, "learning_rate": 1.6148981026001406e-05, "loss": 0.0373, "step": 3833 }, { "epoch": 2.6943078004216443, "grad_norm": 0.409629225730896, "learning_rate": 1.6153197470133523e-05, "loss": 0.076, "step": 3834 }, { "epoch": 2.6950105411103302, "grad_norm": 0.28521814942359924, "learning_rate": 1.6157413914265636e-05, "loss": 0.0339, "step": 3835 }, { "epoch": 2.695713281799016, "grad_norm": 0.28144580125808716, "learning_rate": 1.6161630358397753e-05, "loss": 0.0353, "step": 3836 }, { "epoch": 2.696416022487702, "grad_norm": 0.33057746291160583, "learning_rate": 1.6165846802529866e-05, "loss": 0.0702, "step": 3837 }, { "epoch": 2.697118763176388, "grad_norm": 0.3193962574005127, "learning_rate": 1.6170063246661983e-05, "loss": 0.0378, "step": 3838 }, { "epoch": 2.697821503865074, "grad_norm": 0.4716837406158447, "learning_rate": 1.6174279690794096e-05, "loss": 0.0749, "step": 3839 }, { "epoch": 2.6985242445537594, "grad_norm": 0.37651553750038147, "learning_rate": 1.6178496134926213e-05, "loss": 0.0627, "step": 3840 }, { "epoch": 2.6992269852424453, "grad_norm": 0.515383243560791, "learning_rate": 1.6182712579058326e-05, "loss": 0.0707, "step": 3841 }, { "epoch": 2.6999297259311312, "grad_norm": 0.630672812461853, "learning_rate": 1.6186929023190443e-05, "loss": 0.134, "step": 3842 }, { "epoch": 2.700632466619817, "grad_norm": 0.8876658082008362, "learning_rate": 1.6191145467322556e-05, "loss": 0.1822, "step": 3843 }, { "epoch": 2.701335207308503, "grad_norm": 0.8082993030548096, "learning_rate": 1.6195361911454672e-05, "loss": 0.2089, "step": 3844 }, { "epoch": 2.702037947997189, "grad_norm": 1.2929019927978516, "learning_rate": 1.619957835558679e-05, "loss": 0.2968, "step": 3845 }, { "epoch": 2.702740688685875, "grad_norm": 2.029362201690674, "learning_rate": 1.6203794799718902e-05, "loss": 0.3248, "step": 3846 }, { "epoch": 2.703443429374561, "grad_norm": 0.4056956171989441, "learning_rate": 1.6208011243851022e-05, "loss": 0.1018, "step": 3847 }, { "epoch": 2.7041461700632468, "grad_norm": 0.5095304846763611, "learning_rate": 1.6212227687983136e-05, "loss": 0.0337, "step": 3848 }, { "epoch": 2.7048489107519327, "grad_norm": 0.22996191680431366, "learning_rate": 1.6216444132115252e-05, "loss": 0.0362, "step": 3849 }, { "epoch": 2.7055516514406186, "grad_norm": 0.23256181180477142, "learning_rate": 1.6220660576247365e-05, "loss": 0.0326, "step": 3850 }, { "epoch": 2.7062543921293045, "grad_norm": 0.19155313074588776, "learning_rate": 1.6224877020379482e-05, "loss": 0.0339, "step": 3851 }, { "epoch": 2.70695713281799, "grad_norm": 0.20772884786128998, "learning_rate": 1.6229093464511595e-05, "loss": 0.0232, "step": 3852 }, { "epoch": 2.707659873506676, "grad_norm": 0.3411765396595001, "learning_rate": 1.6233309908643712e-05, "loss": 0.0375, "step": 3853 }, { "epoch": 2.708362614195362, "grad_norm": 0.2469371259212494, "learning_rate": 1.6237526352775825e-05, "loss": 0.0355, "step": 3854 }, { "epoch": 2.7090653548840478, "grad_norm": 0.20752376317977905, "learning_rate": 1.6241742796907942e-05, "loss": 0.0206, "step": 3855 }, { "epoch": 2.7097680955727337, "grad_norm": 0.24420087039470673, "learning_rate": 1.6245959241040055e-05, "loss": 0.031, "step": 3856 }, { "epoch": 2.7104708362614196, "grad_norm": 0.5426177978515625, "learning_rate": 1.6250175685172172e-05, "loss": 0.0373, "step": 3857 }, { "epoch": 2.7111735769501055, "grad_norm": 0.14102156460285187, "learning_rate": 1.6254392129304285e-05, "loss": 0.0125, "step": 3858 }, { "epoch": 2.7118763176387914, "grad_norm": 0.2855796217918396, "learning_rate": 1.62586085734364e-05, "loss": 0.0449, "step": 3859 }, { "epoch": 2.712579058327477, "grad_norm": 0.45389270782470703, "learning_rate": 1.626282501756852e-05, "loss": 0.0448, "step": 3860 }, { "epoch": 2.713281799016163, "grad_norm": 0.2492777407169342, "learning_rate": 1.6267041461700635e-05, "loss": 0.0331, "step": 3861 }, { "epoch": 2.7139845397048488, "grad_norm": 0.2964981198310852, "learning_rate": 1.6271257905832748e-05, "loss": 0.0614, "step": 3862 }, { "epoch": 2.7146872803935347, "grad_norm": 0.30735233426094055, "learning_rate": 1.6275474349964865e-05, "loss": 0.0399, "step": 3863 }, { "epoch": 2.7153900210822206, "grad_norm": 0.3121742606163025, "learning_rate": 1.6279690794096978e-05, "loss": 0.0435, "step": 3864 }, { "epoch": 2.7160927617709065, "grad_norm": 0.42264658212661743, "learning_rate": 1.6283907238229095e-05, "loss": 0.0842, "step": 3865 }, { "epoch": 2.7167955024595924, "grad_norm": 0.45863932371139526, "learning_rate": 1.6288123682361208e-05, "loss": 0.0789, "step": 3866 }, { "epoch": 2.7174982431482784, "grad_norm": 0.4108208417892456, "learning_rate": 1.6292340126493325e-05, "loss": 0.1126, "step": 3867 }, { "epoch": 2.7182009838369643, "grad_norm": 0.8813245892524719, "learning_rate": 1.6296556570625438e-05, "loss": 0.1436, "step": 3868 }, { "epoch": 2.71890372452565, "grad_norm": 0.8876163363456726, "learning_rate": 1.6300773014757555e-05, "loss": 0.2725, "step": 3869 }, { "epoch": 2.719606465214336, "grad_norm": 1.236005187034607, "learning_rate": 1.6304989458889668e-05, "loss": 0.3203, "step": 3870 }, { "epoch": 2.720309205903022, "grad_norm": 1.642998456954956, "learning_rate": 1.6309205903021784e-05, "loss": 0.3651, "step": 3871 }, { "epoch": 2.7210119465917075, "grad_norm": 0.30181002616882324, "learning_rate": 1.6313422347153898e-05, "loss": 0.0997, "step": 3872 }, { "epoch": 2.7217146872803935, "grad_norm": 0.3344082534313202, "learning_rate": 1.6317638791286014e-05, "loss": 0.0261, "step": 3873 }, { "epoch": 2.7224174279690794, "grad_norm": 0.34548014402389526, "learning_rate": 1.6321855235418134e-05, "loss": 0.0454, "step": 3874 }, { "epoch": 2.7231201686577653, "grad_norm": 0.1967104971408844, "learning_rate": 1.6326071679550248e-05, "loss": 0.0217, "step": 3875 }, { "epoch": 2.723822909346451, "grad_norm": 0.4013204574584961, "learning_rate": 1.6330288123682364e-05, "loss": 0.0297, "step": 3876 }, { "epoch": 2.724525650035137, "grad_norm": 0.22314894199371338, "learning_rate": 1.6334504567814477e-05, "loss": 0.0312, "step": 3877 }, { "epoch": 2.725228390723823, "grad_norm": 0.6630922555923462, "learning_rate": 1.6338721011946594e-05, "loss": 0.0372, "step": 3878 }, { "epoch": 2.7259311314125085, "grad_norm": 0.20455050468444824, "learning_rate": 1.6342937456078707e-05, "loss": 0.0315, "step": 3879 }, { "epoch": 2.7266338721011945, "grad_norm": 0.2894061803817749, "learning_rate": 1.6347153900210824e-05, "loss": 0.0391, "step": 3880 }, { "epoch": 2.7273366127898804, "grad_norm": 0.22628752887248993, "learning_rate": 1.6351370344342937e-05, "loss": 0.0249, "step": 3881 }, { "epoch": 2.7280393534785663, "grad_norm": 0.2816225588321686, "learning_rate": 1.6355586788475054e-05, "loss": 0.0555, "step": 3882 }, { "epoch": 2.728742094167252, "grad_norm": 0.2813699245452881, "learning_rate": 1.6359803232607167e-05, "loss": 0.0382, "step": 3883 }, { "epoch": 2.729444834855938, "grad_norm": 0.28460678458213806, "learning_rate": 1.6364019676739284e-05, "loss": 0.0406, "step": 3884 }, { "epoch": 2.730147575544624, "grad_norm": 0.2265988141298294, "learning_rate": 1.6368236120871397e-05, "loss": 0.0284, "step": 3885 }, { "epoch": 2.73085031623331, "grad_norm": 0.2934742569923401, "learning_rate": 1.6372452565003514e-05, "loss": 0.0507, "step": 3886 }, { "epoch": 2.731553056921996, "grad_norm": 1.6915788650512695, "learning_rate": 1.637666900913563e-05, "loss": 0.0503, "step": 3887 }, { "epoch": 2.732255797610682, "grad_norm": 0.30963075160980225, "learning_rate": 1.6380885453267747e-05, "loss": 0.0487, "step": 3888 }, { "epoch": 2.7329585382993677, "grad_norm": 0.35871681571006775, "learning_rate": 1.638510189739986e-05, "loss": 0.0655, "step": 3889 }, { "epoch": 2.7336612789880537, "grad_norm": 0.3661816120147705, "learning_rate": 1.6389318341531977e-05, "loss": 0.0718, "step": 3890 }, { "epoch": 2.734364019676739, "grad_norm": 0.7460487484931946, "learning_rate": 1.639353478566409e-05, "loss": 0.1098, "step": 3891 }, { "epoch": 2.735066760365425, "grad_norm": 0.6634302735328674, "learning_rate": 1.6397751229796207e-05, "loss": 0.1023, "step": 3892 }, { "epoch": 2.735769501054111, "grad_norm": 0.6264200210571289, "learning_rate": 1.640196767392832e-05, "loss": 0.138, "step": 3893 }, { "epoch": 2.736472241742797, "grad_norm": 1.5298765897750854, "learning_rate": 1.6406184118060437e-05, "loss": 0.2097, "step": 3894 }, { "epoch": 2.737174982431483, "grad_norm": 1.6702492237091064, "learning_rate": 1.641040056219255e-05, "loss": 0.2894, "step": 3895 }, { "epoch": 2.7378777231201687, "grad_norm": 1.7160829305648804, "learning_rate": 1.6414617006324666e-05, "loss": 0.3705, "step": 3896 }, { "epoch": 2.7385804638088547, "grad_norm": 0.42076390981674194, "learning_rate": 1.641883345045678e-05, "loss": 0.1104, "step": 3897 }, { "epoch": 2.73928320449754, "grad_norm": 0.18211588263511658, "learning_rate": 1.6423049894588896e-05, "loss": 0.0333, "step": 3898 }, { "epoch": 2.739985945186226, "grad_norm": 0.3890867829322815, "learning_rate": 1.642726633872101e-05, "loss": 0.0423, "step": 3899 }, { "epoch": 2.740688685874912, "grad_norm": 0.16405260562896729, "learning_rate": 1.6431482782853126e-05, "loss": 0.0304, "step": 3900 }, { "epoch": 2.741391426563598, "grad_norm": 0.5750017166137695, "learning_rate": 1.6435699226985243e-05, "loss": 0.045, "step": 3901 }, { "epoch": 2.742094167252284, "grad_norm": 0.20878338813781738, "learning_rate": 1.643991567111736e-05, "loss": 0.0216, "step": 3902 }, { "epoch": 2.7427969079409698, "grad_norm": 0.210970938205719, "learning_rate": 1.6444132115249476e-05, "loss": 0.0378, "step": 3903 }, { "epoch": 2.7434996486296557, "grad_norm": 0.2222418338060379, "learning_rate": 1.644834855938159e-05, "loss": 0.0518, "step": 3904 }, { "epoch": 2.7442023893183416, "grad_norm": 0.2652226388454437, "learning_rate": 1.6452565003513706e-05, "loss": 0.0241, "step": 3905 }, { "epoch": 2.7449051300070275, "grad_norm": 0.28473472595214844, "learning_rate": 1.645678144764582e-05, "loss": 0.0297, "step": 3906 }, { "epoch": 2.7456078706957134, "grad_norm": 0.2637348175048828, "learning_rate": 1.6460997891777936e-05, "loss": 0.0412, "step": 3907 }, { "epoch": 2.7463106113843994, "grad_norm": 0.39591458439826965, "learning_rate": 1.646521433591005e-05, "loss": 0.0481, "step": 3908 }, { "epoch": 2.7470133520730853, "grad_norm": 0.2347865253686905, "learning_rate": 1.6469430780042166e-05, "loss": 0.0305, "step": 3909 }, { "epoch": 2.7477160927617708, "grad_norm": 0.2582826614379883, "learning_rate": 1.647364722417428e-05, "loss": 0.0405, "step": 3910 }, { "epoch": 2.7484188334504567, "grad_norm": 0.3020707368850708, "learning_rate": 1.6477863668306396e-05, "loss": 0.026, "step": 3911 }, { "epoch": 2.7491215741391426, "grad_norm": 0.35837215185165405, "learning_rate": 1.648208011243851e-05, "loss": 0.0656, "step": 3912 }, { "epoch": 2.7498243148278285, "grad_norm": 0.30356377363204956, "learning_rate": 1.6486296556570626e-05, "loss": 0.0435, "step": 3913 }, { "epoch": 2.7505270555165144, "grad_norm": 0.8049944043159485, "learning_rate": 1.649051300070274e-05, "loss": 0.067, "step": 3914 }, { "epoch": 2.7512297962052004, "grad_norm": 0.29667428135871887, "learning_rate": 1.649472944483486e-05, "loss": 0.0489, "step": 3915 }, { "epoch": 2.7519325368938863, "grad_norm": 0.4893750548362732, "learning_rate": 1.6498945888966972e-05, "loss": 0.0732, "step": 3916 }, { "epoch": 2.7526352775825718, "grad_norm": 0.49753445386886597, "learning_rate": 1.650316233309909e-05, "loss": 0.1148, "step": 3917 }, { "epoch": 2.7533380182712577, "grad_norm": 0.7996389269828796, "learning_rate": 1.6507378777231202e-05, "loss": 0.1981, "step": 3918 }, { "epoch": 2.7540407589599436, "grad_norm": 0.8342161774635315, "learning_rate": 1.651159522136332e-05, "loss": 0.1891, "step": 3919 }, { "epoch": 2.7547434996486295, "grad_norm": 3.4832963943481445, "learning_rate": 1.6515811665495432e-05, "loss": 0.2314, "step": 3920 }, { "epoch": 2.7554462403373154, "grad_norm": 4.343364715576172, "learning_rate": 1.652002810962755e-05, "loss": 0.3643, "step": 3921 }, { "epoch": 2.7561489810260014, "grad_norm": 0.49526873230934143, "learning_rate": 1.6524244553759662e-05, "loss": 0.1115, "step": 3922 }, { "epoch": 2.7568517217146873, "grad_norm": 0.21080340445041656, "learning_rate": 1.652846099789178e-05, "loss": 0.0372, "step": 3923 }, { "epoch": 2.757554462403373, "grad_norm": 0.22503603994846344, "learning_rate": 1.653267744202389e-05, "loss": 0.0439, "step": 3924 }, { "epoch": 2.758257203092059, "grad_norm": 0.1966310292482376, "learning_rate": 1.6536893886156008e-05, "loss": 0.0307, "step": 3925 }, { "epoch": 2.758959943780745, "grad_norm": 0.24477028846740723, "learning_rate": 1.654111033028812e-05, "loss": 0.0258, "step": 3926 }, { "epoch": 2.759662684469431, "grad_norm": 0.23945017158985138, "learning_rate": 1.6545326774420238e-05, "loss": 0.0302, "step": 3927 }, { "epoch": 2.760365425158117, "grad_norm": 0.279860258102417, "learning_rate": 1.6549543218552355e-05, "loss": 0.0319, "step": 3928 }, { "epoch": 2.7610681658468024, "grad_norm": 0.32130467891693115, "learning_rate": 1.655375966268447e-05, "loss": 0.0474, "step": 3929 }, { "epoch": 2.7617709065354883, "grad_norm": 0.25345268845558167, "learning_rate": 1.6557976106816585e-05, "loss": 0.0296, "step": 3930 }, { "epoch": 2.762473647224174, "grad_norm": 0.25132516026496887, "learning_rate": 1.65621925509487e-05, "loss": 0.0257, "step": 3931 }, { "epoch": 2.76317638791286, "grad_norm": 0.3501403033733368, "learning_rate": 1.6566408995080815e-05, "loss": 0.0545, "step": 3932 }, { "epoch": 2.763879128601546, "grad_norm": 0.21105186641216278, "learning_rate": 1.657062543921293e-05, "loss": 0.0251, "step": 3933 }, { "epoch": 2.764581869290232, "grad_norm": 0.3316367566585541, "learning_rate": 1.6574841883345048e-05, "loss": 0.0382, "step": 3934 }, { "epoch": 2.765284609978918, "grad_norm": 0.26892581582069397, "learning_rate": 1.657905832747716e-05, "loss": 0.0314, "step": 3935 }, { "epoch": 2.765987350667604, "grad_norm": 0.2851744294166565, "learning_rate": 1.6583274771609278e-05, "loss": 0.031, "step": 3936 }, { "epoch": 2.7666900913562893, "grad_norm": 0.3164467513561249, "learning_rate": 1.658749121574139e-05, "loss": 0.0336, "step": 3937 }, { "epoch": 2.767392832044975, "grad_norm": 0.30714836716651917, "learning_rate": 1.6591707659873508e-05, "loss": 0.0485, "step": 3938 }, { "epoch": 2.768095572733661, "grad_norm": 0.3794756233692169, "learning_rate": 1.659592410400562e-05, "loss": 0.0647, "step": 3939 }, { "epoch": 2.768798313422347, "grad_norm": 0.31659239530563354, "learning_rate": 1.6600140548137738e-05, "loss": 0.0472, "step": 3940 }, { "epoch": 2.769501054111033, "grad_norm": 0.3753455877304077, "learning_rate": 1.660435699226985e-05, "loss": 0.0627, "step": 3941 }, { "epoch": 2.770203794799719, "grad_norm": 0.5518520474433899, "learning_rate": 1.660857343640197e-05, "loss": 0.0983, "step": 3942 }, { "epoch": 2.770906535488405, "grad_norm": 0.6283957958221436, "learning_rate": 1.6612789880534084e-05, "loss": 0.1259, "step": 3943 }, { "epoch": 2.7716092761770907, "grad_norm": 1.2821624279022217, "learning_rate": 1.66170063246662e-05, "loss": 0.2069, "step": 3944 }, { "epoch": 2.7723120168657767, "grad_norm": 1.1204475164413452, "learning_rate": 1.6621222768798314e-05, "loss": 0.2317, "step": 3945 }, { "epoch": 2.7730147575544626, "grad_norm": 1.8008426427841187, "learning_rate": 1.662543921293043e-05, "loss": 0.3502, "step": 3946 }, { "epoch": 2.7737174982431485, "grad_norm": 0.42342379689216614, "learning_rate": 1.6629655657062544e-05, "loss": 0.1092, "step": 3947 }, { "epoch": 2.7744202389318344, "grad_norm": 0.3771456480026245, "learning_rate": 1.663387210119466e-05, "loss": 0.0679, "step": 3948 }, { "epoch": 2.77512297962052, "grad_norm": 0.19568456709384918, "learning_rate": 1.6638088545326774e-05, "loss": 0.025, "step": 3949 }, { "epoch": 2.775825720309206, "grad_norm": 0.23217462003231049, "learning_rate": 1.664230498945889e-05, "loss": 0.0271, "step": 3950 }, { "epoch": 2.7765284609978917, "grad_norm": 0.2411489188671112, "learning_rate": 1.6646521433591004e-05, "loss": 0.0391, "step": 3951 }, { "epoch": 2.7772312016865777, "grad_norm": 0.2116561233997345, "learning_rate": 1.665073787772312e-05, "loss": 0.0166, "step": 3952 }, { "epoch": 2.7779339423752636, "grad_norm": 0.20157457888126373, "learning_rate": 1.6654954321855233e-05, "loss": 0.0265, "step": 3953 }, { "epoch": 2.7786366830639495, "grad_norm": 0.21044619381427765, "learning_rate": 1.665917076598735e-05, "loss": 0.042, "step": 3954 }, { "epoch": 2.7793394237526354, "grad_norm": 0.20343837141990662, "learning_rate": 1.6663387210119467e-05, "loss": 0.0278, "step": 3955 }, { "epoch": 2.780042164441321, "grad_norm": 0.21762089431285858, "learning_rate": 1.6667603654251583e-05, "loss": 0.0217, "step": 3956 }, { "epoch": 2.780744905130007, "grad_norm": 0.2762684226036072, "learning_rate": 1.6671820098383697e-05, "loss": 0.041, "step": 3957 }, { "epoch": 2.7814476458186927, "grad_norm": 0.19060342013835907, "learning_rate": 1.6676036542515813e-05, "loss": 0.0211, "step": 3958 }, { "epoch": 2.7821503865073787, "grad_norm": 0.2060365378856659, "learning_rate": 1.6680252986647927e-05, "loss": 0.0323, "step": 3959 }, { "epoch": 2.7828531271960646, "grad_norm": 0.2615872323513031, "learning_rate": 1.6684469430780043e-05, "loss": 0.0263, "step": 3960 }, { "epoch": 2.7835558678847505, "grad_norm": 0.4437665641307831, "learning_rate": 1.6688685874912156e-05, "loss": 0.0296, "step": 3961 }, { "epoch": 2.7842586085734364, "grad_norm": 0.37532687187194824, "learning_rate": 1.6692902319044273e-05, "loss": 0.0598, "step": 3962 }, { "epoch": 2.7849613492621224, "grad_norm": 0.2879076302051544, "learning_rate": 1.669711876317639e-05, "loss": 0.0371, "step": 3963 }, { "epoch": 2.7856640899508083, "grad_norm": 0.34577062726020813, "learning_rate": 1.6701335207308503e-05, "loss": 0.0491, "step": 3964 }, { "epoch": 2.786366830639494, "grad_norm": 0.4714205861091614, "learning_rate": 1.670555165144062e-05, "loss": 0.0696, "step": 3965 }, { "epoch": 2.78706957132818, "grad_norm": 0.5102375745773315, "learning_rate": 1.6709768095572733e-05, "loss": 0.0949, "step": 3966 }, { "epoch": 2.787772312016866, "grad_norm": 0.5321997404098511, "learning_rate": 1.671398453970485e-05, "loss": 0.1054, "step": 3967 }, { "epoch": 2.7884750527055515, "grad_norm": 0.8655381798744202, "learning_rate": 1.6718200983836963e-05, "loss": 0.1492, "step": 3968 }, { "epoch": 2.7891777933942374, "grad_norm": 0.9950162172317505, "learning_rate": 1.6722417427969083e-05, "loss": 0.1844, "step": 3969 }, { "epoch": 2.7898805340829234, "grad_norm": 1.21217942237854, "learning_rate": 1.6726633872101196e-05, "loss": 0.3018, "step": 3970 }, { "epoch": 2.7905832747716093, "grad_norm": 1.3774375915527344, "learning_rate": 1.6730850316233313e-05, "loss": 0.3625, "step": 3971 }, { "epoch": 2.791286015460295, "grad_norm": 0.3933236300945282, "learning_rate": 1.6735066760365426e-05, "loss": 0.1055, "step": 3972 }, { "epoch": 2.791988756148981, "grad_norm": 0.29058390855789185, "learning_rate": 1.6739283204497543e-05, "loss": 0.0308, "step": 3973 }, { "epoch": 2.792691496837667, "grad_norm": 0.29900646209716797, "learning_rate": 1.6743499648629656e-05, "loss": 0.0433, "step": 3974 }, { "epoch": 2.7933942375263525, "grad_norm": 0.39251673221588135, "learning_rate": 1.6747716092761772e-05, "loss": 0.0574, "step": 3975 }, { "epoch": 2.7940969782150384, "grad_norm": 0.2571686804294586, "learning_rate": 1.6751932536893886e-05, "loss": 0.0265, "step": 3976 }, { "epoch": 2.7947997189037244, "grad_norm": 0.3146427571773529, "learning_rate": 1.6756148981026002e-05, "loss": 0.021, "step": 3977 }, { "epoch": 2.7955024595924103, "grad_norm": 0.25785788893699646, "learning_rate": 1.6760365425158116e-05, "loss": 0.0322, "step": 3978 }, { "epoch": 2.796205200281096, "grad_norm": 0.3901096284389496, "learning_rate": 1.6764581869290232e-05, "loss": 0.0438, "step": 3979 }, { "epoch": 2.796907940969782, "grad_norm": 0.2600664496421814, "learning_rate": 1.6768798313422345e-05, "loss": 0.0439, "step": 3980 }, { "epoch": 2.797610681658468, "grad_norm": 0.2549424171447754, "learning_rate": 1.6773014757554462e-05, "loss": 0.0507, "step": 3981 }, { "epoch": 2.798313422347154, "grad_norm": 0.21184298396110535, "learning_rate": 1.6777231201686575e-05, "loss": 0.0335, "step": 3982 }, { "epoch": 2.79901616303584, "grad_norm": 0.2327057421207428, "learning_rate": 1.6781447645818695e-05, "loss": 0.0334, "step": 3983 }, { "epoch": 2.799718903724526, "grad_norm": 0.3347474932670593, "learning_rate": 1.678566408995081e-05, "loss": 0.0508, "step": 3984 }, { "epoch": 2.8004216444132117, "grad_norm": 0.4974227249622345, "learning_rate": 1.6789880534082925e-05, "loss": 0.0496, "step": 3985 }, { "epoch": 2.8011243851018977, "grad_norm": 0.2321881651878357, "learning_rate": 1.679409697821504e-05, "loss": 0.0256, "step": 3986 }, { "epoch": 2.801827125790583, "grad_norm": 0.2835923135280609, "learning_rate": 1.6798313422347155e-05, "loss": 0.042, "step": 3987 }, { "epoch": 2.802529866479269, "grad_norm": 0.34869304299354553, "learning_rate": 1.680252986647927e-05, "loss": 0.0506, "step": 3988 }, { "epoch": 2.803232607167955, "grad_norm": 0.44030672311782837, "learning_rate": 1.6806746310611385e-05, "loss": 0.0769, "step": 3989 }, { "epoch": 2.803935347856641, "grad_norm": 0.3351471424102783, "learning_rate": 1.6810962754743498e-05, "loss": 0.0464, "step": 3990 }, { "epoch": 2.804638088545327, "grad_norm": 0.5093134045600891, "learning_rate": 1.6815179198875615e-05, "loss": 0.0698, "step": 3991 }, { "epoch": 2.8053408292340127, "grad_norm": 0.6832724809646606, "learning_rate": 1.6819395643007728e-05, "loss": 0.0843, "step": 3992 }, { "epoch": 2.8060435699226987, "grad_norm": 0.6505734920501709, "learning_rate": 1.6823612087139845e-05, "loss": 0.153, "step": 3993 }, { "epoch": 2.806746310611384, "grad_norm": 0.7677228450775146, "learning_rate": 1.682782853127196e-05, "loss": 0.2053, "step": 3994 }, { "epoch": 2.80744905130007, "grad_norm": 1.2152132987976074, "learning_rate": 1.6832044975404075e-05, "loss": 0.2788, "step": 3995 }, { "epoch": 2.808151791988756, "grad_norm": 1.6697115898132324, "learning_rate": 1.6836261419536195e-05, "loss": 0.345, "step": 3996 }, { "epoch": 2.808854532677442, "grad_norm": 0.4315633773803711, "learning_rate": 1.6840477863668308e-05, "loss": 0.1071, "step": 3997 }, { "epoch": 2.809557273366128, "grad_norm": 0.17828573286533356, "learning_rate": 1.6844694307800425e-05, "loss": 0.03, "step": 3998 }, { "epoch": 2.8102600140548137, "grad_norm": 0.2435149848461151, "learning_rate": 1.6848910751932538e-05, "loss": 0.0399, "step": 3999 }, { "epoch": 2.8109627547434997, "grad_norm": 0.2058778703212738, "learning_rate": 1.6853127196064654e-05, "loss": 0.0304, "step": 4000 }, { "epoch": 2.8109627547434997, "eval_cer": 0.19956248022432435, "eval_loss": 0.3093189001083374, "eval_runtime": 25.771, "eval_samples_per_second": 176.09, "eval_steps_per_second": 0.582, "eval_wer": 0.3681740733499064, "step": 4000 }, { "epoch": 2.8116654954321856, "grad_norm": 0.2753070592880249, "learning_rate": 1.6857343640196768e-05, "loss": 0.0491, "step": 4001 }, { "epoch": 2.8123682361208715, "grad_norm": 0.20775632560253143, "learning_rate": 1.6861560084328884e-05, "loss": 0.0183, "step": 4002 }, { "epoch": 2.8130709768095574, "grad_norm": 0.2270822376012802, "learning_rate": 1.6865776528460998e-05, "loss": 0.0378, "step": 4003 }, { "epoch": 2.8137737174982433, "grad_norm": 0.18353138864040375, "learning_rate": 1.6869992972593114e-05, "loss": 0.0294, "step": 4004 }, { "epoch": 2.8144764581869293, "grad_norm": 0.34310734272003174, "learning_rate": 1.6874209416725227e-05, "loss": 0.0502, "step": 4005 }, { "epoch": 2.8151791988756147, "grad_norm": 0.2620091438293457, "learning_rate": 1.6878425860857344e-05, "loss": 0.0324, "step": 4006 }, { "epoch": 2.8158819395643007, "grad_norm": 0.3820318579673767, "learning_rate": 1.6882642304989457e-05, "loss": 0.0424, "step": 4007 }, { "epoch": 2.8165846802529866, "grad_norm": 0.29846397042274475, "learning_rate": 1.6886858749121574e-05, "loss": 0.0526, "step": 4008 }, { "epoch": 2.8172874209416725, "grad_norm": 0.4400661289691925, "learning_rate": 1.6891075193253687e-05, "loss": 0.0374, "step": 4009 }, { "epoch": 2.8179901616303584, "grad_norm": 0.29874733090400696, "learning_rate": 1.6895291637385807e-05, "loss": 0.0315, "step": 4010 }, { "epoch": 2.8186929023190443, "grad_norm": 0.43165266513824463, "learning_rate": 1.689950808151792e-05, "loss": 0.0587, "step": 4011 }, { "epoch": 2.8193956430077303, "grad_norm": 0.430558443069458, "learning_rate": 1.6903724525650037e-05, "loss": 0.0587, "step": 4012 }, { "epoch": 2.820098383696416, "grad_norm": 0.3691946268081665, "learning_rate": 1.690794096978215e-05, "loss": 0.0293, "step": 4013 }, { "epoch": 2.8208011243851017, "grad_norm": 0.511918842792511, "learning_rate": 1.6912157413914267e-05, "loss": 0.0519, "step": 4014 }, { "epoch": 2.8215038650737876, "grad_norm": 0.4195179045200348, "learning_rate": 1.691637385804638e-05, "loss": 0.0766, "step": 4015 }, { "epoch": 2.8222066057624735, "grad_norm": 0.6635159254074097, "learning_rate": 1.6920590302178497e-05, "loss": 0.0696, "step": 4016 }, { "epoch": 2.8229093464511594, "grad_norm": 0.5400606989860535, "learning_rate": 1.692480674631061e-05, "loss": 0.0974, "step": 4017 }, { "epoch": 2.8236120871398454, "grad_norm": 0.8323060274124146, "learning_rate": 1.6929023190442727e-05, "loss": 0.1578, "step": 4018 }, { "epoch": 2.8243148278285313, "grad_norm": 0.8582503199577332, "learning_rate": 1.693323963457484e-05, "loss": 0.2102, "step": 4019 }, { "epoch": 2.825017568517217, "grad_norm": 1.4247965812683105, "learning_rate": 1.6937456078706957e-05, "loss": 0.2868, "step": 4020 }, { "epoch": 2.825720309205903, "grad_norm": 2.364654779434204, "learning_rate": 1.694167252283907e-05, "loss": 0.3822, "step": 4021 }, { "epoch": 2.826423049894589, "grad_norm": 0.3990311324596405, "learning_rate": 1.6945888966971187e-05, "loss": 0.1058, "step": 4022 }, { "epoch": 2.827125790583275, "grad_norm": 0.2932252883911133, "learning_rate": 1.6950105411103303e-05, "loss": 0.0557, "step": 4023 }, { "epoch": 2.827828531271961, "grad_norm": 0.3219039738178253, "learning_rate": 1.695432185523542e-05, "loss": 0.0342, "step": 4024 }, { "epoch": 2.828531271960647, "grad_norm": 0.2244015634059906, "learning_rate": 1.6958538299367537e-05, "loss": 0.0302, "step": 4025 }, { "epoch": 2.8292340126493323, "grad_norm": 0.24938887357711792, "learning_rate": 1.696275474349965e-05, "loss": 0.0398, "step": 4026 }, { "epoch": 2.829936753338018, "grad_norm": 0.21739919483661652, "learning_rate": 1.6966971187631766e-05, "loss": 0.0291, "step": 4027 }, { "epoch": 2.830639494026704, "grad_norm": 0.1792844533920288, "learning_rate": 1.697118763176388e-05, "loss": 0.0297, "step": 4028 }, { "epoch": 2.83134223471539, "grad_norm": 0.6116797924041748, "learning_rate": 1.6975404075895996e-05, "loss": 0.0224, "step": 4029 }, { "epoch": 2.832044975404076, "grad_norm": 0.24251393973827362, "learning_rate": 1.697962052002811e-05, "loss": 0.0351, "step": 4030 }, { "epoch": 2.832747716092762, "grad_norm": 0.20289157330989838, "learning_rate": 1.6983836964160226e-05, "loss": 0.0407, "step": 4031 }, { "epoch": 2.833450456781448, "grad_norm": 0.36273908615112305, "learning_rate": 1.698805340829234e-05, "loss": 0.04, "step": 4032 }, { "epoch": 2.8341531974701333, "grad_norm": 0.2910057008266449, "learning_rate": 1.6992269852424456e-05, "loss": 0.0262, "step": 4033 }, { "epoch": 2.834855938158819, "grad_norm": 0.33950111269950867, "learning_rate": 1.699648629655657e-05, "loss": 0.0389, "step": 4034 }, { "epoch": 2.835558678847505, "grad_norm": 0.29564574360847473, "learning_rate": 1.7000702740688686e-05, "loss": 0.0375, "step": 4035 }, { "epoch": 2.836261419536191, "grad_norm": 0.3175852596759796, "learning_rate": 1.70049191848208e-05, "loss": 0.0462, "step": 4036 }, { "epoch": 2.836964160224877, "grad_norm": 0.3345243036746979, "learning_rate": 1.700913562895292e-05, "loss": 0.0473, "step": 4037 }, { "epoch": 2.837666900913563, "grad_norm": 0.37426114082336426, "learning_rate": 1.7013352073085032e-05, "loss": 0.0368, "step": 4038 }, { "epoch": 2.838369641602249, "grad_norm": 0.42149287462234497, "learning_rate": 1.701756851721715e-05, "loss": 0.0461, "step": 4039 }, { "epoch": 2.8390723822909347, "grad_norm": 0.4602663218975067, "learning_rate": 1.7021784961349262e-05, "loss": 0.0584, "step": 4040 }, { "epoch": 2.8397751229796206, "grad_norm": 0.6664544939994812, "learning_rate": 1.702600140548138e-05, "loss": 0.0576, "step": 4041 }, { "epoch": 2.8404778636683066, "grad_norm": 0.47187933325767517, "learning_rate": 1.7030217849613492e-05, "loss": 0.0913, "step": 4042 }, { "epoch": 2.8411806043569925, "grad_norm": 0.5480758547782898, "learning_rate": 1.703443429374561e-05, "loss": 0.1472, "step": 4043 }, { "epoch": 2.8418833450456784, "grad_norm": 1.199647068977356, "learning_rate": 1.7038650737877722e-05, "loss": 0.2457, "step": 4044 }, { "epoch": 2.842586085734364, "grad_norm": 1.2339099645614624, "learning_rate": 1.704286718200984e-05, "loss": 0.2756, "step": 4045 }, { "epoch": 2.84328882642305, "grad_norm": 1.8140337467193604, "learning_rate": 1.7047083626141952e-05, "loss": 0.3614, "step": 4046 }, { "epoch": 2.8439915671117357, "grad_norm": 0.5377654433250427, "learning_rate": 1.705130007027407e-05, "loss": 0.1286, "step": 4047 }, { "epoch": 2.8446943078004217, "grad_norm": 0.28934353590011597, "learning_rate": 1.7055516514406182e-05, "loss": 0.0583, "step": 4048 }, { "epoch": 2.8453970484891076, "grad_norm": 0.2587830722332001, "learning_rate": 1.70597329585383e-05, "loss": 0.0557, "step": 4049 }, { "epoch": 2.8460997891777935, "grad_norm": 0.2760067880153656, "learning_rate": 1.7063949402670412e-05, "loss": 0.0418, "step": 4050 }, { "epoch": 2.8468025298664794, "grad_norm": 0.273993581533432, "learning_rate": 1.7068165846802532e-05, "loss": 0.0312, "step": 4051 }, { "epoch": 2.847505270555165, "grad_norm": 0.20590750873088837, "learning_rate": 1.707238229093465e-05, "loss": 0.0325, "step": 4052 }, { "epoch": 2.848208011243851, "grad_norm": 0.19160783290863037, "learning_rate": 1.7076598735066762e-05, "loss": 0.0303, "step": 4053 }, { "epoch": 2.8489107519325367, "grad_norm": 0.29712003469467163, "learning_rate": 1.708081517919888e-05, "loss": 0.0388, "step": 4054 }, { "epoch": 2.8496134926212227, "grad_norm": 0.19166842103004456, "learning_rate": 1.708503162333099e-05, "loss": 0.0317, "step": 4055 }, { "epoch": 2.8503162333099086, "grad_norm": 0.3889882266521454, "learning_rate": 1.7089248067463108e-05, "loss": 0.0247, "step": 4056 }, { "epoch": 2.8510189739985945, "grad_norm": 0.24464020133018494, "learning_rate": 1.709346451159522e-05, "loss": 0.0281, "step": 4057 }, { "epoch": 2.8517217146872804, "grad_norm": 0.25808340311050415, "learning_rate": 1.7097680955727338e-05, "loss": 0.0343, "step": 4058 }, { "epoch": 2.8524244553759663, "grad_norm": 0.3600194752216339, "learning_rate": 1.710189739985945e-05, "loss": 0.0324, "step": 4059 }, { "epoch": 2.8531271960646523, "grad_norm": 0.2960624396800995, "learning_rate": 1.7106113843991568e-05, "loss": 0.0419, "step": 4060 }, { "epoch": 2.853829936753338, "grad_norm": 0.31902027130126953, "learning_rate": 1.711033028812368e-05, "loss": 0.0361, "step": 4061 }, { "epoch": 2.854532677442024, "grad_norm": 0.40450260043144226, "learning_rate": 1.7114546732255798e-05, "loss": 0.0509, "step": 4062 }, { "epoch": 2.85523541813071, "grad_norm": 0.34487879276275635, "learning_rate": 1.711876317638791e-05, "loss": 0.0599, "step": 4063 }, { "epoch": 2.8559381588193955, "grad_norm": 0.2875678837299347, "learning_rate": 1.712297962052003e-05, "loss": 0.0316, "step": 4064 }, { "epoch": 2.8566408995080814, "grad_norm": 0.48357126116752625, "learning_rate": 1.7127196064652144e-05, "loss": 0.0918, "step": 4065 }, { "epoch": 2.8573436401967673, "grad_norm": 0.8646196722984314, "learning_rate": 1.713141250878426e-05, "loss": 0.0725, "step": 4066 }, { "epoch": 2.8580463808854533, "grad_norm": 0.528296172618866, "learning_rate": 1.7135628952916374e-05, "loss": 0.1221, "step": 4067 }, { "epoch": 2.858749121574139, "grad_norm": 0.8046467304229736, "learning_rate": 1.713984539704849e-05, "loss": 0.1526, "step": 4068 }, { "epoch": 2.859451862262825, "grad_norm": 0.7230379581451416, "learning_rate": 1.7144061841180604e-05, "loss": 0.2171, "step": 4069 }, { "epoch": 2.860154602951511, "grad_norm": 1.2948588132858276, "learning_rate": 1.714827828531272e-05, "loss": 0.245, "step": 4070 }, { "epoch": 2.8608573436401965, "grad_norm": 2.342724561691284, "learning_rate": 1.7152494729444834e-05, "loss": 0.4074, "step": 4071 }, { "epoch": 2.8615600843288824, "grad_norm": 0.5538323521614075, "learning_rate": 1.715671117357695e-05, "loss": 0.0944, "step": 4072 }, { "epoch": 2.8622628250175683, "grad_norm": 0.26994583010673523, "learning_rate": 1.7160927617709064e-05, "loss": 0.041, "step": 4073 }, { "epoch": 2.8629655657062543, "grad_norm": 0.28881019353866577, "learning_rate": 1.716514406184118e-05, "loss": 0.0607, "step": 4074 }, { "epoch": 2.86366830639494, "grad_norm": 0.30534157156944275, "learning_rate": 1.7169360505973294e-05, "loss": 0.0381, "step": 4075 }, { "epoch": 2.864371047083626, "grad_norm": 0.24393662810325623, "learning_rate": 1.717357695010541e-05, "loss": 0.0284, "step": 4076 }, { "epoch": 2.865073787772312, "grad_norm": 0.2041911631822586, "learning_rate": 1.7177793394237524e-05, "loss": 0.0371, "step": 4077 }, { "epoch": 2.865776528460998, "grad_norm": 0.23960037529468536, "learning_rate": 1.7182009838369644e-05, "loss": 0.0424, "step": 4078 }, { "epoch": 2.866479269149684, "grad_norm": 0.18925617635250092, "learning_rate": 1.7186226282501757e-05, "loss": 0.0212, "step": 4079 }, { "epoch": 2.86718200983837, "grad_norm": 0.25859057903289795, "learning_rate": 1.7190442726633874e-05, "loss": 0.045, "step": 4080 }, { "epoch": 2.8678847505270557, "grad_norm": 0.28959181904792786, "learning_rate": 1.7194659170765987e-05, "loss": 0.0286, "step": 4081 }, { "epoch": 2.8685874912157416, "grad_norm": 0.294876366853714, "learning_rate": 1.7198875614898104e-05, "loss": 0.0393, "step": 4082 }, { "epoch": 2.869290231904427, "grad_norm": 0.2282397598028183, "learning_rate": 1.720309205903022e-05, "loss": 0.0297, "step": 4083 }, { "epoch": 2.869992972593113, "grad_norm": 0.3018393814563751, "learning_rate": 1.7207308503162333e-05, "loss": 0.0378, "step": 4084 }, { "epoch": 2.870695713281799, "grad_norm": 0.3146160840988159, "learning_rate": 1.721152494729445e-05, "loss": 0.0391, "step": 4085 }, { "epoch": 2.871398453970485, "grad_norm": 0.32630908489227295, "learning_rate": 1.7215741391426563e-05, "loss": 0.0465, "step": 4086 }, { "epoch": 2.872101194659171, "grad_norm": 0.3905276954174042, "learning_rate": 1.721995783555868e-05, "loss": 0.0364, "step": 4087 }, { "epoch": 2.8728039353478567, "grad_norm": 0.2957315742969513, "learning_rate": 1.7224174279690793e-05, "loss": 0.052, "step": 4088 }, { "epoch": 2.8735066760365426, "grad_norm": 0.34404897689819336, "learning_rate": 1.722839072382291e-05, "loss": 0.0705, "step": 4089 }, { "epoch": 2.874209416725228, "grad_norm": 0.3743372857570648, "learning_rate": 1.7232607167955023e-05, "loss": 0.0636, "step": 4090 }, { "epoch": 2.874912157413914, "grad_norm": 0.3106851875782013, "learning_rate": 1.723682361208714e-05, "loss": 0.0657, "step": 4091 }, { "epoch": 2.8756148981026, "grad_norm": 0.47288405895233154, "learning_rate": 1.7241040056219256e-05, "loss": 0.1007, "step": 4092 }, { "epoch": 2.876317638791286, "grad_norm": 0.8163753151893616, "learning_rate": 1.7245256500351373e-05, "loss": 0.1818, "step": 4093 }, { "epoch": 2.877020379479972, "grad_norm": 0.6846240758895874, "learning_rate": 1.7249472944483486e-05, "loss": 0.2097, "step": 4094 }, { "epoch": 2.8777231201686577, "grad_norm": 1.207876443862915, "learning_rate": 1.7253689388615603e-05, "loss": 0.2879, "step": 4095 }, { "epoch": 2.8784258608573436, "grad_norm": 1.1877498626708984, "learning_rate": 1.7257905832747716e-05, "loss": 0.343, "step": 4096 }, { "epoch": 2.8791286015460296, "grad_norm": 0.3008863627910614, "learning_rate": 1.7262122276879833e-05, "loss": 0.0737, "step": 4097 }, { "epoch": 2.8798313422347155, "grad_norm": 0.25775429606437683, "learning_rate": 1.7266338721011946e-05, "loss": 0.0352, "step": 4098 }, { "epoch": 2.8805340829234014, "grad_norm": 0.25326287746429443, "learning_rate": 1.7270555165144063e-05, "loss": 0.0345, "step": 4099 }, { "epoch": 2.8812368236120873, "grad_norm": 0.25777292251586914, "learning_rate": 1.7274771609276176e-05, "loss": 0.0341, "step": 4100 }, { "epoch": 2.8819395643007732, "grad_norm": 0.22834767401218414, "learning_rate": 1.7278988053408293e-05, "loss": 0.0285, "step": 4101 }, { "epoch": 2.8826423049894587, "grad_norm": 0.3622364401817322, "learning_rate": 1.7283204497540406e-05, "loss": 0.0224, "step": 4102 }, { "epoch": 2.8833450456781446, "grad_norm": 0.5690155029296875, "learning_rate": 1.7287420941672522e-05, "loss": 0.0607, "step": 4103 }, { "epoch": 2.8840477863668306, "grad_norm": 0.2481377124786377, "learning_rate": 1.7291637385804636e-05, "loss": 0.0525, "step": 4104 }, { "epoch": 2.8847505270555165, "grad_norm": 0.27876952290534973, "learning_rate": 1.7295853829936756e-05, "loss": 0.03, "step": 4105 }, { "epoch": 2.8854532677442024, "grad_norm": 0.23863095045089722, "learning_rate": 1.730007027406887e-05, "loss": 0.0282, "step": 4106 }, { "epoch": 2.8861560084328883, "grad_norm": 0.4471389949321747, "learning_rate": 1.7304286718200986e-05, "loss": 0.0417, "step": 4107 }, { "epoch": 2.8868587491215743, "grad_norm": 0.6822311282157898, "learning_rate": 1.73085031623331e-05, "loss": 0.0265, "step": 4108 }, { "epoch": 2.88756148981026, "grad_norm": 0.3132081925868988, "learning_rate": 1.7312719606465216e-05, "loss": 0.0576, "step": 4109 }, { "epoch": 2.8882642304989457, "grad_norm": 0.21550527215003967, "learning_rate": 1.731693605059733e-05, "loss": 0.0277, "step": 4110 }, { "epoch": 2.8889669711876316, "grad_norm": 0.8567318916320801, "learning_rate": 1.7321152494729445e-05, "loss": 0.0714, "step": 4111 }, { "epoch": 2.8896697118763175, "grad_norm": 0.8131165504455566, "learning_rate": 1.7325368938861562e-05, "loss": 0.0436, "step": 4112 }, { "epoch": 2.8903724525650034, "grad_norm": 0.24022114276885986, "learning_rate": 1.7329585382993675e-05, "loss": 0.0345, "step": 4113 }, { "epoch": 2.8910751932536893, "grad_norm": 0.6392289400100708, "learning_rate": 1.7333801827125792e-05, "loss": 0.0641, "step": 4114 }, { "epoch": 2.8917779339423753, "grad_norm": 0.37370607256889343, "learning_rate": 1.7338018271257905e-05, "loss": 0.0519, "step": 4115 }, { "epoch": 2.892480674631061, "grad_norm": 0.569402277469635, "learning_rate": 1.7342234715390022e-05, "loss": 0.1091, "step": 4116 }, { "epoch": 2.893183415319747, "grad_norm": 0.5105663537979126, "learning_rate": 1.7346451159522135e-05, "loss": 0.1, "step": 4117 }, { "epoch": 2.893886156008433, "grad_norm": 0.7658608555793762, "learning_rate": 1.7350667603654252e-05, "loss": 0.187, "step": 4118 }, { "epoch": 2.894588896697119, "grad_norm": 1.2550493478775024, "learning_rate": 1.735488404778637e-05, "loss": 0.2462, "step": 4119 }, { "epoch": 2.895291637385805, "grad_norm": 1.5095943212509155, "learning_rate": 1.7359100491918485e-05, "loss": 0.2934, "step": 4120 }, { "epoch": 2.895994378074491, "grad_norm": 1.7128047943115234, "learning_rate": 1.7363316936050598e-05, "loss": 0.3734, "step": 4121 }, { "epoch": 2.8966971187631763, "grad_norm": 0.41152358055114746, "learning_rate": 1.7367533380182715e-05, "loss": 0.0849, "step": 4122 }, { "epoch": 2.897399859451862, "grad_norm": 0.26442453265190125, "learning_rate": 1.7371749824314828e-05, "loss": 0.0409, "step": 4123 }, { "epoch": 2.898102600140548, "grad_norm": 0.293966144323349, "learning_rate": 1.7375966268446945e-05, "loss": 0.0435, "step": 4124 }, { "epoch": 2.898805340829234, "grad_norm": 0.34718263149261475, "learning_rate": 1.7380182712579058e-05, "loss": 0.0316, "step": 4125 }, { "epoch": 2.89950808151792, "grad_norm": 0.327533483505249, "learning_rate": 1.7384399156711175e-05, "loss": 0.0276, "step": 4126 }, { "epoch": 2.900210822206606, "grad_norm": 0.23437069356441498, "learning_rate": 1.7388615600843288e-05, "loss": 0.0306, "step": 4127 }, { "epoch": 2.900913562895292, "grad_norm": 0.295108824968338, "learning_rate": 1.7392832044975405e-05, "loss": 0.0392, "step": 4128 }, { "epoch": 2.9016163035839773, "grad_norm": 0.32614511251449585, "learning_rate": 1.7397048489107518e-05, "loss": 0.0484, "step": 4129 }, { "epoch": 2.902319044272663, "grad_norm": 0.2565227746963501, "learning_rate": 1.7401264933239634e-05, "loss": 0.0284, "step": 4130 }, { "epoch": 2.903021784961349, "grad_norm": 0.21524423360824585, "learning_rate": 1.7405481377371748e-05, "loss": 0.0305, "step": 4131 }, { "epoch": 2.903724525650035, "grad_norm": 0.2375345528125763, "learning_rate": 1.7409697821503868e-05, "loss": 0.0445, "step": 4132 }, { "epoch": 2.904427266338721, "grad_norm": 0.23719212412834167, "learning_rate": 1.741391426563598e-05, "loss": 0.041, "step": 4133 }, { "epoch": 2.905130007027407, "grad_norm": 0.3121221959590912, "learning_rate": 1.7418130709768098e-05, "loss": 0.0622, "step": 4134 }, { "epoch": 2.905832747716093, "grad_norm": 0.2222660779953003, "learning_rate": 1.742234715390021e-05, "loss": 0.0327, "step": 4135 }, { "epoch": 2.9065354884047787, "grad_norm": 0.38209637999534607, "learning_rate": 1.7426563598032327e-05, "loss": 0.0677, "step": 4136 }, { "epoch": 2.9072382290934646, "grad_norm": 0.2624346613883972, "learning_rate": 1.743078004216444e-05, "loss": 0.0527, "step": 4137 }, { "epoch": 2.9079409697821506, "grad_norm": 0.21214567124843597, "learning_rate": 1.7434996486296557e-05, "loss": 0.0324, "step": 4138 }, { "epoch": 2.9086437104708365, "grad_norm": 0.34572115540504456, "learning_rate": 1.743921293042867e-05, "loss": 0.0748, "step": 4139 }, { "epoch": 2.9093464511595224, "grad_norm": 0.4972560703754425, "learning_rate": 1.7443429374560787e-05, "loss": 0.0575, "step": 4140 }, { "epoch": 2.910049191848208, "grad_norm": 0.5477784872055054, "learning_rate": 1.74476458186929e-05, "loss": 0.088, "step": 4141 }, { "epoch": 2.910751932536894, "grad_norm": 0.519290566444397, "learning_rate": 1.7451862262825017e-05, "loss": 0.1092, "step": 4142 }, { "epoch": 2.9114546732255797, "grad_norm": 0.6678764224052429, "learning_rate": 1.7456078706957134e-05, "loss": 0.1728, "step": 4143 }, { "epoch": 2.9121574139142656, "grad_norm": 0.816444993019104, "learning_rate": 1.7460295151089247e-05, "loss": 0.2766, "step": 4144 }, { "epoch": 2.9128601546029516, "grad_norm": 1.230553388595581, "learning_rate": 1.7464511595221364e-05, "loss": 0.2656, "step": 4145 }, { "epoch": 2.9135628952916375, "grad_norm": 1.921717882156372, "learning_rate": 1.746872803935348e-05, "loss": 0.3602, "step": 4146 }, { "epoch": 2.9142656359803234, "grad_norm": 0.5074396133422852, "learning_rate": 1.7472944483485597e-05, "loss": 0.1068, "step": 4147 }, { "epoch": 2.914968376669009, "grad_norm": 0.27557373046875, "learning_rate": 1.747716092761771e-05, "loss": 0.0734, "step": 4148 }, { "epoch": 2.915671117357695, "grad_norm": 0.25607383251190186, "learning_rate": 1.7481377371749827e-05, "loss": 0.0366, "step": 4149 }, { "epoch": 2.9163738580463807, "grad_norm": 0.2557310461997986, "learning_rate": 1.748559381588194e-05, "loss": 0.0331, "step": 4150 }, { "epoch": 2.9170765987350666, "grad_norm": 0.25396212935447693, "learning_rate": 1.7489810260014057e-05, "loss": 0.0238, "step": 4151 }, { "epoch": 2.9177793394237526, "grad_norm": 0.3025567829608917, "learning_rate": 1.749402670414617e-05, "loss": 0.0277, "step": 4152 }, { "epoch": 2.9184820801124385, "grad_norm": 0.23176439106464386, "learning_rate": 1.7498243148278287e-05, "loss": 0.025, "step": 4153 }, { "epoch": 2.9191848208011244, "grad_norm": 0.1611310839653015, "learning_rate": 1.75024595924104e-05, "loss": 0.0196, "step": 4154 }, { "epoch": 2.9198875614898103, "grad_norm": 0.24509619176387787, "learning_rate": 1.7506676036542516e-05, "loss": 0.0422, "step": 4155 }, { "epoch": 2.9205903021784962, "grad_norm": 0.2876393496990204, "learning_rate": 1.751089248067463e-05, "loss": 0.0342, "step": 4156 }, { "epoch": 2.921293042867182, "grad_norm": 0.22622232139110565, "learning_rate": 1.7515108924806746e-05, "loss": 0.0175, "step": 4157 }, { "epoch": 2.921995783555868, "grad_norm": 0.3211851418018341, "learning_rate": 1.751932536893886e-05, "loss": 0.0347, "step": 4158 }, { "epoch": 2.922698524244554, "grad_norm": 0.26673588156700134, "learning_rate": 1.7523541813070976e-05, "loss": 0.0501, "step": 4159 }, { "epoch": 2.9234012649332395, "grad_norm": 0.21337878704071045, "learning_rate": 1.7527758257203093e-05, "loss": 0.0215, "step": 4160 }, { "epoch": 2.9241040056219254, "grad_norm": 0.3171830475330353, "learning_rate": 1.753197470133521e-05, "loss": 0.0529, "step": 4161 }, { "epoch": 2.9248067463106113, "grad_norm": 0.3296900987625122, "learning_rate": 1.7536191145467323e-05, "loss": 0.0477, "step": 4162 }, { "epoch": 2.9255094869992972, "grad_norm": 0.24112588167190552, "learning_rate": 1.754040758959944e-05, "loss": 0.0255, "step": 4163 }, { "epoch": 2.926212227687983, "grad_norm": 0.6166280508041382, "learning_rate": 1.7544624033731553e-05, "loss": 0.0531, "step": 4164 }, { "epoch": 2.926914968376669, "grad_norm": 0.40905919671058655, "learning_rate": 1.754884047786367e-05, "loss": 0.0815, "step": 4165 }, { "epoch": 2.927617709065355, "grad_norm": 0.452171266078949, "learning_rate": 1.7553056921995783e-05, "loss": 0.0965, "step": 4166 }, { "epoch": 2.9283204497540405, "grad_norm": 0.5555853843688965, "learning_rate": 1.75572733661279e-05, "loss": 0.1089, "step": 4167 }, { "epoch": 2.9290231904427264, "grad_norm": 0.9916336536407471, "learning_rate": 1.7561489810260012e-05, "loss": 0.1434, "step": 4168 }, { "epoch": 2.9297259311314123, "grad_norm": 0.8164917230606079, "learning_rate": 1.756570625439213e-05, "loss": 0.2347, "step": 4169 }, { "epoch": 2.9304286718200983, "grad_norm": 0.8897045254707336, "learning_rate": 1.7569922698524242e-05, "loss": 0.2365, "step": 4170 }, { "epoch": 2.931131412508784, "grad_norm": 2.185574531555176, "learning_rate": 1.757413914265636e-05, "loss": 0.3941, "step": 4171 }, { "epoch": 2.93183415319747, "grad_norm": 0.6639603972434998, "learning_rate": 1.7578355586788476e-05, "loss": 0.1085, "step": 4172 }, { "epoch": 2.932536893886156, "grad_norm": 0.48190829157829285, "learning_rate": 1.7582572030920592e-05, "loss": 0.0468, "step": 4173 }, { "epoch": 2.933239634574842, "grad_norm": 0.21963590383529663, "learning_rate": 1.758678847505271e-05, "loss": 0.0303, "step": 4174 }, { "epoch": 2.933942375263528, "grad_norm": 0.2445535510778427, "learning_rate": 1.7591004919184822e-05, "loss": 0.0308, "step": 4175 }, { "epoch": 2.934645115952214, "grad_norm": 0.43476930260658264, "learning_rate": 1.759522136331694e-05, "loss": 0.0305, "step": 4176 }, { "epoch": 2.9353478566408997, "grad_norm": 0.15036779642105103, "learning_rate": 1.7599437807449052e-05, "loss": 0.0217, "step": 4177 }, { "epoch": 2.9360505973295856, "grad_norm": 0.26365745067596436, "learning_rate": 1.760365425158117e-05, "loss": 0.032, "step": 4178 }, { "epoch": 2.936753338018271, "grad_norm": 0.5539330840110779, "learning_rate": 1.7607870695713282e-05, "loss": 0.0419, "step": 4179 }, { "epoch": 2.937456078706957, "grad_norm": 0.3161226809024811, "learning_rate": 1.76120871398454e-05, "loss": 0.0317, "step": 4180 }, { "epoch": 2.938158819395643, "grad_norm": 0.2552627921104431, "learning_rate": 1.7616303583977512e-05, "loss": 0.0271, "step": 4181 }, { "epoch": 2.938861560084329, "grad_norm": 0.22245700657367706, "learning_rate": 1.762052002810963e-05, "loss": 0.0317, "step": 4182 }, { "epoch": 2.939564300773015, "grad_norm": 0.28176742792129517, "learning_rate": 1.762473647224174e-05, "loss": 0.0379, "step": 4183 }, { "epoch": 2.9402670414617007, "grad_norm": 0.28870072960853577, "learning_rate": 1.762895291637386e-05, "loss": 0.0387, "step": 4184 }, { "epoch": 2.9409697821503866, "grad_norm": 0.2507544755935669, "learning_rate": 1.763316936050597e-05, "loss": 0.0324, "step": 4185 }, { "epoch": 2.9416725228390725, "grad_norm": 0.303130179643631, "learning_rate": 1.7637385804638088e-05, "loss": 0.0495, "step": 4186 }, { "epoch": 2.942375263527758, "grad_norm": 0.2593928575515747, "learning_rate": 1.7641602248770205e-05, "loss": 0.0508, "step": 4187 }, { "epoch": 2.943078004216444, "grad_norm": 0.2784077525138855, "learning_rate": 1.764581869290232e-05, "loss": 0.0314, "step": 4188 }, { "epoch": 2.94378074490513, "grad_norm": 0.357710063457489, "learning_rate": 1.7650035137034435e-05, "loss": 0.0646, "step": 4189 }, { "epoch": 2.944483485593816, "grad_norm": 0.39464929699897766, "learning_rate": 1.765425158116655e-05, "loss": 0.0782, "step": 4190 }, { "epoch": 2.9451862262825017, "grad_norm": 0.39321988821029663, "learning_rate": 1.7658468025298665e-05, "loss": 0.0922, "step": 4191 }, { "epoch": 2.9458889669711876, "grad_norm": 0.5958985090255737, "learning_rate": 1.766268446943078e-05, "loss": 0.1047, "step": 4192 }, { "epoch": 2.9465917076598735, "grad_norm": 0.642065703868866, "learning_rate": 1.7666900913562894e-05, "loss": 0.1497, "step": 4193 }, { "epoch": 2.9472944483485595, "grad_norm": 0.9999011158943176, "learning_rate": 1.767111735769501e-05, "loss": 0.2492, "step": 4194 }, { "epoch": 2.9479971890372454, "grad_norm": 0.9471238255500793, "learning_rate": 1.7675333801827124e-05, "loss": 0.259, "step": 4195 }, { "epoch": 2.9486999297259313, "grad_norm": 1.670981526374817, "learning_rate": 1.767955024595924e-05, "loss": 0.3378, "step": 4196 }, { "epoch": 2.9494026704146172, "grad_norm": 0.2795892655849457, "learning_rate": 1.7683766690091354e-05, "loss": 0.0826, "step": 4197 }, { "epoch": 2.950105411103303, "grad_norm": 0.2216757833957672, "learning_rate": 1.768798313422347e-05, "loss": 0.0315, "step": 4198 }, { "epoch": 2.9508081517919886, "grad_norm": 0.23034365475177765, "learning_rate": 1.7692199578355584e-05, "loss": 0.0316, "step": 4199 }, { "epoch": 2.9515108924806746, "grad_norm": 0.2967493534088135, "learning_rate": 1.7696416022487704e-05, "loss": 0.0414, "step": 4200 }, { "epoch": 2.9522136331693605, "grad_norm": 0.2055150717496872, "learning_rate": 1.770063246661982e-05, "loss": 0.0246, "step": 4201 }, { "epoch": 2.9529163738580464, "grad_norm": 0.1753024458885193, "learning_rate": 1.7704848910751934e-05, "loss": 0.0188, "step": 4202 }, { "epoch": 2.9536191145467323, "grad_norm": 0.29369768500328064, "learning_rate": 1.770906535488405e-05, "loss": 0.0384, "step": 4203 }, { "epoch": 2.9543218552354182, "grad_norm": 0.3024703562259674, "learning_rate": 1.7713281799016164e-05, "loss": 0.0475, "step": 4204 }, { "epoch": 2.955024595924104, "grad_norm": 0.27930545806884766, "learning_rate": 1.771749824314828e-05, "loss": 0.0442, "step": 4205 }, { "epoch": 2.9557273366127896, "grad_norm": 0.2267659604549408, "learning_rate": 1.7721714687280394e-05, "loss": 0.0315, "step": 4206 }, { "epoch": 2.9564300773014756, "grad_norm": 0.22774210572242737, "learning_rate": 1.772593113141251e-05, "loss": 0.0281, "step": 4207 }, { "epoch": 2.9571328179901615, "grad_norm": 0.2635183036327362, "learning_rate": 1.7730147575544624e-05, "loss": 0.0321, "step": 4208 }, { "epoch": 2.9578355586788474, "grad_norm": 0.23834773898124695, "learning_rate": 1.773436401967674e-05, "loss": 0.0408, "step": 4209 }, { "epoch": 2.9585382993675333, "grad_norm": 0.24937954545021057, "learning_rate": 1.7738580463808854e-05, "loss": 0.0175, "step": 4210 }, { "epoch": 2.9592410400562192, "grad_norm": 0.43003788590431213, "learning_rate": 1.774279690794097e-05, "loss": 0.0707, "step": 4211 }, { "epoch": 2.959943780744905, "grad_norm": 0.2223757952451706, "learning_rate": 1.7747013352073084e-05, "loss": 0.0366, "step": 4212 }, { "epoch": 2.960646521433591, "grad_norm": 0.30501383543014526, "learning_rate": 1.77512297962052e-05, "loss": 0.0355, "step": 4213 }, { "epoch": 2.961349262122277, "grad_norm": 0.5337935090065002, "learning_rate": 1.7755446240337317e-05, "loss": 0.0534, "step": 4214 }, { "epoch": 2.962052002810963, "grad_norm": 0.39270010590553284, "learning_rate": 1.7759662684469433e-05, "loss": 0.0492, "step": 4215 }, { "epoch": 2.962754743499649, "grad_norm": 0.3119359314441681, "learning_rate": 1.7763879128601547e-05, "loss": 0.0582, "step": 4216 }, { "epoch": 2.9634574841883348, "grad_norm": 0.4212530553340912, "learning_rate": 1.7768095572733663e-05, "loss": 0.1107, "step": 4217 }, { "epoch": 2.9641602248770202, "grad_norm": 0.532664954662323, "learning_rate": 1.7772312016865777e-05, "loss": 0.1391, "step": 4218 }, { "epoch": 2.964862965565706, "grad_norm": 0.8664277195930481, "learning_rate": 1.7776528460997893e-05, "loss": 0.2183, "step": 4219 }, { "epoch": 2.965565706254392, "grad_norm": 2.4939494132995605, "learning_rate": 1.7780744905130006e-05, "loss": 0.2889, "step": 4220 }, { "epoch": 2.966268446943078, "grad_norm": 1.495265007019043, "learning_rate": 1.7784961349262123e-05, "loss": 0.366, "step": 4221 }, { "epoch": 2.966971187631764, "grad_norm": 0.34825438261032104, "learning_rate": 1.7789177793394236e-05, "loss": 0.1067, "step": 4222 }, { "epoch": 2.96767392832045, "grad_norm": 0.19704844057559967, "learning_rate": 1.7793394237526353e-05, "loss": 0.0477, "step": 4223 }, { "epoch": 2.9683766690091358, "grad_norm": 0.260747492313385, "learning_rate": 1.7797610681658466e-05, "loss": 0.0305, "step": 4224 }, { "epoch": 2.9690794096978212, "grad_norm": 0.23855842649936676, "learning_rate": 1.7801827125790583e-05, "loss": 0.0319, "step": 4225 }, { "epoch": 2.969782150386507, "grad_norm": 0.19119422137737274, "learning_rate": 1.7806043569922696e-05, "loss": 0.023, "step": 4226 }, { "epoch": 2.970484891075193, "grad_norm": 0.2205006629228592, "learning_rate": 1.7810260014054813e-05, "loss": 0.0282, "step": 4227 }, { "epoch": 2.971187631763879, "grad_norm": 0.19122414290905, "learning_rate": 1.781447645818693e-05, "loss": 0.0249, "step": 4228 }, { "epoch": 2.971890372452565, "grad_norm": 0.1697464883327484, "learning_rate": 1.7818692902319046e-05, "loss": 0.0188, "step": 4229 }, { "epoch": 2.972593113141251, "grad_norm": 0.2213418185710907, "learning_rate": 1.782290934645116e-05, "loss": 0.0333, "step": 4230 }, { "epoch": 2.9732958538299368, "grad_norm": 0.27556121349334717, "learning_rate": 1.7827125790583276e-05, "loss": 0.0326, "step": 4231 }, { "epoch": 2.9739985945186227, "grad_norm": 0.27498406171798706, "learning_rate": 1.7831342234715393e-05, "loss": 0.0464, "step": 4232 }, { "epoch": 2.9747013352073086, "grad_norm": 0.1374395489692688, "learning_rate": 1.7835558678847506e-05, "loss": 0.0232, "step": 4233 }, { "epoch": 2.9754040758959945, "grad_norm": 0.24788405001163483, "learning_rate": 1.7839775122979622e-05, "loss": 0.038, "step": 4234 }, { "epoch": 2.9761068165846805, "grad_norm": 0.21682381629943848, "learning_rate": 1.7843991567111736e-05, "loss": 0.028, "step": 4235 }, { "epoch": 2.9768095572733664, "grad_norm": 0.35808074474334717, "learning_rate": 1.7848208011243852e-05, "loss": 0.0591, "step": 4236 }, { "epoch": 2.977512297962052, "grad_norm": 0.2641446590423584, "learning_rate": 1.7852424455375966e-05, "loss": 0.0438, "step": 4237 }, { "epoch": 2.978215038650738, "grad_norm": 0.2592465281486511, "learning_rate": 1.7856640899508082e-05, "loss": 0.0424, "step": 4238 }, { "epoch": 2.9789177793394237, "grad_norm": 0.41857632994651794, "learning_rate": 1.7860857343640195e-05, "loss": 0.0734, "step": 4239 }, { "epoch": 2.9796205200281096, "grad_norm": 0.30451181530952454, "learning_rate": 1.7865073787772312e-05, "loss": 0.0441, "step": 4240 }, { "epoch": 2.9803232607167955, "grad_norm": 0.5163853764533997, "learning_rate": 1.786929023190443e-05, "loss": 0.0993, "step": 4241 }, { "epoch": 2.9810260014054815, "grad_norm": 0.4267529845237732, "learning_rate": 1.7873506676036545e-05, "loss": 0.1067, "step": 4242 }, { "epoch": 2.9817287420941674, "grad_norm": 1.2275044918060303, "learning_rate": 1.787772312016866e-05, "loss": 0.1631, "step": 4243 }, { "epoch": 2.982431482782853, "grad_norm": 0.6848306655883789, "learning_rate": 1.7881939564300775e-05, "loss": 0.2541, "step": 4244 }, { "epoch": 2.983134223471539, "grad_norm": 1.2737798690795898, "learning_rate": 1.788615600843289e-05, "loss": 0.2729, "step": 4245 }, { "epoch": 2.9838369641602247, "grad_norm": 1.423606514930725, "learning_rate": 1.7890372452565005e-05, "loss": 0.338, "step": 4246 }, { "epoch": 2.9845397048489106, "grad_norm": 0.36173292994499207, "learning_rate": 1.789458889669712e-05, "loss": 0.0782, "step": 4247 }, { "epoch": 2.9852424455375965, "grad_norm": 0.18440274894237518, "learning_rate": 1.7898805340829235e-05, "loss": 0.0398, "step": 4248 }, { "epoch": 2.9859451862262825, "grad_norm": 0.24267520010471344, "learning_rate": 1.7903021784961348e-05, "loss": 0.0359, "step": 4249 }, { "epoch": 2.9866479269149684, "grad_norm": 0.2378108948469162, "learning_rate": 1.7907238229093465e-05, "loss": 0.025, "step": 4250 }, { "epoch": 2.9873506676036543, "grad_norm": 0.175214022397995, "learning_rate": 1.7911454673225578e-05, "loss": 0.0189, "step": 4251 }, { "epoch": 2.9880534082923402, "grad_norm": 0.2939990758895874, "learning_rate": 1.7915671117357695e-05, "loss": 0.0295, "step": 4252 }, { "epoch": 2.988756148981026, "grad_norm": 0.20822037756443024, "learning_rate": 1.7919887561489808e-05, "loss": 0.0196, "step": 4253 }, { "epoch": 2.989458889669712, "grad_norm": 0.3206077814102173, "learning_rate": 1.7924104005621925e-05, "loss": 0.0514, "step": 4254 }, { "epoch": 2.990161630358398, "grad_norm": 0.209259495139122, "learning_rate": 1.792832044975404e-05, "loss": 0.0276, "step": 4255 }, { "epoch": 2.9908643710470835, "grad_norm": 0.19959436357021332, "learning_rate": 1.7932536893886158e-05, "loss": 0.0231, "step": 4256 }, { "epoch": 2.9915671117357694, "grad_norm": 0.3577663004398346, "learning_rate": 1.793675333801827e-05, "loss": 0.0428, "step": 4257 }, { "epoch": 2.9922698524244553, "grad_norm": 0.22099272906780243, "learning_rate": 1.7940969782150388e-05, "loss": 0.0428, "step": 4258 }, { "epoch": 2.9929725931131412, "grad_norm": 0.21761272847652435, "learning_rate": 1.79451862262825e-05, "loss": 0.0311, "step": 4259 }, { "epoch": 2.993675333801827, "grad_norm": 0.30670398473739624, "learning_rate": 1.7949402670414618e-05, "loss": 0.0481, "step": 4260 }, { "epoch": 2.994378074490513, "grad_norm": 0.2828093469142914, "learning_rate": 1.7953619114546734e-05, "loss": 0.0379, "step": 4261 }, { "epoch": 2.995080815179199, "grad_norm": 0.3066260516643524, "learning_rate": 1.7957835558678848e-05, "loss": 0.0484, "step": 4262 }, { "epoch": 2.9957835558678845, "grad_norm": 0.3665352463722229, "learning_rate": 1.7962052002810964e-05, "loss": 0.0701, "step": 4263 }, { "epoch": 2.9964862965565704, "grad_norm": 0.7882843613624573, "learning_rate": 1.7966268446943078e-05, "loss": 0.1094, "step": 4264 }, { "epoch": 2.9971890372452563, "grad_norm": 0.5286230444908142, "learning_rate": 1.7970484891075194e-05, "loss": 0.102, "step": 4265 }, { "epoch": 2.9978917779339422, "grad_norm": 0.9186925292015076, "learning_rate": 1.7974701335207307e-05, "loss": 0.177, "step": 4266 }, { "epoch": 2.998594518622628, "grad_norm": 0.9348119497299194, "learning_rate": 1.7978917779339424e-05, "loss": 0.2769, "step": 4267 }, { "epoch": 2.999297259311314, "grad_norm": 2.245248317718506, "learning_rate": 1.798313422347154e-05, "loss": 0.2762, "step": 4268 }, { "epoch": 3.0, "grad_norm": 2.1521389484405518, "learning_rate": 1.7987350667603657e-05, "loss": 0.2389, "step": 4269 }, { "epoch": 3.000702740688686, "grad_norm": 0.7015858292579651, "learning_rate": 1.799156711173577e-05, "loss": 0.1141, "step": 4270 }, { "epoch": 3.001405481377372, "grad_norm": 0.5097327828407288, "learning_rate": 1.7995783555867887e-05, "loss": 0.0533, "step": 4271 }, { "epoch": 3.0021082220660578, "grad_norm": 0.214578315615654, "learning_rate": 1.8e-05, "loss": 0.0283, "step": 4272 }, { "epoch": 3.0028109627547437, "grad_norm": 0.3622356653213501, "learning_rate": 1.8004216444132117e-05, "loss": 0.0326, "step": 4273 }, { "epoch": 3.003513703443429, "grad_norm": 0.23802407085895538, "learning_rate": 1.800843288826423e-05, "loss": 0.0289, "step": 4274 }, { "epoch": 3.004216444132115, "grad_norm": 0.32354551553726196, "learning_rate": 1.8012649332396347e-05, "loss": 0.0353, "step": 4275 }, { "epoch": 3.004919184820801, "grad_norm": 0.22655846178531647, "learning_rate": 1.801686577652846e-05, "loss": 0.0325, "step": 4276 }, { "epoch": 3.005621925509487, "grad_norm": 0.3578720986843109, "learning_rate": 1.8021082220660577e-05, "loss": 0.0297, "step": 4277 }, { "epoch": 3.006324666198173, "grad_norm": 0.3207983672618866, "learning_rate": 1.802529866479269e-05, "loss": 0.0391, "step": 4278 }, { "epoch": 3.0070274068868588, "grad_norm": 0.31718146800994873, "learning_rate": 1.8029515108924807e-05, "loss": 0.0268, "step": 4279 }, { "epoch": 3.0077301475755447, "grad_norm": 0.2668655514717102, "learning_rate": 1.803373155305692e-05, "loss": 0.04, "step": 4280 }, { "epoch": 3.0084328882642306, "grad_norm": 0.28919750452041626, "learning_rate": 1.8037947997189037e-05, "loss": 0.0538, "step": 4281 }, { "epoch": 3.0091356289529165, "grad_norm": 0.3490549325942993, "learning_rate": 1.8042164441321153e-05, "loss": 0.0602, "step": 4282 }, { "epoch": 3.0098383696416025, "grad_norm": 0.4816582202911377, "learning_rate": 1.804638088545327e-05, "loss": 0.0397, "step": 4283 }, { "epoch": 3.010541110330288, "grad_norm": 0.42314061522483826, "learning_rate": 1.8050597329585383e-05, "loss": 0.0553, "step": 4284 }, { "epoch": 3.011243851018974, "grad_norm": 0.3773159682750702, "learning_rate": 1.80548137737175e-05, "loss": 0.0439, "step": 4285 }, { "epoch": 3.0119465917076598, "grad_norm": 0.3570081293582916, "learning_rate": 1.8059030217849613e-05, "loss": 0.0274, "step": 4286 }, { "epoch": 3.0126493323963457, "grad_norm": 1.3671510219573975, "learning_rate": 1.806324666198173e-05, "loss": 0.055, "step": 4287 }, { "epoch": 3.0133520730850316, "grad_norm": 0.37132254242897034, "learning_rate": 1.8067463106113843e-05, "loss": 0.0624, "step": 4288 }, { "epoch": 3.0140548137737175, "grad_norm": 0.4834282696247101, "learning_rate": 1.807167955024596e-05, "loss": 0.1047, "step": 4289 }, { "epoch": 3.0147575544624035, "grad_norm": 0.8288926482200623, "learning_rate": 1.8075895994378076e-05, "loss": 0.1103, "step": 4290 }, { "epoch": 3.0154602951510894, "grad_norm": 0.5914306044578552, "learning_rate": 1.808011243851019e-05, "loss": 0.1492, "step": 4291 }, { "epoch": 3.0161630358397753, "grad_norm": 0.9180856347084045, "learning_rate": 1.8084328882642306e-05, "loss": 0.1853, "step": 4292 }, { "epoch": 3.016865776528461, "grad_norm": 1.9289944171905518, "learning_rate": 1.808854532677442e-05, "loss": 0.3079, "step": 4293 }, { "epoch": 3.0175685172171467, "grad_norm": 2.5329387187957764, "learning_rate": 1.8092761770906536e-05, "loss": 0.3225, "step": 4294 }, { "epoch": 3.0182712579058326, "grad_norm": 0.4330596625804901, "learning_rate": 1.809697821503865e-05, "loss": 0.0792, "step": 4295 }, { "epoch": 3.0189739985945185, "grad_norm": 0.3286757171154022, "learning_rate": 1.810119465917077e-05, "loss": 0.0306, "step": 4296 }, { "epoch": 3.0196767392832045, "grad_norm": 0.2130342721939087, "learning_rate": 1.8105411103302883e-05, "loss": 0.0405, "step": 4297 }, { "epoch": 3.0203794799718904, "grad_norm": 0.1985965371131897, "learning_rate": 1.8109627547435e-05, "loss": 0.0318, "step": 4298 }, { "epoch": 3.0210822206605763, "grad_norm": 0.32181501388549805, "learning_rate": 1.8113843991567112e-05, "loss": 0.0501, "step": 4299 }, { "epoch": 3.021784961349262, "grad_norm": 0.15460221469402313, "learning_rate": 1.811806043569923e-05, "loss": 0.0187, "step": 4300 }, { "epoch": 3.022487702037948, "grad_norm": 0.19138284027576447, "learning_rate": 1.8122276879831342e-05, "loss": 0.0222, "step": 4301 }, { "epoch": 3.023190442726634, "grad_norm": 0.49673640727996826, "learning_rate": 1.812649332396346e-05, "loss": 0.0477, "step": 4302 }, { "epoch": 3.0238931834153195, "grad_norm": 0.18878310918807983, "learning_rate": 1.8130709768095572e-05, "loss": 0.0223, "step": 4303 }, { "epoch": 3.0245959241040055, "grad_norm": 0.25051116943359375, "learning_rate": 1.813492621222769e-05, "loss": 0.0325, "step": 4304 }, { "epoch": 3.0252986647926914, "grad_norm": 0.439968466758728, "learning_rate": 1.8139142656359802e-05, "loss": 0.0367, "step": 4305 }, { "epoch": 3.0260014054813773, "grad_norm": 0.2133023738861084, "learning_rate": 1.814335910049192e-05, "loss": 0.0176, "step": 4306 }, { "epoch": 3.0267041461700632, "grad_norm": 0.25502854585647583, "learning_rate": 1.8147575544624032e-05, "loss": 0.0302, "step": 4307 }, { "epoch": 3.027406886858749, "grad_norm": 0.18659743666648865, "learning_rate": 1.815179198875615e-05, "loss": 0.032, "step": 4308 }, { "epoch": 3.028109627547435, "grad_norm": 0.3318876624107361, "learning_rate": 1.8156008432888265e-05, "loss": 0.0455, "step": 4309 }, { "epoch": 3.028812368236121, "grad_norm": 0.3436364233493805, "learning_rate": 1.8160224877020382e-05, "loss": 0.036, "step": 4310 }, { "epoch": 3.029515108924807, "grad_norm": 0.3272637128829956, "learning_rate": 1.8164441321152495e-05, "loss": 0.0407, "step": 4311 }, { "epoch": 3.030217849613493, "grad_norm": 0.3340786099433899, "learning_rate": 1.8168657765284612e-05, "loss": 0.0606, "step": 4312 }, { "epoch": 3.0309205903021783, "grad_norm": 0.3764190375804901, "learning_rate": 1.8172874209416725e-05, "loss": 0.0725, "step": 4313 }, { "epoch": 3.0316233309908642, "grad_norm": 0.3827941119670868, "learning_rate": 1.817709065354884e-05, "loss": 0.0608, "step": 4314 }, { "epoch": 3.03232607167955, "grad_norm": 0.47185850143432617, "learning_rate": 1.8181307097680955e-05, "loss": 0.1045, "step": 4315 }, { "epoch": 3.033028812368236, "grad_norm": 0.9083229303359985, "learning_rate": 1.818552354181307e-05, "loss": 0.1482, "step": 4316 }, { "epoch": 3.033731553056922, "grad_norm": 1.0071195363998413, "learning_rate": 1.8189739985945185e-05, "loss": 0.1785, "step": 4317 }, { "epoch": 3.034434293745608, "grad_norm": 1.2379341125488281, "learning_rate": 1.81939564300773e-05, "loss": 0.284, "step": 4318 }, { "epoch": 3.035137034434294, "grad_norm": 2.2861545085906982, "learning_rate": 1.8198172874209415e-05, "loss": 0.3699, "step": 4319 }, { "epoch": 3.0358397751229798, "grad_norm": 0.2713855803012848, "learning_rate": 1.820238931834153e-05, "loss": 0.1034, "step": 4320 }, { "epoch": 3.0365425158116657, "grad_norm": 0.20887579023838043, "learning_rate": 1.8206605762473648e-05, "loss": 0.0308, "step": 4321 }, { "epoch": 3.037245256500351, "grad_norm": 0.8694881200790405, "learning_rate": 1.821082220660576e-05, "loss": 0.0444, "step": 4322 }, { "epoch": 3.037947997189037, "grad_norm": 0.20705732703208923, "learning_rate": 1.821503865073788e-05, "loss": 0.0345, "step": 4323 }, { "epoch": 3.038650737877723, "grad_norm": 0.30842164158821106, "learning_rate": 1.8219255094869994e-05, "loss": 0.0316, "step": 4324 }, { "epoch": 3.039353478566409, "grad_norm": 0.1805116981267929, "learning_rate": 1.822347153900211e-05, "loss": 0.0257, "step": 4325 }, { "epoch": 3.040056219255095, "grad_norm": 0.2309892773628235, "learning_rate": 1.8227687983134224e-05, "loss": 0.0247, "step": 4326 }, { "epoch": 3.0407589599437808, "grad_norm": 0.40781643986701965, "learning_rate": 1.823190442726634e-05, "loss": 0.046, "step": 4327 }, { "epoch": 3.0414617006324667, "grad_norm": 0.21106302738189697, "learning_rate": 1.8236120871398454e-05, "loss": 0.0295, "step": 4328 }, { "epoch": 3.0421644413211526, "grad_norm": 0.2819582521915436, "learning_rate": 1.824033731553057e-05, "loss": 0.0343, "step": 4329 }, { "epoch": 3.0428671820098385, "grad_norm": 0.2609867453575134, "learning_rate": 1.8244553759662684e-05, "loss": 0.0574, "step": 4330 }, { "epoch": 3.0435699226985244, "grad_norm": 0.29772233963012695, "learning_rate": 1.82487702037948e-05, "loss": 0.0289, "step": 4331 }, { "epoch": 3.04427266338721, "grad_norm": 1.051084280014038, "learning_rate": 1.8252986647926914e-05, "loss": 0.0453, "step": 4332 }, { "epoch": 3.044975404075896, "grad_norm": 0.3991084098815918, "learning_rate": 1.825720309205903e-05, "loss": 0.0256, "step": 4333 }, { "epoch": 3.0456781447645818, "grad_norm": 0.26215726137161255, "learning_rate": 1.8261419536191144e-05, "loss": 0.0351, "step": 4334 }, { "epoch": 3.0463808854532677, "grad_norm": 0.606090247631073, "learning_rate": 1.826563598032326e-05, "loss": 0.0527, "step": 4335 }, { "epoch": 3.0470836261419536, "grad_norm": 0.30667251348495483, "learning_rate": 1.8269852424455377e-05, "loss": 0.0264, "step": 4336 }, { "epoch": 3.0477863668306395, "grad_norm": 0.4189833700656891, "learning_rate": 1.8274068868587494e-05, "loss": 0.0633, "step": 4337 }, { "epoch": 3.0484891075193254, "grad_norm": 0.4346824884414673, "learning_rate": 1.8278285312719607e-05, "loss": 0.0506, "step": 4338 }, { "epoch": 3.0491918482080114, "grad_norm": 0.6377344131469727, "learning_rate": 1.8282501756851724e-05, "loss": 0.0645, "step": 4339 }, { "epoch": 3.0498945888966973, "grad_norm": 0.4980333149433136, "learning_rate": 1.8286718200983837e-05, "loss": 0.1024, "step": 4340 }, { "epoch": 3.050597329585383, "grad_norm": 0.8758746981620789, "learning_rate": 1.8290934645115954e-05, "loss": 0.2086, "step": 4341 }, { "epoch": 3.0513000702740687, "grad_norm": 0.9167750477790833, "learning_rate": 1.8295151089248067e-05, "loss": 0.2076, "step": 4342 }, { "epoch": 3.0520028109627546, "grad_norm": 2.2977712154388428, "learning_rate": 1.8299367533380183e-05, "loss": 0.267, "step": 4343 }, { "epoch": 3.0527055516514405, "grad_norm": 1.917407751083374, "learning_rate": 1.8303583977512297e-05, "loss": 0.3913, "step": 4344 }, { "epoch": 3.0534082923401265, "grad_norm": 0.40892666578292847, "learning_rate": 1.8307800421644413e-05, "loss": 0.11, "step": 4345 }, { "epoch": 3.0541110330288124, "grad_norm": 0.2463347613811493, "learning_rate": 1.8312016865776527e-05, "loss": 0.0383, "step": 4346 }, { "epoch": 3.0548137737174983, "grad_norm": 0.34201177954673767, "learning_rate": 1.8316233309908643e-05, "loss": 0.0362, "step": 4347 }, { "epoch": 3.055516514406184, "grad_norm": 0.2928333878517151, "learning_rate": 1.8320449754040757e-05, "loss": 0.0285, "step": 4348 }, { "epoch": 3.05621925509487, "grad_norm": 0.3525705933570862, "learning_rate": 1.8324666198172873e-05, "loss": 0.0384, "step": 4349 }, { "epoch": 3.056921995783556, "grad_norm": 0.19673864543437958, "learning_rate": 1.8328882642304993e-05, "loss": 0.0195, "step": 4350 }, { "epoch": 3.0576247364722415, "grad_norm": 0.17760661244392395, "learning_rate": 1.8333099086437106e-05, "loss": 0.0248, "step": 4351 }, { "epoch": 3.0583274771609275, "grad_norm": 0.21658405661582947, "learning_rate": 1.8337315530569223e-05, "loss": 0.0357, "step": 4352 }, { "epoch": 3.0590302178496134, "grad_norm": 0.21723893284797668, "learning_rate": 1.8341531974701336e-05, "loss": 0.0303, "step": 4353 }, { "epoch": 3.0597329585382993, "grad_norm": 0.9194353818893433, "learning_rate": 1.8345748418833453e-05, "loss": 0.0387, "step": 4354 }, { "epoch": 3.060435699226985, "grad_norm": 0.2206496000289917, "learning_rate": 1.8349964862965566e-05, "loss": 0.0304, "step": 4355 }, { "epoch": 3.061138439915671, "grad_norm": 0.2816212773323059, "learning_rate": 1.8354181307097683e-05, "loss": 0.0257, "step": 4356 }, { "epoch": 3.061841180604357, "grad_norm": 0.2826598584651947, "learning_rate": 1.8358397751229796e-05, "loss": 0.0289, "step": 4357 }, { "epoch": 3.062543921293043, "grad_norm": 0.25789958238601685, "learning_rate": 1.8362614195361913e-05, "loss": 0.0287, "step": 4358 }, { "epoch": 3.063246661981729, "grad_norm": 0.2504256069660187, "learning_rate": 1.8366830639494026e-05, "loss": 0.0433, "step": 4359 }, { "epoch": 3.063949402670415, "grad_norm": 0.28680121898651123, "learning_rate": 1.8371047083626143e-05, "loss": 0.0406, "step": 4360 }, { "epoch": 3.0646521433591003, "grad_norm": 1.1035540103912354, "learning_rate": 1.8375263527758256e-05, "loss": 0.0397, "step": 4361 }, { "epoch": 3.065354884047786, "grad_norm": 0.5342115759849548, "learning_rate": 1.8379479971890372e-05, "loss": 0.0553, "step": 4362 }, { "epoch": 3.066057624736472, "grad_norm": 0.3685019612312317, "learning_rate": 1.8383696416022486e-05, "loss": 0.0606, "step": 4363 }, { "epoch": 3.066760365425158, "grad_norm": 0.48540374636650085, "learning_rate": 1.8387912860154606e-05, "loss": 0.0781, "step": 4364 }, { "epoch": 3.067463106113844, "grad_norm": 0.5606775283813477, "learning_rate": 1.839212930428672e-05, "loss": 0.1476, "step": 4365 }, { "epoch": 3.06816584680253, "grad_norm": 0.7513205409049988, "learning_rate": 1.8396345748418836e-05, "loss": 0.1565, "step": 4366 }, { "epoch": 3.068868587491216, "grad_norm": 0.8222866654396057, "learning_rate": 1.840056219255095e-05, "loss": 0.2196, "step": 4367 }, { "epoch": 3.0695713281799017, "grad_norm": 1.301911473274231, "learning_rate": 1.8404778636683066e-05, "loss": 0.2326, "step": 4368 }, { "epoch": 3.0702740688685877, "grad_norm": 2.003066301345825, "learning_rate": 1.840899508081518e-05, "loss": 0.3079, "step": 4369 }, { "epoch": 3.0709768095572736, "grad_norm": 0.4375755488872528, "learning_rate": 1.8413211524947295e-05, "loss": 0.0956, "step": 4370 }, { "epoch": 3.071679550245959, "grad_norm": 0.30160239338874817, "learning_rate": 1.841742796907941e-05, "loss": 0.0572, "step": 4371 }, { "epoch": 3.072382290934645, "grad_norm": 0.2575681805610657, "learning_rate": 1.8421644413211525e-05, "loss": 0.0374, "step": 4372 }, { "epoch": 3.073085031623331, "grad_norm": 0.27277064323425293, "learning_rate": 1.842586085734364e-05, "loss": 0.0335, "step": 4373 }, { "epoch": 3.073787772312017, "grad_norm": 0.19365644454956055, "learning_rate": 1.8430077301475755e-05, "loss": 0.0275, "step": 4374 }, { "epoch": 3.0744905130007028, "grad_norm": 0.16689954698085785, "learning_rate": 1.843429374560787e-05, "loss": 0.0212, "step": 4375 }, { "epoch": 3.0751932536893887, "grad_norm": 0.18012875318527222, "learning_rate": 1.8438510189739985e-05, "loss": 0.0237, "step": 4376 }, { "epoch": 3.0758959943780746, "grad_norm": 0.2712500989437103, "learning_rate": 1.8442726633872102e-05, "loss": 0.0275, "step": 4377 }, { "epoch": 3.0765987350667605, "grad_norm": 0.3120836913585663, "learning_rate": 1.844694307800422e-05, "loss": 0.0273, "step": 4378 }, { "epoch": 3.0773014757554464, "grad_norm": 0.25836610794067383, "learning_rate": 1.845115952213633e-05, "loss": 0.0174, "step": 4379 }, { "epoch": 3.078004216444132, "grad_norm": 0.23307666182518005, "learning_rate": 1.8455375966268448e-05, "loss": 0.0347, "step": 4380 }, { "epoch": 3.078706957132818, "grad_norm": 0.18919312953948975, "learning_rate": 1.8459592410400565e-05, "loss": 0.0184, "step": 4381 }, { "epoch": 3.0794096978215038, "grad_norm": 0.24000784754753113, "learning_rate": 1.8463808854532678e-05, "loss": 0.0344, "step": 4382 }, { "epoch": 3.0801124385101897, "grad_norm": 0.24490614235401154, "learning_rate": 1.8468025298664795e-05, "loss": 0.0321, "step": 4383 }, { "epoch": 3.0808151791988756, "grad_norm": 0.390081524848938, "learning_rate": 1.8472241742796908e-05, "loss": 0.0565, "step": 4384 }, { "epoch": 3.0815179198875615, "grad_norm": 0.2944202721118927, "learning_rate": 1.8476458186929025e-05, "loss": 0.0552, "step": 4385 }, { "epoch": 3.0822206605762474, "grad_norm": 0.3291133940219879, "learning_rate": 1.8480674631061138e-05, "loss": 0.0295, "step": 4386 }, { "epoch": 3.0829234012649334, "grad_norm": 0.3316352367401123, "learning_rate": 1.8484891075193255e-05, "loss": 0.0507, "step": 4387 }, { "epoch": 3.0836261419536193, "grad_norm": 0.3690555989742279, "learning_rate": 1.8489107519325368e-05, "loss": 0.0687, "step": 4388 }, { "epoch": 3.084328882642305, "grad_norm": 0.564662754535675, "learning_rate": 1.8493323963457484e-05, "loss": 0.0895, "step": 4389 }, { "epoch": 3.0850316233309907, "grad_norm": 0.41173896193504333, "learning_rate": 1.8497540407589598e-05, "loss": 0.1019, "step": 4390 }, { "epoch": 3.0857343640196766, "grad_norm": 0.5073986053466797, "learning_rate": 1.8501756851721718e-05, "loss": 0.1481, "step": 4391 }, { "epoch": 3.0864371047083625, "grad_norm": 0.8247963786125183, "learning_rate": 1.850597329585383e-05, "loss": 0.2148, "step": 4392 }, { "epoch": 3.0871398453970484, "grad_norm": 1.7063214778900146, "learning_rate": 1.8510189739985948e-05, "loss": 0.2725, "step": 4393 }, { "epoch": 3.0878425860857344, "grad_norm": 2.6154768466949463, "learning_rate": 1.851440618411806e-05, "loss": 0.2937, "step": 4394 }, { "epoch": 3.0885453267744203, "grad_norm": 0.45984509587287903, "learning_rate": 1.8518622628250177e-05, "loss": 0.1387, "step": 4395 }, { "epoch": 3.089248067463106, "grad_norm": 0.3270753026008606, "learning_rate": 1.852283907238229e-05, "loss": 0.0463, "step": 4396 }, { "epoch": 3.089950808151792, "grad_norm": 0.23569658398628235, "learning_rate": 1.8527055516514407e-05, "loss": 0.0339, "step": 4397 }, { "epoch": 3.090653548840478, "grad_norm": 0.26545512676239014, "learning_rate": 1.853127196064652e-05, "loss": 0.0268, "step": 4398 }, { "epoch": 3.0913562895291635, "grad_norm": 0.24679598212242126, "learning_rate": 1.8535488404778637e-05, "loss": 0.0291, "step": 4399 }, { "epoch": 3.0920590302178494, "grad_norm": 0.18416467308998108, "learning_rate": 1.853970484891075e-05, "loss": 0.0246, "step": 4400 }, { "epoch": 3.0927617709065354, "grad_norm": 0.19315820932388306, "learning_rate": 1.8543921293042867e-05, "loss": 0.0323, "step": 4401 }, { "epoch": 3.0934645115952213, "grad_norm": 0.22056899964809418, "learning_rate": 1.854813773717498e-05, "loss": 0.032, "step": 4402 }, { "epoch": 3.094167252283907, "grad_norm": 0.23795919120311737, "learning_rate": 1.8552354181307097e-05, "loss": 0.0316, "step": 4403 }, { "epoch": 3.094869992972593, "grad_norm": 0.2174765020608902, "learning_rate": 1.8556570625439214e-05, "loss": 0.0239, "step": 4404 }, { "epoch": 3.095572733661279, "grad_norm": 0.29431161284446716, "learning_rate": 1.856078706957133e-05, "loss": 0.037, "step": 4405 }, { "epoch": 3.096275474349965, "grad_norm": 0.19456422328948975, "learning_rate": 1.8565003513703444e-05, "loss": 0.0175, "step": 4406 }, { "epoch": 3.096978215038651, "grad_norm": 0.24183128774166107, "learning_rate": 1.856921995783556e-05, "loss": 0.056, "step": 4407 }, { "epoch": 3.097680955727337, "grad_norm": 0.1915961652994156, "learning_rate": 1.8573436401967673e-05, "loss": 0.0279, "step": 4408 }, { "epoch": 3.0983836964160223, "grad_norm": 0.281662255525589, "learning_rate": 1.857765284609979e-05, "loss": 0.0387, "step": 4409 }, { "epoch": 3.099086437104708, "grad_norm": 0.34251371026039124, "learning_rate": 1.8581869290231907e-05, "loss": 0.054, "step": 4410 }, { "epoch": 3.099789177793394, "grad_norm": 0.24533669650554657, "learning_rate": 1.858608573436402e-05, "loss": 0.0284, "step": 4411 }, { "epoch": 3.10049191848208, "grad_norm": 0.2719441056251526, "learning_rate": 1.8590302178496137e-05, "loss": 0.0674, "step": 4412 }, { "epoch": 3.101194659170766, "grad_norm": 0.34653419256210327, "learning_rate": 1.859451862262825e-05, "loss": 0.0639, "step": 4413 }, { "epoch": 3.101897399859452, "grad_norm": 0.37556904554367065, "learning_rate": 1.8598735066760366e-05, "loss": 0.0875, "step": 4414 }, { "epoch": 3.102600140548138, "grad_norm": 0.378926545381546, "learning_rate": 1.860295151089248e-05, "loss": 0.0857, "step": 4415 }, { "epoch": 3.1033028812368237, "grad_norm": 0.44623714685440063, "learning_rate": 1.8607167955024596e-05, "loss": 0.1441, "step": 4416 }, { "epoch": 3.1040056219255097, "grad_norm": 1.3063666820526123, "learning_rate": 1.861138439915671e-05, "loss": 0.2321, "step": 4417 }, { "epoch": 3.104708362614195, "grad_norm": 1.5565615892410278, "learning_rate": 1.861560084328883e-05, "loss": 0.2333, "step": 4418 }, { "epoch": 3.105411103302881, "grad_norm": 1.5308464765548706, "learning_rate": 1.8619817287420943e-05, "loss": 0.3772, "step": 4419 }, { "epoch": 3.106113843991567, "grad_norm": 0.31352055072784424, "learning_rate": 1.862403373155306e-05, "loss": 0.0768, "step": 4420 }, { "epoch": 3.106816584680253, "grad_norm": 0.21217921376228333, "learning_rate": 1.8628250175685173e-05, "loss": 0.031, "step": 4421 }, { "epoch": 3.107519325368939, "grad_norm": 0.15544818341732025, "learning_rate": 1.863246661981729e-05, "loss": 0.0249, "step": 4422 }, { "epoch": 3.1082220660576247, "grad_norm": 0.278952956199646, "learning_rate": 1.8636683063949403e-05, "loss": 0.0387, "step": 4423 }, { "epoch": 3.1089248067463107, "grad_norm": 0.19782419502735138, "learning_rate": 1.864089950808152e-05, "loss": 0.0225, "step": 4424 }, { "epoch": 3.1096275474349966, "grad_norm": 0.31900554895401, "learning_rate": 1.8645115952213633e-05, "loss": 0.0267, "step": 4425 }, { "epoch": 3.1103302881236825, "grad_norm": 0.23635803163051605, "learning_rate": 1.864933239634575e-05, "loss": 0.0346, "step": 4426 }, { "epoch": 3.1110330288123684, "grad_norm": 0.2011687159538269, "learning_rate": 1.8653548840477862e-05, "loss": 0.0307, "step": 4427 }, { "epoch": 3.111735769501054, "grad_norm": 0.2526068687438965, "learning_rate": 1.865776528460998e-05, "loss": 0.0305, "step": 4428 }, { "epoch": 3.11243851018974, "grad_norm": 0.19856871664524078, "learning_rate": 1.8661981728742092e-05, "loss": 0.02, "step": 4429 }, { "epoch": 3.1131412508784257, "grad_norm": 0.31294500827789307, "learning_rate": 1.866619817287421e-05, "loss": 0.0328, "step": 4430 }, { "epoch": 3.1138439915671117, "grad_norm": 0.23368915915489197, "learning_rate": 1.8670414617006322e-05, "loss": 0.0246, "step": 4431 }, { "epoch": 3.1145467322557976, "grad_norm": 0.2947906255722046, "learning_rate": 1.8674631061138442e-05, "loss": 0.0491, "step": 4432 }, { "epoch": 3.1152494729444835, "grad_norm": 0.20334722101688385, "learning_rate": 1.8678847505270556e-05, "loss": 0.0229, "step": 4433 }, { "epoch": 3.1159522136331694, "grad_norm": 0.32851654291152954, "learning_rate": 1.8683063949402672e-05, "loss": 0.0602, "step": 4434 }, { "epoch": 3.1166549543218554, "grad_norm": 0.29304125905036926, "learning_rate": 1.8687280393534785e-05, "loss": 0.0479, "step": 4435 }, { "epoch": 3.1173576950105413, "grad_norm": 0.2755410969257355, "learning_rate": 1.8691496837666902e-05, "loss": 0.0364, "step": 4436 }, { "epoch": 3.118060435699227, "grad_norm": 0.7611328959465027, "learning_rate": 1.8695713281799015e-05, "loss": 0.0617, "step": 4437 }, { "epoch": 3.1187631763879127, "grad_norm": 0.29637548327445984, "learning_rate": 1.8699929725931132e-05, "loss": 0.0647, "step": 4438 }, { "epoch": 3.1194659170765986, "grad_norm": 0.3373330235481262, "learning_rate": 1.870414617006325e-05, "loss": 0.08, "step": 4439 }, { "epoch": 3.1201686577652845, "grad_norm": 0.6668757796287537, "learning_rate": 1.8708362614195362e-05, "loss": 0.1078, "step": 4440 }, { "epoch": 3.1208713984539704, "grad_norm": 0.593548059463501, "learning_rate": 1.871257905832748e-05, "loss": 0.1344, "step": 4441 }, { "epoch": 3.1215741391426564, "grad_norm": 0.7460417151451111, "learning_rate": 1.8716795502459592e-05, "loss": 0.1986, "step": 4442 }, { "epoch": 3.1222768798313423, "grad_norm": 1.148723840713501, "learning_rate": 1.872101194659171e-05, "loss": 0.2651, "step": 4443 }, { "epoch": 3.122979620520028, "grad_norm": 1.5638017654418945, "learning_rate": 1.872522839072382e-05, "loss": 0.3275, "step": 4444 }, { "epoch": 3.123682361208714, "grad_norm": 0.3816300928592682, "learning_rate": 1.872944483485594e-05, "loss": 0.0945, "step": 4445 }, { "epoch": 3.1243851018974, "grad_norm": 0.20103967189788818, "learning_rate": 1.8733661278988055e-05, "loss": 0.0422, "step": 4446 }, { "epoch": 3.125087842586086, "grad_norm": 0.18052372336387634, "learning_rate": 1.873787772312017e-05, "loss": 0.0252, "step": 4447 }, { "epoch": 3.1257905832747714, "grad_norm": 0.6816064119338989, "learning_rate": 1.8742094167252285e-05, "loss": 0.0513, "step": 4448 }, { "epoch": 3.1264933239634574, "grad_norm": 0.1726643443107605, "learning_rate": 1.87463106113844e-05, "loss": 0.0231, "step": 4449 }, { "epoch": 3.1271960646521433, "grad_norm": 0.31147927045822144, "learning_rate": 1.8750527055516515e-05, "loss": 0.0358, "step": 4450 }, { "epoch": 3.127898805340829, "grad_norm": 0.1975892186164856, "learning_rate": 1.875474349964863e-05, "loss": 0.0294, "step": 4451 }, { "epoch": 3.128601546029515, "grad_norm": 0.2544596791267395, "learning_rate": 1.8758959943780745e-05, "loss": 0.0353, "step": 4452 }, { "epoch": 3.129304286718201, "grad_norm": 0.256497859954834, "learning_rate": 1.876317638791286e-05, "loss": 0.0263, "step": 4453 }, { "epoch": 3.130007027406887, "grad_norm": 0.2056000530719757, "learning_rate": 1.8767392832044974e-05, "loss": 0.0176, "step": 4454 }, { "epoch": 3.130709768095573, "grad_norm": 0.3119157552719116, "learning_rate": 1.877160927617709e-05, "loss": 0.0233, "step": 4455 }, { "epoch": 3.131412508784259, "grad_norm": 0.24780814349651337, "learning_rate": 1.8775825720309204e-05, "loss": 0.0395, "step": 4456 }, { "epoch": 3.1321152494729443, "grad_norm": 0.32991185784339905, "learning_rate": 1.878004216444132e-05, "loss": 0.0253, "step": 4457 }, { "epoch": 3.13281799016163, "grad_norm": 0.31777358055114746, "learning_rate": 1.8784258608573434e-05, "loss": 0.0193, "step": 4458 }, { "epoch": 3.133520730850316, "grad_norm": 0.30310702323913574, "learning_rate": 1.8788475052705554e-05, "loss": 0.0387, "step": 4459 }, { "epoch": 3.134223471539002, "grad_norm": 0.5465342402458191, "learning_rate": 1.8792691496837667e-05, "loss": 0.0531, "step": 4460 }, { "epoch": 3.134926212227688, "grad_norm": 0.2653489112854004, "learning_rate": 1.8796907940969784e-05, "loss": 0.0256, "step": 4461 }, { "epoch": 3.135628952916374, "grad_norm": 0.2724943161010742, "learning_rate": 1.8801124385101897e-05, "loss": 0.0465, "step": 4462 }, { "epoch": 3.13633169360506, "grad_norm": 0.3698204755783081, "learning_rate": 1.8805340829234014e-05, "loss": 0.0693, "step": 4463 }, { "epoch": 3.1370344342937457, "grad_norm": 0.44183051586151123, "learning_rate": 1.8809557273366127e-05, "loss": 0.0582, "step": 4464 }, { "epoch": 3.1377371749824317, "grad_norm": 0.5760088562965393, "learning_rate": 1.8813773717498244e-05, "loss": 0.1415, "step": 4465 }, { "epoch": 3.1384399156711176, "grad_norm": 0.5621898770332336, "learning_rate": 1.8817990161630357e-05, "loss": 0.1529, "step": 4466 }, { "epoch": 3.139142656359803, "grad_norm": 1.278216004371643, "learning_rate": 1.8822206605762474e-05, "loss": 0.2089, "step": 4467 }, { "epoch": 3.139845397048489, "grad_norm": 1.1926189661026, "learning_rate": 1.8826423049894587e-05, "loss": 0.2467, "step": 4468 }, { "epoch": 3.140548137737175, "grad_norm": 1.5129810571670532, "learning_rate": 1.8830639494026704e-05, "loss": 0.2917, "step": 4469 }, { "epoch": 3.141250878425861, "grad_norm": 0.35350242257118225, "learning_rate": 1.883485593815882e-05, "loss": 0.0939, "step": 4470 }, { "epoch": 3.1419536191145467, "grad_norm": 0.20099963247776031, "learning_rate": 1.8839072382290934e-05, "loss": 0.0389, "step": 4471 }, { "epoch": 3.1426563598032327, "grad_norm": 0.26429203152656555, "learning_rate": 1.884328882642305e-05, "loss": 0.0352, "step": 4472 }, { "epoch": 3.1433591004919186, "grad_norm": 0.1882941722869873, "learning_rate": 1.8847505270555167e-05, "loss": 0.0263, "step": 4473 }, { "epoch": 3.1440618411806045, "grad_norm": 0.2182316780090332, "learning_rate": 1.8851721714687283e-05, "loss": 0.0322, "step": 4474 }, { "epoch": 3.1447645818692904, "grad_norm": 0.1990308314561844, "learning_rate": 1.8855938158819397e-05, "loss": 0.016, "step": 4475 }, { "epoch": 3.145467322557976, "grad_norm": 0.15004949271678925, "learning_rate": 1.8860154602951513e-05, "loss": 0.0215, "step": 4476 }, { "epoch": 3.146170063246662, "grad_norm": 0.26254457235336304, "learning_rate": 1.8864371047083627e-05, "loss": 0.0613, "step": 4477 }, { "epoch": 3.1468728039353477, "grad_norm": 0.16809523105621338, "learning_rate": 1.8868587491215743e-05, "loss": 0.028, "step": 4478 }, { "epoch": 3.1475755446240337, "grad_norm": 0.25094738602638245, "learning_rate": 1.8872803935347856e-05, "loss": 0.0299, "step": 4479 }, { "epoch": 3.1482782853127196, "grad_norm": 0.3872305750846863, "learning_rate": 1.8877020379479973e-05, "loss": 0.0461, "step": 4480 }, { "epoch": 3.1489810260014055, "grad_norm": 0.2141571342945099, "learning_rate": 1.8881236823612086e-05, "loss": 0.0169, "step": 4481 }, { "epoch": 3.1496837666900914, "grad_norm": 0.2697291076183319, "learning_rate": 1.8885453267744203e-05, "loss": 0.0411, "step": 4482 }, { "epoch": 3.1503865073787773, "grad_norm": 0.27330631017684937, "learning_rate": 1.8889669711876316e-05, "loss": 0.0284, "step": 4483 }, { "epoch": 3.1510892480674633, "grad_norm": 0.37506353855133057, "learning_rate": 1.8893886156008433e-05, "loss": 0.0418, "step": 4484 }, { "epoch": 3.151791988756149, "grad_norm": 0.22813518345355988, "learning_rate": 1.8898102600140546e-05, "loss": 0.0438, "step": 4485 }, { "epoch": 3.1524947294448347, "grad_norm": 0.27192968130111694, "learning_rate": 1.8902319044272666e-05, "loss": 0.0496, "step": 4486 }, { "epoch": 3.1531974701335206, "grad_norm": 0.23229846358299255, "learning_rate": 1.890653548840478e-05, "loss": 0.0354, "step": 4487 }, { "epoch": 3.1539002108222065, "grad_norm": 0.44041958451271057, "learning_rate": 1.8910751932536896e-05, "loss": 0.0827, "step": 4488 }, { "epoch": 3.1546029515108924, "grad_norm": 0.4334403872489929, "learning_rate": 1.891496837666901e-05, "loss": 0.1071, "step": 4489 }, { "epoch": 3.1553056921995783, "grad_norm": 0.5157772898674011, "learning_rate": 1.8919184820801126e-05, "loss": 0.1169, "step": 4490 }, { "epoch": 3.1560084328882643, "grad_norm": 0.6559209823608398, "learning_rate": 1.892340126493324e-05, "loss": 0.1597, "step": 4491 }, { "epoch": 3.15671117357695, "grad_norm": 0.8445542454719543, "learning_rate": 1.8927617709065356e-05, "loss": 0.1953, "step": 4492 }, { "epoch": 3.157413914265636, "grad_norm": 1.385438323020935, "learning_rate": 1.893183415319747e-05, "loss": 0.2393, "step": 4493 }, { "epoch": 3.158116654954322, "grad_norm": 1.740832805633545, "learning_rate": 1.8936050597329586e-05, "loss": 0.3535, "step": 4494 }, { "epoch": 3.1588193956430075, "grad_norm": 0.33746692538261414, "learning_rate": 1.89402670414617e-05, "loss": 0.0873, "step": 4495 }, { "epoch": 3.1595221363316934, "grad_norm": 0.33655035495758057, "learning_rate": 1.8944483485593816e-05, "loss": 0.0388, "step": 4496 }, { "epoch": 3.1602248770203794, "grad_norm": 0.23918123543262482, "learning_rate": 1.894869992972593e-05, "loss": 0.0352, "step": 4497 }, { "epoch": 3.1609276177090653, "grad_norm": 0.17631889879703522, "learning_rate": 1.8952916373858045e-05, "loss": 0.0223, "step": 4498 }, { "epoch": 3.161630358397751, "grad_norm": 0.1830848753452301, "learning_rate": 1.8957132817990162e-05, "loss": 0.035, "step": 4499 }, { "epoch": 3.162333099086437, "grad_norm": 0.23049825429916382, "learning_rate": 1.896134926212228e-05, "loss": 0.0311, "step": 4500 }, { "epoch": 3.163035839775123, "grad_norm": 0.2774304449558258, "learning_rate": 1.8965565706254395e-05, "loss": 0.0222, "step": 4501 }, { "epoch": 3.163738580463809, "grad_norm": 0.1869937628507614, "learning_rate": 1.896978215038651e-05, "loss": 0.0278, "step": 4502 }, { "epoch": 3.164441321152495, "grad_norm": 0.15922853350639343, "learning_rate": 1.8973998594518625e-05, "loss": 0.0202, "step": 4503 }, { "epoch": 3.165144061841181, "grad_norm": 0.45634955167770386, "learning_rate": 1.897821503865074e-05, "loss": 0.0332, "step": 4504 }, { "epoch": 3.1658468025298663, "grad_norm": 0.2942720651626587, "learning_rate": 1.8982431482782855e-05, "loss": 0.0485, "step": 4505 }, { "epoch": 3.166549543218552, "grad_norm": 0.2357647716999054, "learning_rate": 1.898664792691497e-05, "loss": 0.0235, "step": 4506 }, { "epoch": 3.167252283907238, "grad_norm": 0.26395902037620544, "learning_rate": 1.8990864371047085e-05, "loss": 0.0558, "step": 4507 }, { "epoch": 3.167955024595924, "grad_norm": 0.21713240444660187, "learning_rate": 1.8995080815179198e-05, "loss": 0.0231, "step": 4508 }, { "epoch": 3.16865776528461, "grad_norm": 0.2695079445838928, "learning_rate": 1.8999297259311315e-05, "loss": 0.0626, "step": 4509 }, { "epoch": 3.169360505973296, "grad_norm": 0.25310054421424866, "learning_rate": 1.9003513703443428e-05, "loss": 0.0624, "step": 4510 }, { "epoch": 3.170063246661982, "grad_norm": 0.2661063075065613, "learning_rate": 1.9007730147575545e-05, "loss": 0.0499, "step": 4511 }, { "epoch": 3.1707659873506677, "grad_norm": 0.3299630582332611, "learning_rate": 1.9011946591707658e-05, "loss": 0.0582, "step": 4512 }, { "epoch": 3.1714687280393536, "grad_norm": 0.26649007201194763, "learning_rate": 1.9016163035839778e-05, "loss": 0.0583, "step": 4513 }, { "epoch": 3.172171468728039, "grad_norm": 0.40280523896217346, "learning_rate": 1.902037947997189e-05, "loss": 0.07, "step": 4514 }, { "epoch": 3.172874209416725, "grad_norm": 0.6235186457633972, "learning_rate": 1.9024595924104008e-05, "loss": 0.0996, "step": 4515 }, { "epoch": 3.173576950105411, "grad_norm": 0.6379192471504211, "learning_rate": 1.902881236823612e-05, "loss": 0.1841, "step": 4516 }, { "epoch": 3.174279690794097, "grad_norm": 0.7277299165725708, "learning_rate": 1.9033028812368238e-05, "loss": 0.2103, "step": 4517 }, { "epoch": 3.174982431482783, "grad_norm": 1.053914189338684, "learning_rate": 1.903724525650035e-05, "loss": 0.2696, "step": 4518 }, { "epoch": 3.1756851721714687, "grad_norm": 1.9425016641616821, "learning_rate": 1.9041461700632468e-05, "loss": 0.2944, "step": 4519 }, { "epoch": 3.1763879128601546, "grad_norm": 0.31560373306274414, "learning_rate": 1.904567814476458e-05, "loss": 0.0884, "step": 4520 }, { "epoch": 3.1770906535488406, "grad_norm": 0.18818162381649017, "learning_rate": 1.9049894588896698e-05, "loss": 0.035, "step": 4521 }, { "epoch": 3.1777933942375265, "grad_norm": 0.16799087822437286, "learning_rate": 1.905411103302881e-05, "loss": 0.0227, "step": 4522 }, { "epoch": 3.1784961349262124, "grad_norm": 0.2812698781490326, "learning_rate": 1.9058327477160928e-05, "loss": 0.0328, "step": 4523 }, { "epoch": 3.1791988756148983, "grad_norm": 0.17350056767463684, "learning_rate": 1.906254392129304e-05, "loss": 0.027, "step": 4524 }, { "epoch": 3.179901616303584, "grad_norm": 0.15489748120307922, "learning_rate": 1.9066760365425157e-05, "loss": 0.0168, "step": 4525 }, { "epoch": 3.1806043569922697, "grad_norm": 0.6328380703926086, "learning_rate": 1.907097680955727e-05, "loss": 0.0356, "step": 4526 }, { "epoch": 3.1813070976809557, "grad_norm": 0.2093203216791153, "learning_rate": 1.907519325368939e-05, "loss": 0.0295, "step": 4527 }, { "epoch": 3.1820098383696416, "grad_norm": 0.2618841826915741, "learning_rate": 1.9079409697821504e-05, "loss": 0.035, "step": 4528 }, { "epoch": 3.1827125790583275, "grad_norm": 0.41609007120132446, "learning_rate": 1.908362614195362e-05, "loss": 0.0241, "step": 4529 }, { "epoch": 3.1834153197470134, "grad_norm": 0.27193939685821533, "learning_rate": 1.9087842586085737e-05, "loss": 0.059, "step": 4530 }, { "epoch": 3.1841180604356993, "grad_norm": 0.2643492817878723, "learning_rate": 1.909205903021785e-05, "loss": 0.0285, "step": 4531 }, { "epoch": 3.1848208011243853, "grad_norm": 0.26017123460769653, "learning_rate": 1.9096275474349967e-05, "loss": 0.058, "step": 4532 }, { "epoch": 3.185523541813071, "grad_norm": 0.3106161952018738, "learning_rate": 1.910049191848208e-05, "loss": 0.0472, "step": 4533 }, { "epoch": 3.1862262825017567, "grad_norm": 0.37729907035827637, "learning_rate": 1.9104708362614197e-05, "loss": 0.0453, "step": 4534 }, { "epoch": 3.1869290231904426, "grad_norm": 0.26060163974761963, "learning_rate": 1.910892480674631e-05, "loss": 0.0435, "step": 4535 }, { "epoch": 3.1876317638791285, "grad_norm": 0.3170732259750366, "learning_rate": 1.9113141250878427e-05, "loss": 0.0563, "step": 4536 }, { "epoch": 3.1883345045678144, "grad_norm": 0.26352962851524353, "learning_rate": 1.911735769501054e-05, "loss": 0.034, "step": 4537 }, { "epoch": 3.1890372452565003, "grad_norm": 0.7608757019042969, "learning_rate": 1.9121574139142657e-05, "loss": 0.0941, "step": 4538 }, { "epoch": 3.1897399859451863, "grad_norm": 0.5121880173683167, "learning_rate": 1.912579058327477e-05, "loss": 0.091, "step": 4539 }, { "epoch": 3.190442726633872, "grad_norm": 0.5765411257743835, "learning_rate": 1.9130007027406887e-05, "loss": 0.0856, "step": 4540 }, { "epoch": 3.191145467322558, "grad_norm": 0.5562862157821655, "learning_rate": 1.9134223471539003e-05, "loss": 0.1574, "step": 4541 }, { "epoch": 3.191848208011244, "grad_norm": 1.0601391792297363, "learning_rate": 1.913843991567112e-05, "loss": 0.2251, "step": 4542 }, { "epoch": 3.19255094869993, "grad_norm": 0.9397532939910889, "learning_rate": 1.9142656359803233e-05, "loss": 0.2785, "step": 4543 }, { "epoch": 3.1932536893886154, "grad_norm": 1.5407410860061646, "learning_rate": 1.914687280393535e-05, "loss": 0.2634, "step": 4544 }, { "epoch": 3.1939564300773013, "grad_norm": 0.3744645118713379, "learning_rate": 1.9151089248067463e-05, "loss": 0.0755, "step": 4545 }, { "epoch": 3.1946591707659873, "grad_norm": 0.26957032084465027, "learning_rate": 1.915530569219958e-05, "loss": 0.0397, "step": 4546 }, { "epoch": 3.195361911454673, "grad_norm": 0.19324228167533875, "learning_rate": 1.9159522136331693e-05, "loss": 0.0317, "step": 4547 }, { "epoch": 3.196064652143359, "grad_norm": 0.23435264825820923, "learning_rate": 1.916373858046381e-05, "loss": 0.0286, "step": 4548 }, { "epoch": 3.196767392832045, "grad_norm": 0.2432665079832077, "learning_rate": 1.9167955024595923e-05, "loss": 0.0264, "step": 4549 }, { "epoch": 3.197470133520731, "grad_norm": 0.21417918801307678, "learning_rate": 1.917217146872804e-05, "loss": 0.0327, "step": 4550 }, { "epoch": 3.198172874209417, "grad_norm": 0.20887446403503418, "learning_rate": 1.9176387912860153e-05, "loss": 0.0246, "step": 4551 }, { "epoch": 3.198875614898103, "grad_norm": 0.24981409311294556, "learning_rate": 1.918060435699227e-05, "loss": 0.0386, "step": 4552 }, { "epoch": 3.1995783555867883, "grad_norm": 0.374336302280426, "learning_rate": 1.9184820801124383e-05, "loss": 0.0341, "step": 4553 }, { "epoch": 3.200281096275474, "grad_norm": 0.17803139984607697, "learning_rate": 1.9189037245256503e-05, "loss": 0.0213, "step": 4554 }, { "epoch": 3.20098383696416, "grad_norm": 0.4674331247806549, "learning_rate": 1.9193253689388616e-05, "loss": 0.0457, "step": 4555 }, { "epoch": 3.201686577652846, "grad_norm": 0.4018625319004059, "learning_rate": 1.9197470133520733e-05, "loss": 0.0244, "step": 4556 }, { "epoch": 3.202389318341532, "grad_norm": 0.34366464614868164, "learning_rate": 1.9201686577652846e-05, "loss": 0.0477, "step": 4557 }, { "epoch": 3.203092059030218, "grad_norm": 0.25277179479599, "learning_rate": 1.9205903021784962e-05, "loss": 0.0266, "step": 4558 }, { "epoch": 3.203794799718904, "grad_norm": 0.2159193903207779, "learning_rate": 1.921011946591708e-05, "loss": 0.0422, "step": 4559 }, { "epoch": 3.2044975404075897, "grad_norm": 0.3291636109352112, "learning_rate": 1.9214335910049192e-05, "loss": 0.0435, "step": 4560 }, { "epoch": 3.2052002810962756, "grad_norm": 0.2501731812953949, "learning_rate": 1.921855235418131e-05, "loss": 0.0297, "step": 4561 }, { "epoch": 3.2059030217849616, "grad_norm": 0.759652316570282, "learning_rate": 1.9222768798313422e-05, "loss": 0.0808, "step": 4562 }, { "epoch": 3.206605762473647, "grad_norm": 0.3180261254310608, "learning_rate": 1.922698524244554e-05, "loss": 0.0578, "step": 4563 }, { "epoch": 3.207308503162333, "grad_norm": 0.49726685881614685, "learning_rate": 1.9231201686577652e-05, "loss": 0.0902, "step": 4564 }, { "epoch": 3.208011243851019, "grad_norm": 0.46553584933280945, "learning_rate": 1.923541813070977e-05, "loss": 0.1015, "step": 4565 }, { "epoch": 3.208713984539705, "grad_norm": 1.3476389646530151, "learning_rate": 1.9239634574841882e-05, "loss": 0.1318, "step": 4566 }, { "epoch": 3.2094167252283907, "grad_norm": 1.981300711631775, "learning_rate": 1.9243851018974e-05, "loss": 0.2187, "step": 4567 }, { "epoch": 3.2101194659170766, "grad_norm": 1.398253321647644, "learning_rate": 1.9248067463106115e-05, "loss": 0.2343, "step": 4568 }, { "epoch": 3.2108222066057626, "grad_norm": 8.365918159484863, "learning_rate": 1.9252283907238232e-05, "loss": 0.3661, "step": 4569 }, { "epoch": 3.2115249472944485, "grad_norm": 0.37322142720222473, "learning_rate": 1.9256500351370345e-05, "loss": 0.0882, "step": 4570 }, { "epoch": 3.2122276879831344, "grad_norm": 0.16662311553955078, "learning_rate": 1.9260716795502462e-05, "loss": 0.0328, "step": 4571 }, { "epoch": 3.21293042867182, "grad_norm": 0.16112020611763, "learning_rate": 1.9264933239634575e-05, "loss": 0.0279, "step": 4572 }, { "epoch": 3.213633169360506, "grad_norm": 0.26998060941696167, "learning_rate": 1.926914968376669e-05, "loss": 0.0269, "step": 4573 }, { "epoch": 3.2143359100491917, "grad_norm": 0.29662469029426575, "learning_rate": 1.9273366127898805e-05, "loss": 0.0294, "step": 4574 }, { "epoch": 3.2150386507378776, "grad_norm": 0.3350490629673004, "learning_rate": 1.927758257203092e-05, "loss": 0.0194, "step": 4575 }, { "epoch": 3.2157413914265636, "grad_norm": 0.2351471185684204, "learning_rate": 1.9281799016163035e-05, "loss": 0.0337, "step": 4576 }, { "epoch": 3.2164441321152495, "grad_norm": 0.20528174936771393, "learning_rate": 1.928601546029515e-05, "loss": 0.0218, "step": 4577 }, { "epoch": 3.2171468728039354, "grad_norm": 0.29645851254463196, "learning_rate": 1.9290231904427265e-05, "loss": 0.0524, "step": 4578 }, { "epoch": 3.2178496134926213, "grad_norm": 0.37516453862190247, "learning_rate": 1.929444834855938e-05, "loss": 0.0199, "step": 4579 }, { "epoch": 3.2185523541813073, "grad_norm": 0.5742937326431274, "learning_rate": 1.9298664792691495e-05, "loss": 0.0331, "step": 4580 }, { "epoch": 3.219255094869993, "grad_norm": 0.14615756273269653, "learning_rate": 1.9302881236823615e-05, "loss": 0.0154, "step": 4581 }, { "epoch": 3.2199578355586786, "grad_norm": 0.21923699975013733, "learning_rate": 1.9307097680955728e-05, "loss": 0.0307, "step": 4582 }, { "epoch": 3.2206605762473646, "grad_norm": 0.2546778917312622, "learning_rate": 1.9311314125087844e-05, "loss": 0.0322, "step": 4583 }, { "epoch": 3.2213633169360505, "grad_norm": 0.2652933895587921, "learning_rate": 1.9315530569219958e-05, "loss": 0.0498, "step": 4584 }, { "epoch": 3.2220660576247364, "grad_norm": 0.27342918515205383, "learning_rate": 1.9319747013352074e-05, "loss": 0.0527, "step": 4585 }, { "epoch": 3.2227687983134223, "grad_norm": 0.2516772150993347, "learning_rate": 1.9323963457484188e-05, "loss": 0.0357, "step": 4586 }, { "epoch": 3.2234715390021083, "grad_norm": 0.4089854955673218, "learning_rate": 1.9328179901616304e-05, "loss": 0.0516, "step": 4587 }, { "epoch": 3.224174279690794, "grad_norm": 0.2734333574771881, "learning_rate": 1.933239634574842e-05, "loss": 0.0427, "step": 4588 }, { "epoch": 3.22487702037948, "grad_norm": 0.4099595248699188, "learning_rate": 1.9336612789880534e-05, "loss": 0.0756, "step": 4589 }, { "epoch": 3.225579761068166, "grad_norm": 0.7617831230163574, "learning_rate": 1.934082923401265e-05, "loss": 0.1221, "step": 4590 }, { "epoch": 3.2262825017568515, "grad_norm": 0.645359992980957, "learning_rate": 1.9345045678144764e-05, "loss": 0.1563, "step": 4591 }, { "epoch": 3.2269852424455374, "grad_norm": 1.0860811471939087, "learning_rate": 1.934926212227688e-05, "loss": 0.2052, "step": 4592 }, { "epoch": 3.2276879831342233, "grad_norm": 1.4939484596252441, "learning_rate": 1.9353478566408994e-05, "loss": 0.2895, "step": 4593 }, { "epoch": 3.2283907238229093, "grad_norm": 2.0597503185272217, "learning_rate": 1.935769501054111e-05, "loss": 0.2596, "step": 4594 }, { "epoch": 3.229093464511595, "grad_norm": 0.35735949873924255, "learning_rate": 1.9361911454673227e-05, "loss": 0.086, "step": 4595 }, { "epoch": 3.229796205200281, "grad_norm": 0.22792255878448486, "learning_rate": 1.9366127898805344e-05, "loss": 0.0389, "step": 4596 }, { "epoch": 3.230498945888967, "grad_norm": 0.24961943924427032, "learning_rate": 1.9370344342937457e-05, "loss": 0.0379, "step": 4597 }, { "epoch": 3.231201686577653, "grad_norm": 0.15440386533737183, "learning_rate": 1.9374560787069574e-05, "loss": 0.019, "step": 4598 }, { "epoch": 3.231904427266339, "grad_norm": 0.20500150322914124, "learning_rate": 1.9378777231201687e-05, "loss": 0.0271, "step": 4599 }, { "epoch": 3.232607167955025, "grad_norm": 0.1876499205827713, "learning_rate": 1.9382993675333804e-05, "loss": 0.0184, "step": 4600 }, { "epoch": 3.2333099086437107, "grad_norm": 0.23798328638076782, "learning_rate": 1.9387210119465917e-05, "loss": 0.0394, "step": 4601 }, { "epoch": 3.234012649332396, "grad_norm": 0.1667119562625885, "learning_rate": 1.9391426563598033e-05, "loss": 0.0263, "step": 4602 }, { "epoch": 3.234715390021082, "grad_norm": 0.23356720805168152, "learning_rate": 1.9395643007730147e-05, "loss": 0.032, "step": 4603 }, { "epoch": 3.235418130709768, "grad_norm": 0.22254745662212372, "learning_rate": 1.9399859451862263e-05, "loss": 0.0231, "step": 4604 }, { "epoch": 3.236120871398454, "grad_norm": 0.6851292848587036, "learning_rate": 1.9404075895994377e-05, "loss": 0.0455, "step": 4605 }, { "epoch": 3.23682361208714, "grad_norm": 0.2406516820192337, "learning_rate": 1.9408292340126493e-05, "loss": 0.033, "step": 4606 }, { "epoch": 3.237526352775826, "grad_norm": 0.32400310039520264, "learning_rate": 1.9412508784258607e-05, "loss": 0.0366, "step": 4607 }, { "epoch": 3.2382290934645117, "grad_norm": 0.21908558905124664, "learning_rate": 1.9416725228390723e-05, "loss": 0.0294, "step": 4608 }, { "epoch": 3.2389318341531976, "grad_norm": 0.2247781902551651, "learning_rate": 1.942094167252284e-05, "loss": 0.0471, "step": 4609 }, { "epoch": 3.2396345748418836, "grad_norm": 0.25338220596313477, "learning_rate": 1.9425158116654956e-05, "loss": 0.0317, "step": 4610 }, { "epoch": 3.240337315530569, "grad_norm": 0.23431368172168732, "learning_rate": 1.942937456078707e-05, "loss": 0.0372, "step": 4611 }, { "epoch": 3.241040056219255, "grad_norm": 0.344915509223938, "learning_rate": 1.9433591004919186e-05, "loss": 0.0574, "step": 4612 }, { "epoch": 3.241742796907941, "grad_norm": 0.9076451659202576, "learning_rate": 1.94378074490513e-05, "loss": 0.0455, "step": 4613 }, { "epoch": 3.242445537596627, "grad_norm": 0.4626850187778473, "learning_rate": 1.9442023893183416e-05, "loss": 0.0705, "step": 4614 }, { "epoch": 3.2431482782853127, "grad_norm": 0.49799373745918274, "learning_rate": 1.944624033731553e-05, "loss": 0.114, "step": 4615 }, { "epoch": 3.2438510189739986, "grad_norm": 0.6138284802436829, "learning_rate": 1.9450456781447646e-05, "loss": 0.1384, "step": 4616 }, { "epoch": 3.2445537596626846, "grad_norm": 0.6805697679519653, "learning_rate": 1.945467322557976e-05, "loss": 0.2002, "step": 4617 }, { "epoch": 3.2452565003513705, "grad_norm": 1.0175628662109375, "learning_rate": 1.9458889669711876e-05, "loss": 0.2362, "step": 4618 }, { "epoch": 3.2459592410400564, "grad_norm": 1.8538085222244263, "learning_rate": 1.9463106113843993e-05, "loss": 0.3051, "step": 4619 }, { "epoch": 3.2466619817287423, "grad_norm": 0.3088902235031128, "learning_rate": 1.9467322557976106e-05, "loss": 0.0982, "step": 4620 }, { "epoch": 3.247364722417428, "grad_norm": 0.4140017628669739, "learning_rate": 1.9471539002108223e-05, "loss": 0.0589, "step": 4621 }, { "epoch": 3.2480674631061137, "grad_norm": 0.20275840163230896, "learning_rate": 1.947575544624034e-05, "loss": 0.0289, "step": 4622 }, { "epoch": 3.2487702037947996, "grad_norm": 0.8168703317642212, "learning_rate": 1.9479971890372456e-05, "loss": 0.0276, "step": 4623 }, { "epoch": 3.2494729444834856, "grad_norm": 0.21800053119659424, "learning_rate": 1.948418833450457e-05, "loss": 0.0304, "step": 4624 }, { "epoch": 3.2501756851721715, "grad_norm": 0.20498435199260712, "learning_rate": 1.9488404778636686e-05, "loss": 0.0258, "step": 4625 }, { "epoch": 3.2508784258608574, "grad_norm": 0.2770930528640747, "learning_rate": 1.94926212227688e-05, "loss": 0.0287, "step": 4626 }, { "epoch": 3.2515811665495433, "grad_norm": 0.203762024641037, "learning_rate": 1.9496837666900916e-05, "loss": 0.025, "step": 4627 }, { "epoch": 3.2522839072382292, "grad_norm": 0.2184915840625763, "learning_rate": 1.950105411103303e-05, "loss": 0.0284, "step": 4628 }, { "epoch": 3.2529866479269147, "grad_norm": 0.20192360877990723, "learning_rate": 1.9505270555165145e-05, "loss": 0.0218, "step": 4629 }, { "epoch": 3.2536893886156006, "grad_norm": 0.3684536814689636, "learning_rate": 1.950948699929726e-05, "loss": 0.0402, "step": 4630 }, { "epoch": 3.2543921293042866, "grad_norm": 0.2271924912929535, "learning_rate": 1.9513703443429375e-05, "loss": 0.0208, "step": 4631 }, { "epoch": 3.2550948699929725, "grad_norm": 0.24559952318668365, "learning_rate": 1.951791988756149e-05, "loss": 0.0365, "step": 4632 }, { "epoch": 3.2557976106816584, "grad_norm": 0.16306759417057037, "learning_rate": 1.9522136331693605e-05, "loss": 0.0194, "step": 4633 }, { "epoch": 3.2565003513703443, "grad_norm": 0.35676342248916626, "learning_rate": 1.952635277582572e-05, "loss": 0.039, "step": 4634 }, { "epoch": 3.2572030920590302, "grad_norm": 0.3611392676830292, "learning_rate": 1.9530569219957835e-05, "loss": 0.036, "step": 4635 }, { "epoch": 3.257905832747716, "grad_norm": 0.2284347414970398, "learning_rate": 1.9534785664089952e-05, "loss": 0.0346, "step": 4636 }, { "epoch": 3.258608573436402, "grad_norm": 0.26671409606933594, "learning_rate": 1.953900210822207e-05, "loss": 0.0491, "step": 4637 }, { "epoch": 3.259311314125088, "grad_norm": 0.3179350793361664, "learning_rate": 1.954321855235418e-05, "loss": 0.0885, "step": 4638 }, { "epoch": 3.260014054813774, "grad_norm": 0.28863972425460815, "learning_rate": 1.9547434996486298e-05, "loss": 0.0857, "step": 4639 }, { "epoch": 3.2607167955024594, "grad_norm": 0.7698166370391846, "learning_rate": 1.955165144061841e-05, "loss": 0.1218, "step": 4640 }, { "epoch": 3.2614195361911453, "grad_norm": 0.566651463508606, "learning_rate": 1.9555867884750528e-05, "loss": 0.1669, "step": 4641 }, { "epoch": 3.2621222768798313, "grad_norm": 0.6219834685325623, "learning_rate": 1.956008432888264e-05, "loss": 0.2082, "step": 4642 }, { "epoch": 3.262825017568517, "grad_norm": 0.9049077033996582, "learning_rate": 1.9564300773014758e-05, "loss": 0.3161, "step": 4643 }, { "epoch": 3.263527758257203, "grad_norm": 6.139340400695801, "learning_rate": 1.956851721714687e-05, "loss": 0.315, "step": 4644 }, { "epoch": 3.264230498945889, "grad_norm": 0.3517777919769287, "learning_rate": 1.9572733661278988e-05, "loss": 0.0905, "step": 4645 }, { "epoch": 3.264933239634575, "grad_norm": 0.3289374113082886, "learning_rate": 1.95769501054111e-05, "loss": 0.0522, "step": 4646 }, { "epoch": 3.265635980323261, "grad_norm": 0.2615046799182892, "learning_rate": 1.9581166549543218e-05, "loss": 0.0258, "step": 4647 }, { "epoch": 3.2663387210119468, "grad_norm": 0.25187772512435913, "learning_rate": 1.9585382993675334e-05, "loss": 0.0442, "step": 4648 }, { "epoch": 3.2670414617006323, "grad_norm": 0.15528257191181183, "learning_rate": 1.958959943780745e-05, "loss": 0.0315, "step": 4649 }, { "epoch": 3.267744202389318, "grad_norm": 0.3679131269454956, "learning_rate": 1.9593815881939568e-05, "loss": 0.027, "step": 4650 }, { "epoch": 3.268446943078004, "grad_norm": 2.475911855697632, "learning_rate": 1.959803232607168e-05, "loss": 0.0337, "step": 4651 }, { "epoch": 3.26914968376669, "grad_norm": 0.33570805191993713, "learning_rate": 1.9602248770203798e-05, "loss": 0.028, "step": 4652 }, { "epoch": 3.269852424455376, "grad_norm": 0.2528342604637146, "learning_rate": 1.960646521433591e-05, "loss": 0.0278, "step": 4653 }, { "epoch": 3.270555165144062, "grad_norm": 0.18190115690231323, "learning_rate": 1.9610681658468028e-05, "loss": 0.0233, "step": 4654 }, { "epoch": 3.271257905832748, "grad_norm": 0.1737910658121109, "learning_rate": 1.961489810260014e-05, "loss": 0.0221, "step": 4655 }, { "epoch": 3.2719606465214337, "grad_norm": 0.23928546905517578, "learning_rate": 1.9619114546732257e-05, "loss": 0.0215, "step": 4656 }, { "epoch": 3.2726633872101196, "grad_norm": 0.43114736676216125, "learning_rate": 1.962333099086437e-05, "loss": 0.039, "step": 4657 }, { "epoch": 3.2733661278988055, "grad_norm": 0.36232465505599976, "learning_rate": 1.9627547434996487e-05, "loss": 0.0173, "step": 4658 }, { "epoch": 3.2740688685874915, "grad_norm": 0.3506028652191162, "learning_rate": 1.96317638791286e-05, "loss": 0.0388, "step": 4659 }, { "epoch": 3.274771609276177, "grad_norm": 0.2866400182247162, "learning_rate": 1.9635980323260717e-05, "loss": 0.0518, "step": 4660 }, { "epoch": 3.275474349964863, "grad_norm": 0.3240189850330353, "learning_rate": 1.964019676739283e-05, "loss": 0.0343, "step": 4661 }, { "epoch": 3.276177090653549, "grad_norm": 0.2917158901691437, "learning_rate": 1.9644413211524947e-05, "loss": 0.0495, "step": 4662 }, { "epoch": 3.2768798313422347, "grad_norm": 0.32664915919303894, "learning_rate": 1.9648629655657064e-05, "loss": 0.0507, "step": 4663 }, { "epoch": 3.2775825720309206, "grad_norm": 0.33747997879981995, "learning_rate": 1.965284609978918e-05, "loss": 0.05, "step": 4664 }, { "epoch": 3.2782853127196065, "grad_norm": 0.8199382424354553, "learning_rate": 1.9657062543921294e-05, "loss": 0.1328, "step": 4665 }, { "epoch": 3.2789880534082925, "grad_norm": 0.578426718711853, "learning_rate": 1.966127898805341e-05, "loss": 0.128, "step": 4666 }, { "epoch": 3.2796907940969784, "grad_norm": 0.6332907676696777, "learning_rate": 1.9665495432185523e-05, "loss": 0.1859, "step": 4667 }, { "epoch": 3.280393534785664, "grad_norm": 4.149335861206055, "learning_rate": 1.966971187631764e-05, "loss": 0.2412, "step": 4668 }, { "epoch": 3.28109627547435, "grad_norm": 1.7737078666687012, "learning_rate": 1.9673928320449753e-05, "loss": 0.3208, "step": 4669 }, { "epoch": 3.2817990161630357, "grad_norm": 0.4870518147945404, "learning_rate": 1.967814476458187e-05, "loss": 0.0728, "step": 4670 }, { "epoch": 3.2825017568517216, "grad_norm": 0.20232714712619781, "learning_rate": 1.9682361208713983e-05, "loss": 0.0337, "step": 4671 }, { "epoch": 3.2832044975404076, "grad_norm": 0.1838066577911377, "learning_rate": 1.96865776528461e-05, "loss": 0.0257, "step": 4672 }, { "epoch": 3.2839072382290935, "grad_norm": 0.18879574537277222, "learning_rate": 1.9690794096978213e-05, "loss": 0.0288, "step": 4673 }, { "epoch": 3.2846099789177794, "grad_norm": 0.23742637038230896, "learning_rate": 1.969501054111033e-05, "loss": 0.0263, "step": 4674 }, { "epoch": 3.2853127196064653, "grad_norm": 0.15260688960552216, "learning_rate": 1.9699226985242443e-05, "loss": 0.014, "step": 4675 }, { "epoch": 3.2860154602951512, "grad_norm": 0.3929078280925751, "learning_rate": 1.970344342937456e-05, "loss": 0.0516, "step": 4676 }, { "epoch": 3.286718200983837, "grad_norm": 0.1997491419315338, "learning_rate": 1.970765987350668e-05, "loss": 0.0318, "step": 4677 }, { "epoch": 3.287420941672523, "grad_norm": 0.27641433477401733, "learning_rate": 1.9711876317638793e-05, "loss": 0.049, "step": 4678 }, { "epoch": 3.2881236823612086, "grad_norm": 0.43815848231315613, "learning_rate": 1.971609276177091e-05, "loss": 0.0165, "step": 4679 }, { "epoch": 3.2888264230498945, "grad_norm": 0.29391250014305115, "learning_rate": 1.9720309205903023e-05, "loss": 0.0301, "step": 4680 }, { "epoch": 3.2895291637385804, "grad_norm": 0.21864762902259827, "learning_rate": 1.972452565003514e-05, "loss": 0.015, "step": 4681 }, { "epoch": 3.2902319044272663, "grad_norm": 0.2124241441488266, "learning_rate": 1.9728742094167253e-05, "loss": 0.044, "step": 4682 }, { "epoch": 3.2909346451159522, "grad_norm": 0.1961028277873993, "learning_rate": 1.973295853829937e-05, "loss": 0.0199, "step": 4683 }, { "epoch": 3.291637385804638, "grad_norm": 0.3154448866844177, "learning_rate": 1.9737174982431483e-05, "loss": 0.0517, "step": 4684 }, { "epoch": 3.292340126493324, "grad_norm": 0.31062155961990356, "learning_rate": 1.97413914265636e-05, "loss": 0.0679, "step": 4685 }, { "epoch": 3.29304286718201, "grad_norm": 0.3488081991672516, "learning_rate": 1.9745607870695712e-05, "loss": 0.0318, "step": 4686 }, { "epoch": 3.2937456078706955, "grad_norm": 0.262979120016098, "learning_rate": 1.974982431482783e-05, "loss": 0.0503, "step": 4687 }, { "epoch": 3.2944483485593814, "grad_norm": 0.32411274313926697, "learning_rate": 1.9754040758959942e-05, "loss": 0.0456, "step": 4688 }, { "epoch": 3.2951510892480673, "grad_norm": 0.32645416259765625, "learning_rate": 1.975825720309206e-05, "loss": 0.071, "step": 4689 }, { "epoch": 3.2958538299367532, "grad_norm": 0.495504766702652, "learning_rate": 1.9762473647224176e-05, "loss": 0.1194, "step": 4690 }, { "epoch": 3.296556570625439, "grad_norm": 0.9984771013259888, "learning_rate": 1.9766690091356292e-05, "loss": 0.1631, "step": 4691 }, { "epoch": 3.297259311314125, "grad_norm": 0.7217867374420166, "learning_rate": 1.9770906535488406e-05, "loss": 0.1798, "step": 4692 }, { "epoch": 3.297962052002811, "grad_norm": 0.9082524180412292, "learning_rate": 1.9775122979620522e-05, "loss": 0.2553, "step": 4693 }, { "epoch": 3.298664792691497, "grad_norm": 1.691991925239563, "learning_rate": 1.9779339423752635e-05, "loss": 0.3477, "step": 4694 }, { "epoch": 3.299367533380183, "grad_norm": 0.42545586824417114, "learning_rate": 1.9783555867884752e-05, "loss": 0.1316, "step": 4695 }, { "epoch": 3.3000702740688688, "grad_norm": 0.20087802410125732, "learning_rate": 1.9787772312016865e-05, "loss": 0.0478, "step": 4696 }, { "epoch": 3.3007730147575547, "grad_norm": 0.2088746428489685, "learning_rate": 1.9791988756148982e-05, "loss": 0.0356, "step": 4697 }, { "epoch": 3.30147575544624, "grad_norm": 0.18445949256420135, "learning_rate": 1.9796205200281095e-05, "loss": 0.0297, "step": 4698 }, { "epoch": 3.302178496134926, "grad_norm": 0.2341349571943283, "learning_rate": 1.9800421644413212e-05, "loss": 0.0363, "step": 4699 }, { "epoch": 3.302881236823612, "grad_norm": 0.17785540223121643, "learning_rate": 1.9804638088545325e-05, "loss": 0.0158, "step": 4700 }, { "epoch": 3.303583977512298, "grad_norm": 0.18100330233573914, "learning_rate": 1.9808854532677442e-05, "loss": 0.023, "step": 4701 }, { "epoch": 3.304286718200984, "grad_norm": 0.21393078565597534, "learning_rate": 1.9813070976809555e-05, "loss": 0.0341, "step": 4702 }, { "epoch": 3.3049894588896698, "grad_norm": 0.1662742793560028, "learning_rate": 1.981728742094167e-05, "loss": 0.0258, "step": 4703 }, { "epoch": 3.3056921995783557, "grad_norm": 0.19625510275363922, "learning_rate": 1.9821503865073788e-05, "loss": 0.0233, "step": 4704 }, { "epoch": 3.3063949402670416, "grad_norm": 0.2077314555644989, "learning_rate": 1.9825720309205905e-05, "loss": 0.0462, "step": 4705 }, { "epoch": 3.307097680955727, "grad_norm": 0.23742662370204926, "learning_rate": 1.9829936753338018e-05, "loss": 0.0308, "step": 4706 }, { "epoch": 3.307800421644413, "grad_norm": 0.2960609495639801, "learning_rate": 1.9834153197470135e-05, "loss": 0.0615, "step": 4707 }, { "epoch": 3.308503162333099, "grad_norm": 0.2904791831970215, "learning_rate": 1.983836964160225e-05, "loss": 0.026, "step": 4708 }, { "epoch": 3.309205903021785, "grad_norm": 0.6755877733230591, "learning_rate": 1.9842586085734365e-05, "loss": 0.0666, "step": 4709 }, { "epoch": 3.3099086437104708, "grad_norm": 0.27602967619895935, "learning_rate": 1.984680252986648e-05, "loss": 0.048, "step": 4710 }, { "epoch": 3.3106113843991567, "grad_norm": 0.20714536309242249, "learning_rate": 1.9851018973998595e-05, "loss": 0.0266, "step": 4711 }, { "epoch": 3.3113141250878426, "grad_norm": 0.48232972621917725, "learning_rate": 1.985523541813071e-05, "loss": 0.05, "step": 4712 }, { "epoch": 3.3120168657765285, "grad_norm": 0.34546589851379395, "learning_rate": 1.9859451862262824e-05, "loss": 0.0559, "step": 4713 }, { "epoch": 3.3127196064652145, "grad_norm": 0.7465872764587402, "learning_rate": 1.986366830639494e-05, "loss": 0.067, "step": 4714 }, { "epoch": 3.3134223471539004, "grad_norm": 0.5436322689056396, "learning_rate": 1.9867884750527054e-05, "loss": 0.1217, "step": 4715 }, { "epoch": 3.3141250878425863, "grad_norm": 0.5306801199913025, "learning_rate": 1.987210119465917e-05, "loss": 0.1468, "step": 4716 }, { "epoch": 3.314827828531272, "grad_norm": 0.846674919128418, "learning_rate": 1.9876317638791288e-05, "loss": 0.2218, "step": 4717 }, { "epoch": 3.3155305692199577, "grad_norm": 0.967602550983429, "learning_rate": 1.9880534082923404e-05, "loss": 0.2352, "step": 4718 }, { "epoch": 3.3162333099086436, "grad_norm": 1.3746294975280762, "learning_rate": 1.9884750527055517e-05, "loss": 0.3631, "step": 4719 }, { "epoch": 3.3169360505973295, "grad_norm": 0.28574520349502563, "learning_rate": 1.9888966971187634e-05, "loss": 0.0785, "step": 4720 }, { "epoch": 3.3176387912860155, "grad_norm": 0.33903396129608154, "learning_rate": 1.9893183415319747e-05, "loss": 0.024, "step": 4721 }, { "epoch": 3.3183415319747014, "grad_norm": 0.28768786787986755, "learning_rate": 1.9897399859451864e-05, "loss": 0.037, "step": 4722 }, { "epoch": 3.3190442726633873, "grad_norm": 0.19145867228507996, "learning_rate": 1.9901616303583977e-05, "loss": 0.0283, "step": 4723 }, { "epoch": 3.3197470133520732, "grad_norm": 0.3023488223552704, "learning_rate": 1.9905832747716094e-05, "loss": 0.0429, "step": 4724 }, { "epoch": 3.320449754040759, "grad_norm": 0.18776047229766846, "learning_rate": 1.9910049191848207e-05, "loss": 0.0193, "step": 4725 }, { "epoch": 3.3211524947294446, "grad_norm": 0.20203113555908203, "learning_rate": 1.9914265635980324e-05, "loss": 0.0208, "step": 4726 }, { "epoch": 3.3218552354181305, "grad_norm": 0.17342695593833923, "learning_rate": 1.9918482080112437e-05, "loss": 0.0199, "step": 4727 }, { "epoch": 3.3225579761068165, "grad_norm": 0.19052597880363464, "learning_rate": 1.9922698524244554e-05, "loss": 0.0224, "step": 4728 }, { "epoch": 3.3232607167955024, "grad_norm": 0.15983638167381287, "learning_rate": 1.9926914968376667e-05, "loss": 0.0169, "step": 4729 }, { "epoch": 3.3239634574841883, "grad_norm": 0.24170362949371338, "learning_rate": 1.9931131412508784e-05, "loss": 0.0373, "step": 4730 }, { "epoch": 3.3246661981728742, "grad_norm": 0.201654314994812, "learning_rate": 1.99353478566409e-05, "loss": 0.0262, "step": 4731 }, { "epoch": 3.32536893886156, "grad_norm": 0.2790301442146301, "learning_rate": 1.9939564300773017e-05, "loss": 0.0577, "step": 4732 }, { "epoch": 3.326071679550246, "grad_norm": 0.22234155237674713, "learning_rate": 1.994378074490513e-05, "loss": 0.025, "step": 4733 }, { "epoch": 3.326774420238932, "grad_norm": 0.36039453744888306, "learning_rate": 1.9947997189037247e-05, "loss": 0.042, "step": 4734 }, { "epoch": 3.327477160927618, "grad_norm": 0.5423803925514221, "learning_rate": 1.995221363316936e-05, "loss": 0.0669, "step": 4735 }, { "epoch": 3.3281799016163034, "grad_norm": 0.3158084452152252, "learning_rate": 1.9956430077301477e-05, "loss": 0.0432, "step": 4736 }, { "epoch": 3.3288826423049893, "grad_norm": 0.5156298279762268, "learning_rate": 1.9960646521433593e-05, "loss": 0.0761, "step": 4737 }, { "epoch": 3.3295853829936752, "grad_norm": 0.39989954233169556, "learning_rate": 1.9964862965565706e-05, "loss": 0.0604, "step": 4738 }, { "epoch": 3.330288123682361, "grad_norm": 0.4342139959335327, "learning_rate": 1.9969079409697823e-05, "loss": 0.0794, "step": 4739 }, { "epoch": 3.330990864371047, "grad_norm": 0.7243716716766357, "learning_rate": 1.9973295853829936e-05, "loss": 0.1331, "step": 4740 }, { "epoch": 3.331693605059733, "grad_norm": 0.7433497309684753, "learning_rate": 1.9977512297962053e-05, "loss": 0.1705, "step": 4741 }, { "epoch": 3.332396345748419, "grad_norm": 1.5507901906967163, "learning_rate": 1.9981728742094166e-05, "loss": 0.2562, "step": 4742 }, { "epoch": 3.333099086437105, "grad_norm": 1.880271315574646, "learning_rate": 1.9985945186226283e-05, "loss": 0.2663, "step": 4743 }, { "epoch": 3.3338018271257908, "grad_norm": 1.4806931018829346, "learning_rate": 1.9990161630358396e-05, "loss": 0.3001, "step": 4744 }, { "epoch": 3.3345045678144762, "grad_norm": 0.34717440605163574, "learning_rate": 1.9994378074490516e-05, "loss": 0.1148, "step": 4745 }, { "epoch": 3.335207308503162, "grad_norm": 0.2759299874305725, "learning_rate": 1.999859451862263e-05, "loss": 0.0521, "step": 4746 }, { "epoch": 3.335910049191848, "grad_norm": 0.21186023950576782, "learning_rate": 2.0002810962754746e-05, "loss": 0.0343, "step": 4747 }, { "epoch": 3.336612789880534, "grad_norm": 0.18263404071331024, "learning_rate": 2.000702740688686e-05, "loss": 0.0263, "step": 4748 }, { "epoch": 3.33731553056922, "grad_norm": 0.15306612849235535, "learning_rate": 2.0011243851018976e-05, "loss": 0.0213, "step": 4749 }, { "epoch": 3.338018271257906, "grad_norm": 0.3308313488960266, "learning_rate": 2.001546029515109e-05, "loss": 0.0214, "step": 4750 }, { "epoch": 3.3387210119465918, "grad_norm": 0.23058737814426422, "learning_rate": 2.0019676739283206e-05, "loss": 0.0363, "step": 4751 }, { "epoch": 3.3394237526352777, "grad_norm": 0.1841321736574173, "learning_rate": 2.002389318341532e-05, "loss": 0.0247, "step": 4752 }, { "epoch": 3.3401264933239636, "grad_norm": 0.22611230611801147, "learning_rate": 2.0028109627547436e-05, "loss": 0.0395, "step": 4753 }, { "epoch": 3.3408292340126495, "grad_norm": 0.34592846035957336, "learning_rate": 2.003232607167955e-05, "loss": 0.0292, "step": 4754 }, { "epoch": 3.3415319747013355, "grad_norm": 0.2881113588809967, "learning_rate": 2.0036542515811666e-05, "loss": 0.0384, "step": 4755 }, { "epoch": 3.342234715390021, "grad_norm": 0.2333046942949295, "learning_rate": 2.004075895994378e-05, "loss": 0.0337, "step": 4756 }, { "epoch": 3.342937456078707, "grad_norm": 0.20958644151687622, "learning_rate": 2.0044975404075895e-05, "loss": 0.0467, "step": 4757 }, { "epoch": 3.3436401967673928, "grad_norm": 0.23977874219417572, "learning_rate": 2.0049191848208012e-05, "loss": 0.0405, "step": 4758 }, { "epoch": 3.3443429374560787, "grad_norm": 0.293446809053421, "learning_rate": 2.005340829234013e-05, "loss": 0.0481, "step": 4759 }, { "epoch": 3.3450456781447646, "grad_norm": 0.2723807394504547, "learning_rate": 2.0057624736472242e-05, "loss": 0.0446, "step": 4760 }, { "epoch": 3.3457484188334505, "grad_norm": 0.38203194737434387, "learning_rate": 2.006184118060436e-05, "loss": 0.0465, "step": 4761 }, { "epoch": 3.3464511595221365, "grad_norm": 0.2245408594608307, "learning_rate": 2.0066057624736472e-05, "loss": 0.0343, "step": 4762 }, { "epoch": 3.3471539002108224, "grad_norm": 0.47497522830963135, "learning_rate": 2.007027406886859e-05, "loss": 0.0976, "step": 4763 }, { "epoch": 3.347856640899508, "grad_norm": 0.6546820402145386, "learning_rate": 2.0074490513000702e-05, "loss": 0.0707, "step": 4764 }, { "epoch": 3.3485593815881938, "grad_norm": 0.45942962169647217, "learning_rate": 2.007870695713282e-05, "loss": 0.0789, "step": 4765 }, { "epoch": 3.3492621222768797, "grad_norm": 1.3191083669662476, "learning_rate": 2.008292340126493e-05, "loss": 0.166, "step": 4766 }, { "epoch": 3.3499648629655656, "grad_norm": 0.7566773295402527, "learning_rate": 2.008713984539705e-05, "loss": 0.2048, "step": 4767 }, { "epoch": 3.3506676036542515, "grad_norm": 1.3044896125793457, "learning_rate": 2.0091356289529165e-05, "loss": 0.2565, "step": 4768 }, { "epoch": 3.3513703443429375, "grad_norm": 1.716511607170105, "learning_rate": 2.0095572733661278e-05, "loss": 0.3026, "step": 4769 }, { "epoch": 3.3520730850316234, "grad_norm": 0.23059386014938354, "learning_rate": 2.0099789177793395e-05, "loss": 0.0818, "step": 4770 }, { "epoch": 3.3527758257203093, "grad_norm": 0.33371517062187195, "learning_rate": 2.0104005621925508e-05, "loss": 0.0246, "step": 4771 }, { "epoch": 3.353478566408995, "grad_norm": 0.775212287902832, "learning_rate": 2.0108222066057628e-05, "loss": 0.0444, "step": 4772 }, { "epoch": 3.354181307097681, "grad_norm": 0.29019254446029663, "learning_rate": 2.011243851018974e-05, "loss": 0.0283, "step": 4773 }, { "epoch": 3.354884047786367, "grad_norm": 0.2004019320011139, "learning_rate": 2.0116654954321858e-05, "loss": 0.0315, "step": 4774 }, { "epoch": 3.3555867884750525, "grad_norm": 0.20569555461406708, "learning_rate": 2.012087139845397e-05, "loss": 0.0203, "step": 4775 }, { "epoch": 3.3562895291637385, "grad_norm": 0.1912693828344345, "learning_rate": 2.0125087842586088e-05, "loss": 0.0262, "step": 4776 }, { "epoch": 3.3569922698524244, "grad_norm": 0.1856427788734436, "learning_rate": 2.01293042867182e-05, "loss": 0.0322, "step": 4777 }, { "epoch": 3.3576950105411103, "grad_norm": 0.3977762758731842, "learning_rate": 2.0133520730850318e-05, "loss": 0.0287, "step": 4778 }, { "epoch": 3.3583977512297962, "grad_norm": 0.18589860200881958, "learning_rate": 2.013773717498243e-05, "loss": 0.0215, "step": 4779 }, { "epoch": 3.359100491918482, "grad_norm": 0.23974670469760895, "learning_rate": 2.0141953619114548e-05, "loss": 0.0323, "step": 4780 }, { "epoch": 3.359803232607168, "grad_norm": 0.19573870301246643, "learning_rate": 2.014617006324666e-05, "loss": 0.0195, "step": 4781 }, { "epoch": 3.360505973295854, "grad_norm": 0.26694315671920776, "learning_rate": 2.0150386507378778e-05, "loss": 0.0247, "step": 4782 }, { "epoch": 3.3612087139845395, "grad_norm": 0.24678918719291687, "learning_rate": 2.015460295151089e-05, "loss": 0.0288, "step": 4783 }, { "epoch": 3.3619114546732254, "grad_norm": 0.24840615689754486, "learning_rate": 2.0158819395643007e-05, "loss": 0.0457, "step": 4784 }, { "epoch": 3.3626141953619113, "grad_norm": 0.28224462270736694, "learning_rate": 2.0163035839775124e-05, "loss": 0.0529, "step": 4785 }, { "epoch": 3.3633169360505972, "grad_norm": 0.1952710747718811, "learning_rate": 2.016725228390724e-05, "loss": 0.031, "step": 4786 }, { "epoch": 3.364019676739283, "grad_norm": 0.25915923714637756, "learning_rate": 2.0171468728039354e-05, "loss": 0.06, "step": 4787 }, { "epoch": 3.364722417427969, "grad_norm": 0.35592979192733765, "learning_rate": 2.017568517217147e-05, "loss": 0.0687, "step": 4788 }, { "epoch": 3.365425158116655, "grad_norm": 0.5373016595840454, "learning_rate": 2.0179901616303584e-05, "loss": 0.0687, "step": 4789 }, { "epoch": 3.366127898805341, "grad_norm": 0.34356293082237244, "learning_rate": 2.01841180604357e-05, "loss": 0.0907, "step": 4790 }, { "epoch": 3.366830639494027, "grad_norm": 0.5949551463127136, "learning_rate": 2.0188334504567814e-05, "loss": 0.1365, "step": 4791 }, { "epoch": 3.3675333801827128, "grad_norm": 0.8699778318405151, "learning_rate": 2.019255094869993e-05, "loss": 0.1755, "step": 4792 }, { "epoch": 3.3682361208713987, "grad_norm": 1.192894458770752, "learning_rate": 2.0196767392832044e-05, "loss": 0.2431, "step": 4793 }, { "epoch": 3.368938861560084, "grad_norm": 1.6536880731582642, "learning_rate": 2.020098383696416e-05, "loss": 0.3437, "step": 4794 }, { "epoch": 3.36964160224877, "grad_norm": 0.2467363476753235, "learning_rate": 2.0205200281096274e-05, "loss": 0.0827, "step": 4795 }, { "epoch": 3.370344342937456, "grad_norm": 0.23784108459949493, "learning_rate": 2.020941672522839e-05, "loss": 0.0383, "step": 4796 }, { "epoch": 3.371047083626142, "grad_norm": 0.16574151813983917, "learning_rate": 2.0213633169360507e-05, "loss": 0.0216, "step": 4797 }, { "epoch": 3.371749824314828, "grad_norm": 0.22498922049999237, "learning_rate": 2.021784961349262e-05, "loss": 0.034, "step": 4798 }, { "epoch": 3.3724525650035138, "grad_norm": 0.2460751086473465, "learning_rate": 2.022206605762474e-05, "loss": 0.0325, "step": 4799 }, { "epoch": 3.3731553056921997, "grad_norm": 0.27596843242645264, "learning_rate": 2.0226282501756853e-05, "loss": 0.0238, "step": 4800 }, { "epoch": 3.3738580463808856, "grad_norm": 0.7845429182052612, "learning_rate": 2.023049894588897e-05, "loss": 0.0295, "step": 4801 }, { "epoch": 3.3745607870695715, "grad_norm": 0.19281522929668427, "learning_rate": 2.0234715390021083e-05, "loss": 0.0281, "step": 4802 }, { "epoch": 3.375263527758257, "grad_norm": 0.22508175671100616, "learning_rate": 2.02389318341532e-05, "loss": 0.0324, "step": 4803 }, { "epoch": 3.375966268446943, "grad_norm": 0.18086552619934082, "learning_rate": 2.0243148278285313e-05, "loss": 0.0243, "step": 4804 }, { "epoch": 3.376669009135629, "grad_norm": 0.3426417410373688, "learning_rate": 2.024736472241743e-05, "loss": 0.0432, "step": 4805 }, { "epoch": 3.3773717498243148, "grad_norm": 0.18461699783802032, "learning_rate": 2.0251581166549543e-05, "loss": 0.0289, "step": 4806 }, { "epoch": 3.3780744905130007, "grad_norm": 0.2592089772224426, "learning_rate": 2.025579761068166e-05, "loss": 0.0413, "step": 4807 }, { "epoch": 3.3787772312016866, "grad_norm": 0.19518619775772095, "learning_rate": 2.0260014054813773e-05, "loss": 0.0289, "step": 4808 }, { "epoch": 3.3794799718903725, "grad_norm": 0.24238628149032593, "learning_rate": 2.026423049894589e-05, "loss": 0.0383, "step": 4809 }, { "epoch": 3.3801827125790584, "grad_norm": 0.25423189997673035, "learning_rate": 2.0268446943078003e-05, "loss": 0.0379, "step": 4810 }, { "epoch": 3.3808854532677444, "grad_norm": 0.3204158842563629, "learning_rate": 2.027266338721012e-05, "loss": 0.0312, "step": 4811 }, { "epoch": 3.3815881939564303, "grad_norm": 0.44576942920684814, "learning_rate": 2.0276879831342233e-05, "loss": 0.0636, "step": 4812 }, { "epoch": 3.3822909346451158, "grad_norm": 0.25132936239242554, "learning_rate": 2.0281096275474353e-05, "loss": 0.0483, "step": 4813 }, { "epoch": 3.3829936753338017, "grad_norm": 0.3517901599407196, "learning_rate": 2.0285312719606466e-05, "loss": 0.0793, "step": 4814 }, { "epoch": 3.3836964160224876, "grad_norm": 0.6265081167221069, "learning_rate": 2.0289529163738583e-05, "loss": 0.1082, "step": 4815 }, { "epoch": 3.3843991567111735, "grad_norm": 0.6331154704093933, "learning_rate": 2.0293745607870696e-05, "loss": 0.1463, "step": 4816 }, { "epoch": 3.3851018973998594, "grad_norm": 1.4835642576217651, "learning_rate": 2.0297962052002812e-05, "loss": 0.1901, "step": 4817 }, { "epoch": 3.3858046380885454, "grad_norm": 0.927204966545105, "learning_rate": 2.0302178496134926e-05, "loss": 0.2544, "step": 4818 }, { "epoch": 3.3865073787772313, "grad_norm": 1.7409626245498657, "learning_rate": 2.0306394940267042e-05, "loss": 0.3115, "step": 4819 }, { "epoch": 3.387210119465917, "grad_norm": 0.4854714572429657, "learning_rate": 2.0310611384399156e-05, "loss": 0.0847, "step": 4820 }, { "epoch": 3.387912860154603, "grad_norm": 0.22702980041503906, "learning_rate": 2.0314827828531272e-05, "loss": 0.0309, "step": 4821 }, { "epoch": 3.3886156008432886, "grad_norm": 0.259718656539917, "learning_rate": 2.0319044272663385e-05, "loss": 0.0307, "step": 4822 }, { "epoch": 3.3893183415319745, "grad_norm": 0.23275603353977203, "learning_rate": 2.0323260716795502e-05, "loss": 0.0376, "step": 4823 }, { "epoch": 3.3900210822206605, "grad_norm": 0.3569774329662323, "learning_rate": 2.0327477160927615e-05, "loss": 0.0322, "step": 4824 }, { "epoch": 3.3907238229093464, "grad_norm": 0.21617810428142548, "learning_rate": 2.0331693605059732e-05, "loss": 0.024, "step": 4825 }, { "epoch": 3.3914265635980323, "grad_norm": 0.24824799597263336, "learning_rate": 2.0335910049191852e-05, "loss": 0.0263, "step": 4826 }, { "epoch": 3.392129304286718, "grad_norm": 0.2229757159948349, "learning_rate": 2.0340126493323965e-05, "loss": 0.0439, "step": 4827 }, { "epoch": 3.392832044975404, "grad_norm": 0.38400745391845703, "learning_rate": 2.0344342937456082e-05, "loss": 0.0326, "step": 4828 }, { "epoch": 3.39353478566409, "grad_norm": 0.15150058269500732, "learning_rate": 2.0348559381588195e-05, "loss": 0.0205, "step": 4829 }, { "epoch": 3.394237526352776, "grad_norm": 0.26946040987968445, "learning_rate": 2.0352775825720312e-05, "loss": 0.0168, "step": 4830 }, { "epoch": 3.394940267041462, "grad_norm": 0.15476752817630768, "learning_rate": 2.0356992269852425e-05, "loss": 0.0155, "step": 4831 }, { "epoch": 3.395643007730148, "grad_norm": 0.250354140996933, "learning_rate": 2.036120871398454e-05, "loss": 0.0401, "step": 4832 }, { "epoch": 3.3963457484188333, "grad_norm": 0.27862322330474854, "learning_rate": 2.0365425158116655e-05, "loss": 0.0394, "step": 4833 }, { "epoch": 3.397048489107519, "grad_norm": 0.35009709000587463, "learning_rate": 2.036964160224877e-05, "loss": 0.0254, "step": 4834 }, { "epoch": 3.397751229796205, "grad_norm": 0.2502034604549408, "learning_rate": 2.0373858046380885e-05, "loss": 0.0511, "step": 4835 }, { "epoch": 3.398453970484891, "grad_norm": 0.3447626233100891, "learning_rate": 2.0378074490513e-05, "loss": 0.0478, "step": 4836 }, { "epoch": 3.399156711173577, "grad_norm": 0.32192766666412354, "learning_rate": 2.0382290934645115e-05, "loss": 0.0602, "step": 4837 }, { "epoch": 3.399859451862263, "grad_norm": 0.29316839575767517, "learning_rate": 2.038650737877723e-05, "loss": 0.0536, "step": 4838 }, { "epoch": 3.400562192550949, "grad_norm": 0.3854798674583435, "learning_rate": 2.0390723822909345e-05, "loss": 0.065, "step": 4839 }, { "epoch": 3.4012649332396347, "grad_norm": 0.3583129048347473, "learning_rate": 2.0394940267041465e-05, "loss": 0.0877, "step": 4840 }, { "epoch": 3.4019676739283202, "grad_norm": 0.7570658922195435, "learning_rate": 2.0399156711173578e-05, "loss": 0.1461, "step": 4841 }, { "epoch": 3.402670414617006, "grad_norm": 0.7400066256523132, "learning_rate": 2.0403373155305695e-05, "loss": 0.154, "step": 4842 }, { "epoch": 3.403373155305692, "grad_norm": 1.1203707456588745, "learning_rate": 2.0407589599437808e-05, "loss": 0.2974, "step": 4843 }, { "epoch": 3.404075895994378, "grad_norm": 1.4860661029815674, "learning_rate": 2.0411806043569924e-05, "loss": 0.3235, "step": 4844 }, { "epoch": 3.404778636683064, "grad_norm": 0.3233364224433899, "learning_rate": 2.0416022487702038e-05, "loss": 0.0786, "step": 4845 }, { "epoch": 3.40548137737175, "grad_norm": 0.19585683941841125, "learning_rate": 2.0420238931834154e-05, "loss": 0.0376, "step": 4846 }, { "epoch": 3.4061841180604358, "grad_norm": 0.18975670635700226, "learning_rate": 2.0424455375966268e-05, "loss": 0.0312, "step": 4847 }, { "epoch": 3.4068868587491217, "grad_norm": 0.9302873015403748, "learning_rate": 2.0428671820098384e-05, "loss": 0.0268, "step": 4848 }, { "epoch": 3.4075895994378076, "grad_norm": 0.22225835919380188, "learning_rate": 2.0432888264230497e-05, "loss": 0.0406, "step": 4849 }, { "epoch": 3.4082923401264935, "grad_norm": 0.22022423148155212, "learning_rate": 2.0437104708362614e-05, "loss": 0.025, "step": 4850 }, { "epoch": 3.4089950808151794, "grad_norm": 0.1879122108221054, "learning_rate": 2.0441321152494727e-05, "loss": 0.0246, "step": 4851 }, { "epoch": 3.409697821503865, "grad_norm": 0.5869065523147583, "learning_rate": 2.0445537596626844e-05, "loss": 0.0352, "step": 4852 }, { "epoch": 3.410400562192551, "grad_norm": 0.28227901458740234, "learning_rate": 2.044975404075896e-05, "loss": 0.0383, "step": 4853 }, { "epoch": 3.4111033028812368, "grad_norm": 0.22075772285461426, "learning_rate": 2.0453970484891077e-05, "loss": 0.0283, "step": 4854 }, { "epoch": 3.4118060435699227, "grad_norm": 0.19792573153972626, "learning_rate": 2.045818692902319e-05, "loss": 0.0346, "step": 4855 }, { "epoch": 3.4125087842586086, "grad_norm": 0.1499890238046646, "learning_rate": 2.0462403373155307e-05, "loss": 0.0187, "step": 4856 }, { "epoch": 3.4132115249472945, "grad_norm": 0.3334595561027527, "learning_rate": 2.0466619817287424e-05, "loss": 0.0335, "step": 4857 }, { "epoch": 3.4139142656359804, "grad_norm": 0.5210283994674683, "learning_rate": 2.0470836261419537e-05, "loss": 0.0211, "step": 4858 }, { "epoch": 3.4146170063246664, "grad_norm": 0.27388492226600647, "learning_rate": 2.0475052705551654e-05, "loss": 0.0313, "step": 4859 }, { "epoch": 3.415319747013352, "grad_norm": 0.3659123480319977, "learning_rate": 2.0479269149683767e-05, "loss": 0.0593, "step": 4860 }, { "epoch": 3.4160224877020378, "grad_norm": 0.2486974596977234, "learning_rate": 2.0483485593815884e-05, "loss": 0.0439, "step": 4861 }, { "epoch": 3.4167252283907237, "grad_norm": 0.46117594838142395, "learning_rate": 2.0487702037947997e-05, "loss": 0.0563, "step": 4862 }, { "epoch": 3.4174279690794096, "grad_norm": 0.4902755320072174, "learning_rate": 2.0491918482080113e-05, "loss": 0.0788, "step": 4863 }, { "epoch": 3.4181307097680955, "grad_norm": 0.26971063017845154, "learning_rate": 2.0496134926212227e-05, "loss": 0.0522, "step": 4864 }, { "epoch": 3.4188334504567814, "grad_norm": 0.43916869163513184, "learning_rate": 2.0500351370344343e-05, "loss": 0.0959, "step": 4865 }, { "epoch": 3.4195361911454674, "grad_norm": 0.7396826148033142, "learning_rate": 2.0504567814476457e-05, "loss": 0.1472, "step": 4866 }, { "epoch": 3.4202389318341533, "grad_norm": 0.8297461271286011, "learning_rate": 2.0508784258608577e-05, "loss": 0.2378, "step": 4867 }, { "epoch": 3.420941672522839, "grad_norm": 0.9364542961120605, "learning_rate": 2.051300070274069e-05, "loss": 0.2477, "step": 4868 }, { "epoch": 3.421644413211525, "grad_norm": 1.7476061582565308, "learning_rate": 2.0517217146872806e-05, "loss": 0.317, "step": 4869 }, { "epoch": 3.422347153900211, "grad_norm": 0.3010025918483734, "learning_rate": 2.052143359100492e-05, "loss": 0.0837, "step": 4870 }, { "epoch": 3.4230498945888965, "grad_norm": 0.286772221326828, "learning_rate": 2.0525650035137036e-05, "loss": 0.0294, "step": 4871 }, { "epoch": 3.4237526352775824, "grad_norm": 0.1823003739118576, "learning_rate": 2.052986647926915e-05, "loss": 0.0403, "step": 4872 }, { "epoch": 3.4244553759662684, "grad_norm": 0.21795466542243958, "learning_rate": 2.0534082923401266e-05, "loss": 0.0299, "step": 4873 }, { "epoch": 3.4251581166549543, "grad_norm": 0.2762085497379303, "learning_rate": 2.053829936753338e-05, "loss": 0.0457, "step": 4874 }, { "epoch": 3.42586085734364, "grad_norm": 0.24893896281719208, "learning_rate": 2.0542515811665496e-05, "loss": 0.0358, "step": 4875 }, { "epoch": 3.426563598032326, "grad_norm": 0.22475425899028778, "learning_rate": 2.054673225579761e-05, "loss": 0.0185, "step": 4876 }, { "epoch": 3.427266338721012, "grad_norm": 0.2425612509250641, "learning_rate": 2.0550948699929726e-05, "loss": 0.0382, "step": 4877 }, { "epoch": 3.427969079409698, "grad_norm": 0.20786258578300476, "learning_rate": 2.055516514406184e-05, "loss": 0.0312, "step": 4878 }, { "epoch": 3.4286718200983834, "grad_norm": 0.2297382354736328, "learning_rate": 2.0559381588193956e-05, "loss": 0.0334, "step": 4879 }, { "epoch": 3.4293745607870694, "grad_norm": 0.34097689390182495, "learning_rate": 2.056359803232607e-05, "loss": 0.0312, "step": 4880 }, { "epoch": 3.4300773014757553, "grad_norm": 0.2065698206424713, "learning_rate": 2.056781447645819e-05, "loss": 0.0321, "step": 4881 }, { "epoch": 3.430780042164441, "grad_norm": 0.2900940179824829, "learning_rate": 2.0572030920590302e-05, "loss": 0.0345, "step": 4882 }, { "epoch": 3.431482782853127, "grad_norm": 0.21749840676784515, "learning_rate": 2.057624736472242e-05, "loss": 0.0152, "step": 4883 }, { "epoch": 3.432185523541813, "grad_norm": 0.33838456869125366, "learning_rate": 2.0580463808854532e-05, "loss": 0.0513, "step": 4884 }, { "epoch": 3.432888264230499, "grad_norm": 0.3711647093296051, "learning_rate": 2.058468025298665e-05, "loss": 0.0368, "step": 4885 }, { "epoch": 3.433591004919185, "grad_norm": 0.510604739189148, "learning_rate": 2.0588896697118766e-05, "loss": 0.0573, "step": 4886 }, { "epoch": 3.434293745607871, "grad_norm": 0.2817387878894806, "learning_rate": 2.059311314125088e-05, "loss": 0.0393, "step": 4887 }, { "epoch": 3.4349964862965567, "grad_norm": 0.4289171099662781, "learning_rate": 2.0597329585382995e-05, "loss": 0.0711, "step": 4888 }, { "epoch": 3.4356992269852427, "grad_norm": 0.40042027831077576, "learning_rate": 2.060154602951511e-05, "loss": 0.0917, "step": 4889 }, { "epoch": 3.436401967673928, "grad_norm": 2.334625482559204, "learning_rate": 2.0605762473647225e-05, "loss": 0.1336, "step": 4890 }, { "epoch": 3.437104708362614, "grad_norm": 1.9435731172561646, "learning_rate": 2.060997891777934e-05, "loss": 0.1266, "step": 4891 }, { "epoch": 3.4378074490513, "grad_norm": 0.8078992962837219, "learning_rate": 2.0614195361911455e-05, "loss": 0.2223, "step": 4892 }, { "epoch": 3.438510189739986, "grad_norm": 1.2712478637695312, "learning_rate": 2.061841180604357e-05, "loss": 0.2612, "step": 4893 }, { "epoch": 3.439212930428672, "grad_norm": 1.4094070196151733, "learning_rate": 2.062262825017569e-05, "loss": 0.2698, "step": 4894 }, { "epoch": 3.4399156711173577, "grad_norm": 0.30292677879333496, "learning_rate": 2.0626844694307802e-05, "loss": 0.0824, "step": 4895 }, { "epoch": 3.4406184118060437, "grad_norm": 0.1656624674797058, "learning_rate": 2.063106113843992e-05, "loss": 0.029, "step": 4896 }, { "epoch": 3.4413211524947296, "grad_norm": 0.2695988416671753, "learning_rate": 2.063527758257203e-05, "loss": 0.0382, "step": 4897 }, { "epoch": 3.4420238931834155, "grad_norm": 0.21493878960609436, "learning_rate": 2.0639494026704148e-05, "loss": 0.0264, "step": 4898 }, { "epoch": 3.442726633872101, "grad_norm": 0.304718941450119, "learning_rate": 2.064371047083626e-05, "loss": 0.0253, "step": 4899 }, { "epoch": 3.443429374560787, "grad_norm": 0.18053241074085236, "learning_rate": 2.0647926914968378e-05, "loss": 0.0303, "step": 4900 }, { "epoch": 3.444132115249473, "grad_norm": 0.2508079707622528, "learning_rate": 2.065214335910049e-05, "loss": 0.0302, "step": 4901 }, { "epoch": 3.4448348559381587, "grad_norm": 0.21042482554912567, "learning_rate": 2.0656359803232608e-05, "loss": 0.035, "step": 4902 }, { "epoch": 3.4455375966268447, "grad_norm": 0.21632209420204163, "learning_rate": 2.066057624736472e-05, "loss": 0.022, "step": 4903 }, { "epoch": 3.4462403373155306, "grad_norm": 0.20617228746414185, "learning_rate": 2.0664792691496838e-05, "loss": 0.0182, "step": 4904 }, { "epoch": 3.4469430780042165, "grad_norm": 0.4344894587993622, "learning_rate": 2.066900913562895e-05, "loss": 0.0279, "step": 4905 }, { "epoch": 3.4476458186929024, "grad_norm": 0.570482075214386, "learning_rate": 2.0673225579761068e-05, "loss": 0.0259, "step": 4906 }, { "epoch": 3.4483485593815884, "grad_norm": 0.3761458098888397, "learning_rate": 2.067744202389318e-05, "loss": 0.0498, "step": 4907 }, { "epoch": 3.4490513000702743, "grad_norm": 0.330501914024353, "learning_rate": 2.06816584680253e-05, "loss": 0.0438, "step": 4908 }, { "epoch": 3.4497540407589597, "grad_norm": 0.2227536290884018, "learning_rate": 2.0685874912157414e-05, "loss": 0.0332, "step": 4909 }, { "epoch": 3.4504567814476457, "grad_norm": 0.4435538351535797, "learning_rate": 2.069009135628953e-05, "loss": 0.0509, "step": 4910 }, { "epoch": 3.4511595221363316, "grad_norm": 0.36433032155036926, "learning_rate": 2.0694307800421644e-05, "loss": 0.049, "step": 4911 }, { "epoch": 3.4518622628250175, "grad_norm": 0.48463699221611023, "learning_rate": 2.069852424455376e-05, "loss": 0.0705, "step": 4912 }, { "epoch": 3.4525650035137034, "grad_norm": 0.3699181377887726, "learning_rate": 2.0702740688685874e-05, "loss": 0.0464, "step": 4913 }, { "epoch": 3.4532677442023894, "grad_norm": 0.3236810863018036, "learning_rate": 2.070695713281799e-05, "loss": 0.0388, "step": 4914 }, { "epoch": 3.4539704848910753, "grad_norm": 1.2037711143493652, "learning_rate": 2.0711173576950104e-05, "loss": 0.1072, "step": 4915 }, { "epoch": 3.454673225579761, "grad_norm": 0.6744419932365417, "learning_rate": 2.071539002108222e-05, "loss": 0.1442, "step": 4916 }, { "epoch": 3.455375966268447, "grad_norm": 0.9574888348579407, "learning_rate": 2.0719606465214337e-05, "loss": 0.1981, "step": 4917 }, { "epoch": 3.4560787069571326, "grad_norm": 1.0138800144195557, "learning_rate": 2.072382290934645e-05, "loss": 0.236, "step": 4918 }, { "epoch": 3.4567814476458185, "grad_norm": 2.367706298828125, "learning_rate": 2.0728039353478567e-05, "loss": 0.3026, "step": 4919 }, { "epoch": 3.4574841883345044, "grad_norm": 0.3173237442970276, "learning_rate": 2.073225579761068e-05, "loss": 0.1019, "step": 4920 }, { "epoch": 3.4581869290231904, "grad_norm": 0.38912397623062134, "learning_rate": 2.0736472241742797e-05, "loss": 0.0493, "step": 4921 }, { "epoch": 3.4588896697118763, "grad_norm": 0.2766917049884796, "learning_rate": 2.0740688685874914e-05, "loss": 0.036, "step": 4922 }, { "epoch": 3.459592410400562, "grad_norm": 0.19961531460285187, "learning_rate": 2.074490513000703e-05, "loss": 0.0319, "step": 4923 }, { "epoch": 3.460295151089248, "grad_norm": 2.310133695602417, "learning_rate": 2.0749121574139144e-05, "loss": 0.0264, "step": 4924 }, { "epoch": 3.460997891777934, "grad_norm": 0.2657177448272705, "learning_rate": 2.075333801827126e-05, "loss": 0.0254, "step": 4925 }, { "epoch": 3.46170063246662, "grad_norm": 0.19944927096366882, "learning_rate": 2.0757554462403373e-05, "loss": 0.019, "step": 4926 }, { "epoch": 3.462403373155306, "grad_norm": 0.2547244429588318, "learning_rate": 2.076177090653549e-05, "loss": 0.0326, "step": 4927 }, { "epoch": 3.463106113843992, "grad_norm": 0.29434749484062195, "learning_rate": 2.0765987350667603e-05, "loss": 0.0454, "step": 4928 }, { "epoch": 3.4638088545326773, "grad_norm": 0.2144019603729248, "learning_rate": 2.077020379479972e-05, "loss": 0.0452, "step": 4929 }, { "epoch": 3.464511595221363, "grad_norm": 0.2904396653175354, "learning_rate": 2.0774420238931833e-05, "loss": 0.0487, "step": 4930 }, { "epoch": 3.465214335910049, "grad_norm": 0.23266395926475525, "learning_rate": 2.077863668306395e-05, "loss": 0.027, "step": 4931 }, { "epoch": 3.465917076598735, "grad_norm": 0.19184263050556183, "learning_rate": 2.0782853127196063e-05, "loss": 0.0314, "step": 4932 }, { "epoch": 3.466619817287421, "grad_norm": 0.18574173748493195, "learning_rate": 2.078706957132818e-05, "loss": 0.0214, "step": 4933 }, { "epoch": 3.467322557976107, "grad_norm": 0.26876959204673767, "learning_rate": 2.0791286015460293e-05, "loss": 0.0396, "step": 4934 }, { "epoch": 3.468025298664793, "grad_norm": 0.2826419770717621, "learning_rate": 2.0795502459592413e-05, "loss": 0.052, "step": 4935 }, { "epoch": 3.4687280393534787, "grad_norm": 0.2587614357471466, "learning_rate": 2.0799718903724526e-05, "loss": 0.0392, "step": 4936 }, { "epoch": 3.469430780042164, "grad_norm": 0.23064278066158295, "learning_rate": 2.0803935347856643e-05, "loss": 0.0341, "step": 4937 }, { "epoch": 3.47013352073085, "grad_norm": 0.5991947054862976, "learning_rate": 2.0808151791988756e-05, "loss": 0.0762, "step": 4938 }, { "epoch": 3.470836261419536, "grad_norm": 0.5088192820549011, "learning_rate": 2.0812368236120873e-05, "loss": 0.0931, "step": 4939 }, { "epoch": 3.471539002108222, "grad_norm": 1.19285249710083, "learning_rate": 2.0816584680252986e-05, "loss": 0.1092, "step": 4940 }, { "epoch": 3.472241742796908, "grad_norm": 0.5120694041252136, "learning_rate": 2.0820801124385103e-05, "loss": 0.1541, "step": 4941 }, { "epoch": 3.472944483485594, "grad_norm": 0.7887358069419861, "learning_rate": 2.0825017568517216e-05, "loss": 0.2134, "step": 4942 }, { "epoch": 3.4736472241742797, "grad_norm": 1.8364312648773193, "learning_rate": 2.0829234012649333e-05, "loss": 0.2668, "step": 4943 }, { "epoch": 3.4743499648629657, "grad_norm": 1.712075114250183, "learning_rate": 2.0833450456781446e-05, "loss": 0.345, "step": 4944 }, { "epoch": 3.4750527055516516, "grad_norm": 0.3064524233341217, "learning_rate": 2.0837666900913562e-05, "loss": 0.0769, "step": 4945 }, { "epoch": 3.4757554462403375, "grad_norm": 0.2672661542892456, "learning_rate": 2.084188334504568e-05, "loss": 0.0543, "step": 4946 }, { "epoch": 3.4764581869290234, "grad_norm": 0.155535027384758, "learning_rate": 2.0846099789177792e-05, "loss": 0.0196, "step": 4947 }, { "epoch": 3.477160927617709, "grad_norm": 0.26709431409835815, "learning_rate": 2.085031623330991e-05, "loss": 0.0245, "step": 4948 }, { "epoch": 3.477863668306395, "grad_norm": 0.19551531970500946, "learning_rate": 2.0854532677442026e-05, "loss": 0.0274, "step": 4949 }, { "epoch": 3.4785664089950807, "grad_norm": 0.26557064056396484, "learning_rate": 2.0858749121574142e-05, "loss": 0.0278, "step": 4950 }, { "epoch": 3.4792691496837667, "grad_norm": 0.23506051301956177, "learning_rate": 2.0862965565706256e-05, "loss": 0.0297, "step": 4951 }, { "epoch": 3.4799718903724526, "grad_norm": 0.2798542082309723, "learning_rate": 2.0867182009838372e-05, "loss": 0.0558, "step": 4952 }, { "epoch": 3.4806746310611385, "grad_norm": 0.2557480037212372, "learning_rate": 2.0871398453970485e-05, "loss": 0.0333, "step": 4953 }, { "epoch": 3.4813773717498244, "grad_norm": 0.25562891364097595, "learning_rate": 2.0875614898102602e-05, "loss": 0.0292, "step": 4954 }, { "epoch": 3.4820801124385103, "grad_norm": 0.3534550666809082, "learning_rate": 2.0879831342234715e-05, "loss": 0.026, "step": 4955 }, { "epoch": 3.482782853127196, "grad_norm": 0.15009891986846924, "learning_rate": 2.0884047786366832e-05, "loss": 0.0243, "step": 4956 }, { "epoch": 3.4834855938158817, "grad_norm": 0.34236323833465576, "learning_rate": 2.0888264230498945e-05, "loss": 0.0494, "step": 4957 }, { "epoch": 3.4841883345045677, "grad_norm": 0.21518172323703766, "learning_rate": 2.0892480674631062e-05, "loss": 0.0207, "step": 4958 }, { "epoch": 3.4848910751932536, "grad_norm": 0.23734557628631592, "learning_rate": 2.0896697118763175e-05, "loss": 0.0255, "step": 4959 }, { "epoch": 3.4855938158819395, "grad_norm": 0.25061675906181335, "learning_rate": 2.0900913562895292e-05, "loss": 0.0437, "step": 4960 }, { "epoch": 3.4862965565706254, "grad_norm": 0.23610377311706543, "learning_rate": 2.0905130007027405e-05, "loss": 0.0315, "step": 4961 }, { "epoch": 3.4869992972593113, "grad_norm": 0.3198849856853485, "learning_rate": 2.0909346451159525e-05, "loss": 0.045, "step": 4962 }, { "epoch": 3.4877020379479973, "grad_norm": 0.41998931765556335, "learning_rate": 2.0913562895291638e-05, "loss": 0.0783, "step": 4963 }, { "epoch": 3.488404778636683, "grad_norm": 0.31571468710899353, "learning_rate": 2.0917779339423755e-05, "loss": 0.0721, "step": 4964 }, { "epoch": 3.489107519325369, "grad_norm": 0.4366775155067444, "learning_rate": 2.0921995783555868e-05, "loss": 0.0785, "step": 4965 }, { "epoch": 3.489810260014055, "grad_norm": 0.7341553568840027, "learning_rate": 2.0926212227687985e-05, "loss": 0.188, "step": 4966 }, { "epoch": 3.4905130007027405, "grad_norm": 0.8269462585449219, "learning_rate": 2.0930428671820098e-05, "loss": 0.2137, "step": 4967 }, { "epoch": 3.4912157413914264, "grad_norm": 0.9632827043533325, "learning_rate": 2.0934645115952215e-05, "loss": 0.2784, "step": 4968 }, { "epoch": 3.4919184820801124, "grad_norm": 1.2161204814910889, "learning_rate": 2.0938861560084328e-05, "loss": 0.3129, "step": 4969 }, { "epoch": 3.4926212227687983, "grad_norm": 0.39094117283821106, "learning_rate": 2.0943078004216445e-05, "loss": 0.0993, "step": 4970 }, { "epoch": 3.493323963457484, "grad_norm": 0.2557234466075897, "learning_rate": 2.0947294448348558e-05, "loss": 0.0352, "step": 4971 }, { "epoch": 3.49402670414617, "grad_norm": 0.20675994455814362, "learning_rate": 2.0951510892480674e-05, "loss": 0.027, "step": 4972 }, { "epoch": 3.494729444834856, "grad_norm": 0.18814696371555328, "learning_rate": 2.0955727336612788e-05, "loss": 0.0307, "step": 4973 }, { "epoch": 3.495432185523542, "grad_norm": 0.17914997041225433, "learning_rate": 2.0959943780744904e-05, "loss": 0.0236, "step": 4974 }, { "epoch": 3.496134926212228, "grad_norm": 0.5583809614181519, "learning_rate": 2.096416022487702e-05, "loss": 0.0219, "step": 4975 }, { "epoch": 3.4968376669009134, "grad_norm": 0.286325603723526, "learning_rate": 2.0968376669009138e-05, "loss": 0.0408, "step": 4976 }, { "epoch": 3.4975404075895993, "grad_norm": 0.2758176326751709, "learning_rate": 2.0972593113141254e-05, "loss": 0.0318, "step": 4977 }, { "epoch": 3.498243148278285, "grad_norm": 0.2904849946498871, "learning_rate": 2.0976809557273367e-05, "loss": 0.0259, "step": 4978 }, { "epoch": 3.498945888966971, "grad_norm": 0.4676221013069153, "learning_rate": 2.0981026001405484e-05, "loss": 0.0245, "step": 4979 }, { "epoch": 3.499648629655657, "grad_norm": 0.22538135945796967, "learning_rate": 2.0985242445537597e-05, "loss": 0.0289, "step": 4980 }, { "epoch": 3.500351370344343, "grad_norm": 0.1797700971364975, "learning_rate": 2.0989458889669714e-05, "loss": 0.0266, "step": 4981 }, { "epoch": 3.501054111033029, "grad_norm": 0.2545415163040161, "learning_rate": 2.0993675333801827e-05, "loss": 0.0315, "step": 4982 }, { "epoch": 3.501756851721715, "grad_norm": 0.31491002440452576, "learning_rate": 2.0997891777933944e-05, "loss": 0.0356, "step": 4983 }, { "epoch": 3.5024595924104007, "grad_norm": 0.3851839005947113, "learning_rate": 2.1002108222066057e-05, "loss": 0.042, "step": 4984 }, { "epoch": 3.5031623330990866, "grad_norm": 0.18259680271148682, "learning_rate": 2.1006324666198174e-05, "loss": 0.0253, "step": 4985 }, { "epoch": 3.5038650737877726, "grad_norm": 0.22696168720722198, "learning_rate": 2.1010541110330287e-05, "loss": 0.0277, "step": 4986 }, { "epoch": 3.504567814476458, "grad_norm": 0.35071730613708496, "learning_rate": 2.1014757554462404e-05, "loss": 0.0525, "step": 4987 }, { "epoch": 3.505270555165144, "grad_norm": 0.4298151731491089, "learning_rate": 2.1018973998594517e-05, "loss": 0.0819, "step": 4988 }, { "epoch": 3.50597329585383, "grad_norm": 0.38183116912841797, "learning_rate": 2.1023190442726634e-05, "loss": 0.0688, "step": 4989 }, { "epoch": 3.506676036542516, "grad_norm": 0.5162882208824158, "learning_rate": 2.102740688685875e-05, "loss": 0.0921, "step": 4990 }, { "epoch": 3.5073787772312017, "grad_norm": 1.0811532735824585, "learning_rate": 2.1031623330990867e-05, "loss": 0.139, "step": 4991 }, { "epoch": 3.5080815179198876, "grad_norm": 0.6766942143440247, "learning_rate": 2.103583977512298e-05, "loss": 0.2309, "step": 4992 }, { "epoch": 3.5087842586085736, "grad_norm": 1.4571679830551147, "learning_rate": 2.1040056219255097e-05, "loss": 0.244, "step": 4993 }, { "epoch": 3.509486999297259, "grad_norm": 1.4621001482009888, "learning_rate": 2.104427266338721e-05, "loss": 0.2871, "step": 4994 }, { "epoch": 3.510189739985945, "grad_norm": 0.45943206548690796, "learning_rate": 2.1048489107519327e-05, "loss": 0.1125, "step": 4995 }, { "epoch": 3.510892480674631, "grad_norm": 0.631712794303894, "learning_rate": 2.105270555165144e-05, "loss": 0.0438, "step": 4996 }, { "epoch": 3.511595221363317, "grad_norm": 0.260985791683197, "learning_rate": 2.1056921995783557e-05, "loss": 0.0403, "step": 4997 }, { "epoch": 3.5122979620520027, "grad_norm": 0.2211351841688156, "learning_rate": 2.106113843991567e-05, "loss": 0.032, "step": 4998 }, { "epoch": 3.5130007027406887, "grad_norm": 0.1962992399930954, "learning_rate": 2.1065354884047786e-05, "loss": 0.016, "step": 4999 }, { "epoch": 3.5137034434293746, "grad_norm": 0.30039146542549133, "learning_rate": 2.10695713281799e-05, "loss": 0.0284, "step": 5000 }, { "epoch": 3.5137034434293746, "eval_cer": 0.2009481413592572, "eval_loss": 0.33390572667121887, "eval_runtime": 18.4132, "eval_samples_per_second": 246.454, "eval_steps_per_second": 0.815, "eval_wer": 0.37510125415491186, "step": 5000 }, { "epoch": 3.5144061841180605, "grad_norm": 0.17115353047847748, "learning_rate": 2.1073787772312016e-05, "loss": 0.023, "step": 5001 }, { "epoch": 3.5151089248067464, "grad_norm": 0.27681228518486023, "learning_rate": 2.107800421644413e-05, "loss": 0.0417, "step": 5002 }, { "epoch": 3.5158116654954323, "grad_norm": 0.18351465463638306, "learning_rate": 2.108222066057625e-05, "loss": 0.0219, "step": 5003 }, { "epoch": 3.5165144061841183, "grad_norm": 0.2486037164926529, "learning_rate": 2.1086437104708363e-05, "loss": 0.0248, "step": 5004 }, { "epoch": 3.517217146872804, "grad_norm": 0.23285850882530212, "learning_rate": 2.109065354884048e-05, "loss": 0.0536, "step": 5005 }, { "epoch": 3.5179198875614897, "grad_norm": 0.207634836435318, "learning_rate": 2.1094869992972596e-05, "loss": 0.0264, "step": 5006 }, { "epoch": 3.5186226282501756, "grad_norm": 0.2181103527545929, "learning_rate": 2.109908643710471e-05, "loss": 0.0303, "step": 5007 }, { "epoch": 3.5193253689388615, "grad_norm": 0.244887113571167, "learning_rate": 2.1103302881236826e-05, "loss": 0.0183, "step": 5008 }, { "epoch": 3.5200281096275474, "grad_norm": 0.6088599562644958, "learning_rate": 2.110751932536894e-05, "loss": 0.0647, "step": 5009 }, { "epoch": 3.5207308503162333, "grad_norm": 0.44697147607803345, "learning_rate": 2.1111735769501056e-05, "loss": 0.0556, "step": 5010 }, { "epoch": 3.5214335910049193, "grad_norm": 0.2656745910644531, "learning_rate": 2.111595221363317e-05, "loss": 0.023, "step": 5011 }, { "epoch": 3.522136331693605, "grad_norm": 0.2714621126651764, "learning_rate": 2.1120168657765286e-05, "loss": 0.0615, "step": 5012 }, { "epoch": 3.5228390723822907, "grad_norm": 0.23674842715263367, "learning_rate": 2.11243851018974e-05, "loss": 0.0423, "step": 5013 }, { "epoch": 3.5235418130709766, "grad_norm": 0.487941712141037, "learning_rate": 2.1128601546029516e-05, "loss": 0.086, "step": 5014 }, { "epoch": 3.5242445537596625, "grad_norm": 0.6301090717315674, "learning_rate": 2.113281799016163e-05, "loss": 0.0883, "step": 5015 }, { "epoch": 3.5249472944483484, "grad_norm": 0.5781604647636414, "learning_rate": 2.1137034434293746e-05, "loss": 0.1341, "step": 5016 }, { "epoch": 3.5256500351370343, "grad_norm": 0.6042354702949524, "learning_rate": 2.1141250878425862e-05, "loss": 0.1583, "step": 5017 }, { "epoch": 3.5263527758257203, "grad_norm": 1.0709307193756104, "learning_rate": 2.114546732255798e-05, "loss": 0.2602, "step": 5018 }, { "epoch": 3.527055516514406, "grad_norm": 1.8688088655471802, "learning_rate": 2.1149683766690092e-05, "loss": 0.2771, "step": 5019 }, { "epoch": 3.527758257203092, "grad_norm": 0.3381069004535675, "learning_rate": 2.115390021082221e-05, "loss": 0.081, "step": 5020 }, { "epoch": 3.528460997891778, "grad_norm": 0.1919216364622116, "learning_rate": 2.1158116654954322e-05, "loss": 0.0373, "step": 5021 }, { "epoch": 3.529163738580464, "grad_norm": 0.34431859850883484, "learning_rate": 2.116233309908644e-05, "loss": 0.0475, "step": 5022 }, { "epoch": 3.52986647926915, "grad_norm": 0.1787198930978775, "learning_rate": 2.1166549543218552e-05, "loss": 0.0178, "step": 5023 }, { "epoch": 3.530569219957836, "grad_norm": 0.22479332983493805, "learning_rate": 2.117076598735067e-05, "loss": 0.027, "step": 5024 }, { "epoch": 3.5312719606465213, "grad_norm": 0.24612225592136383, "learning_rate": 2.1174982431482782e-05, "loss": 0.0275, "step": 5025 }, { "epoch": 3.531974701335207, "grad_norm": 0.16651205718517303, "learning_rate": 2.11791988756149e-05, "loss": 0.0204, "step": 5026 }, { "epoch": 3.532677442023893, "grad_norm": 0.24270574748516083, "learning_rate": 2.118341531974701e-05, "loss": 0.0457, "step": 5027 }, { "epoch": 3.533380182712579, "grad_norm": 0.1993403434753418, "learning_rate": 2.1187631763879128e-05, "loss": 0.0395, "step": 5028 }, { "epoch": 3.534082923401265, "grad_norm": 0.172560453414917, "learning_rate": 2.119184820801124e-05, "loss": 0.0221, "step": 5029 }, { "epoch": 3.534785664089951, "grad_norm": 0.22527751326560974, "learning_rate": 2.119606465214336e-05, "loss": 0.0402, "step": 5030 }, { "epoch": 3.535488404778637, "grad_norm": 0.23177894949913025, "learning_rate": 2.1200281096275475e-05, "loss": 0.016, "step": 5031 }, { "epoch": 3.5361911454673227, "grad_norm": 0.21584677696228027, "learning_rate": 2.120449754040759e-05, "loss": 0.0367, "step": 5032 }, { "epoch": 3.536893886156008, "grad_norm": 0.17832964658737183, "learning_rate": 2.1208713984539705e-05, "loss": 0.018, "step": 5033 }, { "epoch": 3.537596626844694, "grad_norm": 0.35614490509033203, "learning_rate": 2.121293042867182e-05, "loss": 0.0385, "step": 5034 }, { "epoch": 3.53829936753338, "grad_norm": 0.2781495153903961, "learning_rate": 2.1217146872803938e-05, "loss": 0.0454, "step": 5035 }, { "epoch": 3.539002108222066, "grad_norm": 0.23546187579631805, "learning_rate": 2.122136331693605e-05, "loss": 0.025, "step": 5036 }, { "epoch": 3.539704848910752, "grad_norm": 0.31863096356391907, "learning_rate": 2.1225579761068168e-05, "loss": 0.0514, "step": 5037 }, { "epoch": 3.540407589599438, "grad_norm": 0.4346301555633545, "learning_rate": 2.122979620520028e-05, "loss": 0.0572, "step": 5038 }, { "epoch": 3.5411103302881237, "grad_norm": 0.5832288861274719, "learning_rate": 2.1234012649332398e-05, "loss": 0.0643, "step": 5039 }, { "epoch": 3.5418130709768096, "grad_norm": 0.4745121896266937, "learning_rate": 2.123822909346451e-05, "loss": 0.0764, "step": 5040 }, { "epoch": 3.5425158116654956, "grad_norm": 1.302137851715088, "learning_rate": 2.1242445537596628e-05, "loss": 0.1417, "step": 5041 }, { "epoch": 3.5432185523541815, "grad_norm": 0.8309231400489807, "learning_rate": 2.124666198172874e-05, "loss": 0.2195, "step": 5042 }, { "epoch": 3.5439212930428674, "grad_norm": 0.8195456862449646, "learning_rate": 2.1250878425860857e-05, "loss": 0.2337, "step": 5043 }, { "epoch": 3.5446240337315533, "grad_norm": 1.8756372928619385, "learning_rate": 2.1255094869992974e-05, "loss": 0.348, "step": 5044 }, { "epoch": 3.545326774420239, "grad_norm": 0.36158284544944763, "learning_rate": 2.125931131412509e-05, "loss": 0.1003, "step": 5045 }, { "epoch": 3.5460295151089247, "grad_norm": 0.26613184809684753, "learning_rate": 2.1263527758257204e-05, "loss": 0.0471, "step": 5046 }, { "epoch": 3.5467322557976106, "grad_norm": 0.2023162841796875, "learning_rate": 2.126774420238932e-05, "loss": 0.0341, "step": 5047 }, { "epoch": 3.5474349964862966, "grad_norm": 0.2582640051841736, "learning_rate": 2.1271960646521434e-05, "loss": 0.0579, "step": 5048 }, { "epoch": 3.5481377371749825, "grad_norm": 0.17165733873844147, "learning_rate": 2.127617709065355e-05, "loss": 0.0251, "step": 5049 }, { "epoch": 3.5488404778636684, "grad_norm": 0.2205306887626648, "learning_rate": 2.1280393534785664e-05, "loss": 0.0484, "step": 5050 }, { "epoch": 3.5495432185523543, "grad_norm": 0.2259361445903778, "learning_rate": 2.128460997891778e-05, "loss": 0.0308, "step": 5051 }, { "epoch": 3.55024595924104, "grad_norm": 0.19058474898338318, "learning_rate": 2.1288826423049894e-05, "loss": 0.0172, "step": 5052 }, { "epoch": 3.5509486999297257, "grad_norm": 0.2576451897621155, "learning_rate": 2.129304286718201e-05, "loss": 0.0515, "step": 5053 }, { "epoch": 3.5516514406184116, "grad_norm": 0.37388238310813904, "learning_rate": 2.1297259311314124e-05, "loss": 0.0248, "step": 5054 }, { "epoch": 3.5523541813070976, "grad_norm": 0.25897738337516785, "learning_rate": 2.130147575544624e-05, "loss": 0.0427, "step": 5055 }, { "epoch": 3.5530569219957835, "grad_norm": 0.2293960601091385, "learning_rate": 2.1305692199578353e-05, "loss": 0.0307, "step": 5056 }, { "epoch": 3.5537596626844694, "grad_norm": 0.3169863522052765, "learning_rate": 2.1309908643710473e-05, "loss": 0.0734, "step": 5057 }, { "epoch": 3.5544624033731553, "grad_norm": 0.25742053985595703, "learning_rate": 2.1314125087842587e-05, "loss": 0.0378, "step": 5058 }, { "epoch": 3.5551651440618413, "grad_norm": 0.2286502718925476, "learning_rate": 2.1318341531974703e-05, "loss": 0.0394, "step": 5059 }, { "epoch": 3.555867884750527, "grad_norm": 0.2509063184261322, "learning_rate": 2.1322557976106817e-05, "loss": 0.0444, "step": 5060 }, { "epoch": 3.556570625439213, "grad_norm": 0.25936850905418396, "learning_rate": 2.1326774420238933e-05, "loss": 0.0331, "step": 5061 }, { "epoch": 3.557273366127899, "grad_norm": 0.34176260232925415, "learning_rate": 2.1330990864371046e-05, "loss": 0.0431, "step": 5062 }, { "epoch": 3.557976106816585, "grad_norm": 0.388602614402771, "learning_rate": 2.1335207308503163e-05, "loss": 0.0474, "step": 5063 }, { "epoch": 3.5586788475052704, "grad_norm": 0.5731221437454224, "learning_rate": 2.133942375263528e-05, "loss": 0.0635, "step": 5064 }, { "epoch": 3.5593815881939563, "grad_norm": 0.6902585029602051, "learning_rate": 2.1343640196767393e-05, "loss": 0.1308, "step": 5065 }, { "epoch": 3.5600843288826423, "grad_norm": 0.5581376552581787, "learning_rate": 2.134785664089951e-05, "loss": 0.1668, "step": 5066 }, { "epoch": 3.560787069571328, "grad_norm": 2.3877410888671875, "learning_rate": 2.1352073085031623e-05, "loss": 0.19, "step": 5067 }, { "epoch": 3.561489810260014, "grad_norm": 0.9853572845458984, "learning_rate": 2.135628952916374e-05, "loss": 0.2613, "step": 5068 }, { "epoch": 3.5621925509487, "grad_norm": 2.0581459999084473, "learning_rate": 2.1360505973295853e-05, "loss": 0.2929, "step": 5069 }, { "epoch": 3.562895291637386, "grad_norm": 0.37926414608955383, "learning_rate": 2.136472241742797e-05, "loss": 0.0942, "step": 5070 }, { "epoch": 3.5635980323260714, "grad_norm": 0.2016747146844864, "learning_rate": 2.1368938861560086e-05, "loss": 0.0335, "step": 5071 }, { "epoch": 3.5643007730147573, "grad_norm": 0.32560157775878906, "learning_rate": 2.1373155305692203e-05, "loss": 0.0415, "step": 5072 }, { "epoch": 3.5650035137034433, "grad_norm": 0.2383374124765396, "learning_rate": 2.1377371749824316e-05, "loss": 0.0334, "step": 5073 }, { "epoch": 3.565706254392129, "grad_norm": 0.3666781783103943, "learning_rate": 2.1381588193956433e-05, "loss": 0.0355, "step": 5074 }, { "epoch": 3.566408995080815, "grad_norm": 0.32492420077323914, "learning_rate": 2.1385804638088546e-05, "loss": 0.0217, "step": 5075 }, { "epoch": 3.567111735769501, "grad_norm": 0.18403665721416473, "learning_rate": 2.1390021082220662e-05, "loss": 0.0358, "step": 5076 }, { "epoch": 3.567814476458187, "grad_norm": 0.36129432916641235, "learning_rate": 2.1394237526352776e-05, "loss": 0.0306, "step": 5077 }, { "epoch": 3.568517217146873, "grad_norm": 0.2025419920682907, "learning_rate": 2.1398453970484892e-05, "loss": 0.0387, "step": 5078 }, { "epoch": 3.569219957835559, "grad_norm": 0.4684699475765228, "learning_rate": 2.1402670414617006e-05, "loss": 0.0193, "step": 5079 }, { "epoch": 3.5699226985242447, "grad_norm": 0.1989433914422989, "learning_rate": 2.1406886858749122e-05, "loss": 0.0334, "step": 5080 }, { "epoch": 3.5706254392129306, "grad_norm": 0.2523752748966217, "learning_rate": 2.1411103302881235e-05, "loss": 0.0202, "step": 5081 }, { "epoch": 3.5713281799016166, "grad_norm": 0.25121060013771057, "learning_rate": 2.1415319747013352e-05, "loss": 0.0448, "step": 5082 }, { "epoch": 3.572030920590302, "grad_norm": 0.37030908465385437, "learning_rate": 2.1419536191145465e-05, "loss": 0.0338, "step": 5083 }, { "epoch": 3.572733661278988, "grad_norm": 0.28336790204048157, "learning_rate": 2.1423752635277582e-05, "loss": 0.0472, "step": 5084 }, { "epoch": 3.573436401967674, "grad_norm": 0.19821499288082123, "learning_rate": 2.14279690794097e-05, "loss": 0.0354, "step": 5085 }, { "epoch": 3.57413914265636, "grad_norm": 0.15880997478961945, "learning_rate": 2.1432185523541815e-05, "loss": 0.0213, "step": 5086 }, { "epoch": 3.5748418833450457, "grad_norm": 0.4218341112136841, "learning_rate": 2.143640196767393e-05, "loss": 0.062, "step": 5087 }, { "epoch": 3.5755446240337316, "grad_norm": 0.5368338227272034, "learning_rate": 2.1440618411806045e-05, "loss": 0.0665, "step": 5088 }, { "epoch": 3.5762473647224176, "grad_norm": 0.41571012139320374, "learning_rate": 2.144483485593816e-05, "loss": 0.0921, "step": 5089 }, { "epoch": 3.576950105411103, "grad_norm": 0.5354587435722351, "learning_rate": 2.1449051300070275e-05, "loss": 0.1057, "step": 5090 }, { "epoch": 3.577652846099789, "grad_norm": 0.6102652549743652, "learning_rate": 2.145326774420239e-05, "loss": 0.1351, "step": 5091 }, { "epoch": 3.578355586788475, "grad_norm": 1.0494805574417114, "learning_rate": 2.1457484188334505e-05, "loss": 0.2244, "step": 5092 }, { "epoch": 3.579058327477161, "grad_norm": 0.8864296078681946, "learning_rate": 2.1461700632466618e-05, "loss": 0.287, "step": 5093 }, { "epoch": 3.5797610681658467, "grad_norm": 2.2455780506134033, "learning_rate": 2.1465917076598735e-05, "loss": 0.3488, "step": 5094 }, { "epoch": 3.5804638088545326, "grad_norm": 0.462919682264328, "learning_rate": 2.147013352073085e-05, "loss": 0.0863, "step": 5095 }, { "epoch": 3.5811665495432186, "grad_norm": 0.1955917626619339, "learning_rate": 2.1474349964862965e-05, "loss": 0.0468, "step": 5096 }, { "epoch": 3.5818692902319045, "grad_norm": 0.3292374610900879, "learning_rate": 2.147856640899508e-05, "loss": 0.0545, "step": 5097 }, { "epoch": 3.5825720309205904, "grad_norm": 0.19047634303569794, "learning_rate": 2.1482782853127198e-05, "loss": 0.0421, "step": 5098 }, { "epoch": 3.5832747716092763, "grad_norm": 0.16957023739814758, "learning_rate": 2.1486999297259315e-05, "loss": 0.0246, "step": 5099 }, { "epoch": 3.5839775122979622, "grad_norm": 0.11069336533546448, "learning_rate": 2.1491215741391428e-05, "loss": 0.0145, "step": 5100 }, { "epoch": 3.584680252986648, "grad_norm": 0.1963312327861786, "learning_rate": 2.1495432185523545e-05, "loss": 0.0204, "step": 5101 }, { "epoch": 3.5853829936753336, "grad_norm": 0.17258979380130768, "learning_rate": 2.1499648629655658e-05, "loss": 0.031, "step": 5102 }, { "epoch": 3.5860857343640196, "grad_norm": 0.1789485663175583, "learning_rate": 2.1503865073787774e-05, "loss": 0.0303, "step": 5103 }, { "epoch": 3.5867884750527055, "grad_norm": 0.25712743401527405, "learning_rate": 2.1508081517919888e-05, "loss": 0.0203, "step": 5104 }, { "epoch": 3.5874912157413914, "grad_norm": 0.32617712020874023, "learning_rate": 2.1512297962052004e-05, "loss": 0.0374, "step": 5105 }, { "epoch": 3.5881939564300773, "grad_norm": 0.16806142032146454, "learning_rate": 2.1516514406184118e-05, "loss": 0.0192, "step": 5106 }, { "epoch": 3.5888966971187632, "grad_norm": 0.20026028156280518, "learning_rate": 2.1520730850316234e-05, "loss": 0.0283, "step": 5107 }, { "epoch": 3.589599437807449, "grad_norm": 0.24319332838058472, "learning_rate": 2.1524947294448347e-05, "loss": 0.027, "step": 5108 }, { "epoch": 3.590302178496135, "grad_norm": 0.6533092260360718, "learning_rate": 2.1529163738580464e-05, "loss": 0.062, "step": 5109 }, { "epoch": 3.5910049191848206, "grad_norm": 0.29215702414512634, "learning_rate": 2.1533380182712577e-05, "loss": 0.0575, "step": 5110 }, { "epoch": 3.5917076598735065, "grad_norm": 0.3255958557128906, "learning_rate": 2.1537596626844694e-05, "loss": 0.0533, "step": 5111 }, { "epoch": 3.5924104005621924, "grad_norm": 0.21018388867378235, "learning_rate": 2.154181307097681e-05, "loss": 0.0225, "step": 5112 }, { "epoch": 3.5931131412508783, "grad_norm": 0.5891090035438538, "learning_rate": 2.1546029515108927e-05, "loss": 0.0751, "step": 5113 }, { "epoch": 3.5938158819395642, "grad_norm": 0.37539440393447876, "learning_rate": 2.155024595924104e-05, "loss": 0.0729, "step": 5114 }, { "epoch": 3.59451862262825, "grad_norm": 0.5223250985145569, "learning_rate": 2.1554462403373157e-05, "loss": 0.1498, "step": 5115 }, { "epoch": 3.595221363316936, "grad_norm": 0.6119455099105835, "learning_rate": 2.155867884750527e-05, "loss": 0.1423, "step": 5116 }, { "epoch": 3.595924104005622, "grad_norm": 0.79588383436203, "learning_rate": 2.1562895291637387e-05, "loss": 0.2304, "step": 5117 }, { "epoch": 3.596626844694308, "grad_norm": 0.9275707006454468, "learning_rate": 2.15671117357695e-05, "loss": 0.2342, "step": 5118 }, { "epoch": 3.597329585382994, "grad_norm": 1.483830213546753, "learning_rate": 2.1571328179901617e-05, "loss": 0.3309, "step": 5119 }, { "epoch": 3.5980323260716798, "grad_norm": 0.2815403342247009, "learning_rate": 2.157554462403373e-05, "loss": 0.0697, "step": 5120 }, { "epoch": 3.5987350667603657, "grad_norm": 0.1880340576171875, "learning_rate": 2.1579761068165847e-05, "loss": 0.0411, "step": 5121 }, { "epoch": 3.599437807449051, "grad_norm": 0.2527788281440735, "learning_rate": 2.158397751229796e-05, "loss": 0.0439, "step": 5122 }, { "epoch": 3.600140548137737, "grad_norm": 0.2613259553909302, "learning_rate": 2.1588193956430077e-05, "loss": 0.0341, "step": 5123 }, { "epoch": 3.600843288826423, "grad_norm": 0.25378158688545227, "learning_rate": 2.1592410400562193e-05, "loss": 0.0284, "step": 5124 }, { "epoch": 3.601546029515109, "grad_norm": 0.27427905797958374, "learning_rate": 2.159662684469431e-05, "loss": 0.0323, "step": 5125 }, { "epoch": 3.602248770203795, "grad_norm": 0.2229623645544052, "learning_rate": 2.1600843288826427e-05, "loss": 0.04, "step": 5126 }, { "epoch": 3.602951510892481, "grad_norm": 0.16206711530685425, "learning_rate": 2.160505973295854e-05, "loss": 0.0226, "step": 5127 }, { "epoch": 3.6036542515811667, "grad_norm": 0.35396265983581543, "learning_rate": 2.1609276177090656e-05, "loss": 0.0263, "step": 5128 }, { "epoch": 3.604356992269852, "grad_norm": 0.12624898552894592, "learning_rate": 2.161349262122277e-05, "loss": 0.0176, "step": 5129 }, { "epoch": 3.605059732958538, "grad_norm": 0.2342374175786972, "learning_rate": 2.1617709065354886e-05, "loss": 0.0342, "step": 5130 }, { "epoch": 3.605762473647224, "grad_norm": 0.24063163995742798, "learning_rate": 2.1621925509487e-05, "loss": 0.0239, "step": 5131 }, { "epoch": 3.60646521433591, "grad_norm": 0.23058930039405823, "learning_rate": 2.1626141953619116e-05, "loss": 0.0481, "step": 5132 }, { "epoch": 3.607167955024596, "grad_norm": 0.19547216594219208, "learning_rate": 2.163035839775123e-05, "loss": 0.0312, "step": 5133 }, { "epoch": 3.607870695713282, "grad_norm": 0.26748454570770264, "learning_rate": 2.1634574841883346e-05, "loss": 0.0416, "step": 5134 }, { "epoch": 3.6085734364019677, "grad_norm": 0.34708520770072937, "learning_rate": 2.163879128601546e-05, "loss": 0.0516, "step": 5135 }, { "epoch": 3.6092761770906536, "grad_norm": 0.23226186633110046, "learning_rate": 2.1643007730147576e-05, "loss": 0.0409, "step": 5136 }, { "epoch": 3.6099789177793395, "grad_norm": 0.23976068198680878, "learning_rate": 2.164722417427969e-05, "loss": 0.0459, "step": 5137 }, { "epoch": 3.6106816584680255, "grad_norm": 0.4054390490055084, "learning_rate": 2.1651440618411806e-05, "loss": 0.0632, "step": 5138 }, { "epoch": 3.6113843991567114, "grad_norm": 0.32437780499458313, "learning_rate": 2.1655657062543923e-05, "loss": 0.064, "step": 5139 }, { "epoch": 3.6120871398453973, "grad_norm": 0.3978639245033264, "learning_rate": 2.165987350667604e-05, "loss": 0.0857, "step": 5140 }, { "epoch": 3.612789880534083, "grad_norm": 0.5234559774398804, "learning_rate": 2.1664089950808152e-05, "loss": 0.1675, "step": 5141 }, { "epoch": 3.6134926212227687, "grad_norm": 0.6735255718231201, "learning_rate": 2.166830639494027e-05, "loss": 0.1887, "step": 5142 }, { "epoch": 3.6141953619114546, "grad_norm": 1.550973892211914, "learning_rate": 2.1672522839072382e-05, "loss": 0.266, "step": 5143 }, { "epoch": 3.6148981026001406, "grad_norm": 3.585019588470459, "learning_rate": 2.16767392832045e-05, "loss": 0.3303, "step": 5144 }, { "epoch": 3.6156008432888265, "grad_norm": 0.3660460114479065, "learning_rate": 2.1680955727336612e-05, "loss": 0.1133, "step": 5145 }, { "epoch": 3.6163035839775124, "grad_norm": 0.1983361393213272, "learning_rate": 2.168517217146873e-05, "loss": 0.0368, "step": 5146 }, { "epoch": 3.6170063246661983, "grad_norm": 0.2455582320690155, "learning_rate": 2.1689388615600842e-05, "loss": 0.0366, "step": 5147 }, { "epoch": 3.617709065354884, "grad_norm": 0.24626930058002472, "learning_rate": 2.169360505973296e-05, "loss": 0.0228, "step": 5148 }, { "epoch": 3.6184118060435697, "grad_norm": 0.15181492269039154, "learning_rate": 2.1697821503865072e-05, "loss": 0.02, "step": 5149 }, { "epoch": 3.6191145467322556, "grad_norm": 0.18299488723278046, "learning_rate": 2.170203794799719e-05, "loss": 0.0242, "step": 5150 }, { "epoch": 3.6198172874209416, "grad_norm": 0.24685384333133698, "learning_rate": 2.1706254392129302e-05, "loss": 0.0228, "step": 5151 }, { "epoch": 3.6205200281096275, "grad_norm": 0.20956718921661377, "learning_rate": 2.171047083626142e-05, "loss": 0.0356, "step": 5152 }, { "epoch": 3.6212227687983134, "grad_norm": 0.22386322915554047, "learning_rate": 2.1714687280393535e-05, "loss": 0.0325, "step": 5153 }, { "epoch": 3.6219255094869993, "grad_norm": 0.19480350613594055, "learning_rate": 2.1718903724525652e-05, "loss": 0.0194, "step": 5154 }, { "epoch": 3.6226282501756852, "grad_norm": 0.18869371712207794, "learning_rate": 2.172312016865777e-05, "loss": 0.0259, "step": 5155 }, { "epoch": 3.623330990864371, "grad_norm": 0.17862652242183685, "learning_rate": 2.172733661278988e-05, "loss": 0.0301, "step": 5156 }, { "epoch": 3.624033731553057, "grad_norm": 0.5167891383171082, "learning_rate": 2.1731553056922e-05, "loss": 0.0364, "step": 5157 }, { "epoch": 3.624736472241743, "grad_norm": 0.23154325783252716, "learning_rate": 2.173576950105411e-05, "loss": 0.0258, "step": 5158 }, { "epoch": 3.625439212930429, "grad_norm": 0.23399439454078674, "learning_rate": 2.1739985945186228e-05, "loss": 0.0395, "step": 5159 }, { "epoch": 3.6261419536191144, "grad_norm": 0.27498626708984375, "learning_rate": 2.174420238931834e-05, "loss": 0.034, "step": 5160 }, { "epoch": 3.6268446943078003, "grad_norm": 0.3370746374130249, "learning_rate": 2.1748418833450458e-05, "loss": 0.0426, "step": 5161 }, { "epoch": 3.6275474349964862, "grad_norm": 0.2939833104610443, "learning_rate": 2.175263527758257e-05, "loss": 0.0453, "step": 5162 }, { "epoch": 3.628250175685172, "grad_norm": 0.507665753364563, "learning_rate": 2.1756851721714688e-05, "loss": 0.053, "step": 5163 }, { "epoch": 3.628952916373858, "grad_norm": 0.3340475559234619, "learning_rate": 2.17610681658468e-05, "loss": 0.0802, "step": 5164 }, { "epoch": 3.629655657062544, "grad_norm": 0.46259570121765137, "learning_rate": 2.1765284609978918e-05, "loss": 0.1169, "step": 5165 }, { "epoch": 3.63035839775123, "grad_norm": 0.7015411853790283, "learning_rate": 2.1769501054111034e-05, "loss": 0.1716, "step": 5166 }, { "epoch": 3.6310611384399154, "grad_norm": 1.4842655658721924, "learning_rate": 2.177371749824315e-05, "loss": 0.1998, "step": 5167 }, { "epoch": 3.6317638791286013, "grad_norm": 1.253490686416626, "learning_rate": 2.1777933942375264e-05, "loss": 0.2693, "step": 5168 }, { "epoch": 3.6324666198172872, "grad_norm": 2.4805123805999756, "learning_rate": 2.178215038650738e-05, "loss": 0.3078, "step": 5169 }, { "epoch": 3.633169360505973, "grad_norm": 0.27956756949424744, "learning_rate": 2.1786366830639494e-05, "loss": 0.0936, "step": 5170 }, { "epoch": 3.633872101194659, "grad_norm": 0.20166072249412537, "learning_rate": 2.179058327477161e-05, "loss": 0.0508, "step": 5171 }, { "epoch": 3.634574841883345, "grad_norm": 0.3846874237060547, "learning_rate": 2.1794799718903724e-05, "loss": 0.056, "step": 5172 }, { "epoch": 3.635277582572031, "grad_norm": 0.21454888582229614, "learning_rate": 2.179901616303584e-05, "loss": 0.0273, "step": 5173 }, { "epoch": 3.635980323260717, "grad_norm": 0.18610632419586182, "learning_rate": 2.1803232607167954e-05, "loss": 0.0303, "step": 5174 }, { "epoch": 3.6366830639494028, "grad_norm": 0.1498817354440689, "learning_rate": 2.180744905130007e-05, "loss": 0.0198, "step": 5175 }, { "epoch": 3.6373858046380887, "grad_norm": 0.15263435244560242, "learning_rate": 2.1811665495432184e-05, "loss": 0.0233, "step": 5176 }, { "epoch": 3.6380885453267746, "grad_norm": 0.21104159951210022, "learning_rate": 2.18158819395643e-05, "loss": 0.0334, "step": 5177 }, { "epoch": 3.6387912860154605, "grad_norm": 0.22144654393196106, "learning_rate": 2.1820098383696414e-05, "loss": 0.039, "step": 5178 }, { "epoch": 3.639494026704146, "grad_norm": 0.20818392932415009, "learning_rate": 2.182431482782853e-05, "loss": 0.016, "step": 5179 }, { "epoch": 3.640196767392832, "grad_norm": 0.12987227737903595, "learning_rate": 2.1828531271960647e-05, "loss": 0.0176, "step": 5180 }, { "epoch": 3.640899508081518, "grad_norm": 0.275678813457489, "learning_rate": 2.1832747716092764e-05, "loss": 0.0431, "step": 5181 }, { "epoch": 3.6416022487702038, "grad_norm": 0.2133159637451172, "learning_rate": 2.1836964160224877e-05, "loss": 0.0334, "step": 5182 }, { "epoch": 3.6423049894588897, "grad_norm": 0.21754160523414612, "learning_rate": 2.1841180604356994e-05, "loss": 0.0298, "step": 5183 }, { "epoch": 3.6430077301475756, "grad_norm": 0.21904391050338745, "learning_rate": 2.184539704848911e-05, "loss": 0.0332, "step": 5184 }, { "epoch": 3.6437104708362615, "grad_norm": 0.2517940104007721, "learning_rate": 2.1849613492621224e-05, "loss": 0.0375, "step": 5185 }, { "epoch": 3.6444132115249475, "grad_norm": 0.34337419271469116, "learning_rate": 2.185382993675334e-05, "loss": 0.0415, "step": 5186 }, { "epoch": 3.645115952213633, "grad_norm": 0.22748062014579773, "learning_rate": 2.1858046380885453e-05, "loss": 0.0401, "step": 5187 }, { "epoch": 3.645818692902319, "grad_norm": 0.27081963419914246, "learning_rate": 2.186226282501757e-05, "loss": 0.0598, "step": 5188 }, { "epoch": 3.646521433591005, "grad_norm": 0.3566763699054718, "learning_rate": 2.1866479269149683e-05, "loss": 0.071, "step": 5189 }, { "epoch": 3.6472241742796907, "grad_norm": 0.4274652302265167, "learning_rate": 2.18706957132818e-05, "loss": 0.1082, "step": 5190 }, { "epoch": 3.6479269149683766, "grad_norm": 0.5464333295822144, "learning_rate": 2.1874912157413913e-05, "loss": 0.1366, "step": 5191 }, { "epoch": 3.6486296556570625, "grad_norm": 0.707046389579773, "learning_rate": 2.187912860154603e-05, "loss": 0.2049, "step": 5192 }, { "epoch": 3.6493323963457485, "grad_norm": 0.8568105697631836, "learning_rate": 2.1883345045678146e-05, "loss": 0.2609, "step": 5193 }, { "epoch": 3.6500351370344344, "grad_norm": 0.9518351554870605, "learning_rate": 2.1887561489810263e-05, "loss": 0.2849, "step": 5194 }, { "epoch": 3.6507378777231203, "grad_norm": 0.24580302834510803, "learning_rate": 2.1891777933942376e-05, "loss": 0.0771, "step": 5195 }, { "epoch": 3.6514406184118062, "grad_norm": 0.25869977474212646, "learning_rate": 2.1895994378074493e-05, "loss": 0.0411, "step": 5196 }, { "epoch": 3.652143359100492, "grad_norm": 0.1969788670539856, "learning_rate": 2.1900210822206606e-05, "loss": 0.0343, "step": 5197 }, { "epoch": 3.652846099789178, "grad_norm": 0.2625882923603058, "learning_rate": 2.1904427266338723e-05, "loss": 0.0188, "step": 5198 }, { "epoch": 3.6535488404778635, "grad_norm": 0.14491067826747894, "learning_rate": 2.1908643710470836e-05, "loss": 0.0228, "step": 5199 }, { "epoch": 3.6542515811665495, "grad_norm": 0.1554509401321411, "learning_rate": 2.1912860154602953e-05, "loss": 0.0183, "step": 5200 }, { "epoch": 3.6549543218552354, "grad_norm": 0.25044965744018555, "learning_rate": 2.1917076598735066e-05, "loss": 0.0289, "step": 5201 }, { "epoch": 3.6556570625439213, "grad_norm": 0.19685626029968262, "learning_rate": 2.1921293042867183e-05, "loss": 0.0236, "step": 5202 }, { "epoch": 3.6563598032326072, "grad_norm": 0.22725605964660645, "learning_rate": 2.1925509486999296e-05, "loss": 0.0225, "step": 5203 }, { "epoch": 3.657062543921293, "grad_norm": 0.22452928125858307, "learning_rate": 2.1929725931131413e-05, "loss": 0.0164, "step": 5204 }, { "epoch": 3.657765284609979, "grad_norm": 0.3982760012149811, "learning_rate": 2.1933942375263526e-05, "loss": 0.0499, "step": 5205 }, { "epoch": 3.6584680252986645, "grad_norm": 0.20753978192806244, "learning_rate": 2.1938158819395642e-05, "loss": 0.0227, "step": 5206 }, { "epoch": 3.6591707659873505, "grad_norm": 0.2558160126209259, "learning_rate": 2.194237526352776e-05, "loss": 0.0395, "step": 5207 }, { "epoch": 3.6598735066760364, "grad_norm": 0.22165721654891968, "learning_rate": 2.1946591707659876e-05, "loss": 0.0212, "step": 5208 }, { "epoch": 3.6605762473647223, "grad_norm": 0.32638904452323914, "learning_rate": 2.195080815179199e-05, "loss": 0.0542, "step": 5209 }, { "epoch": 3.6612789880534082, "grad_norm": 0.40339145064353943, "learning_rate": 2.1955024595924106e-05, "loss": 0.0428, "step": 5210 }, { "epoch": 3.661981728742094, "grad_norm": 0.23565639555454254, "learning_rate": 2.195924104005622e-05, "loss": 0.0324, "step": 5211 }, { "epoch": 3.66268446943078, "grad_norm": 0.22792008519172668, "learning_rate": 2.1963457484188335e-05, "loss": 0.0568, "step": 5212 }, { "epoch": 3.663387210119466, "grad_norm": 0.6523830890655518, "learning_rate": 2.1967673928320452e-05, "loss": 0.049, "step": 5213 }, { "epoch": 3.664089950808152, "grad_norm": 0.27003368735313416, "learning_rate": 2.1971890372452565e-05, "loss": 0.0655, "step": 5214 }, { "epoch": 3.664792691496838, "grad_norm": 0.3614289164543152, "learning_rate": 2.1976106816584682e-05, "loss": 0.093, "step": 5215 }, { "epoch": 3.6654954321855238, "grad_norm": 0.4409782886505127, "learning_rate": 2.1980323260716795e-05, "loss": 0.1582, "step": 5216 }, { "epoch": 3.6661981728742097, "grad_norm": 0.8161969184875488, "learning_rate": 2.1984539704848912e-05, "loss": 0.188, "step": 5217 }, { "epoch": 3.666900913562895, "grad_norm": 1.1249969005584717, "learning_rate": 2.1988756148981025e-05, "loss": 0.2641, "step": 5218 }, { "epoch": 3.667603654251581, "grad_norm": 1.4363610744476318, "learning_rate": 2.1992972593113142e-05, "loss": 0.3437, "step": 5219 }, { "epoch": 3.668306394940267, "grad_norm": 0.2539732754230499, "learning_rate": 2.1997189037245255e-05, "loss": 0.093, "step": 5220 }, { "epoch": 3.669009135628953, "grad_norm": 0.18539737164974213, "learning_rate": 2.2001405481377375e-05, "loss": 0.041, "step": 5221 }, { "epoch": 3.669711876317639, "grad_norm": 0.3562217652797699, "learning_rate": 2.2005621925509488e-05, "loss": 0.0515, "step": 5222 }, { "epoch": 3.6704146170063248, "grad_norm": 0.14100250601768494, "learning_rate": 2.2009838369641605e-05, "loss": 0.0241, "step": 5223 }, { "epoch": 3.6711173576950107, "grad_norm": 0.2777837812900543, "learning_rate": 2.2014054813773718e-05, "loss": 0.0303, "step": 5224 }, { "epoch": 3.671820098383696, "grad_norm": 0.8077452778816223, "learning_rate": 2.2018271257905835e-05, "loss": 0.0177, "step": 5225 }, { "epoch": 3.672522839072382, "grad_norm": 0.1626090407371521, "learning_rate": 2.2022487702037948e-05, "loss": 0.0206, "step": 5226 }, { "epoch": 3.673225579761068, "grad_norm": 0.17670951783657074, "learning_rate": 2.2026704146170065e-05, "loss": 0.0269, "step": 5227 }, { "epoch": 3.673928320449754, "grad_norm": 0.33320382237434387, "learning_rate": 2.2030920590302178e-05, "loss": 0.0516, "step": 5228 }, { "epoch": 3.67463106113844, "grad_norm": 0.18733559548854828, "learning_rate": 2.2035137034434295e-05, "loss": 0.0192, "step": 5229 }, { "epoch": 3.6753338018271258, "grad_norm": 0.22460609674453735, "learning_rate": 2.2039353478566408e-05, "loss": 0.0291, "step": 5230 }, { "epoch": 3.6760365425158117, "grad_norm": 0.36786767840385437, "learning_rate": 2.2043569922698524e-05, "loss": 0.0332, "step": 5231 }, { "epoch": 3.6767392832044976, "grad_norm": 0.1726643294095993, "learning_rate": 2.2047786366830638e-05, "loss": 0.0302, "step": 5232 }, { "epoch": 3.6774420238931835, "grad_norm": 0.41537705063819885, "learning_rate": 2.2052002810962754e-05, "loss": 0.0266, "step": 5233 }, { "epoch": 3.6781447645818695, "grad_norm": 0.3212544023990631, "learning_rate": 2.205621925509487e-05, "loss": 0.0481, "step": 5234 }, { "epoch": 3.6788475052705554, "grad_norm": 0.38291135430336, "learning_rate": 2.2060435699226988e-05, "loss": 0.0465, "step": 5235 }, { "epoch": 3.6795502459592413, "grad_norm": 0.35947558283805847, "learning_rate": 2.20646521433591e-05, "loss": 0.0398, "step": 5236 }, { "epoch": 3.6802529866479268, "grad_norm": 0.29804033041000366, "learning_rate": 2.2068868587491218e-05, "loss": 0.0535, "step": 5237 }, { "epoch": 3.6809557273366127, "grad_norm": 0.3748781681060791, "learning_rate": 2.207308503162333e-05, "loss": 0.0752, "step": 5238 }, { "epoch": 3.6816584680252986, "grad_norm": 0.2908816635608673, "learning_rate": 2.2077301475755447e-05, "loss": 0.0499, "step": 5239 }, { "epoch": 3.6823612087139845, "grad_norm": 0.5713027119636536, "learning_rate": 2.208151791988756e-05, "loss": 0.098, "step": 5240 }, { "epoch": 3.6830639494026705, "grad_norm": 0.5312762260437012, "learning_rate": 2.2085734364019677e-05, "loss": 0.1417, "step": 5241 }, { "epoch": 3.6837666900913564, "grad_norm": 0.8612864017486572, "learning_rate": 2.208995080815179e-05, "loss": 0.2613, "step": 5242 }, { "epoch": 3.6844694307800423, "grad_norm": 1.0644559860229492, "learning_rate": 2.2094167252283907e-05, "loss": 0.2853, "step": 5243 }, { "epoch": 3.6851721714687278, "grad_norm": 1.208101749420166, "learning_rate": 2.2098383696416024e-05, "loss": 0.3052, "step": 5244 }, { "epoch": 3.6858749121574137, "grad_norm": 0.3323592245578766, "learning_rate": 2.2102600140548137e-05, "loss": 0.0809, "step": 5245 }, { "epoch": 3.6865776528460996, "grad_norm": 0.2785886526107788, "learning_rate": 2.2106816584680254e-05, "loss": 0.0328, "step": 5246 }, { "epoch": 3.6872803935347855, "grad_norm": 0.30616268515586853, "learning_rate": 2.2111033028812367e-05, "loss": 0.0412, "step": 5247 }, { "epoch": 3.6879831342234715, "grad_norm": 0.22340410947799683, "learning_rate": 2.2115249472944487e-05, "loss": 0.0281, "step": 5248 }, { "epoch": 3.6886858749121574, "grad_norm": 0.18394187092781067, "learning_rate": 2.21194659170766e-05, "loss": 0.0245, "step": 5249 }, { "epoch": 3.6893886156008433, "grad_norm": 0.20772463083267212, "learning_rate": 2.2123682361208717e-05, "loss": 0.0246, "step": 5250 }, { "epoch": 3.6900913562895292, "grad_norm": 0.21138642728328705, "learning_rate": 2.212789880534083e-05, "loss": 0.03, "step": 5251 }, { "epoch": 3.690794096978215, "grad_norm": 0.20376046001911163, "learning_rate": 2.2132115249472947e-05, "loss": 0.0365, "step": 5252 }, { "epoch": 3.691496837666901, "grad_norm": 0.22013118863105774, "learning_rate": 2.213633169360506e-05, "loss": 0.0357, "step": 5253 }, { "epoch": 3.692199578355587, "grad_norm": 0.23364536464214325, "learning_rate": 2.2140548137737177e-05, "loss": 0.0273, "step": 5254 }, { "epoch": 3.692902319044273, "grad_norm": 0.22350522875785828, "learning_rate": 2.214476458186929e-05, "loss": 0.0201, "step": 5255 }, { "epoch": 3.6936050597329584, "grad_norm": 0.2220441848039627, "learning_rate": 2.2148981026001407e-05, "loss": 0.0244, "step": 5256 }, { "epoch": 3.6943078004216443, "grad_norm": 0.3509344160556793, "learning_rate": 2.215319747013352e-05, "loss": 0.0312, "step": 5257 }, { "epoch": 3.6950105411103302, "grad_norm": 0.3689238727092743, "learning_rate": 2.2157413914265636e-05, "loss": 0.0305, "step": 5258 }, { "epoch": 3.695713281799016, "grad_norm": 0.2810169458389282, "learning_rate": 2.216163035839775e-05, "loss": 0.0318, "step": 5259 }, { "epoch": 3.696416022487702, "grad_norm": 0.2852126955986023, "learning_rate": 2.2165846802529866e-05, "loss": 0.0457, "step": 5260 }, { "epoch": 3.697118763176388, "grad_norm": 0.3279528319835663, "learning_rate": 2.2170063246661983e-05, "loss": 0.0338, "step": 5261 }, { "epoch": 3.697821503865074, "grad_norm": 0.2983741760253906, "learning_rate": 2.21742796907941e-05, "loss": 0.0611, "step": 5262 }, { "epoch": 3.6985242445537594, "grad_norm": 0.4387800991535187, "learning_rate": 2.2178496134926213e-05, "loss": 0.0603, "step": 5263 }, { "epoch": 3.6992269852424453, "grad_norm": 1.1231496334075928, "learning_rate": 2.218271257905833e-05, "loss": 0.1045, "step": 5264 }, { "epoch": 3.6999297259311312, "grad_norm": 0.5977063179016113, "learning_rate": 2.2186929023190443e-05, "loss": 0.1052, "step": 5265 }, { "epoch": 3.700632466619817, "grad_norm": 0.596340000629425, "learning_rate": 2.219114546732256e-05, "loss": 0.161, "step": 5266 }, { "epoch": 3.701335207308503, "grad_norm": 0.7114410400390625, "learning_rate": 2.2195361911454673e-05, "loss": 0.2056, "step": 5267 }, { "epoch": 3.702037947997189, "grad_norm": 1.1553226709365845, "learning_rate": 2.219957835558679e-05, "loss": 0.2557, "step": 5268 }, { "epoch": 3.702740688685875, "grad_norm": 3.7691330909729004, "learning_rate": 2.2203794799718902e-05, "loss": 0.3007, "step": 5269 }, { "epoch": 3.703443429374561, "grad_norm": 0.2650023400783539, "learning_rate": 2.220801124385102e-05, "loss": 0.0836, "step": 5270 }, { "epoch": 3.7041461700632468, "grad_norm": 0.32882052659988403, "learning_rate": 2.2212227687983132e-05, "loss": 0.0283, "step": 5271 }, { "epoch": 3.7048489107519327, "grad_norm": 0.27304700016975403, "learning_rate": 2.221644413211525e-05, "loss": 0.0316, "step": 5272 }, { "epoch": 3.7055516514406186, "grad_norm": 0.2762203514575958, "learning_rate": 2.2220660576247366e-05, "loss": 0.022, "step": 5273 }, { "epoch": 3.7062543921293045, "grad_norm": 0.1077122911810875, "learning_rate": 2.222487702037948e-05, "loss": 0.0126, "step": 5274 }, { "epoch": 3.70695713281799, "grad_norm": 0.14695535600185394, "learning_rate": 2.22290934645116e-05, "loss": 0.0254, "step": 5275 }, { "epoch": 3.707659873506676, "grad_norm": 0.21090634167194366, "learning_rate": 2.2233309908643712e-05, "loss": 0.0239, "step": 5276 }, { "epoch": 3.708362614195362, "grad_norm": 0.2522670328617096, "learning_rate": 2.223752635277583e-05, "loss": 0.04, "step": 5277 }, { "epoch": 3.7090653548840478, "grad_norm": 0.41660892963409424, "learning_rate": 2.2241742796907942e-05, "loss": 0.0307, "step": 5278 }, { "epoch": 3.7097680955727337, "grad_norm": 0.18200980126857758, "learning_rate": 2.224595924104006e-05, "loss": 0.0223, "step": 5279 }, { "epoch": 3.7104708362614196, "grad_norm": 0.369926393032074, "learning_rate": 2.2250175685172172e-05, "loss": 0.056, "step": 5280 }, { "epoch": 3.7111735769501055, "grad_norm": 0.16948264837265015, "learning_rate": 2.225439212930429e-05, "loss": 0.0242, "step": 5281 }, { "epoch": 3.7118763176387914, "grad_norm": 0.3080485165119171, "learning_rate": 2.2258608573436402e-05, "loss": 0.0404, "step": 5282 }, { "epoch": 3.712579058327477, "grad_norm": 0.2812427878379822, "learning_rate": 2.226282501756852e-05, "loss": 0.0337, "step": 5283 }, { "epoch": 3.713281799016163, "grad_norm": 0.2686638832092285, "learning_rate": 2.2267041461700632e-05, "loss": 0.059, "step": 5284 }, { "epoch": 3.7139845397048488, "grad_norm": 0.24848149716854095, "learning_rate": 2.227125790583275e-05, "loss": 0.0548, "step": 5285 }, { "epoch": 3.7146872803935347, "grad_norm": 0.23529009521007538, "learning_rate": 2.227547434996486e-05, "loss": 0.034, "step": 5286 }, { "epoch": 3.7153900210822206, "grad_norm": 0.2231675386428833, "learning_rate": 2.2279690794096978e-05, "loss": 0.0373, "step": 5287 }, { "epoch": 3.7160927617709065, "grad_norm": 0.23999401926994324, "learning_rate": 2.228390723822909e-05, "loss": 0.0439, "step": 5288 }, { "epoch": 3.7167955024595924, "grad_norm": 0.3626449406147003, "learning_rate": 2.228812368236121e-05, "loss": 0.0843, "step": 5289 }, { "epoch": 3.7174982431482784, "grad_norm": 0.8681626915931702, "learning_rate": 2.2292340126493325e-05, "loss": 0.0889, "step": 5290 }, { "epoch": 3.7182009838369643, "grad_norm": 0.6702003479003906, "learning_rate": 2.229655657062544e-05, "loss": 0.1555, "step": 5291 }, { "epoch": 3.71890372452565, "grad_norm": 0.6910040378570557, "learning_rate": 2.2300773014757555e-05, "loss": 0.177, "step": 5292 }, { "epoch": 3.719606465214336, "grad_norm": 1.093438744544983, "learning_rate": 2.230498945888967e-05, "loss": 0.2405, "step": 5293 }, { "epoch": 3.720309205903022, "grad_norm": 1.3630062341690063, "learning_rate": 2.2309205903021785e-05, "loss": 0.294, "step": 5294 }, { "epoch": 3.7210119465917075, "grad_norm": 0.24253617227077484, "learning_rate": 2.23134223471539e-05, "loss": 0.07, "step": 5295 }, { "epoch": 3.7217146872803935, "grad_norm": 0.22001199424266815, "learning_rate": 2.2317638791286014e-05, "loss": 0.0425, "step": 5296 }, { "epoch": 3.7224174279690794, "grad_norm": 0.22891508042812347, "learning_rate": 2.232185523541813e-05, "loss": 0.04, "step": 5297 }, { "epoch": 3.7231201686577653, "grad_norm": 0.33067259192466736, "learning_rate": 2.2326071679550244e-05, "loss": 0.0258, "step": 5298 }, { "epoch": 3.723822909346451, "grad_norm": 0.36185139417648315, "learning_rate": 2.233028812368236e-05, "loss": 0.0247, "step": 5299 }, { "epoch": 3.724525650035137, "grad_norm": 0.1450439989566803, "learning_rate": 2.2334504567814474e-05, "loss": 0.0158, "step": 5300 }, { "epoch": 3.725228390723823, "grad_norm": 0.37940099835395813, "learning_rate": 2.233872101194659e-05, "loss": 0.0277, "step": 5301 }, { "epoch": 3.7259311314125085, "grad_norm": 0.20544061064720154, "learning_rate": 2.2342937456078707e-05, "loss": 0.0321, "step": 5302 }, { "epoch": 3.7266338721011945, "grad_norm": 0.29997289180755615, "learning_rate": 2.2347153900210824e-05, "loss": 0.0385, "step": 5303 }, { "epoch": 3.7273366127898804, "grad_norm": 0.20167718827724457, "learning_rate": 2.235137034434294e-05, "loss": 0.0217, "step": 5304 }, { "epoch": 3.7280393534785663, "grad_norm": 0.2401350736618042, "learning_rate": 2.2355586788475054e-05, "loss": 0.0519, "step": 5305 }, { "epoch": 3.728742094167252, "grad_norm": 0.2545068562030792, "learning_rate": 2.235980323260717e-05, "loss": 0.0173, "step": 5306 }, { "epoch": 3.729444834855938, "grad_norm": 0.277039110660553, "learning_rate": 2.2364019676739284e-05, "loss": 0.0432, "step": 5307 }, { "epoch": 3.730147575544624, "grad_norm": 0.230488121509552, "learning_rate": 2.23682361208714e-05, "loss": 0.0172, "step": 5308 }, { "epoch": 3.73085031623331, "grad_norm": 0.2038886696100235, "learning_rate": 2.2372452565003514e-05, "loss": 0.0345, "step": 5309 }, { "epoch": 3.731553056921996, "grad_norm": 0.27271437644958496, "learning_rate": 2.237666900913563e-05, "loss": 0.0672, "step": 5310 }, { "epoch": 3.732255797610682, "grad_norm": 0.3283292353153229, "learning_rate": 2.2380885453267744e-05, "loss": 0.0524, "step": 5311 }, { "epoch": 3.7329585382993677, "grad_norm": 0.2794245183467865, "learning_rate": 2.238510189739986e-05, "loss": 0.0476, "step": 5312 }, { "epoch": 3.7336612789880537, "grad_norm": 0.33764398097991943, "learning_rate": 2.2389318341531974e-05, "loss": 0.0517, "step": 5313 }, { "epoch": 3.734364019676739, "grad_norm": 0.3638205826282501, "learning_rate": 2.239353478566409e-05, "loss": 0.0856, "step": 5314 }, { "epoch": 3.735066760365425, "grad_norm": 0.3653195798397064, "learning_rate": 2.2397751229796203e-05, "loss": 0.0871, "step": 5315 }, { "epoch": 3.735769501054111, "grad_norm": 0.5294289588928223, "learning_rate": 2.2401967673928323e-05, "loss": 0.1607, "step": 5316 }, { "epoch": 3.736472241742797, "grad_norm": 4.691779136657715, "learning_rate": 2.2406184118060437e-05, "loss": 0.2089, "step": 5317 }, { "epoch": 3.737174982431483, "grad_norm": 0.908261775970459, "learning_rate": 2.2410400562192553e-05, "loss": 0.2427, "step": 5318 }, { "epoch": 3.7378777231201687, "grad_norm": 2.184739589691162, "learning_rate": 2.2414617006324667e-05, "loss": 0.296, "step": 5319 }, { "epoch": 3.7385804638088547, "grad_norm": 0.28900378942489624, "learning_rate": 2.2418833450456783e-05, "loss": 0.0744, "step": 5320 }, { "epoch": 3.73928320449754, "grad_norm": 0.23741652071475983, "learning_rate": 2.2423049894588896e-05, "loss": 0.0509, "step": 5321 }, { "epoch": 3.739985945186226, "grad_norm": 0.2627622187137604, "learning_rate": 2.2427266338721013e-05, "loss": 0.0431, "step": 5322 }, { "epoch": 3.740688685874912, "grad_norm": 0.18095727264881134, "learning_rate": 2.2431482782853126e-05, "loss": 0.0231, "step": 5323 }, { "epoch": 3.741391426563598, "grad_norm": 0.14998474717140198, "learning_rate": 2.2435699226985243e-05, "loss": 0.0215, "step": 5324 }, { "epoch": 3.742094167252284, "grad_norm": 0.1909657120704651, "learning_rate": 2.2439915671117356e-05, "loss": 0.0149, "step": 5325 }, { "epoch": 3.7427969079409698, "grad_norm": 0.23990002274513245, "learning_rate": 2.2444132115249473e-05, "loss": 0.0247, "step": 5326 }, { "epoch": 3.7434996486296557, "grad_norm": 0.3402603566646576, "learning_rate": 2.2448348559381586e-05, "loss": 0.0363, "step": 5327 }, { "epoch": 3.7442023893183416, "grad_norm": 0.20066390931606293, "learning_rate": 2.2452565003513703e-05, "loss": 0.0282, "step": 5328 }, { "epoch": 3.7449051300070275, "grad_norm": 0.1595800817012787, "learning_rate": 2.245678144764582e-05, "loss": 0.0181, "step": 5329 }, { "epoch": 3.7456078706957134, "grad_norm": 0.1986156702041626, "learning_rate": 2.2460997891777936e-05, "loss": 0.0324, "step": 5330 }, { "epoch": 3.7463106113843994, "grad_norm": 0.15228085219860077, "learning_rate": 2.246521433591005e-05, "loss": 0.0186, "step": 5331 }, { "epoch": 3.7470133520730853, "grad_norm": 0.20825889706611633, "learning_rate": 2.2469430780042166e-05, "loss": 0.0325, "step": 5332 }, { "epoch": 3.7477160927617708, "grad_norm": 0.2252357006072998, "learning_rate": 2.2473647224174283e-05, "loss": 0.0283, "step": 5333 }, { "epoch": 3.7484188334504567, "grad_norm": 0.28949448466300964, "learning_rate": 2.2477863668306396e-05, "loss": 0.0468, "step": 5334 }, { "epoch": 3.7491215741391426, "grad_norm": 0.30678942799568176, "learning_rate": 2.2482080112438512e-05, "loss": 0.0682, "step": 5335 }, { "epoch": 3.7498243148278285, "grad_norm": 0.2616114020347595, "learning_rate": 2.2486296556570626e-05, "loss": 0.0301, "step": 5336 }, { "epoch": 3.7505270555165144, "grad_norm": 0.28641626238822937, "learning_rate": 2.2490513000702742e-05, "loss": 0.0547, "step": 5337 }, { "epoch": 3.7512297962052004, "grad_norm": 0.34959739446640015, "learning_rate": 2.2494729444834856e-05, "loss": 0.0472, "step": 5338 }, { "epoch": 3.7519325368938863, "grad_norm": 0.3651126027107239, "learning_rate": 2.2498945888966972e-05, "loss": 0.1058, "step": 5339 }, { "epoch": 3.7526352775825718, "grad_norm": 0.483527272939682, "learning_rate": 2.2503162333099086e-05, "loss": 0.087, "step": 5340 }, { "epoch": 3.7533380182712577, "grad_norm": 0.5852931141853333, "learning_rate": 2.2507378777231202e-05, "loss": 0.167, "step": 5341 }, { "epoch": 3.7540407589599436, "grad_norm": 0.7350825667381287, "learning_rate": 2.2511595221363315e-05, "loss": 0.2037, "step": 5342 }, { "epoch": 3.7547434996486295, "grad_norm": 0.833188533782959, "learning_rate": 2.2515811665495435e-05, "loss": 0.2347, "step": 5343 }, { "epoch": 3.7554462403373154, "grad_norm": 1.5488547086715698, "learning_rate": 2.252002810962755e-05, "loss": 0.3013, "step": 5344 }, { "epoch": 3.7561489810260014, "grad_norm": 0.2254878580570221, "learning_rate": 2.2524244553759665e-05, "loss": 0.0773, "step": 5345 }, { "epoch": 3.7568517217146873, "grad_norm": 0.2120024710893631, "learning_rate": 2.252846099789178e-05, "loss": 0.0379, "step": 5346 }, { "epoch": 3.757554462403373, "grad_norm": 0.23359616100788116, "learning_rate": 2.2532677442023895e-05, "loss": 0.0294, "step": 5347 }, { "epoch": 3.758257203092059, "grad_norm": 0.17165569961071014, "learning_rate": 2.253689388615601e-05, "loss": 0.0239, "step": 5348 }, { "epoch": 3.758959943780745, "grad_norm": 0.21269884705543518, "learning_rate": 2.2541110330288125e-05, "loss": 0.0271, "step": 5349 }, { "epoch": 3.759662684469431, "grad_norm": 0.18584571778774261, "learning_rate": 2.254532677442024e-05, "loss": 0.0134, "step": 5350 }, { "epoch": 3.760365425158117, "grad_norm": 0.16336363554000854, "learning_rate": 2.2549543218552355e-05, "loss": 0.0382, "step": 5351 }, { "epoch": 3.7610681658468024, "grad_norm": 0.18199479579925537, "learning_rate": 2.2553759662684468e-05, "loss": 0.0282, "step": 5352 }, { "epoch": 3.7617709065354883, "grad_norm": 0.27301743626594543, "learning_rate": 2.2557976106816585e-05, "loss": 0.0435, "step": 5353 }, { "epoch": 3.762473647224174, "grad_norm": 0.23707301914691925, "learning_rate": 2.2562192550948698e-05, "loss": 0.0433, "step": 5354 }, { "epoch": 3.76317638791286, "grad_norm": 0.21476273238658905, "learning_rate": 2.2566408995080815e-05, "loss": 0.0318, "step": 5355 }, { "epoch": 3.763879128601546, "grad_norm": 0.19190622866153717, "learning_rate": 2.2570625439212928e-05, "loss": 0.0149, "step": 5356 }, { "epoch": 3.764581869290232, "grad_norm": 0.344618558883667, "learning_rate": 2.2574841883345048e-05, "loss": 0.0503, "step": 5357 }, { "epoch": 3.765284609978918, "grad_norm": 0.3138766884803772, "learning_rate": 2.257905832747716e-05, "loss": 0.0373, "step": 5358 }, { "epoch": 3.765987350667604, "grad_norm": 0.33579084277153015, "learning_rate": 2.2583274771609278e-05, "loss": 0.0431, "step": 5359 }, { "epoch": 3.7666900913562893, "grad_norm": 0.33202147483825684, "learning_rate": 2.258749121574139e-05, "loss": 0.0493, "step": 5360 }, { "epoch": 3.767392832044975, "grad_norm": 0.23544932901859283, "learning_rate": 2.2591707659873508e-05, "loss": 0.0409, "step": 5361 }, { "epoch": 3.768095572733661, "grad_norm": 0.29994717240333557, "learning_rate": 2.2595924104005624e-05, "loss": 0.0524, "step": 5362 }, { "epoch": 3.768798313422347, "grad_norm": 0.3458114564418793, "learning_rate": 2.2600140548137738e-05, "loss": 0.0561, "step": 5363 }, { "epoch": 3.769501054111033, "grad_norm": 0.34447911381721497, "learning_rate": 2.2604356992269854e-05, "loss": 0.0742, "step": 5364 }, { "epoch": 3.770203794799719, "grad_norm": 0.44489073753356934, "learning_rate": 2.2608573436401968e-05, "loss": 0.0881, "step": 5365 }, { "epoch": 3.770906535488405, "grad_norm": 0.623231828212738, "learning_rate": 2.2612789880534084e-05, "loss": 0.1296, "step": 5366 }, { "epoch": 3.7716092761770907, "grad_norm": 0.6847935318946838, "learning_rate": 2.2617006324666197e-05, "loss": 0.1825, "step": 5367 }, { "epoch": 3.7723120168657767, "grad_norm": 1.0509109497070312, "learning_rate": 2.2621222768798314e-05, "loss": 0.2807, "step": 5368 }, { "epoch": 3.7730147575544626, "grad_norm": 4.376037120819092, "learning_rate": 2.2625439212930427e-05, "loss": 0.3398, "step": 5369 }, { "epoch": 3.7737174982431485, "grad_norm": 0.3096683621406555, "learning_rate": 2.2629655657062547e-05, "loss": 0.0868, "step": 5370 }, { "epoch": 3.7744202389318344, "grad_norm": 0.17387637495994568, "learning_rate": 2.263387210119466e-05, "loss": 0.0266, "step": 5371 }, { "epoch": 3.77512297962052, "grad_norm": 0.14166544377803802, "learning_rate": 2.2638088545326777e-05, "loss": 0.0269, "step": 5372 }, { "epoch": 3.775825720309206, "grad_norm": 0.19292275607585907, "learning_rate": 2.264230498945889e-05, "loss": 0.0258, "step": 5373 }, { "epoch": 3.7765284609978917, "grad_norm": 0.28968745470046997, "learning_rate": 2.2646521433591007e-05, "loss": 0.0379, "step": 5374 }, { "epoch": 3.7772312016865777, "grad_norm": 0.13523022830486298, "learning_rate": 2.265073787772312e-05, "loss": 0.0179, "step": 5375 }, { "epoch": 3.7779339423752636, "grad_norm": 0.2043144255876541, "learning_rate": 2.2654954321855237e-05, "loss": 0.0299, "step": 5376 }, { "epoch": 3.7786366830639495, "grad_norm": 0.21024960279464722, "learning_rate": 2.265917076598735e-05, "loss": 0.0341, "step": 5377 }, { "epoch": 3.7793394237526354, "grad_norm": 0.232976034283638, "learning_rate": 2.2663387210119467e-05, "loss": 0.0382, "step": 5378 }, { "epoch": 3.780042164441321, "grad_norm": 0.1266322284936905, "learning_rate": 2.266760365425158e-05, "loss": 0.0168, "step": 5379 }, { "epoch": 3.780744905130007, "grad_norm": 0.21456950902938843, "learning_rate": 2.2671820098383697e-05, "loss": 0.0333, "step": 5380 }, { "epoch": 3.7814476458186927, "grad_norm": 0.3537557125091553, "learning_rate": 2.267603654251581e-05, "loss": 0.0368, "step": 5381 }, { "epoch": 3.7821503865073787, "grad_norm": 0.2699544131755829, "learning_rate": 2.2680252986647927e-05, "loss": 0.0525, "step": 5382 }, { "epoch": 3.7828531271960646, "grad_norm": 0.19758108258247375, "learning_rate": 2.268446943078004e-05, "loss": 0.0246, "step": 5383 }, { "epoch": 3.7835558678847505, "grad_norm": 0.18664687871932983, "learning_rate": 2.268868587491216e-05, "loss": 0.0269, "step": 5384 }, { "epoch": 3.7842586085734364, "grad_norm": 0.23974548280239105, "learning_rate": 2.2692902319044273e-05, "loss": 0.0481, "step": 5385 }, { "epoch": 3.7849613492621224, "grad_norm": 0.23853714764118195, "learning_rate": 2.269711876317639e-05, "loss": 0.0389, "step": 5386 }, { "epoch": 3.7856640899508083, "grad_norm": 0.16500157117843628, "learning_rate": 2.2701335207308503e-05, "loss": 0.0327, "step": 5387 }, { "epoch": 3.786366830639494, "grad_norm": 0.24917490780353546, "learning_rate": 2.270555165144062e-05, "loss": 0.058, "step": 5388 }, { "epoch": 3.78706957132818, "grad_norm": 0.4486284554004669, "learning_rate": 2.2709768095572733e-05, "loss": 0.071, "step": 5389 }, { "epoch": 3.787772312016866, "grad_norm": 0.42946240305900574, "learning_rate": 2.271398453970485e-05, "loss": 0.0835, "step": 5390 }, { "epoch": 3.7884750527055515, "grad_norm": 0.5832136273384094, "learning_rate": 2.2718200983836963e-05, "loss": 0.1403, "step": 5391 }, { "epoch": 3.7891777933942374, "grad_norm": 0.5644333362579346, "learning_rate": 2.272241742796908e-05, "loss": 0.1741, "step": 5392 }, { "epoch": 3.7898805340829234, "grad_norm": 1.1602733135223389, "learning_rate": 2.2726633872101196e-05, "loss": 0.2441, "step": 5393 }, { "epoch": 3.7905832747716093, "grad_norm": 1.8683383464813232, "learning_rate": 2.273085031623331e-05, "loss": 0.3879, "step": 5394 }, { "epoch": 3.791286015460295, "grad_norm": 0.2807663083076477, "learning_rate": 2.2735066760365426e-05, "loss": 0.0873, "step": 5395 }, { "epoch": 3.791988756148981, "grad_norm": 0.17705470323562622, "learning_rate": 2.273928320449754e-05, "loss": 0.04, "step": 5396 }, { "epoch": 3.792691496837667, "grad_norm": 0.198775514960289, "learning_rate": 2.2743499648629656e-05, "loss": 0.0342, "step": 5397 }, { "epoch": 3.7933942375263525, "grad_norm": 0.2847193777561188, "learning_rate": 2.2747716092761773e-05, "loss": 0.0238, "step": 5398 }, { "epoch": 3.7940969782150384, "grad_norm": 0.18557865917682648, "learning_rate": 2.275193253689389e-05, "loss": 0.0231, "step": 5399 }, { "epoch": 3.7947997189037244, "grad_norm": 0.18087808787822723, "learning_rate": 2.2756148981026002e-05, "loss": 0.0202, "step": 5400 }, { "epoch": 3.7955024595924103, "grad_norm": 0.16027069091796875, "learning_rate": 2.276036542515812e-05, "loss": 0.0343, "step": 5401 }, { "epoch": 3.796205200281096, "grad_norm": 0.19208787381649017, "learning_rate": 2.2764581869290232e-05, "loss": 0.0171, "step": 5402 }, { "epoch": 3.796907940969782, "grad_norm": 0.14529027044773102, "learning_rate": 2.276879831342235e-05, "loss": 0.0197, "step": 5403 }, { "epoch": 3.797610681658468, "grad_norm": 0.1966599076986313, "learning_rate": 2.2773014757554462e-05, "loss": 0.0268, "step": 5404 }, { "epoch": 3.798313422347154, "grad_norm": 0.4414336681365967, "learning_rate": 2.277723120168658e-05, "loss": 0.0521, "step": 5405 }, { "epoch": 3.79901616303584, "grad_norm": 0.19029474258422852, "learning_rate": 2.2781447645818692e-05, "loss": 0.0185, "step": 5406 }, { "epoch": 3.799718903724526, "grad_norm": 0.1866113841533661, "learning_rate": 2.278566408995081e-05, "loss": 0.0324, "step": 5407 }, { "epoch": 3.8004216444132117, "grad_norm": 0.33949196338653564, "learning_rate": 2.2789880534082922e-05, "loss": 0.0416, "step": 5408 }, { "epoch": 3.8011243851018977, "grad_norm": 0.23349671065807343, "learning_rate": 2.279409697821504e-05, "loss": 0.0442, "step": 5409 }, { "epoch": 3.801827125790583, "grad_norm": 0.5612595677375793, "learning_rate": 2.2798313422347152e-05, "loss": 0.0597, "step": 5410 }, { "epoch": 3.802529866479269, "grad_norm": 0.2758747935295105, "learning_rate": 2.2802529866479272e-05, "loss": 0.034, "step": 5411 }, { "epoch": 3.803232607167955, "grad_norm": 0.36783772706985474, "learning_rate": 2.2806746310611385e-05, "loss": 0.0504, "step": 5412 }, { "epoch": 3.803935347856641, "grad_norm": 0.3047374188899994, "learning_rate": 2.2810962754743502e-05, "loss": 0.0707, "step": 5413 }, { "epoch": 3.804638088545327, "grad_norm": 0.3779538571834564, "learning_rate": 2.2815179198875615e-05, "loss": 0.0765, "step": 5414 }, { "epoch": 3.8053408292340127, "grad_norm": 0.4952682852745056, "learning_rate": 2.281939564300773e-05, "loss": 0.1094, "step": 5415 }, { "epoch": 3.8060435699226987, "grad_norm": 0.74312824010849, "learning_rate": 2.2823612087139845e-05, "loss": 0.1471, "step": 5416 }, { "epoch": 3.806746310611384, "grad_norm": 0.7981728911399841, "learning_rate": 2.282782853127196e-05, "loss": 0.1917, "step": 5417 }, { "epoch": 3.80744905130007, "grad_norm": 1.533632516860962, "learning_rate": 2.2832044975404075e-05, "loss": 0.2355, "step": 5418 }, { "epoch": 3.808151791988756, "grad_norm": 1.9526195526123047, "learning_rate": 2.283626141953619e-05, "loss": 0.314, "step": 5419 }, { "epoch": 3.808854532677442, "grad_norm": 0.5614333748817444, "learning_rate": 2.2840477863668305e-05, "loss": 0.1289, "step": 5420 }, { "epoch": 3.809557273366128, "grad_norm": 0.23672296106815338, "learning_rate": 2.284469430780042e-05, "loss": 0.0309, "step": 5421 }, { "epoch": 3.8102600140548137, "grad_norm": 0.27944985032081604, "learning_rate": 2.2848910751932538e-05, "loss": 0.0301, "step": 5422 }, { "epoch": 3.8109627547434997, "grad_norm": 0.17716015875339508, "learning_rate": 2.285312719606465e-05, "loss": 0.0282, "step": 5423 }, { "epoch": 3.8116654954321856, "grad_norm": 0.2625274062156677, "learning_rate": 2.2857343640196768e-05, "loss": 0.0491, "step": 5424 }, { "epoch": 3.8123682361208715, "grad_norm": 0.19851231575012207, "learning_rate": 2.2861560084328885e-05, "loss": 0.0195, "step": 5425 }, { "epoch": 3.8130709768095574, "grad_norm": 0.2000732719898224, "learning_rate": 2.2865776528461e-05, "loss": 0.0296, "step": 5426 }, { "epoch": 3.8137737174982433, "grad_norm": 0.20743228495121002, "learning_rate": 2.2869992972593114e-05, "loss": 0.0297, "step": 5427 }, { "epoch": 3.8144764581869293, "grad_norm": 0.15095356106758118, "learning_rate": 2.287420941672523e-05, "loss": 0.0311, "step": 5428 }, { "epoch": 3.8151791988756147, "grad_norm": 0.2507191300392151, "learning_rate": 2.2878425860857344e-05, "loss": 0.023, "step": 5429 }, { "epoch": 3.8158819395643007, "grad_norm": 0.3522813618183136, "learning_rate": 2.288264230498946e-05, "loss": 0.0445, "step": 5430 }, { "epoch": 3.8165846802529866, "grad_norm": 0.20644380152225494, "learning_rate": 2.2886858749121574e-05, "loss": 0.0277, "step": 5431 }, { "epoch": 3.8172874209416725, "grad_norm": 0.479446142911911, "learning_rate": 2.289107519325369e-05, "loss": 0.0436, "step": 5432 }, { "epoch": 3.8179901616303584, "grad_norm": 0.2138461470603943, "learning_rate": 2.2895291637385804e-05, "loss": 0.0243, "step": 5433 }, { "epoch": 3.8186929023190443, "grad_norm": 0.23565459251403809, "learning_rate": 2.289950808151792e-05, "loss": 0.0399, "step": 5434 }, { "epoch": 3.8193956430077303, "grad_norm": 0.25548064708709717, "learning_rate": 2.2903724525650034e-05, "loss": 0.0401, "step": 5435 }, { "epoch": 3.820098383696416, "grad_norm": 0.22588105499744415, "learning_rate": 2.290794096978215e-05, "loss": 0.0335, "step": 5436 }, { "epoch": 3.8208011243851017, "grad_norm": 0.401957243680954, "learning_rate": 2.2912157413914264e-05, "loss": 0.0687, "step": 5437 }, { "epoch": 3.8215038650737876, "grad_norm": 0.4987495541572571, "learning_rate": 2.2916373858046384e-05, "loss": 0.0504, "step": 5438 }, { "epoch": 3.8222066057624735, "grad_norm": 0.47230181097984314, "learning_rate": 2.2920590302178497e-05, "loss": 0.0964, "step": 5439 }, { "epoch": 3.8229093464511594, "grad_norm": 0.5369580984115601, "learning_rate": 2.2924806746310614e-05, "loss": 0.1101, "step": 5440 }, { "epoch": 3.8236120871398454, "grad_norm": 0.6878937482833862, "learning_rate": 2.2929023190442727e-05, "loss": 0.156, "step": 5441 }, { "epoch": 3.8243148278285313, "grad_norm": 0.9382386207580566, "learning_rate": 2.2933239634574844e-05, "loss": 0.2329, "step": 5442 }, { "epoch": 3.825017568517217, "grad_norm": 1.5364481210708618, "learning_rate": 2.2937456078706957e-05, "loss": 0.2365, "step": 5443 }, { "epoch": 3.825720309205903, "grad_norm": 1.7630913257598877, "learning_rate": 2.2941672522839074e-05, "loss": 0.3247, "step": 5444 }, { "epoch": 3.826423049894589, "grad_norm": 0.3231941759586334, "learning_rate": 2.2945888966971187e-05, "loss": 0.0846, "step": 5445 }, { "epoch": 3.827125790583275, "grad_norm": 0.2768932282924652, "learning_rate": 2.2950105411103303e-05, "loss": 0.0245, "step": 5446 }, { "epoch": 3.827828531271961, "grad_norm": 0.2076907604932785, "learning_rate": 2.2954321855235417e-05, "loss": 0.0349, "step": 5447 }, { "epoch": 3.828531271960647, "grad_norm": 0.20646411180496216, "learning_rate": 2.2958538299367533e-05, "loss": 0.0285, "step": 5448 }, { "epoch": 3.8292340126493323, "grad_norm": 0.19273625314235687, "learning_rate": 2.2962754743499647e-05, "loss": 0.0261, "step": 5449 }, { "epoch": 3.829936753338018, "grad_norm": 0.19318009912967682, "learning_rate": 2.2966971187631763e-05, "loss": 0.0179, "step": 5450 }, { "epoch": 3.830639494026704, "grad_norm": 0.2944647967815399, "learning_rate": 2.297118763176388e-05, "loss": 0.0293, "step": 5451 }, { "epoch": 3.83134223471539, "grad_norm": 0.19939754903316498, "learning_rate": 2.2975404075895996e-05, "loss": 0.0337, "step": 5452 }, { "epoch": 3.832044975404076, "grad_norm": 0.22424380481243134, "learning_rate": 2.2979620520028113e-05, "loss": 0.0306, "step": 5453 }, { "epoch": 3.832747716092762, "grad_norm": 0.19995737075805664, "learning_rate": 2.2983836964160226e-05, "loss": 0.0202, "step": 5454 }, { "epoch": 3.833450456781448, "grad_norm": 0.24937167763710022, "learning_rate": 2.2988053408292343e-05, "loss": 0.0387, "step": 5455 }, { "epoch": 3.8341531974701333, "grad_norm": 0.20146284997463226, "learning_rate": 2.2992269852424456e-05, "loss": 0.0189, "step": 5456 }, { "epoch": 3.834855938158819, "grad_norm": 0.2182566225528717, "learning_rate": 2.2996486296556573e-05, "loss": 0.0193, "step": 5457 }, { "epoch": 3.835558678847505, "grad_norm": 0.39839494228363037, "learning_rate": 2.3000702740688686e-05, "loss": 0.0297, "step": 5458 }, { "epoch": 3.836261419536191, "grad_norm": 0.2865401804447174, "learning_rate": 2.3004919184820803e-05, "loss": 0.0389, "step": 5459 }, { "epoch": 3.836964160224877, "grad_norm": 0.33653444051742554, "learning_rate": 2.3009135628952916e-05, "loss": 0.0674, "step": 5460 }, { "epoch": 3.837666900913563, "grad_norm": 0.19215139746665955, "learning_rate": 2.3013352073085033e-05, "loss": 0.0306, "step": 5461 }, { "epoch": 3.838369641602249, "grad_norm": 0.24809083342552185, "learning_rate": 2.3017568517217146e-05, "loss": 0.0477, "step": 5462 }, { "epoch": 3.8390723822909347, "grad_norm": 0.29083821177482605, "learning_rate": 2.3021784961349263e-05, "loss": 0.0366, "step": 5463 }, { "epoch": 3.8397751229796206, "grad_norm": 0.48780444264411926, "learning_rate": 2.3026001405481376e-05, "loss": 0.0778, "step": 5464 }, { "epoch": 3.8404778636683066, "grad_norm": 0.5911641120910645, "learning_rate": 2.3030217849613492e-05, "loss": 0.1054, "step": 5465 }, { "epoch": 3.8411806043569925, "grad_norm": 0.5046696662902832, "learning_rate": 2.303443429374561e-05, "loss": 0.1171, "step": 5466 }, { "epoch": 3.8418833450456784, "grad_norm": 1.0832455158233643, "learning_rate": 2.3038650737877726e-05, "loss": 0.2249, "step": 5467 }, { "epoch": 3.842586085734364, "grad_norm": 0.9410293102264404, "learning_rate": 2.304286718200984e-05, "loss": 0.2133, "step": 5468 }, { "epoch": 3.84328882642305, "grad_norm": 2.9671058654785156, "learning_rate": 2.3047083626141956e-05, "loss": 0.3146, "step": 5469 }, { "epoch": 3.8439915671117357, "grad_norm": 0.3464115262031555, "learning_rate": 2.305130007027407e-05, "loss": 0.0923, "step": 5470 }, { "epoch": 3.8446943078004217, "grad_norm": 0.23416242003440857, "learning_rate": 2.3055516514406185e-05, "loss": 0.0358, "step": 5471 }, { "epoch": 3.8453970484891076, "grad_norm": 0.28202158212661743, "learning_rate": 2.30597329585383e-05, "loss": 0.0317, "step": 5472 }, { "epoch": 3.8460997891777935, "grad_norm": 0.15844617784023285, "learning_rate": 2.3063949402670415e-05, "loss": 0.0266, "step": 5473 }, { "epoch": 3.8468025298664794, "grad_norm": 0.2061835080385208, "learning_rate": 2.306816584680253e-05, "loss": 0.0203, "step": 5474 }, { "epoch": 3.847505270555165, "grad_norm": 0.16832570731639862, "learning_rate": 2.3072382290934645e-05, "loss": 0.0168, "step": 5475 }, { "epoch": 3.848208011243851, "grad_norm": 0.17806577682495117, "learning_rate": 2.307659873506676e-05, "loss": 0.0216, "step": 5476 }, { "epoch": 3.8489107519325367, "grad_norm": 0.14324042201042175, "learning_rate": 2.3080815179198875e-05, "loss": 0.0228, "step": 5477 }, { "epoch": 3.8496134926212227, "grad_norm": 0.26031193137168884, "learning_rate": 2.308503162333099e-05, "loss": 0.0355, "step": 5478 }, { "epoch": 3.8503162333099086, "grad_norm": 0.2502703368663788, "learning_rate": 2.308924806746311e-05, "loss": 0.0211, "step": 5479 }, { "epoch": 3.8510189739985945, "grad_norm": 0.33695223927497864, "learning_rate": 2.309346451159522e-05, "loss": 0.0297, "step": 5480 }, { "epoch": 3.8517217146872804, "grad_norm": 0.18769821524620056, "learning_rate": 2.3097680955727338e-05, "loss": 0.0344, "step": 5481 }, { "epoch": 3.8524244553759663, "grad_norm": 0.27770349383354187, "learning_rate": 2.3101897399859455e-05, "loss": 0.0465, "step": 5482 }, { "epoch": 3.8531271960646523, "grad_norm": 0.17073681950569153, "learning_rate": 2.3106113843991568e-05, "loss": 0.0255, "step": 5483 }, { "epoch": 3.853829936753338, "grad_norm": 0.20479589700698853, "learning_rate": 2.3110330288123685e-05, "loss": 0.0292, "step": 5484 }, { "epoch": 3.854532677442024, "grad_norm": 0.3193359971046448, "learning_rate": 2.3114546732255798e-05, "loss": 0.042, "step": 5485 }, { "epoch": 3.85523541813071, "grad_norm": 0.2047632932662964, "learning_rate": 2.3118763176387915e-05, "loss": 0.033, "step": 5486 }, { "epoch": 3.8559381588193955, "grad_norm": 0.39569178223609924, "learning_rate": 2.3122979620520028e-05, "loss": 0.0703, "step": 5487 }, { "epoch": 3.8566408995080814, "grad_norm": 0.3255167603492737, "learning_rate": 2.3127196064652145e-05, "loss": 0.0784, "step": 5488 }, { "epoch": 3.8573436401967673, "grad_norm": 0.351468950510025, "learning_rate": 2.3131412508784258e-05, "loss": 0.057, "step": 5489 }, { "epoch": 3.8580463808854533, "grad_norm": 0.3760177493095398, "learning_rate": 2.3135628952916374e-05, "loss": 0.0913, "step": 5490 }, { "epoch": 3.858749121574139, "grad_norm": 0.6555721759796143, "learning_rate": 2.3139845397048488e-05, "loss": 0.1492, "step": 5491 }, { "epoch": 3.859451862262825, "grad_norm": 0.8202782273292542, "learning_rate": 2.3144061841180604e-05, "loss": 0.1844, "step": 5492 }, { "epoch": 3.860154602951511, "grad_norm": 1.1005629301071167, "learning_rate": 2.314827828531272e-05, "loss": 0.2675, "step": 5493 }, { "epoch": 3.8608573436401965, "grad_norm": 1.070753812789917, "learning_rate": 2.3152494729444838e-05, "loss": 0.2564, "step": 5494 }, { "epoch": 3.8615600843288824, "grad_norm": 0.2723650336265564, "learning_rate": 2.315671117357695e-05, "loss": 0.0993, "step": 5495 }, { "epoch": 3.8622628250175683, "grad_norm": 0.2609221637248993, "learning_rate": 2.3160927617709068e-05, "loss": 0.0424, "step": 5496 }, { "epoch": 3.8629655657062543, "grad_norm": 0.2090911567211151, "learning_rate": 2.316514406184118e-05, "loss": 0.021, "step": 5497 }, { "epoch": 3.86366830639494, "grad_norm": 0.22513240575790405, "learning_rate": 2.3169360505973297e-05, "loss": 0.0253, "step": 5498 }, { "epoch": 3.864371047083626, "grad_norm": 0.15241840481758118, "learning_rate": 2.317357695010541e-05, "loss": 0.0251, "step": 5499 }, { "epoch": 3.865073787772312, "grad_norm": 0.3427753746509552, "learning_rate": 2.3177793394237527e-05, "loss": 0.0281, "step": 5500 }, { "epoch": 3.865776528460998, "grad_norm": 0.27169954776763916, "learning_rate": 2.318200983836964e-05, "loss": 0.0305, "step": 5501 }, { "epoch": 3.866479269149684, "grad_norm": 0.1729104369878769, "learning_rate": 2.3186226282501757e-05, "loss": 0.0288, "step": 5502 }, { "epoch": 3.86718200983837, "grad_norm": 0.21419422328472137, "learning_rate": 2.319044272663387e-05, "loss": 0.0401, "step": 5503 }, { "epoch": 3.8678847505270557, "grad_norm": 0.18035274744033813, "learning_rate": 2.3194659170765987e-05, "loss": 0.0181, "step": 5504 }, { "epoch": 3.8685874912157416, "grad_norm": 0.19724339246749878, "learning_rate": 2.31988756148981e-05, "loss": 0.0342, "step": 5505 }, { "epoch": 3.869290231904427, "grad_norm": 0.48664864897727966, "learning_rate": 2.320309205903022e-05, "loss": 0.0247, "step": 5506 }, { "epoch": 3.869992972593113, "grad_norm": 0.35833534598350525, "learning_rate": 2.3207308503162334e-05, "loss": 0.0305, "step": 5507 }, { "epoch": 3.870695713281799, "grad_norm": 0.3246305286884308, "learning_rate": 2.321152494729445e-05, "loss": 0.0194, "step": 5508 }, { "epoch": 3.871398453970485, "grad_norm": 0.30557170510292053, "learning_rate": 2.3215741391426563e-05, "loss": 0.0586, "step": 5509 }, { "epoch": 3.872101194659171, "grad_norm": 0.2373645156621933, "learning_rate": 2.321995783555868e-05, "loss": 0.0476, "step": 5510 }, { "epoch": 3.8728039353478567, "grad_norm": 0.2297508716583252, "learning_rate": 2.3224174279690797e-05, "loss": 0.028, "step": 5511 }, { "epoch": 3.8735066760365426, "grad_norm": 0.2187754362821579, "learning_rate": 2.322839072382291e-05, "loss": 0.0571, "step": 5512 }, { "epoch": 3.874209416725228, "grad_norm": 0.494881272315979, "learning_rate": 2.3232607167955027e-05, "loss": 0.0709, "step": 5513 }, { "epoch": 3.874912157413914, "grad_norm": 0.31391188502311707, "learning_rate": 2.323682361208714e-05, "loss": 0.0501, "step": 5514 }, { "epoch": 3.8756148981026, "grad_norm": 0.5196157693862915, "learning_rate": 2.3241040056219257e-05, "loss": 0.0966, "step": 5515 }, { "epoch": 3.876317638791286, "grad_norm": 0.6986380219459534, "learning_rate": 2.324525650035137e-05, "loss": 0.1656, "step": 5516 }, { "epoch": 3.877020379479972, "grad_norm": 1.1772911548614502, "learning_rate": 2.3249472944483486e-05, "loss": 0.2053, "step": 5517 }, { "epoch": 3.8777231201686577, "grad_norm": 1.828576922416687, "learning_rate": 2.32536893886156e-05, "loss": 0.2458, "step": 5518 }, { "epoch": 3.8784258608573436, "grad_norm": 2.0854170322418213, "learning_rate": 2.3257905832747716e-05, "loss": 0.3085, "step": 5519 }, { "epoch": 3.8791286015460296, "grad_norm": 0.3684026300907135, "learning_rate": 2.3262122276879833e-05, "loss": 0.0845, "step": 5520 }, { "epoch": 3.8798313422347155, "grad_norm": 0.18952852487564087, "learning_rate": 2.326633872101195e-05, "loss": 0.0354, "step": 5521 }, { "epoch": 3.8805340829234014, "grad_norm": 0.20248860120773315, "learning_rate": 2.3270555165144063e-05, "loss": 0.0559, "step": 5522 }, { "epoch": 3.8812368236120873, "grad_norm": 0.22562247514724731, "learning_rate": 2.327477160927618e-05, "loss": 0.0434, "step": 5523 }, { "epoch": 3.8819395643007732, "grad_norm": 0.23826903104782104, "learning_rate": 2.3278988053408293e-05, "loss": 0.0339, "step": 5524 }, { "epoch": 3.8826423049894587, "grad_norm": 0.20016634464263916, "learning_rate": 2.328320449754041e-05, "loss": 0.0233, "step": 5525 }, { "epoch": 3.8833450456781446, "grad_norm": 0.21495629847049713, "learning_rate": 2.3287420941672523e-05, "loss": 0.0323, "step": 5526 }, { "epoch": 3.8840477863668306, "grad_norm": 0.1511337012052536, "learning_rate": 2.329163738580464e-05, "loss": 0.0197, "step": 5527 }, { "epoch": 3.8847505270555165, "grad_norm": 0.2053236961364746, "learning_rate": 2.3295853829936753e-05, "loss": 0.0414, "step": 5528 }, { "epoch": 3.8854532677442024, "grad_norm": 0.20574449002742767, "learning_rate": 2.330007027406887e-05, "loss": 0.0176, "step": 5529 }, { "epoch": 3.8861560084328883, "grad_norm": 0.18871068954467773, "learning_rate": 2.3304286718200982e-05, "loss": 0.0223, "step": 5530 }, { "epoch": 3.8868587491215743, "grad_norm": 0.2673763036727905, "learning_rate": 2.33085031623331e-05, "loss": 0.0226, "step": 5531 }, { "epoch": 3.88756148981026, "grad_norm": 0.22596007585525513, "learning_rate": 2.3312719606465212e-05, "loss": 0.04, "step": 5532 }, { "epoch": 3.8882642304989457, "grad_norm": 0.3572591543197632, "learning_rate": 2.331693605059733e-05, "loss": 0.0308, "step": 5533 }, { "epoch": 3.8889669711876316, "grad_norm": 0.29402151703834534, "learning_rate": 2.3321152494729446e-05, "loss": 0.0458, "step": 5534 }, { "epoch": 3.8896697118763175, "grad_norm": 0.2689647972583771, "learning_rate": 2.3325368938861562e-05, "loss": 0.0479, "step": 5535 }, { "epoch": 3.8903724525650034, "grad_norm": 0.37348324060440063, "learning_rate": 2.3329585382993675e-05, "loss": 0.0307, "step": 5536 }, { "epoch": 3.8910751932536893, "grad_norm": 0.3924490809440613, "learning_rate": 2.3333801827125792e-05, "loss": 0.0788, "step": 5537 }, { "epoch": 3.8917779339423753, "grad_norm": 0.6329536437988281, "learning_rate": 2.3338018271257905e-05, "loss": 0.0444, "step": 5538 }, { "epoch": 3.892480674631061, "grad_norm": 0.4914315342903137, "learning_rate": 2.3342234715390022e-05, "loss": 0.0729, "step": 5539 }, { "epoch": 3.893183415319747, "grad_norm": 0.5291758179664612, "learning_rate": 2.3346451159522135e-05, "loss": 0.1128, "step": 5540 }, { "epoch": 3.893886156008433, "grad_norm": 0.5017212629318237, "learning_rate": 2.3350667603654252e-05, "loss": 0.1522, "step": 5541 }, { "epoch": 3.894588896697119, "grad_norm": 1.9645888805389404, "learning_rate": 2.335488404778637e-05, "loss": 0.19, "step": 5542 }, { "epoch": 3.895291637385805, "grad_norm": 1.3115849494934082, "learning_rate": 2.3359100491918482e-05, "loss": 0.2551, "step": 5543 }, { "epoch": 3.895994378074491, "grad_norm": 1.0024363994598389, "learning_rate": 2.33633169360506e-05, "loss": 0.3202, "step": 5544 }, { "epoch": 3.8966971187631763, "grad_norm": 0.2695102393627167, "learning_rate": 2.336753338018271e-05, "loss": 0.0758, "step": 5545 }, { "epoch": 3.897399859451862, "grad_norm": 0.23811663687229156, "learning_rate": 2.3371749824314828e-05, "loss": 0.0376, "step": 5546 }, { "epoch": 3.898102600140548, "grad_norm": 0.21089868247509003, "learning_rate": 2.3375966268446945e-05, "loss": 0.0606, "step": 5547 }, { "epoch": 3.898805340829234, "grad_norm": 0.18912695348262787, "learning_rate": 2.338018271257906e-05, "loss": 0.0274, "step": 5548 }, { "epoch": 3.89950808151792, "grad_norm": 0.1762237250804901, "learning_rate": 2.3384399156711175e-05, "loss": 0.0395, "step": 5549 }, { "epoch": 3.900210822206606, "grad_norm": 0.17857563495635986, "learning_rate": 2.338861560084329e-05, "loss": 0.0221, "step": 5550 }, { "epoch": 3.900913562895292, "grad_norm": 0.23069962859153748, "learning_rate": 2.3392832044975405e-05, "loss": 0.0206, "step": 5551 }, { "epoch": 3.9016163035839773, "grad_norm": 0.16441747546195984, "learning_rate": 2.339704848910752e-05, "loss": 0.0356, "step": 5552 }, { "epoch": 3.902319044272663, "grad_norm": 0.3123507499694824, "learning_rate": 2.3401264933239635e-05, "loss": 0.0359, "step": 5553 }, { "epoch": 3.903021784961349, "grad_norm": 0.21934039890766144, "learning_rate": 2.340548137737175e-05, "loss": 0.0354, "step": 5554 }, { "epoch": 3.903724525650035, "grad_norm": 0.23133742809295654, "learning_rate": 2.3409697821503864e-05, "loss": 0.0354, "step": 5555 }, { "epoch": 3.904427266338721, "grad_norm": 0.1715240478515625, "learning_rate": 2.341391426563598e-05, "loss": 0.0228, "step": 5556 }, { "epoch": 3.905130007027407, "grad_norm": 0.22369180619716644, "learning_rate": 2.3418130709768094e-05, "loss": 0.0552, "step": 5557 }, { "epoch": 3.905832747716093, "grad_norm": 0.1928553730249405, "learning_rate": 2.342234715390021e-05, "loss": 0.0281, "step": 5558 }, { "epoch": 3.9065354884047787, "grad_norm": 0.22417250275611877, "learning_rate": 2.3426563598032324e-05, "loss": 0.0338, "step": 5559 }, { "epoch": 3.9072382290934646, "grad_norm": 0.43764910101890564, "learning_rate": 2.343078004216444e-05, "loss": 0.05, "step": 5560 }, { "epoch": 3.9079409697821506, "grad_norm": 0.22915184497833252, "learning_rate": 2.3434996486296558e-05, "loss": 0.0204, "step": 5561 }, { "epoch": 3.9086437104708365, "grad_norm": 0.35913243889808655, "learning_rate": 2.3439212930428674e-05, "loss": 0.0564, "step": 5562 }, { "epoch": 3.9093464511595224, "grad_norm": 0.3367360532283783, "learning_rate": 2.3443429374560787e-05, "loss": 0.0593, "step": 5563 }, { "epoch": 3.910049191848208, "grad_norm": 0.30754944682121277, "learning_rate": 2.3447645818692904e-05, "loss": 0.0624, "step": 5564 }, { "epoch": 3.910751932536894, "grad_norm": 0.4062662422657013, "learning_rate": 2.3451862262825017e-05, "loss": 0.089, "step": 5565 }, { "epoch": 3.9114546732255797, "grad_norm": 0.5107595920562744, "learning_rate": 2.3456078706957134e-05, "loss": 0.1482, "step": 5566 }, { "epoch": 3.9121574139142656, "grad_norm": 1.7645946741104126, "learning_rate": 2.3460295151089247e-05, "loss": 0.22, "step": 5567 }, { "epoch": 3.9128601546029516, "grad_norm": 1.2859909534454346, "learning_rate": 2.3464511595221364e-05, "loss": 0.2127, "step": 5568 }, { "epoch": 3.9135628952916375, "grad_norm": 1.552087426185608, "learning_rate": 2.3468728039353477e-05, "loss": 0.3168, "step": 5569 }, { "epoch": 3.9142656359803234, "grad_norm": 0.3758400082588196, "learning_rate": 2.3472944483485594e-05, "loss": 0.1037, "step": 5570 }, { "epoch": 3.914968376669009, "grad_norm": 0.17149890959262848, "learning_rate": 2.347716092761771e-05, "loss": 0.0412, "step": 5571 }, { "epoch": 3.915671117357695, "grad_norm": 0.2580203711986542, "learning_rate": 2.3481377371749824e-05, "loss": 0.044, "step": 5572 }, { "epoch": 3.9163738580463807, "grad_norm": 0.1828463077545166, "learning_rate": 2.348559381588194e-05, "loss": 0.028, "step": 5573 }, { "epoch": 3.9170765987350666, "grad_norm": 0.2560315430164337, "learning_rate": 2.3489810260014057e-05, "loss": 0.0247, "step": 5574 }, { "epoch": 3.9177793394237526, "grad_norm": 0.17676250636577606, "learning_rate": 2.3494026704146173e-05, "loss": 0.0251, "step": 5575 }, { "epoch": 3.9184820801124385, "grad_norm": 0.12750385701656342, "learning_rate": 2.3498243148278287e-05, "loss": 0.0116, "step": 5576 }, { "epoch": 3.9191848208011244, "grad_norm": 0.26905450224876404, "learning_rate": 2.3502459592410403e-05, "loss": 0.0318, "step": 5577 }, { "epoch": 3.9198875614898103, "grad_norm": 0.262381374835968, "learning_rate": 2.3506676036542517e-05, "loss": 0.0301, "step": 5578 }, { "epoch": 3.9205903021784962, "grad_norm": 0.23956198990345, "learning_rate": 2.3510892480674633e-05, "loss": 0.0241, "step": 5579 }, { "epoch": 3.921293042867182, "grad_norm": 0.31683045625686646, "learning_rate": 2.3515108924806747e-05, "loss": 0.0337, "step": 5580 }, { "epoch": 3.921995783555868, "grad_norm": 0.22948701679706573, "learning_rate": 2.3519325368938863e-05, "loss": 0.0298, "step": 5581 }, { "epoch": 3.922698524244554, "grad_norm": 0.1983276754617691, "learning_rate": 2.3523541813070976e-05, "loss": 0.0335, "step": 5582 }, { "epoch": 3.9234012649332395, "grad_norm": 0.16641534864902496, "learning_rate": 2.3527758257203093e-05, "loss": 0.0271, "step": 5583 }, { "epoch": 3.9241040056219254, "grad_norm": 0.3734586834907532, "learning_rate": 2.3531974701335206e-05, "loss": 0.043, "step": 5584 }, { "epoch": 3.9248067463106113, "grad_norm": 0.4725429117679596, "learning_rate": 2.3536191145467323e-05, "loss": 0.0763, "step": 5585 }, { "epoch": 3.9255094869992972, "grad_norm": 0.1783856898546219, "learning_rate": 2.3540407589599436e-05, "loss": 0.0282, "step": 5586 }, { "epoch": 3.926212227687983, "grad_norm": 0.36598441004753113, "learning_rate": 2.3544624033731553e-05, "loss": 0.0635, "step": 5587 }, { "epoch": 3.926914968376669, "grad_norm": 0.26028120517730713, "learning_rate": 2.354884047786367e-05, "loss": 0.0603, "step": 5588 }, { "epoch": 3.927617709065355, "grad_norm": 0.32177314162254333, "learning_rate": 2.3553056921995786e-05, "loss": 0.1045, "step": 5589 }, { "epoch": 3.9283204497540405, "grad_norm": 0.4355942904949188, "learning_rate": 2.35572733661279e-05, "loss": 0.0807, "step": 5590 }, { "epoch": 3.9290231904427264, "grad_norm": 0.5970174074172974, "learning_rate": 2.3561489810260016e-05, "loss": 0.1476, "step": 5591 }, { "epoch": 3.9297259311314123, "grad_norm": 1.1003901958465576, "learning_rate": 2.356570625439213e-05, "loss": 0.2218, "step": 5592 }, { "epoch": 3.9304286718200983, "grad_norm": 1.1164631843566895, "learning_rate": 2.3569922698524246e-05, "loss": 0.2547, "step": 5593 }, { "epoch": 3.931131412508784, "grad_norm": 1.755752444267273, "learning_rate": 2.357413914265636e-05, "loss": 0.3825, "step": 5594 }, { "epoch": 3.93183415319747, "grad_norm": 0.7705883979797363, "learning_rate": 2.3578355586788476e-05, "loss": 0.0821, "step": 5595 }, { "epoch": 3.932536893886156, "grad_norm": 0.1946948915719986, "learning_rate": 2.358257203092059e-05, "loss": 0.0279, "step": 5596 }, { "epoch": 3.933239634574842, "grad_norm": 0.22567018866539001, "learning_rate": 2.3586788475052706e-05, "loss": 0.0394, "step": 5597 }, { "epoch": 3.933942375263528, "grad_norm": 0.13779258728027344, "learning_rate": 2.359100491918482e-05, "loss": 0.027, "step": 5598 }, { "epoch": 3.934645115952214, "grad_norm": 0.18409350514411926, "learning_rate": 2.3595221363316936e-05, "loss": 0.018, "step": 5599 }, { "epoch": 3.9353478566408997, "grad_norm": 0.1426829695701599, "learning_rate": 2.3599437807449052e-05, "loss": 0.0309, "step": 5600 }, { "epoch": 3.9360505973295856, "grad_norm": 0.21060116589069366, "learning_rate": 2.3603654251581165e-05, "loss": 0.0259, "step": 5601 }, { "epoch": 3.936753338018271, "grad_norm": 0.14808130264282227, "learning_rate": 2.3607870695713285e-05, "loss": 0.0199, "step": 5602 }, { "epoch": 3.937456078706957, "grad_norm": 0.3291078209877014, "learning_rate": 2.36120871398454e-05, "loss": 0.0228, "step": 5603 }, { "epoch": 3.938158819395643, "grad_norm": 0.14212395250797272, "learning_rate": 2.3616303583977515e-05, "loss": 0.0156, "step": 5604 }, { "epoch": 3.938861560084329, "grad_norm": 0.28169959783554077, "learning_rate": 2.362052002810963e-05, "loss": 0.0574, "step": 5605 }, { "epoch": 3.939564300773015, "grad_norm": 0.20713305473327637, "learning_rate": 2.3624736472241745e-05, "loss": 0.0233, "step": 5606 }, { "epoch": 3.9402670414617007, "grad_norm": 0.2235964834690094, "learning_rate": 2.362895291637386e-05, "loss": 0.0364, "step": 5607 }, { "epoch": 3.9409697821503866, "grad_norm": 0.22353516519069672, "learning_rate": 2.3633169360505975e-05, "loss": 0.0506, "step": 5608 }, { "epoch": 3.9416725228390725, "grad_norm": 0.21137066185474396, "learning_rate": 2.363738580463809e-05, "loss": 0.028, "step": 5609 }, { "epoch": 3.942375263527758, "grad_norm": 0.25236594676971436, "learning_rate": 2.3641602248770205e-05, "loss": 0.0356, "step": 5610 }, { "epoch": 3.943078004216444, "grad_norm": 0.2737858295440674, "learning_rate": 2.3645818692902318e-05, "loss": 0.0319, "step": 5611 }, { "epoch": 3.94378074490513, "grad_norm": 0.3000853955745697, "learning_rate": 2.3650035137034435e-05, "loss": 0.0443, "step": 5612 }, { "epoch": 3.944483485593816, "grad_norm": 0.30738845467567444, "learning_rate": 2.3654251581166548e-05, "loss": 0.0531, "step": 5613 }, { "epoch": 3.9451862262825017, "grad_norm": 0.451745867729187, "learning_rate": 2.3658468025298665e-05, "loss": 0.0505, "step": 5614 }, { "epoch": 3.9458889669711876, "grad_norm": 0.5938675403594971, "learning_rate": 2.366268446943078e-05, "loss": 0.0956, "step": 5615 }, { "epoch": 3.9465917076598735, "grad_norm": 0.6415624618530273, "learning_rate": 2.3666900913562898e-05, "loss": 0.1481, "step": 5616 }, { "epoch": 3.9472944483485595, "grad_norm": 1.0965895652770996, "learning_rate": 2.367111735769501e-05, "loss": 0.1941, "step": 5617 }, { "epoch": 3.9479971890372454, "grad_norm": 1.0194783210754395, "learning_rate": 2.3675333801827128e-05, "loss": 0.2784, "step": 5618 }, { "epoch": 3.9486999297259313, "grad_norm": 2.829785108566284, "learning_rate": 2.367955024595924e-05, "loss": 0.3056, "step": 5619 }, { "epoch": 3.9494026704146172, "grad_norm": 0.38475802540779114, "learning_rate": 2.3683766690091358e-05, "loss": 0.0882, "step": 5620 }, { "epoch": 3.950105411103303, "grad_norm": 0.23481346666812897, "learning_rate": 2.368798313422347e-05, "loss": 0.0629, "step": 5621 }, { "epoch": 3.9508081517919886, "grad_norm": 0.24625782668590546, "learning_rate": 2.3692199578355588e-05, "loss": 0.0287, "step": 5622 }, { "epoch": 3.9515108924806746, "grad_norm": 0.1844538152217865, "learning_rate": 2.36964160224877e-05, "loss": 0.03, "step": 5623 }, { "epoch": 3.9522136331693605, "grad_norm": 0.22825011610984802, "learning_rate": 2.3700632466619818e-05, "loss": 0.0249, "step": 5624 }, { "epoch": 3.9529163738580464, "grad_norm": 0.13687723875045776, "learning_rate": 2.370484891075193e-05, "loss": 0.0209, "step": 5625 }, { "epoch": 3.9536191145467323, "grad_norm": 0.09043224900960922, "learning_rate": 2.3709065354884047e-05, "loss": 0.0141, "step": 5626 }, { "epoch": 3.9543218552354182, "grad_norm": 0.14619456231594086, "learning_rate": 2.371328179901616e-05, "loss": 0.0218, "step": 5627 }, { "epoch": 3.955024595924104, "grad_norm": 0.16749665141105652, "learning_rate": 2.3717498243148277e-05, "loss": 0.024, "step": 5628 }, { "epoch": 3.9557273366127896, "grad_norm": 0.1587647944688797, "learning_rate": 2.3721714687280394e-05, "loss": 0.0182, "step": 5629 }, { "epoch": 3.9564300773014756, "grad_norm": 0.13859982788562775, "learning_rate": 2.372593113141251e-05, "loss": 0.0309, "step": 5630 }, { "epoch": 3.9571328179901615, "grad_norm": 0.2750456631183624, "learning_rate": 2.3730147575544627e-05, "loss": 0.0216, "step": 5631 }, { "epoch": 3.9578355586788474, "grad_norm": 0.4234717786312103, "learning_rate": 2.373436401967674e-05, "loss": 0.0467, "step": 5632 }, { "epoch": 3.9585382993675333, "grad_norm": 0.22970882058143616, "learning_rate": 2.3738580463808857e-05, "loss": 0.0494, "step": 5633 }, { "epoch": 3.9592410400562192, "grad_norm": 0.36106136441230774, "learning_rate": 2.374279690794097e-05, "loss": 0.0316, "step": 5634 }, { "epoch": 3.959943780744905, "grad_norm": 0.3177037537097931, "learning_rate": 2.3747013352073087e-05, "loss": 0.034, "step": 5635 }, { "epoch": 3.960646521433591, "grad_norm": 0.3133501410484314, "learning_rate": 2.37512297962052e-05, "loss": 0.0349, "step": 5636 }, { "epoch": 3.961349262122277, "grad_norm": 0.23869578540325165, "learning_rate": 2.3755446240337317e-05, "loss": 0.0532, "step": 5637 }, { "epoch": 3.962052002810963, "grad_norm": 0.20226353406906128, "learning_rate": 2.375966268446943e-05, "loss": 0.0352, "step": 5638 }, { "epoch": 3.962754743499649, "grad_norm": 0.3152768015861511, "learning_rate": 2.3763879128601547e-05, "loss": 0.0469, "step": 5639 }, { "epoch": 3.9634574841883348, "grad_norm": 0.4521509110927582, "learning_rate": 2.376809557273366e-05, "loss": 0.084, "step": 5640 }, { "epoch": 3.9641602248770202, "grad_norm": 0.6805924773216248, "learning_rate": 2.3772312016865777e-05, "loss": 0.1299, "step": 5641 }, { "epoch": 3.964862965565706, "grad_norm": 5.745996952056885, "learning_rate": 2.3776528460997893e-05, "loss": 0.2187, "step": 5642 }, { "epoch": 3.965565706254392, "grad_norm": 1.8833423852920532, "learning_rate": 2.378074490513001e-05, "loss": 0.2665, "step": 5643 }, { "epoch": 3.966268446943078, "grad_norm": 1.5503379106521606, "learning_rate": 2.3784961349262123e-05, "loss": 0.3478, "step": 5644 }, { "epoch": 3.966971187631764, "grad_norm": 0.4990382194519043, "learning_rate": 2.378917779339424e-05, "loss": 0.1029, "step": 5645 }, { "epoch": 3.96767392832045, "grad_norm": 0.16400954127311707, "learning_rate": 2.3793394237526353e-05, "loss": 0.0246, "step": 5646 }, { "epoch": 3.9683766690091358, "grad_norm": 0.19265642762184143, "learning_rate": 2.379761068165847e-05, "loss": 0.0347, "step": 5647 }, { "epoch": 3.9690794096978212, "grad_norm": 0.2313363403081894, "learning_rate": 2.3801827125790583e-05, "loss": 0.0348, "step": 5648 }, { "epoch": 3.969782150386507, "grad_norm": 0.30783721804618835, "learning_rate": 2.38060435699227e-05, "loss": 0.0368, "step": 5649 }, { "epoch": 3.970484891075193, "grad_norm": 0.16336867213249207, "learning_rate": 2.3810260014054813e-05, "loss": 0.0213, "step": 5650 }, { "epoch": 3.971187631763879, "grad_norm": 0.19165048003196716, "learning_rate": 2.381447645818693e-05, "loss": 0.0287, "step": 5651 }, { "epoch": 3.971890372452565, "grad_norm": 0.16788288950920105, "learning_rate": 2.3818692902319043e-05, "loss": 0.0238, "step": 5652 }, { "epoch": 3.972593113141251, "grad_norm": 0.43844130635261536, "learning_rate": 2.382290934645116e-05, "loss": 0.04, "step": 5653 }, { "epoch": 3.9732958538299368, "grad_norm": 0.18592913448810577, "learning_rate": 2.3827125790583273e-05, "loss": 0.0155, "step": 5654 }, { "epoch": 3.9739985945186227, "grad_norm": 0.18751682341098785, "learning_rate": 2.383134223471539e-05, "loss": 0.0249, "step": 5655 }, { "epoch": 3.9747013352073086, "grad_norm": 0.2660641372203827, "learning_rate": 2.3835558678847506e-05, "loss": 0.0281, "step": 5656 }, { "epoch": 3.9754040758959945, "grad_norm": 0.22924546897411346, "learning_rate": 2.3839775122979623e-05, "loss": 0.0363, "step": 5657 }, { "epoch": 3.9761068165846805, "grad_norm": 0.24927546083927155, "learning_rate": 2.3843991567111736e-05, "loss": 0.0298, "step": 5658 }, { "epoch": 3.9768095572733664, "grad_norm": 0.2948470413684845, "learning_rate": 2.3848208011243852e-05, "loss": 0.0552, "step": 5659 }, { "epoch": 3.977512297962052, "grad_norm": 0.19099387526512146, "learning_rate": 2.385242445537597e-05, "loss": 0.0374, "step": 5660 }, { "epoch": 3.978215038650738, "grad_norm": 0.30476418137550354, "learning_rate": 2.3856640899508082e-05, "loss": 0.0296, "step": 5661 }, { "epoch": 3.9789177793394237, "grad_norm": 0.2530515193939209, "learning_rate": 2.38608573436402e-05, "loss": 0.0345, "step": 5662 }, { "epoch": 3.9796205200281096, "grad_norm": 0.4880456328392029, "learning_rate": 2.3865073787772312e-05, "loss": 0.0801, "step": 5663 }, { "epoch": 3.9803232607167955, "grad_norm": 0.5975717306137085, "learning_rate": 2.386929023190443e-05, "loss": 0.105, "step": 5664 }, { "epoch": 3.9810260014054815, "grad_norm": 1.7488958835601807, "learning_rate": 2.3873506676036542e-05, "loss": 0.0916, "step": 5665 }, { "epoch": 3.9817287420941674, "grad_norm": 0.9742645621299744, "learning_rate": 2.387772312016866e-05, "loss": 0.1311, "step": 5666 }, { "epoch": 3.982431482782853, "grad_norm": 0.6934530138969421, "learning_rate": 2.3881939564300772e-05, "loss": 0.1916, "step": 5667 }, { "epoch": 3.983134223471539, "grad_norm": 1.5272337198257446, "learning_rate": 2.388615600843289e-05, "loss": 0.2982, "step": 5668 }, { "epoch": 3.9838369641602247, "grad_norm": 1.5825265645980835, "learning_rate": 2.3890372452565002e-05, "loss": 0.3117, "step": 5669 }, { "epoch": 3.9845397048489106, "grad_norm": 0.33664166927337646, "learning_rate": 2.3894588896697122e-05, "loss": 0.0747, "step": 5670 }, { "epoch": 3.9852424455375965, "grad_norm": 0.33920833468437195, "learning_rate": 2.3898805340829235e-05, "loss": 0.0289, "step": 5671 }, { "epoch": 3.9859451862262825, "grad_norm": 0.2687256336212158, "learning_rate": 2.3903021784961352e-05, "loss": 0.0299, "step": 5672 }, { "epoch": 3.9866479269149684, "grad_norm": 0.20163650810718536, "learning_rate": 2.3907238229093465e-05, "loss": 0.0228, "step": 5673 }, { "epoch": 3.9873506676036543, "grad_norm": 0.13547739386558533, "learning_rate": 2.3911454673225582e-05, "loss": 0.0156, "step": 5674 }, { "epoch": 3.9880534082923402, "grad_norm": 0.2794354259967804, "learning_rate": 2.3915671117357695e-05, "loss": 0.0348, "step": 5675 }, { "epoch": 3.988756148981026, "grad_norm": 0.20100873708724976, "learning_rate": 2.391988756148981e-05, "loss": 0.033, "step": 5676 }, { "epoch": 3.989458889669712, "grad_norm": 0.2679266035556793, "learning_rate": 2.3924104005621925e-05, "loss": 0.0466, "step": 5677 }, { "epoch": 3.990161630358398, "grad_norm": 0.2068939507007599, "learning_rate": 2.392832044975404e-05, "loss": 0.0356, "step": 5678 }, { "epoch": 3.9908643710470835, "grad_norm": 0.23751524090766907, "learning_rate": 2.3932536893886155e-05, "loss": 0.0142, "step": 5679 }, { "epoch": 3.9915671117357694, "grad_norm": 0.23514927923679352, "learning_rate": 2.393675333801827e-05, "loss": 0.0465, "step": 5680 }, { "epoch": 3.9922698524244553, "grad_norm": 0.25971654057502747, "learning_rate": 2.3940969782150385e-05, "loss": 0.0439, "step": 5681 }, { "epoch": 3.9929725931131412, "grad_norm": 0.34153836965560913, "learning_rate": 2.39451862262825e-05, "loss": 0.0169, "step": 5682 }, { "epoch": 3.993675333801827, "grad_norm": 0.2963320016860962, "learning_rate": 2.3949402670414618e-05, "loss": 0.0484, "step": 5683 }, { "epoch": 3.994378074490513, "grad_norm": 0.2685246765613556, "learning_rate": 2.3953619114546735e-05, "loss": 0.0288, "step": 5684 }, { "epoch": 3.995080815179199, "grad_norm": 0.3980046212673187, "learning_rate": 2.3957835558678848e-05, "loss": 0.041, "step": 5685 }, { "epoch": 3.9957835558678845, "grad_norm": 1.779954195022583, "learning_rate": 2.3962052002810964e-05, "loss": 0.0753, "step": 5686 }, { "epoch": 3.9964862965565704, "grad_norm": 0.3780330419540405, "learning_rate": 2.3966268446943078e-05, "loss": 0.0613, "step": 5687 }, { "epoch": 3.9971890372452563, "grad_norm": 0.6000464558601379, "learning_rate": 2.3970484891075194e-05, "loss": 0.106, "step": 5688 }, { "epoch": 3.9978917779339422, "grad_norm": 0.7052240371704102, "learning_rate": 2.3974701335207308e-05, "loss": 0.1856, "step": 5689 }, { "epoch": 3.998594518622628, "grad_norm": 0.8494587540626526, "learning_rate": 2.3978917779339424e-05, "loss": 0.2003, "step": 5690 }, { "epoch": 3.999297259311314, "grad_norm": 1.1972899436950684, "learning_rate": 2.398313422347154e-05, "loss": 0.257, "step": 5691 }, { "epoch": 4.0, "grad_norm": 0.8322726488113403, "learning_rate": 2.3987350667603654e-05, "loss": 0.2201, "step": 5692 }, { "epoch": 4.000702740688686, "grad_norm": 0.3011798858642578, "learning_rate": 2.399156711173577e-05, "loss": 0.1194, "step": 5693 }, { "epoch": 4.001405481377372, "grad_norm": 0.2811771631240845, "learning_rate": 2.3995783555867884e-05, "loss": 0.0351, "step": 5694 }, { "epoch": 4.002108222066058, "grad_norm": 0.20208610594272614, "learning_rate": 2.4e-05, "loss": 0.0308, "step": 5695 }, { "epoch": 4.002810962754744, "grad_norm": 0.24961726367473602, "learning_rate": 2.4004216444132114e-05, "loss": 0.0267, "step": 5696 }, { "epoch": 4.00351370344343, "grad_norm": 0.2319362461566925, "learning_rate": 2.4008432888264234e-05, "loss": 0.0378, "step": 5697 }, { "epoch": 4.0042164441321155, "grad_norm": 0.14786484837532043, "learning_rate": 2.4012649332396347e-05, "loss": 0.0154, "step": 5698 }, { "epoch": 4.0049191848208014, "grad_norm": 0.1752682328224182, "learning_rate": 2.4016865776528464e-05, "loss": 0.0218, "step": 5699 }, { "epoch": 4.005621925509487, "grad_norm": 0.23611848056316376, "learning_rate": 2.4021082220660577e-05, "loss": 0.0237, "step": 5700 }, { "epoch": 4.006324666198173, "grad_norm": 0.33474087715148926, "learning_rate": 2.4025298664792694e-05, "loss": 0.0529, "step": 5701 }, { "epoch": 4.007027406886858, "grad_norm": 0.25139686465263367, "learning_rate": 2.4029515108924807e-05, "loss": 0.0185, "step": 5702 }, { "epoch": 4.007730147575544, "grad_norm": 0.2636444568634033, "learning_rate": 2.4033731553056924e-05, "loss": 0.027, "step": 5703 }, { "epoch": 4.00843288826423, "grad_norm": 0.20765770971775055, "learning_rate": 2.4037947997189037e-05, "loss": 0.0221, "step": 5704 }, { "epoch": 4.009135628952916, "grad_norm": 0.30850905179977417, "learning_rate": 2.4042164441321153e-05, "loss": 0.0573, "step": 5705 }, { "epoch": 4.009838369641602, "grad_norm": 0.2707047760486603, "learning_rate": 2.4046380885453267e-05, "loss": 0.0255, "step": 5706 }, { "epoch": 4.010541110330288, "grad_norm": 0.335714727640152, "learning_rate": 2.4050597329585383e-05, "loss": 0.0459, "step": 5707 }, { "epoch": 4.011243851018974, "grad_norm": 0.27301204204559326, "learning_rate": 2.4054813773717497e-05, "loss": 0.0308, "step": 5708 }, { "epoch": 4.01194659170766, "grad_norm": 0.18507854640483856, "learning_rate": 2.4059030217849613e-05, "loss": 0.0249, "step": 5709 }, { "epoch": 4.012649332396346, "grad_norm": 0.29122716188430786, "learning_rate": 2.406324666198173e-05, "loss": 0.0518, "step": 5710 }, { "epoch": 4.013352073085032, "grad_norm": 0.26051729917526245, "learning_rate": 2.4067463106113846e-05, "loss": 0.0509, "step": 5711 }, { "epoch": 4.0140548137737175, "grad_norm": 0.34060660004615784, "learning_rate": 2.407167955024596e-05, "loss": 0.0687, "step": 5712 }, { "epoch": 4.0147575544624035, "grad_norm": 0.6871298551559448, "learning_rate": 2.4075895994378076e-05, "loss": 0.0918, "step": 5713 }, { "epoch": 4.015460295151089, "grad_norm": 0.5908634066581726, "learning_rate": 2.408011243851019e-05, "loss": 0.1077, "step": 5714 }, { "epoch": 4.016163035839775, "grad_norm": 0.6040441393852234, "learning_rate": 2.4084328882642306e-05, "loss": 0.1698, "step": 5715 }, { "epoch": 4.016865776528461, "grad_norm": 0.8708987236022949, "learning_rate": 2.408854532677442e-05, "loss": 0.231, "step": 5716 }, { "epoch": 4.017568517217147, "grad_norm": 1.211300253868103, "learning_rate": 2.4092761770906536e-05, "loss": 0.2759, "step": 5717 }, { "epoch": 4.018271257905833, "grad_norm": 0.747405469417572, "learning_rate": 2.409697821503865e-05, "loss": 0.0997, "step": 5718 }, { "epoch": 4.018973998594519, "grad_norm": 0.23777303099632263, "learning_rate": 2.4101194659170766e-05, "loss": 0.0473, "step": 5719 }, { "epoch": 4.019676739283205, "grad_norm": 0.18655472993850708, "learning_rate": 2.4105411103302883e-05, "loss": 0.0268, "step": 5720 }, { "epoch": 4.02037947997189, "grad_norm": 0.16082070767879486, "learning_rate": 2.4109627547434996e-05, "loss": 0.0251, "step": 5721 }, { "epoch": 4.021082220660576, "grad_norm": 0.1678813397884369, "learning_rate": 2.4113843991567113e-05, "loss": 0.0249, "step": 5722 }, { "epoch": 4.021784961349262, "grad_norm": 0.17894549667835236, "learning_rate": 2.4118060435699226e-05, "loss": 0.0171, "step": 5723 }, { "epoch": 4.022487702037948, "grad_norm": 0.18323376774787903, "learning_rate": 2.4122276879831346e-05, "loss": 0.018, "step": 5724 }, { "epoch": 4.023190442726634, "grad_norm": 0.2271697223186493, "learning_rate": 2.412649332396346e-05, "loss": 0.0409, "step": 5725 }, { "epoch": 4.0238931834153195, "grad_norm": 0.1739700734615326, "learning_rate": 2.4130709768095576e-05, "loss": 0.0142, "step": 5726 }, { "epoch": 4.0245959241040055, "grad_norm": 0.2405744343996048, "learning_rate": 2.413492621222769e-05, "loss": 0.0186, "step": 5727 }, { "epoch": 4.025298664792691, "grad_norm": 0.21662817895412445, "learning_rate": 2.4139142656359806e-05, "loss": 0.0257, "step": 5728 }, { "epoch": 4.026001405481377, "grad_norm": 0.231779545545578, "learning_rate": 2.414335910049192e-05, "loss": 0.0383, "step": 5729 }, { "epoch": 4.026704146170063, "grad_norm": 0.3879590630531311, "learning_rate": 2.4147575544624035e-05, "loss": 0.0579, "step": 5730 }, { "epoch": 4.027406886858749, "grad_norm": 0.23210594058036804, "learning_rate": 2.415179198875615e-05, "loss": 0.0314, "step": 5731 }, { "epoch": 4.028109627547435, "grad_norm": 0.33598390221595764, "learning_rate": 2.4156008432888265e-05, "loss": 0.0313, "step": 5732 }, { "epoch": 4.028812368236121, "grad_norm": 0.21158179640769958, "learning_rate": 2.416022487702038e-05, "loss": 0.0506, "step": 5733 }, { "epoch": 4.029515108924807, "grad_norm": 0.2512006461620331, "learning_rate": 2.4164441321152495e-05, "loss": 0.0395, "step": 5734 }, { "epoch": 4.030217849613493, "grad_norm": 0.43939197063446045, "learning_rate": 2.416865776528461e-05, "loss": 0.0567, "step": 5735 }, { "epoch": 4.030920590302179, "grad_norm": 0.3544171452522278, "learning_rate": 2.4172874209416725e-05, "loss": 0.0499, "step": 5736 }, { "epoch": 4.031623330990865, "grad_norm": 0.45996642112731934, "learning_rate": 2.417709065354884e-05, "loss": 0.0646, "step": 5737 }, { "epoch": 4.032326071679551, "grad_norm": 0.44403746724128723, "learning_rate": 2.418130709768096e-05, "loss": 0.1157, "step": 5738 }, { "epoch": 4.0330288123682365, "grad_norm": 0.5272992849349976, "learning_rate": 2.418552354181307e-05, "loss": 0.1276, "step": 5739 }, { "epoch": 4.033731553056922, "grad_norm": 0.8007686138153076, "learning_rate": 2.418973998594519e-05, "loss": 0.1961, "step": 5740 }, { "epoch": 4.0344342937456075, "grad_norm": 1.1886340379714966, "learning_rate": 2.41939564300773e-05, "loss": 0.2273, "step": 5741 }, { "epoch": 4.035137034434293, "grad_norm": 1.8897294998168945, "learning_rate": 2.4198172874209418e-05, "loss": 0.2696, "step": 5742 }, { "epoch": 4.035839775122979, "grad_norm": 0.3379676043987274, "learning_rate": 2.420238931834153e-05, "loss": 0.1039, "step": 5743 }, { "epoch": 4.036542515811665, "grad_norm": 0.23075419664382935, "learning_rate": 2.4206605762473648e-05, "loss": 0.0323, "step": 5744 }, { "epoch": 4.037245256500351, "grad_norm": 0.16036635637283325, "learning_rate": 2.421082220660576e-05, "loss": 0.0323, "step": 5745 }, { "epoch": 4.037947997189037, "grad_norm": 0.1798030287027359, "learning_rate": 2.4215038650737878e-05, "loss": 0.0264, "step": 5746 }, { "epoch": 4.038650737877723, "grad_norm": 0.13940992951393127, "learning_rate": 2.421925509486999e-05, "loss": 0.0261, "step": 5747 }, { "epoch": 4.039353478566409, "grad_norm": 0.2315889596939087, "learning_rate": 2.4223471539002108e-05, "loss": 0.0282, "step": 5748 }, { "epoch": 4.040056219255095, "grad_norm": 0.18109670281410217, "learning_rate": 2.4227687983134225e-05, "loss": 0.0173, "step": 5749 }, { "epoch": 4.040758959943781, "grad_norm": 0.2957742512226105, "learning_rate": 2.4231904427266338e-05, "loss": 0.0244, "step": 5750 }, { "epoch": 4.041461700632467, "grad_norm": 0.1727503389120102, "learning_rate": 2.4236120871398458e-05, "loss": 0.0277, "step": 5751 }, { "epoch": 4.042164441321153, "grad_norm": 0.11172745376825333, "learning_rate": 2.424033731553057e-05, "loss": 0.0227, "step": 5752 }, { "epoch": 4.0428671820098385, "grad_norm": 0.1734478771686554, "learning_rate": 2.4244553759662688e-05, "loss": 0.0261, "step": 5753 }, { "epoch": 4.043569922698524, "grad_norm": 0.492440789937973, "learning_rate": 2.42487702037948e-05, "loss": 0.0282, "step": 5754 }, { "epoch": 4.04427266338721, "grad_norm": 0.35035428404808044, "learning_rate": 2.4252986647926918e-05, "loss": 0.0396, "step": 5755 }, { "epoch": 4.044975404075896, "grad_norm": 0.27169302105903625, "learning_rate": 2.425720309205903e-05, "loss": 0.0225, "step": 5756 }, { "epoch": 4.045678144764582, "grad_norm": 0.3327859938144684, "learning_rate": 2.4261419536191147e-05, "loss": 0.0502, "step": 5757 }, { "epoch": 4.046380885453268, "grad_norm": 0.2394058257341385, "learning_rate": 2.426563598032326e-05, "loss": 0.0425, "step": 5758 }, { "epoch": 4.047083626141954, "grad_norm": 0.21779265999794006, "learning_rate": 2.4269852424455377e-05, "loss": 0.0252, "step": 5759 }, { "epoch": 4.047786366830639, "grad_norm": 0.64183509349823, "learning_rate": 2.427406886858749e-05, "loss": 0.0342, "step": 5760 }, { "epoch": 4.048489107519325, "grad_norm": 0.3201395273208618, "learning_rate": 2.4278285312719607e-05, "loss": 0.0516, "step": 5761 }, { "epoch": 4.049191848208011, "grad_norm": 0.3029296398162842, "learning_rate": 2.428250175685172e-05, "loss": 0.057, "step": 5762 }, { "epoch": 4.049894588896697, "grad_norm": 0.43091362714767456, "learning_rate": 2.4286718200983837e-05, "loss": 0.0903, "step": 5763 }, { "epoch": 4.050597329585383, "grad_norm": 0.6752785444259644, "learning_rate": 2.429093464511595e-05, "loss": 0.1661, "step": 5764 }, { "epoch": 4.051300070274069, "grad_norm": 0.7282573580741882, "learning_rate": 2.429515108924807e-05, "loss": 0.1894, "step": 5765 }, { "epoch": 4.052002810962755, "grad_norm": 1.0034840106964111, "learning_rate": 2.4299367533380184e-05, "loss": 0.2278, "step": 5766 }, { "epoch": 4.0527055516514405, "grad_norm": 1.5724538564682007, "learning_rate": 2.43035839775123e-05, "loss": 0.2829, "step": 5767 }, { "epoch": 4.0534082923401265, "grad_norm": 0.2476585954427719, "learning_rate": 2.4307800421644414e-05, "loss": 0.0873, "step": 5768 }, { "epoch": 4.054111033028812, "grad_norm": 0.25132352113723755, "learning_rate": 2.431201686577653e-05, "loss": 0.0386, "step": 5769 }, { "epoch": 4.054813773717498, "grad_norm": 0.1752585619688034, "learning_rate": 2.4316233309908643e-05, "loss": 0.0435, "step": 5770 }, { "epoch": 4.055516514406184, "grad_norm": 0.19154103100299835, "learning_rate": 2.432044975404076e-05, "loss": 0.0274, "step": 5771 }, { "epoch": 4.05621925509487, "grad_norm": 0.27100151777267456, "learning_rate": 2.4324666198172873e-05, "loss": 0.0262, "step": 5772 }, { "epoch": 4.056921995783556, "grad_norm": 0.17026224732398987, "learning_rate": 2.432888264230499e-05, "loss": 0.0191, "step": 5773 }, { "epoch": 4.057624736472242, "grad_norm": 0.1944853961467743, "learning_rate": 2.4333099086437103e-05, "loss": 0.0317, "step": 5774 }, { "epoch": 4.058327477160928, "grad_norm": 0.22060897946357727, "learning_rate": 2.433731553056922e-05, "loss": 0.0295, "step": 5775 }, { "epoch": 4.059030217849614, "grad_norm": 0.1608905792236328, "learning_rate": 2.4341531974701333e-05, "loss": 0.0247, "step": 5776 }, { "epoch": 4.0597329585383, "grad_norm": 0.23354405164718628, "learning_rate": 2.434574841883345e-05, "loss": 0.0127, "step": 5777 }, { "epoch": 4.060435699226986, "grad_norm": 0.22522851824760437, "learning_rate": 2.4349964862965566e-05, "loss": 0.0529, "step": 5778 }, { "epoch": 4.061138439915671, "grad_norm": 0.2523351013660431, "learning_rate": 2.4354181307097683e-05, "loss": 0.0197, "step": 5779 }, { "epoch": 4.061841180604357, "grad_norm": 0.27462780475616455, "learning_rate": 2.43583977512298e-05, "loss": 0.0393, "step": 5780 }, { "epoch": 4.0625439212930425, "grad_norm": 0.24982798099517822, "learning_rate": 2.4362614195361913e-05, "loss": 0.0246, "step": 5781 }, { "epoch": 4.0632466619817285, "grad_norm": 0.38781440258026123, "learning_rate": 2.436683063949403e-05, "loss": 0.0543, "step": 5782 }, { "epoch": 4.063949402670414, "grad_norm": 0.36518988013267517, "learning_rate": 2.4371047083626143e-05, "loss": 0.0429, "step": 5783 }, { "epoch": 4.0646521433591, "grad_norm": 0.22325319051742554, "learning_rate": 2.437526352775826e-05, "loss": 0.0284, "step": 5784 }, { "epoch": 4.065354884047786, "grad_norm": 0.2665845453739166, "learning_rate": 2.4379479971890373e-05, "loss": 0.0435, "step": 5785 }, { "epoch": 4.066057624736472, "grad_norm": 0.3555753231048584, "learning_rate": 2.438369641602249e-05, "loss": 0.0851, "step": 5786 }, { "epoch": 4.066760365425158, "grad_norm": 0.5192802548408508, "learning_rate": 2.4387912860154603e-05, "loss": 0.0588, "step": 5787 }, { "epoch": 4.067463106113844, "grad_norm": 0.35921886563301086, "learning_rate": 2.439212930428672e-05, "loss": 0.0831, "step": 5788 }, { "epoch": 4.06816584680253, "grad_norm": 0.9497347474098206, "learning_rate": 2.4396345748418832e-05, "loss": 0.1508, "step": 5789 }, { "epoch": 4.068868587491216, "grad_norm": 0.9234259128570557, "learning_rate": 2.440056219255095e-05, "loss": 0.1947, "step": 5790 }, { "epoch": 4.069571328179902, "grad_norm": 1.3581969738006592, "learning_rate": 2.4404778636683062e-05, "loss": 0.2289, "step": 5791 }, { "epoch": 4.070274068868588, "grad_norm": 1.8740829229354858, "learning_rate": 2.4408995080815182e-05, "loss": 0.2876, "step": 5792 }, { "epoch": 4.070976809557274, "grad_norm": 0.25776615738868713, "learning_rate": 2.4413211524947296e-05, "loss": 0.0751, "step": 5793 }, { "epoch": 4.0716795502459595, "grad_norm": 0.27111348509788513, "learning_rate": 2.4417427969079412e-05, "loss": 0.0237, "step": 5794 }, { "epoch": 4.072382290934645, "grad_norm": 0.18715496361255646, "learning_rate": 2.4421644413211525e-05, "loss": 0.0249, "step": 5795 }, { "epoch": 4.073085031623331, "grad_norm": 0.16828183829784393, "learning_rate": 2.4425860857343642e-05, "loss": 0.0352, "step": 5796 }, { "epoch": 4.073787772312017, "grad_norm": 0.24065181612968445, "learning_rate": 2.4430077301475755e-05, "loss": 0.0387, "step": 5797 }, { "epoch": 4.074490513000702, "grad_norm": 0.256186306476593, "learning_rate": 2.4434293745607872e-05, "loss": 0.0252, "step": 5798 }, { "epoch": 4.075193253689388, "grad_norm": 0.2051020860671997, "learning_rate": 2.4438510189739985e-05, "loss": 0.0528, "step": 5799 }, { "epoch": 4.075895994378074, "grad_norm": 0.17071406543254852, "learning_rate": 2.4442726633872102e-05, "loss": 0.0232, "step": 5800 }, { "epoch": 4.07659873506676, "grad_norm": 0.15701580047607422, "learning_rate": 2.4446943078004215e-05, "loss": 0.0237, "step": 5801 }, { "epoch": 4.077301475755446, "grad_norm": 0.1636781394481659, "learning_rate": 2.4451159522136332e-05, "loss": 0.0207, "step": 5802 }, { "epoch": 4.078004216444132, "grad_norm": 0.31471455097198486, "learning_rate": 2.4455375966268445e-05, "loss": 0.034, "step": 5803 }, { "epoch": 4.078706957132818, "grad_norm": 0.16073520481586456, "learning_rate": 2.445959241040056e-05, "loss": 0.023, "step": 5804 }, { "epoch": 4.079409697821504, "grad_norm": 0.22252877056598663, "learning_rate": 2.4463808854532675e-05, "loss": 0.0325, "step": 5805 }, { "epoch": 4.08011243851019, "grad_norm": 0.3583621382713318, "learning_rate": 2.4468025298664795e-05, "loss": 0.0248, "step": 5806 }, { "epoch": 4.080815179198876, "grad_norm": 0.21875914931297302, "learning_rate": 2.4472241742796908e-05, "loss": 0.0498, "step": 5807 }, { "epoch": 4.0815179198875615, "grad_norm": 0.2654307782649994, "learning_rate": 2.4476458186929025e-05, "loss": 0.0344, "step": 5808 }, { "epoch": 4.082220660576247, "grad_norm": 0.22272813320159912, "learning_rate": 2.448067463106114e-05, "loss": 0.0285, "step": 5809 }, { "epoch": 4.082923401264933, "grad_norm": 0.36152175068855286, "learning_rate": 2.4484891075193255e-05, "loss": 0.0671, "step": 5810 }, { "epoch": 4.083626141953619, "grad_norm": 0.26228973269462585, "learning_rate": 2.448910751932537e-05, "loss": 0.0541, "step": 5811 }, { "epoch": 4.084328882642305, "grad_norm": 0.38778427243232727, "learning_rate": 2.4493323963457485e-05, "loss": 0.0629, "step": 5812 }, { "epoch": 4.085031623330991, "grad_norm": 0.3547486364841461, "learning_rate": 2.44975404075896e-05, "loss": 0.0757, "step": 5813 }, { "epoch": 4.085734364019677, "grad_norm": 0.7233961224555969, "learning_rate": 2.4501756851721714e-05, "loss": 0.137, "step": 5814 }, { "epoch": 4.086437104708363, "grad_norm": 0.7274426817893982, "learning_rate": 2.450597329585383e-05, "loss": 0.1925, "step": 5815 }, { "epoch": 4.087139845397049, "grad_norm": 3.69716739654541, "learning_rate": 2.4510189739985944e-05, "loss": 0.2058, "step": 5816 }, { "epoch": 4.087842586085735, "grad_norm": 1.5728464126586914, "learning_rate": 2.451440618411806e-05, "loss": 0.2833, "step": 5817 }, { "epoch": 4.08854532677442, "grad_norm": 0.21938073635101318, "learning_rate": 2.4518622628250174e-05, "loss": 0.0763, "step": 5818 }, { "epoch": 4.089248067463106, "grad_norm": 0.1474905014038086, "learning_rate": 2.4522839072382294e-05, "loss": 0.0318, "step": 5819 }, { "epoch": 4.089950808151792, "grad_norm": 0.20160314440727234, "learning_rate": 2.4527055516514408e-05, "loss": 0.0301, "step": 5820 }, { "epoch": 4.090653548840478, "grad_norm": 0.19436772167682648, "learning_rate": 2.4531271960646524e-05, "loss": 0.0326, "step": 5821 }, { "epoch": 4.0913562895291635, "grad_norm": 0.19497331976890564, "learning_rate": 2.4535488404778637e-05, "loss": 0.0215, "step": 5822 }, { "epoch": 4.0920590302178494, "grad_norm": 0.19918106496334076, "learning_rate": 2.4539704848910754e-05, "loss": 0.0186, "step": 5823 }, { "epoch": 4.092761770906535, "grad_norm": 0.13969887793064117, "learning_rate": 2.4543921293042867e-05, "loss": 0.0184, "step": 5824 }, { "epoch": 4.093464511595221, "grad_norm": 0.22307902574539185, "learning_rate": 2.4548137737174984e-05, "loss": 0.0182, "step": 5825 }, { "epoch": 4.094167252283907, "grad_norm": 0.2320626825094223, "learning_rate": 2.4552354181307097e-05, "loss": 0.0216, "step": 5826 }, { "epoch": 4.094869992972593, "grad_norm": 0.19615423679351807, "learning_rate": 2.4556570625439214e-05, "loss": 0.0318, "step": 5827 }, { "epoch": 4.095572733661279, "grad_norm": 0.3484828770160675, "learning_rate": 2.4560787069571327e-05, "loss": 0.0279, "step": 5828 }, { "epoch": 4.096275474349965, "grad_norm": 0.22715739905834198, "learning_rate": 2.4565003513703444e-05, "loss": 0.0273, "step": 5829 }, { "epoch": 4.096978215038651, "grad_norm": 0.20705130696296692, "learning_rate": 2.4569219957835557e-05, "loss": 0.0319, "step": 5830 }, { "epoch": 4.097680955727337, "grad_norm": 0.22729156911373138, "learning_rate": 2.4573436401967674e-05, "loss": 0.0447, "step": 5831 }, { "epoch": 4.098383696416023, "grad_norm": 0.25601857900619507, "learning_rate": 2.4577652846099787e-05, "loss": 0.0343, "step": 5832 }, { "epoch": 4.099086437104709, "grad_norm": 0.33435919880867004, "learning_rate": 2.4581869290231907e-05, "loss": 0.0797, "step": 5833 }, { "epoch": 4.099789177793395, "grad_norm": 0.2860971987247467, "learning_rate": 2.458608573436402e-05, "loss": 0.0276, "step": 5834 }, { "epoch": 4.1004919184820805, "grad_norm": 0.4051424562931061, "learning_rate": 2.4590302178496137e-05, "loss": 0.0368, "step": 5835 }, { "epoch": 4.101194659170766, "grad_norm": 0.45731282234191895, "learning_rate": 2.459451862262825e-05, "loss": 0.0603, "step": 5836 }, { "epoch": 4.1018973998594515, "grad_norm": 0.33290132880210876, "learning_rate": 2.4598735066760367e-05, "loss": 0.061, "step": 5837 }, { "epoch": 4.102600140548137, "grad_norm": 0.3977328836917877, "learning_rate": 2.4602951510892483e-05, "loss": 0.0916, "step": 5838 }, { "epoch": 4.103302881236823, "grad_norm": 0.572360634803772, "learning_rate": 2.4607167955024597e-05, "loss": 0.1355, "step": 5839 }, { "epoch": 4.104005621925509, "grad_norm": 2.4460856914520264, "learning_rate": 2.4611384399156713e-05, "loss": 0.2267, "step": 5840 }, { "epoch": 4.104708362614195, "grad_norm": 1.2082608938217163, "learning_rate": 2.4615600843288826e-05, "loss": 0.2251, "step": 5841 }, { "epoch": 4.105411103302881, "grad_norm": 4.002320766448975, "learning_rate": 2.4619817287420943e-05, "loss": 0.3042, "step": 5842 }, { "epoch": 4.106113843991567, "grad_norm": 0.34779641032218933, "learning_rate": 2.4624033731553056e-05, "loss": 0.0951, "step": 5843 }, { "epoch": 4.106816584680253, "grad_norm": 0.20253479480743408, "learning_rate": 2.4628250175685173e-05, "loss": 0.0443, "step": 5844 }, { "epoch": 4.107519325368939, "grad_norm": 0.2701273560523987, "learning_rate": 2.4632466619817286e-05, "loss": 0.0372, "step": 5845 }, { "epoch": 4.108222066057625, "grad_norm": 0.24205254018306732, "learning_rate": 2.4636683063949403e-05, "loss": 0.0252, "step": 5846 }, { "epoch": 4.108924806746311, "grad_norm": 0.20782263576984406, "learning_rate": 2.464089950808152e-05, "loss": 0.0212, "step": 5847 }, { "epoch": 4.109627547434997, "grad_norm": 0.16633249819278717, "learning_rate": 2.4645115952213636e-05, "loss": 0.0211, "step": 5848 }, { "epoch": 4.1103302881236825, "grad_norm": 0.16248071193695068, "learning_rate": 2.464933239634575e-05, "loss": 0.0244, "step": 5849 }, { "epoch": 4.111033028812368, "grad_norm": 0.1731555163860321, "learning_rate": 2.4653548840477866e-05, "loss": 0.0362, "step": 5850 }, { "epoch": 4.111735769501054, "grad_norm": 0.25245603919029236, "learning_rate": 2.465776528460998e-05, "loss": 0.0285, "step": 5851 }, { "epoch": 4.11243851018974, "grad_norm": 0.20257392525672913, "learning_rate": 2.4661981728742096e-05, "loss": 0.0132, "step": 5852 }, { "epoch": 4.113141250878426, "grad_norm": 0.2755665183067322, "learning_rate": 2.466619817287421e-05, "loss": 0.0371, "step": 5853 }, { "epoch": 4.113843991567112, "grad_norm": 0.17789773643016815, "learning_rate": 2.4670414617006326e-05, "loss": 0.023, "step": 5854 }, { "epoch": 4.114546732255798, "grad_norm": 0.18871402740478516, "learning_rate": 2.467463106113844e-05, "loss": 0.0342, "step": 5855 }, { "epoch": 4.115249472944483, "grad_norm": 0.9982564449310303, "learning_rate": 2.4678847505270556e-05, "loss": 0.0194, "step": 5856 }, { "epoch": 4.115952213633169, "grad_norm": 0.22151786088943481, "learning_rate": 2.468306394940267e-05, "loss": 0.0366, "step": 5857 }, { "epoch": 4.116654954321855, "grad_norm": 0.4988997280597687, "learning_rate": 2.4687280393534786e-05, "loss": 0.0247, "step": 5858 }, { "epoch": 4.117357695010541, "grad_norm": 0.3281925320625305, "learning_rate": 2.46914968376669e-05, "loss": 0.0519, "step": 5859 }, { "epoch": 4.118060435699227, "grad_norm": 0.2125154286623001, "learning_rate": 2.469571328179902e-05, "loss": 0.0523, "step": 5860 }, { "epoch": 4.118763176387913, "grad_norm": 2.5010159015655518, "learning_rate": 2.4699929725931132e-05, "loss": 0.0558, "step": 5861 }, { "epoch": 4.119465917076599, "grad_norm": 0.692094087600708, "learning_rate": 2.470414617006325e-05, "loss": 0.0955, "step": 5862 }, { "epoch": 4.1201686577652845, "grad_norm": 0.3769412636756897, "learning_rate": 2.4708362614195362e-05, "loss": 0.1065, "step": 5863 }, { "epoch": 4.12087139845397, "grad_norm": 0.5231210589408875, "learning_rate": 2.471257905832748e-05, "loss": 0.1542, "step": 5864 }, { "epoch": 4.121574139142656, "grad_norm": 0.8967016339302063, "learning_rate": 2.4716795502459592e-05, "loss": 0.2123, "step": 5865 }, { "epoch": 4.122276879831342, "grad_norm": 1.0198901891708374, "learning_rate": 2.472101194659171e-05, "loss": 0.2414, "step": 5866 }, { "epoch": 4.122979620520028, "grad_norm": 1.6613761186599731, "learning_rate": 2.4725228390723822e-05, "loss": 0.3175, "step": 5867 }, { "epoch": 4.123682361208714, "grad_norm": 0.38084128499031067, "learning_rate": 2.472944483485594e-05, "loss": 0.0969, "step": 5868 }, { "epoch": 4.1243851018974, "grad_norm": 0.33486801385879517, "learning_rate": 2.4733661278988055e-05, "loss": 0.031, "step": 5869 }, { "epoch": 4.125087842586086, "grad_norm": 0.18617631494998932, "learning_rate": 2.4737877723120168e-05, "loss": 0.0373, "step": 5870 }, { "epoch": 4.125790583274772, "grad_norm": 0.13333001732826233, "learning_rate": 2.4742094167252285e-05, "loss": 0.0198, "step": 5871 }, { "epoch": 4.126493323963458, "grad_norm": 0.1788203865289688, "learning_rate": 2.4746310611384398e-05, "loss": 0.0288, "step": 5872 }, { "epoch": 4.127196064652144, "grad_norm": 0.13327260315418243, "learning_rate": 2.4750527055516515e-05, "loss": 0.0157, "step": 5873 }, { "epoch": 4.12789880534083, "grad_norm": 0.11802535504102707, "learning_rate": 2.475474349964863e-05, "loss": 0.0195, "step": 5874 }, { "epoch": 4.128601546029515, "grad_norm": 0.13990645110607147, "learning_rate": 2.4758959943780748e-05, "loss": 0.0223, "step": 5875 }, { "epoch": 4.129304286718201, "grad_norm": 0.20579741895198822, "learning_rate": 2.476317638791286e-05, "loss": 0.0315, "step": 5876 }, { "epoch": 4.1300070274068865, "grad_norm": 0.1225113496184349, "learning_rate": 2.4767392832044978e-05, "loss": 0.0168, "step": 5877 }, { "epoch": 4.130709768095572, "grad_norm": 0.37623071670532227, "learning_rate": 2.477160927617709e-05, "loss": 0.0305, "step": 5878 }, { "epoch": 4.131412508784258, "grad_norm": 0.39310210943222046, "learning_rate": 2.4775825720309208e-05, "loss": 0.0372, "step": 5879 }, { "epoch": 4.132115249472944, "grad_norm": 0.29010340571403503, "learning_rate": 2.478004216444132e-05, "loss": 0.026, "step": 5880 }, { "epoch": 4.13281799016163, "grad_norm": 0.17024323344230652, "learning_rate": 2.4784258608573438e-05, "loss": 0.0194, "step": 5881 }, { "epoch": 4.133520730850316, "grad_norm": 0.12398488819599152, "learning_rate": 2.478847505270555e-05, "loss": 0.0261, "step": 5882 }, { "epoch": 4.134223471539002, "grad_norm": 0.3004859387874603, "learning_rate": 2.4792691496837668e-05, "loss": 0.0347, "step": 5883 }, { "epoch": 4.134926212227688, "grad_norm": 0.33483219146728516, "learning_rate": 2.479690794096978e-05, "loss": 0.0422, "step": 5884 }, { "epoch": 4.135628952916374, "grad_norm": 0.37071579694747925, "learning_rate": 2.4801124385101897e-05, "loss": 0.064, "step": 5885 }, { "epoch": 4.13633169360506, "grad_norm": 0.4129808247089386, "learning_rate": 2.480534082923401e-05, "loss": 0.0555, "step": 5886 }, { "epoch": 4.137034434293746, "grad_norm": 0.2499161660671234, "learning_rate": 2.480955727336613e-05, "loss": 0.0398, "step": 5887 }, { "epoch": 4.137737174982432, "grad_norm": 0.6584929823875427, "learning_rate": 2.4813773717498244e-05, "loss": 0.1185, "step": 5888 }, { "epoch": 4.138439915671118, "grad_norm": 0.9986823797225952, "learning_rate": 2.481799016163036e-05, "loss": 0.1667, "step": 5889 }, { "epoch": 4.1391426563598035, "grad_norm": 0.8339748978614807, "learning_rate": 2.4822206605762474e-05, "loss": 0.1956, "step": 5890 }, { "epoch": 4.139845397048489, "grad_norm": 1.3059098720550537, "learning_rate": 2.482642304989459e-05, "loss": 0.2745, "step": 5891 }, { "epoch": 4.140548137737175, "grad_norm": 3.460085391998291, "learning_rate": 2.4830639494026704e-05, "loss": 0.3128, "step": 5892 }, { "epoch": 4.141250878425861, "grad_norm": 0.24479202926158905, "learning_rate": 2.483485593815882e-05, "loss": 0.0886, "step": 5893 }, { "epoch": 4.141953619114547, "grad_norm": 0.1482333391904831, "learning_rate": 2.4839072382290934e-05, "loss": 0.0332, "step": 5894 }, { "epoch": 4.142656359803232, "grad_norm": 0.18989680707454681, "learning_rate": 2.484328882642305e-05, "loss": 0.0299, "step": 5895 }, { "epoch": 4.143359100491918, "grad_norm": 0.21658027172088623, "learning_rate": 2.4847505270555164e-05, "loss": 0.0246, "step": 5896 }, { "epoch": 4.144061841180604, "grad_norm": 0.17554853856563568, "learning_rate": 2.485172171468728e-05, "loss": 0.0257, "step": 5897 }, { "epoch": 4.14476458186929, "grad_norm": 0.1614425778388977, "learning_rate": 2.4855938158819397e-05, "loss": 0.03, "step": 5898 }, { "epoch": 4.145467322557976, "grad_norm": 0.26061147451400757, "learning_rate": 2.486015460295151e-05, "loss": 0.0281, "step": 5899 }, { "epoch": 4.146170063246662, "grad_norm": 0.20997320115566254, "learning_rate": 2.4864371047083627e-05, "loss": 0.0223, "step": 5900 }, { "epoch": 4.146872803935348, "grad_norm": 0.1227785050868988, "learning_rate": 2.4868587491215743e-05, "loss": 0.0214, "step": 5901 }, { "epoch": 4.147575544624034, "grad_norm": 0.1738780438899994, "learning_rate": 2.487280393534786e-05, "loss": 0.0238, "step": 5902 }, { "epoch": 4.14827828531272, "grad_norm": 0.18745169043540955, "learning_rate": 2.4877020379479973e-05, "loss": 0.0294, "step": 5903 }, { "epoch": 4.1489810260014055, "grad_norm": 0.35029616951942444, "learning_rate": 2.488123682361209e-05, "loss": 0.037, "step": 5904 }, { "epoch": 4.149683766690091, "grad_norm": 0.2244994342327118, "learning_rate": 2.4885453267744203e-05, "loss": 0.0367, "step": 5905 }, { "epoch": 4.150386507378777, "grad_norm": 0.2779165804386139, "learning_rate": 2.488966971187632e-05, "loss": 0.0338, "step": 5906 }, { "epoch": 4.151089248067463, "grad_norm": 0.25162583589553833, "learning_rate": 2.4893886156008433e-05, "loss": 0.0492, "step": 5907 }, { "epoch": 4.151791988756149, "grad_norm": 0.45657336711883545, "learning_rate": 2.489810260014055e-05, "loss": 0.0322, "step": 5908 }, { "epoch": 4.152494729444835, "grad_norm": 0.22377687692642212, "learning_rate": 2.4902319044272663e-05, "loss": 0.0277, "step": 5909 }, { "epoch": 4.153197470133521, "grad_norm": 0.2667401432991028, "learning_rate": 2.490653548840478e-05, "loss": 0.0523, "step": 5910 }, { "epoch": 4.153900210822207, "grad_norm": 0.33955103158950806, "learning_rate": 2.4910751932536893e-05, "loss": 0.0389, "step": 5911 }, { "epoch": 4.154602951510893, "grad_norm": 0.41601210832595825, "learning_rate": 2.491496837666901e-05, "loss": 0.0644, "step": 5912 }, { "epoch": 4.155305692199578, "grad_norm": 0.5309646725654602, "learning_rate": 2.4919184820801123e-05, "loss": 0.1415, "step": 5913 }, { "epoch": 4.156008432888264, "grad_norm": 0.6834692358970642, "learning_rate": 2.492340126493324e-05, "loss": 0.155, "step": 5914 }, { "epoch": 4.15671117357695, "grad_norm": 0.6824363470077515, "learning_rate": 2.4927617709065356e-05, "loss": 0.2152, "step": 5915 }, { "epoch": 4.157413914265636, "grad_norm": 0.8825266361236572, "learning_rate": 2.4931834153197473e-05, "loss": 0.2552, "step": 5916 }, { "epoch": 4.158116654954322, "grad_norm": 2.8438422679901123, "learning_rate": 2.4936050597329586e-05, "loss": 0.2911, "step": 5917 }, { "epoch": 4.1588193956430075, "grad_norm": 0.25061535835266113, "learning_rate": 2.4940267041461702e-05, "loss": 0.0786, "step": 5918 }, { "epoch": 4.159522136331693, "grad_norm": 0.23174872994422913, "learning_rate": 2.4944483485593816e-05, "loss": 0.0381, "step": 5919 }, { "epoch": 4.160224877020379, "grad_norm": 0.27141138911247253, "learning_rate": 2.4948699929725932e-05, "loss": 0.044, "step": 5920 }, { "epoch": 4.160927617709065, "grad_norm": 0.16523593664169312, "learning_rate": 2.4952916373858046e-05, "loss": 0.0252, "step": 5921 }, { "epoch": 4.161630358397751, "grad_norm": 0.38143065571784973, "learning_rate": 2.4957132817990162e-05, "loss": 0.0393, "step": 5922 }, { "epoch": 4.162333099086437, "grad_norm": 0.14120712876319885, "learning_rate": 2.4961349262122276e-05, "loss": 0.0155, "step": 5923 }, { "epoch": 4.163035839775123, "grad_norm": 0.16967988014221191, "learning_rate": 2.4965565706254392e-05, "loss": 0.0258, "step": 5924 }, { "epoch": 4.163738580463809, "grad_norm": 0.39802733063697815, "learning_rate": 2.4969782150386505e-05, "loss": 0.0527, "step": 5925 }, { "epoch": 4.164441321152495, "grad_norm": 0.20461083948612213, "learning_rate": 2.4973998594518622e-05, "loss": 0.0318, "step": 5926 }, { "epoch": 4.165144061841181, "grad_norm": 0.18219417333602905, "learning_rate": 2.4978215038650735e-05, "loss": 0.021, "step": 5927 }, { "epoch": 4.165846802529867, "grad_norm": 0.2090543806552887, "learning_rate": 2.4982431482782855e-05, "loss": 0.0363, "step": 5928 }, { "epoch": 4.166549543218553, "grad_norm": 0.19517993927001953, "learning_rate": 2.4986647926914972e-05, "loss": 0.0205, "step": 5929 }, { "epoch": 4.167252283907239, "grad_norm": 0.22187519073486328, "learning_rate": 2.4990864371047085e-05, "loss": 0.0308, "step": 5930 }, { "epoch": 4.1679550245959245, "grad_norm": 0.18191096186637878, "learning_rate": 2.4995080815179202e-05, "loss": 0.0232, "step": 5931 }, { "epoch": 4.16865776528461, "grad_norm": 0.41537922620773315, "learning_rate": 2.4999297259311315e-05, "loss": 0.0238, "step": 5932 }, { "epoch": 4.169360505973295, "grad_norm": 0.29026490449905396, "learning_rate": 2.5003513703443432e-05, "loss": 0.0359, "step": 5933 }, { "epoch": 4.170063246661981, "grad_norm": 0.3373505473136902, "learning_rate": 2.5007730147575545e-05, "loss": 0.0226, "step": 5934 }, { "epoch": 4.170765987350667, "grad_norm": 0.2825557291507721, "learning_rate": 2.501194659170766e-05, "loss": 0.043, "step": 5935 }, { "epoch": 4.171468728039353, "grad_norm": 0.31857895851135254, "learning_rate": 2.5016163035839775e-05, "loss": 0.0414, "step": 5936 }, { "epoch": 4.172171468728039, "grad_norm": 0.721918523311615, "learning_rate": 2.502037947997189e-05, "loss": 0.07, "step": 5937 }, { "epoch": 4.172874209416725, "grad_norm": 0.47828394174575806, "learning_rate": 2.5024595924104005e-05, "loss": 0.0928, "step": 5938 }, { "epoch": 4.173576950105411, "grad_norm": 0.7243223786354065, "learning_rate": 2.502881236823612e-05, "loss": 0.119, "step": 5939 }, { "epoch": 4.174279690794097, "grad_norm": 0.7470762133598328, "learning_rate": 2.5033028812368235e-05, "loss": 0.1862, "step": 5940 }, { "epoch": 4.174982431482783, "grad_norm": 0.8347209095954895, "learning_rate": 2.503724525650035e-05, "loss": 0.2422, "step": 5941 }, { "epoch": 4.175685172171469, "grad_norm": 1.6042530536651611, "learning_rate": 2.5041461700632468e-05, "loss": 0.2205, "step": 5942 }, { "epoch": 4.176387912860155, "grad_norm": 0.26025599241256714, "learning_rate": 2.5045678144764585e-05, "loss": 0.0759, "step": 5943 }, { "epoch": 4.177090653548841, "grad_norm": 0.16412003338336945, "learning_rate": 2.5049894588896698e-05, "loss": 0.0324, "step": 5944 }, { "epoch": 4.1777933942375265, "grad_norm": 0.38890546560287476, "learning_rate": 2.5054111033028814e-05, "loss": 0.0191, "step": 5945 }, { "epoch": 4.178496134926212, "grad_norm": 0.16993668675422668, "learning_rate": 2.5058327477160928e-05, "loss": 0.0289, "step": 5946 }, { "epoch": 4.179198875614898, "grad_norm": 0.20431211590766907, "learning_rate": 2.5062543921293044e-05, "loss": 0.0267, "step": 5947 }, { "epoch": 4.179901616303584, "grad_norm": 0.5129315853118896, "learning_rate": 2.5066760365425158e-05, "loss": 0.0211, "step": 5948 }, { "epoch": 4.18060435699227, "grad_norm": 0.18064828217029572, "learning_rate": 2.5070976809557274e-05, "loss": 0.0206, "step": 5949 }, { "epoch": 4.181307097680956, "grad_norm": 0.2053612470626831, "learning_rate": 2.5075193253689387e-05, "loss": 0.0203, "step": 5950 }, { "epoch": 4.182009838369642, "grad_norm": 0.3090868592262268, "learning_rate": 2.5079409697821504e-05, "loss": 0.0478, "step": 5951 }, { "epoch": 4.182712579058327, "grad_norm": 0.19282947480678558, "learning_rate": 2.5083626141953617e-05, "loss": 0.0194, "step": 5952 }, { "epoch": 4.183415319747013, "grad_norm": 0.23522226512432098, "learning_rate": 2.5087842586085734e-05, "loss": 0.0372, "step": 5953 }, { "epoch": 4.184118060435699, "grad_norm": 0.1972799450159073, "learning_rate": 2.5092059030217847e-05, "loss": 0.0284, "step": 5954 }, { "epoch": 4.184820801124385, "grad_norm": 0.40156108140945435, "learning_rate": 2.5096275474349967e-05, "loss": 0.0416, "step": 5955 }, { "epoch": 4.185523541813071, "grad_norm": 0.18011920154094696, "learning_rate": 2.510049191848208e-05, "loss": 0.023, "step": 5956 }, { "epoch": 4.186226282501757, "grad_norm": 0.2022370547056198, "learning_rate": 2.5104708362614197e-05, "loss": 0.0292, "step": 5957 }, { "epoch": 4.186929023190443, "grad_norm": 0.2210931032896042, "learning_rate": 2.5108924806746314e-05, "loss": 0.032, "step": 5958 }, { "epoch": 4.1876317638791285, "grad_norm": 0.22979870438575745, "learning_rate": 2.5113141250878427e-05, "loss": 0.0392, "step": 5959 }, { "epoch": 4.188334504567814, "grad_norm": 0.28756770491600037, "learning_rate": 2.5117357695010544e-05, "loss": 0.0359, "step": 5960 }, { "epoch": 4.1890372452565, "grad_norm": 0.22539633512496948, "learning_rate": 2.5121574139142657e-05, "loss": 0.0377, "step": 5961 }, { "epoch": 4.189739985945186, "grad_norm": 0.5148810744285583, "learning_rate": 2.5125790583274774e-05, "loss": 0.0934, "step": 5962 }, { "epoch": 4.190442726633872, "grad_norm": 0.4132918119430542, "learning_rate": 2.5130007027406887e-05, "loss": 0.0822, "step": 5963 }, { "epoch": 4.191145467322558, "grad_norm": 5.489317893981934, "learning_rate": 2.5134223471539003e-05, "loss": 0.1065, "step": 5964 }, { "epoch": 4.191848208011244, "grad_norm": 0.7350560426712036, "learning_rate": 2.5138439915671117e-05, "loss": 0.1803, "step": 5965 }, { "epoch": 4.19255094869993, "grad_norm": 1.666891098022461, "learning_rate": 2.5142656359803233e-05, "loss": 0.2526, "step": 5966 }, { "epoch": 4.193253689388616, "grad_norm": 1.3390268087387085, "learning_rate": 2.5146872803935347e-05, "loss": 0.306, "step": 5967 }, { "epoch": 4.193956430077302, "grad_norm": 0.6284376978874207, "learning_rate": 2.5151089248067463e-05, "loss": 0.1039, "step": 5968 }, { "epoch": 4.194659170765988, "grad_norm": 0.18181142210960388, "learning_rate": 2.515530569219958e-05, "loss": 0.0366, "step": 5969 }, { "epoch": 4.195361911454674, "grad_norm": 1.7687755823135376, "learning_rate": 2.5159522136331697e-05, "loss": 0.0346, "step": 5970 }, { "epoch": 4.1960646521433596, "grad_norm": 0.21672052145004272, "learning_rate": 2.516373858046381e-05, "loss": 0.0308, "step": 5971 }, { "epoch": 4.196767392832045, "grad_norm": 0.3232060670852661, "learning_rate": 2.5167955024595926e-05, "loss": 0.0264, "step": 5972 }, { "epoch": 4.1974701335207305, "grad_norm": 0.19562208652496338, "learning_rate": 2.517217146872804e-05, "loss": 0.013, "step": 5973 }, { "epoch": 4.198172874209416, "grad_norm": 0.19414101541042328, "learning_rate": 2.5176387912860156e-05, "loss": 0.0262, "step": 5974 }, { "epoch": 4.198875614898102, "grad_norm": 0.9504270553588867, "learning_rate": 2.518060435699227e-05, "loss": 0.0311, "step": 5975 }, { "epoch": 4.199578355586788, "grad_norm": 0.2205485701560974, "learning_rate": 2.5184820801124386e-05, "loss": 0.0287, "step": 5976 }, { "epoch": 4.200281096275474, "grad_norm": 0.21129433810710907, "learning_rate": 2.51890372452565e-05, "loss": 0.0248, "step": 5977 }, { "epoch": 4.20098383696416, "grad_norm": 0.20015913248062134, "learning_rate": 2.5193253689388616e-05, "loss": 0.0175, "step": 5978 }, { "epoch": 4.201686577652846, "grad_norm": 0.20943181216716766, "learning_rate": 2.519747013352073e-05, "loss": 0.0367, "step": 5979 }, { "epoch": 4.202389318341532, "grad_norm": 0.22013628482818604, "learning_rate": 2.5201686577652846e-05, "loss": 0.0428, "step": 5980 }, { "epoch": 4.203092059030218, "grad_norm": 0.241612046957016, "learning_rate": 2.520590302178496e-05, "loss": 0.0324, "step": 5981 }, { "epoch": 4.203794799718904, "grad_norm": 0.43828877806663513, "learning_rate": 2.5210119465917076e-05, "loss": 0.0461, "step": 5982 }, { "epoch": 4.20449754040759, "grad_norm": 0.3790353238582611, "learning_rate": 2.5214335910049192e-05, "loss": 0.0669, "step": 5983 }, { "epoch": 4.205200281096276, "grad_norm": 0.19065077602863312, "learning_rate": 2.521855235418131e-05, "loss": 0.0179, "step": 5984 }, { "epoch": 4.205903021784962, "grad_norm": 0.4650404155254364, "learning_rate": 2.5222768798313422e-05, "loss": 0.0727, "step": 5985 }, { "epoch": 4.2066057624736475, "grad_norm": 0.26084256172180176, "learning_rate": 2.522698524244554e-05, "loss": 0.0378, "step": 5986 }, { "epoch": 4.207308503162333, "grad_norm": 0.36065447330474854, "learning_rate": 2.5231201686577656e-05, "loss": 0.0756, "step": 5987 }, { "epoch": 4.208011243851019, "grad_norm": 0.3670533001422882, "learning_rate": 2.523541813070977e-05, "loss": 0.0852, "step": 5988 }, { "epoch": 4.208713984539705, "grad_norm": 0.5031070709228516, "learning_rate": 2.5239634574841886e-05, "loss": 0.1325, "step": 5989 }, { "epoch": 4.20941672522839, "grad_norm": 0.7930192351341248, "learning_rate": 2.5243851018974e-05, "loss": 0.2073, "step": 5990 }, { "epoch": 4.210119465917076, "grad_norm": 1.0484951734542847, "learning_rate": 2.5248067463106115e-05, "loss": 0.2088, "step": 5991 }, { "epoch": 4.210822206605762, "grad_norm": 1.9300341606140137, "learning_rate": 2.525228390723823e-05, "loss": 0.3008, "step": 5992 }, { "epoch": 4.211524947294448, "grad_norm": 0.44933828711509705, "learning_rate": 2.5256500351370345e-05, "loss": 0.0719, "step": 5993 }, { "epoch": 4.212227687983134, "grad_norm": 0.224019393324852, "learning_rate": 2.526071679550246e-05, "loss": 0.0278, "step": 5994 }, { "epoch": 4.21293042867182, "grad_norm": 0.18422026932239532, "learning_rate": 2.5264933239634575e-05, "loss": 0.0438, "step": 5995 }, { "epoch": 4.213633169360506, "grad_norm": 0.17190198600292206, "learning_rate": 2.5269149683766692e-05, "loss": 0.0212, "step": 5996 }, { "epoch": 4.214335910049192, "grad_norm": 0.23713932931423187, "learning_rate": 2.527336612789881e-05, "loss": 0.0211, "step": 5997 }, { "epoch": 4.215038650737878, "grad_norm": 0.11989707499742508, "learning_rate": 2.5277582572030922e-05, "loss": 0.0128, "step": 5998 }, { "epoch": 4.215741391426564, "grad_norm": 0.3002963662147522, "learning_rate": 2.528179901616304e-05, "loss": 0.0195, "step": 5999 }, { "epoch": 4.2164441321152495, "grad_norm": 0.1792982667684555, "learning_rate": 2.528601546029515e-05, "loss": 0.0201, "step": 6000 }, { "epoch": 4.2164441321152495, "eval_cer": 0.19650202393811442, "eval_loss": 0.3028562366962433, "eval_runtime": 18.2817, "eval_samples_per_second": 248.226, "eval_steps_per_second": 0.82, "eval_wer": 0.3564984218317924, "step": 6000 }, { "epoch": 4.217146872803935, "grad_norm": 0.1974816471338272, "learning_rate": 2.5290231904427268e-05, "loss": 0.0313, "step": 6001 }, { "epoch": 4.217849613492621, "grad_norm": 0.23385538160800934, "learning_rate": 2.529444834855938e-05, "loss": 0.0137, "step": 6002 }, { "epoch": 4.218552354181307, "grad_norm": 0.21226677298545837, "learning_rate": 2.5298664792691498e-05, "loss": 0.0362, "step": 6003 }, { "epoch": 4.219255094869993, "grad_norm": 0.16391994059085846, "learning_rate": 2.530288123682361e-05, "loss": 0.0185, "step": 6004 }, { "epoch": 4.219957835558679, "grad_norm": 0.4091104567050934, "learning_rate": 2.5307097680955728e-05, "loss": 0.0452, "step": 6005 }, { "epoch": 4.220660576247365, "grad_norm": 0.21327315270900726, "learning_rate": 2.531131412508784e-05, "loss": 0.0266, "step": 6006 }, { "epoch": 4.221363316936051, "grad_norm": 0.3018781840801239, "learning_rate": 2.5315530569219958e-05, "loss": 0.0415, "step": 6007 }, { "epoch": 4.222066057624737, "grad_norm": 0.23761942982673645, "learning_rate": 2.531974701335207e-05, "loss": 0.0465, "step": 6008 }, { "epoch": 4.222768798313423, "grad_norm": 0.3800961673259735, "learning_rate": 2.5323963457484188e-05, "loss": 0.0351, "step": 6009 }, { "epoch": 4.223471539002108, "grad_norm": 0.2833493649959564, "learning_rate": 2.5328179901616304e-05, "loss": 0.0425, "step": 6010 }, { "epoch": 4.224174279690794, "grad_norm": 0.37384381890296936, "learning_rate": 2.533239634574842e-05, "loss": 0.0824, "step": 6011 }, { "epoch": 4.22487702037948, "grad_norm": 0.2571490406990051, "learning_rate": 2.5336612789880534e-05, "loss": 0.0772, "step": 6012 }, { "epoch": 4.225579761068166, "grad_norm": 1.1525003910064697, "learning_rate": 2.534082923401265e-05, "loss": 0.1223, "step": 6013 }, { "epoch": 4.2262825017568515, "grad_norm": 0.7023828029632568, "learning_rate": 2.5345045678144764e-05, "loss": 0.1852, "step": 6014 }, { "epoch": 4.226985242445537, "grad_norm": 0.9519362449645996, "learning_rate": 2.534926212227688e-05, "loss": 0.2484, "step": 6015 }, { "epoch": 4.227687983134223, "grad_norm": 1.3007017374038696, "learning_rate": 2.5353478566408994e-05, "loss": 0.2811, "step": 6016 }, { "epoch": 4.228390723822909, "grad_norm": 2.6512935161590576, "learning_rate": 2.535769501054111e-05, "loss": 0.3069, "step": 6017 }, { "epoch": 4.229093464511595, "grad_norm": 0.26508617401123047, "learning_rate": 2.5361911454673227e-05, "loss": 0.0926, "step": 6018 }, { "epoch": 4.229796205200281, "grad_norm": 0.1780354529619217, "learning_rate": 2.536612789880534e-05, "loss": 0.0377, "step": 6019 }, { "epoch": 4.230498945888967, "grad_norm": 0.1568145751953125, "learning_rate": 2.5370344342937457e-05, "loss": 0.0216, "step": 6020 }, { "epoch": 4.231201686577653, "grad_norm": 0.12486403435468674, "learning_rate": 2.537456078706957e-05, "loss": 0.014, "step": 6021 }, { "epoch": 4.231904427266339, "grad_norm": 0.37783676385879517, "learning_rate": 2.5378777231201687e-05, "loss": 0.0488, "step": 6022 }, { "epoch": 4.232607167955025, "grad_norm": 0.17648310959339142, "learning_rate": 2.5382993675333804e-05, "loss": 0.0347, "step": 6023 }, { "epoch": 4.233309908643711, "grad_norm": 0.21980684995651245, "learning_rate": 2.538721011946592e-05, "loss": 0.0183, "step": 6024 }, { "epoch": 4.234012649332397, "grad_norm": 0.1684994399547577, "learning_rate": 2.5391426563598034e-05, "loss": 0.0286, "step": 6025 }, { "epoch": 4.2347153900210825, "grad_norm": 0.7682106494903564, "learning_rate": 2.539564300773015e-05, "loss": 0.0251, "step": 6026 }, { "epoch": 4.2354181307097685, "grad_norm": 0.25559183955192566, "learning_rate": 2.5399859451862264e-05, "loss": 0.0342, "step": 6027 }, { "epoch": 4.236120871398454, "grad_norm": 0.2019188106060028, "learning_rate": 2.540407589599438e-05, "loss": 0.0185, "step": 6028 }, { "epoch": 4.236823612087139, "grad_norm": 0.21513846516609192, "learning_rate": 2.5408292340126493e-05, "loss": 0.0339, "step": 6029 }, { "epoch": 4.237526352775825, "grad_norm": 0.25754037499427795, "learning_rate": 2.541250878425861e-05, "loss": 0.0679, "step": 6030 }, { "epoch": 4.238229093464511, "grad_norm": 0.24359320104122162, "learning_rate": 2.5416725228390723e-05, "loss": 0.0166, "step": 6031 }, { "epoch": 4.238931834153197, "grad_norm": 0.2856464385986328, "learning_rate": 2.542094167252284e-05, "loss": 0.0355, "step": 6032 }, { "epoch": 4.239634574841883, "grad_norm": 0.3750576674938202, "learning_rate": 2.5425158116654953e-05, "loss": 0.0434, "step": 6033 }, { "epoch": 4.240337315530569, "grad_norm": 0.21823973953723907, "learning_rate": 2.542937456078707e-05, "loss": 0.0216, "step": 6034 }, { "epoch": 4.241040056219255, "grad_norm": 0.3282548785209656, "learning_rate": 2.5433591004919183e-05, "loss": 0.0602, "step": 6035 }, { "epoch": 4.241742796907941, "grad_norm": 0.320934921503067, "learning_rate": 2.54378074490513e-05, "loss": 0.0424, "step": 6036 }, { "epoch": 4.242445537596627, "grad_norm": 0.4430222809314728, "learning_rate": 2.5442023893183416e-05, "loss": 0.0757, "step": 6037 }, { "epoch": 4.243148278285313, "grad_norm": 0.49604323506355286, "learning_rate": 2.5446240337315533e-05, "loss": 0.0662, "step": 6038 }, { "epoch": 4.243851018973999, "grad_norm": 0.5601938962936401, "learning_rate": 2.5450456781447646e-05, "loss": 0.1292, "step": 6039 }, { "epoch": 4.2445537596626846, "grad_norm": 2.1832637786865234, "learning_rate": 2.5454673225579763e-05, "loss": 0.1733, "step": 6040 }, { "epoch": 4.2452565003513705, "grad_norm": 1.4057996273040771, "learning_rate": 2.5458889669711876e-05, "loss": 0.2564, "step": 6041 }, { "epoch": 4.245959241040056, "grad_norm": 1.2501461505889893, "learning_rate": 2.5463106113843993e-05, "loss": 0.3217, "step": 6042 }, { "epoch": 4.246661981728742, "grad_norm": 0.2810527980327606, "learning_rate": 2.5467322557976106e-05, "loss": 0.0746, "step": 6043 }, { "epoch": 4.247364722417428, "grad_norm": 1.0202465057373047, "learning_rate": 2.5471539002108223e-05, "loss": 0.0316, "step": 6044 }, { "epoch": 4.248067463106114, "grad_norm": 0.21320906281471252, "learning_rate": 2.5475755446240336e-05, "loss": 0.0287, "step": 6045 }, { "epoch": 4.2487702037948, "grad_norm": 0.2356249839067459, "learning_rate": 2.5479971890372453e-05, "loss": 0.027, "step": 6046 }, { "epoch": 4.249472944483486, "grad_norm": 0.7796327471733093, "learning_rate": 2.548418833450457e-05, "loss": 0.0251, "step": 6047 }, { "epoch": 4.250175685172172, "grad_norm": 0.2009515017271042, "learning_rate": 2.5488404778636682e-05, "loss": 0.0285, "step": 6048 }, { "epoch": 4.250878425860857, "grad_norm": 0.2384355515241623, "learning_rate": 2.54926212227688e-05, "loss": 0.0209, "step": 6049 }, { "epoch": 4.251581166549543, "grad_norm": 0.15517558157444, "learning_rate": 2.5496837666900912e-05, "loss": 0.0251, "step": 6050 }, { "epoch": 4.252283907238229, "grad_norm": 0.2502729594707489, "learning_rate": 2.5501054111033032e-05, "loss": 0.0485, "step": 6051 }, { "epoch": 4.252986647926915, "grad_norm": 0.1409909427165985, "learning_rate": 2.5505270555165146e-05, "loss": 0.0099, "step": 6052 }, { "epoch": 4.253689388615601, "grad_norm": 0.32308343052864075, "learning_rate": 2.5509486999297262e-05, "loss": 0.041, "step": 6053 }, { "epoch": 4.254392129304287, "grad_norm": 0.17614027857780457, "learning_rate": 2.5513703443429375e-05, "loss": 0.0149, "step": 6054 }, { "epoch": 4.2550948699929725, "grad_norm": 0.19372864067554474, "learning_rate": 2.5517919887561492e-05, "loss": 0.0248, "step": 6055 }, { "epoch": 4.255797610681658, "grad_norm": 0.5652133226394653, "learning_rate": 2.5522136331693605e-05, "loss": 0.0461, "step": 6056 }, { "epoch": 4.256500351370344, "grad_norm": 0.2370263785123825, "learning_rate": 2.5526352775825722e-05, "loss": 0.0256, "step": 6057 }, { "epoch": 4.25720309205903, "grad_norm": 0.29587557911872864, "learning_rate": 2.5530569219957835e-05, "loss": 0.0408, "step": 6058 }, { "epoch": 4.257905832747716, "grad_norm": 0.29200446605682373, "learning_rate": 2.5534785664089952e-05, "loss": 0.0436, "step": 6059 }, { "epoch": 4.258608573436402, "grad_norm": 1.1934181451797485, "learning_rate": 2.5539002108222065e-05, "loss": 0.0468, "step": 6060 }, { "epoch": 4.259311314125088, "grad_norm": 0.24048244953155518, "learning_rate": 2.5543218552354182e-05, "loss": 0.0476, "step": 6061 }, { "epoch": 4.260014054813774, "grad_norm": 0.28761157393455505, "learning_rate": 2.5547434996486295e-05, "loss": 0.0708, "step": 6062 }, { "epoch": 4.26071679550246, "grad_norm": 0.532882571220398, "learning_rate": 2.555165144061841e-05, "loss": 0.1159, "step": 6063 }, { "epoch": 4.261419536191146, "grad_norm": 0.6895865797996521, "learning_rate": 2.555586788475053e-05, "loss": 0.1109, "step": 6064 }, { "epoch": 4.262122276879832, "grad_norm": 0.6920594573020935, "learning_rate": 2.5560084328882645e-05, "loss": 0.181, "step": 6065 }, { "epoch": 4.262825017568518, "grad_norm": 0.9373330473899841, "learning_rate": 2.5564300773014758e-05, "loss": 0.2219, "step": 6066 }, { "epoch": 4.263527758257203, "grad_norm": 2.4078502655029297, "learning_rate": 2.5568517217146875e-05, "loss": 0.3172, "step": 6067 }, { "epoch": 4.264230498945889, "grad_norm": 0.3986873924732208, "learning_rate": 2.5572733661278988e-05, "loss": 0.0954, "step": 6068 }, { "epoch": 4.2649332396345745, "grad_norm": 0.29640644788742065, "learning_rate": 2.5576950105411105e-05, "loss": 0.033, "step": 6069 }, { "epoch": 4.26563598032326, "grad_norm": 0.3451928198337555, "learning_rate": 2.5581166549543218e-05, "loss": 0.0374, "step": 6070 }, { "epoch": 4.266338721011946, "grad_norm": 0.248805969953537, "learning_rate": 2.5585382993675335e-05, "loss": 0.0402, "step": 6071 }, { "epoch": 4.267041461700632, "grad_norm": 0.14192698895931244, "learning_rate": 2.5589599437807448e-05, "loss": 0.0231, "step": 6072 }, { "epoch": 4.267744202389318, "grad_norm": 0.24306955933570862, "learning_rate": 2.5593815881939564e-05, "loss": 0.0254, "step": 6073 }, { "epoch": 4.268446943078004, "grad_norm": 0.20444342494010925, "learning_rate": 2.5598032326071678e-05, "loss": 0.0391, "step": 6074 }, { "epoch": 4.26914968376669, "grad_norm": 0.5481185913085938, "learning_rate": 2.5602248770203794e-05, "loss": 0.042, "step": 6075 }, { "epoch": 4.269852424455376, "grad_norm": 0.28482627868652344, "learning_rate": 2.5606465214335908e-05, "loss": 0.0274, "step": 6076 }, { "epoch": 4.270555165144062, "grad_norm": 0.30759504437446594, "learning_rate": 2.5610681658468024e-05, "loss": 0.0251, "step": 6077 }, { "epoch": 4.271257905832748, "grad_norm": 0.6485279202461243, "learning_rate": 2.5614898102600144e-05, "loss": 0.0353, "step": 6078 }, { "epoch": 4.271960646521434, "grad_norm": 0.5700094103813171, "learning_rate": 2.5619114546732258e-05, "loss": 0.0341, "step": 6079 }, { "epoch": 4.27266338721012, "grad_norm": 0.23726969957351685, "learning_rate": 2.5623330990864374e-05, "loss": 0.0363, "step": 6080 }, { "epoch": 4.2733661278988055, "grad_norm": 0.25339287519454956, "learning_rate": 2.5627547434996487e-05, "loss": 0.019, "step": 6081 }, { "epoch": 4.2740688685874915, "grad_norm": 0.46009737253189087, "learning_rate": 2.5631763879128604e-05, "loss": 0.0703, "step": 6082 }, { "epoch": 4.274771609276177, "grad_norm": 0.39743298292160034, "learning_rate": 2.5635980323260717e-05, "loss": 0.0542, "step": 6083 }, { "epoch": 4.275474349964863, "grad_norm": 0.20837676525115967, "learning_rate": 2.5640196767392834e-05, "loss": 0.0264, "step": 6084 }, { "epoch": 4.276177090653549, "grad_norm": 0.5459452271461487, "learning_rate": 2.5644413211524947e-05, "loss": 0.0353, "step": 6085 }, { "epoch": 4.276879831342235, "grad_norm": 0.3284667134284973, "learning_rate": 2.5648629655657064e-05, "loss": 0.042, "step": 6086 }, { "epoch": 4.27758257203092, "grad_norm": 0.3852488398551941, "learning_rate": 2.5652846099789177e-05, "loss": 0.073, "step": 6087 }, { "epoch": 4.278285312719606, "grad_norm": 0.3450556695461273, "learning_rate": 2.5657062543921294e-05, "loss": 0.0709, "step": 6088 }, { "epoch": 4.278988053408292, "grad_norm": 0.6709344387054443, "learning_rate": 2.5661278988053407e-05, "loss": 0.1464, "step": 6089 }, { "epoch": 4.279690794096978, "grad_norm": 1.9621065855026245, "learning_rate": 2.5665495432185524e-05, "loss": 0.2067, "step": 6090 }, { "epoch": 4.280393534785664, "grad_norm": 1.0845074653625488, "learning_rate": 2.566971187631764e-05, "loss": 0.2549, "step": 6091 }, { "epoch": 4.28109627547435, "grad_norm": 1.1847703456878662, "learning_rate": 2.5673928320449757e-05, "loss": 0.2927, "step": 6092 }, { "epoch": 4.281799016163036, "grad_norm": 0.34748703241348267, "learning_rate": 2.567814476458187e-05, "loss": 0.076, "step": 6093 }, { "epoch": 4.282501756851722, "grad_norm": 0.23315276205539703, "learning_rate": 2.5682361208713987e-05, "loss": 0.0527, "step": 6094 }, { "epoch": 4.2832044975404076, "grad_norm": 0.2342822104692459, "learning_rate": 2.56865776528461e-05, "loss": 0.0379, "step": 6095 }, { "epoch": 4.2839072382290935, "grad_norm": 0.18210411071777344, "learning_rate": 2.5690794096978217e-05, "loss": 0.0339, "step": 6096 }, { "epoch": 4.284609978917779, "grad_norm": 0.18193092942237854, "learning_rate": 2.569501054111033e-05, "loss": 0.0218, "step": 6097 }, { "epoch": 4.285312719606465, "grad_norm": 0.16707518696784973, "learning_rate": 2.5699226985242447e-05, "loss": 0.0316, "step": 6098 }, { "epoch": 4.286015460295151, "grad_norm": 0.1813901662826538, "learning_rate": 2.570344342937456e-05, "loss": 0.0258, "step": 6099 }, { "epoch": 4.286718200983837, "grad_norm": 0.23449920117855072, "learning_rate": 2.5707659873506676e-05, "loss": 0.0446, "step": 6100 }, { "epoch": 4.287420941672523, "grad_norm": 0.19437678158283234, "learning_rate": 2.571187631763879e-05, "loss": 0.0261, "step": 6101 }, { "epoch": 4.288123682361209, "grad_norm": 0.18705034255981445, "learning_rate": 2.5716092761770906e-05, "loss": 0.0351, "step": 6102 }, { "epoch": 4.288826423049895, "grad_norm": 0.18873244524002075, "learning_rate": 2.572030920590302e-05, "loss": 0.0212, "step": 6103 }, { "epoch": 4.289529163738581, "grad_norm": 0.2368467003107071, "learning_rate": 2.5724525650035136e-05, "loss": 0.0314, "step": 6104 }, { "epoch": 4.290231904427266, "grad_norm": 0.19594040513038635, "learning_rate": 2.5728742094167253e-05, "loss": 0.0206, "step": 6105 }, { "epoch": 4.290934645115952, "grad_norm": 0.145862877368927, "learning_rate": 2.573295853829937e-05, "loss": 0.0302, "step": 6106 }, { "epoch": 4.291637385804638, "grad_norm": 0.21737109124660492, "learning_rate": 2.5737174982431486e-05, "loss": 0.056, "step": 6107 }, { "epoch": 4.292340126493324, "grad_norm": 0.19896918535232544, "learning_rate": 2.57413914265636e-05, "loss": 0.0461, "step": 6108 }, { "epoch": 4.29304286718201, "grad_norm": 0.44749537110328674, "learning_rate": 2.5745607870695716e-05, "loss": 0.0291, "step": 6109 }, { "epoch": 4.2937456078706955, "grad_norm": 0.3331450819969177, "learning_rate": 2.574982431482783e-05, "loss": 0.0349, "step": 6110 }, { "epoch": 4.294448348559381, "grad_norm": 0.3683828115463257, "learning_rate": 2.5754040758959946e-05, "loss": 0.0451, "step": 6111 }, { "epoch": 4.295151089248067, "grad_norm": 0.8886565566062927, "learning_rate": 2.575825720309206e-05, "loss": 0.0824, "step": 6112 }, { "epoch": 4.295853829936753, "grad_norm": 0.4906656742095947, "learning_rate": 2.5762473647224176e-05, "loss": 0.0977, "step": 6113 }, { "epoch": 4.296556570625439, "grad_norm": 0.5252396464347839, "learning_rate": 2.576669009135629e-05, "loss": 0.1158, "step": 6114 }, { "epoch": 4.297259311314125, "grad_norm": 0.8633352518081665, "learning_rate": 2.5770906535488406e-05, "loss": 0.2663, "step": 6115 }, { "epoch": 4.297962052002811, "grad_norm": 2.7477574348449707, "learning_rate": 2.577512297962052e-05, "loss": 0.2413, "step": 6116 }, { "epoch": 4.298664792691497, "grad_norm": 1.6177010536193848, "learning_rate": 2.5779339423752636e-05, "loss": 0.318, "step": 6117 }, { "epoch": 4.299367533380183, "grad_norm": 0.22713755071163177, "learning_rate": 2.578355586788475e-05, "loss": 0.0776, "step": 6118 }, { "epoch": 4.300070274068869, "grad_norm": 0.21152430772781372, "learning_rate": 2.578777231201687e-05, "loss": 0.0324, "step": 6119 }, { "epoch": 4.300773014757555, "grad_norm": 0.3540782034397125, "learning_rate": 2.5791988756148982e-05, "loss": 0.052, "step": 6120 }, { "epoch": 4.301475755446241, "grad_norm": 0.1482207179069519, "learning_rate": 2.57962052002811e-05, "loss": 0.0215, "step": 6121 }, { "epoch": 4.3021784961349265, "grad_norm": 0.16017338633537292, "learning_rate": 2.5800421644413212e-05, "loss": 0.0212, "step": 6122 }, { "epoch": 4.3028812368236125, "grad_norm": 0.17659388482570648, "learning_rate": 2.580463808854533e-05, "loss": 0.0285, "step": 6123 }, { "epoch": 4.303583977512298, "grad_norm": 0.19866788387298584, "learning_rate": 2.5808854532677442e-05, "loss": 0.0181, "step": 6124 }, { "epoch": 4.304286718200984, "grad_norm": 0.2732897102832794, "learning_rate": 2.581307097680956e-05, "loss": 0.0325, "step": 6125 }, { "epoch": 4.304989458889669, "grad_norm": 0.21642619371414185, "learning_rate": 2.5817287420941672e-05, "loss": 0.0299, "step": 6126 }, { "epoch": 4.305692199578355, "grad_norm": 0.1859620362520218, "learning_rate": 2.582150386507379e-05, "loss": 0.0245, "step": 6127 }, { "epoch": 4.306394940267041, "grad_norm": 0.17101044952869415, "learning_rate": 2.58257203092059e-05, "loss": 0.0231, "step": 6128 }, { "epoch": 4.307097680955727, "grad_norm": 0.15886440873146057, "learning_rate": 2.5829936753338018e-05, "loss": 0.0202, "step": 6129 }, { "epoch": 4.307800421644413, "grad_norm": 0.3299023509025574, "learning_rate": 2.583415319747013e-05, "loss": 0.0371, "step": 6130 }, { "epoch": 4.308503162333099, "grad_norm": 0.24428905546665192, "learning_rate": 2.5838369641602248e-05, "loss": 0.0219, "step": 6131 }, { "epoch": 4.309205903021785, "grad_norm": 0.2809430658817291, "learning_rate": 2.5842586085734365e-05, "loss": 0.0732, "step": 6132 }, { "epoch": 4.309908643710471, "grad_norm": 0.18197095394134521, "learning_rate": 2.584680252986648e-05, "loss": 0.0293, "step": 6133 }, { "epoch": 4.310611384399157, "grad_norm": 0.28427308797836304, "learning_rate": 2.5851018973998595e-05, "loss": 0.0424, "step": 6134 }, { "epoch": 4.311314125087843, "grad_norm": 0.33993300795555115, "learning_rate": 2.585523541813071e-05, "loss": 0.0504, "step": 6135 }, { "epoch": 4.3120168657765285, "grad_norm": 0.3112685978412628, "learning_rate": 2.5859451862262828e-05, "loss": 0.0636, "step": 6136 }, { "epoch": 4.3127196064652145, "grad_norm": 0.6814554333686829, "learning_rate": 2.586366830639494e-05, "loss": 0.0416, "step": 6137 }, { "epoch": 4.3134223471539, "grad_norm": 0.665193498134613, "learning_rate": 2.5867884750527058e-05, "loss": 0.105, "step": 6138 }, { "epoch": 4.314125087842586, "grad_norm": 0.49871915578842163, "learning_rate": 2.587210119465917e-05, "loss": 0.115, "step": 6139 }, { "epoch": 4.314827828531272, "grad_norm": 0.6913740038871765, "learning_rate": 2.5876317638791288e-05, "loss": 0.2009, "step": 6140 }, { "epoch": 4.315530569219958, "grad_norm": 9.671966552734375, "learning_rate": 2.58805340829234e-05, "loss": 0.2347, "step": 6141 }, { "epoch": 4.316233309908644, "grad_norm": 1.4363408088684082, "learning_rate": 2.5884750527055518e-05, "loss": 0.2889, "step": 6142 }, { "epoch": 4.31693605059733, "grad_norm": 0.40655747056007385, "learning_rate": 2.588896697118763e-05, "loss": 0.0844, "step": 6143 }, { "epoch": 4.317638791286015, "grad_norm": 0.13699668645858765, "learning_rate": 2.5893183415319748e-05, "loss": 0.0287, "step": 6144 }, { "epoch": 4.318341531974701, "grad_norm": 0.27639588713645935, "learning_rate": 2.589739985945186e-05, "loss": 0.0313, "step": 6145 }, { "epoch": 4.319044272663387, "grad_norm": 0.4222683906555176, "learning_rate": 2.590161630358398e-05, "loss": 0.0201, "step": 6146 }, { "epoch": 4.319747013352073, "grad_norm": 0.1633801907300949, "learning_rate": 2.5905832747716094e-05, "loss": 0.054, "step": 6147 }, { "epoch": 4.320449754040759, "grad_norm": 0.19978821277618408, "learning_rate": 2.591004919184821e-05, "loss": 0.0208, "step": 6148 }, { "epoch": 4.321152494729445, "grad_norm": 0.14511889219284058, "learning_rate": 2.5914265635980324e-05, "loss": 0.0224, "step": 6149 }, { "epoch": 4.3218552354181305, "grad_norm": 0.1896430402994156, "learning_rate": 2.591848208011244e-05, "loss": 0.0236, "step": 6150 }, { "epoch": 4.3225579761068165, "grad_norm": 0.22787301242351532, "learning_rate": 2.5922698524244554e-05, "loss": 0.035, "step": 6151 }, { "epoch": 4.323260716795502, "grad_norm": 0.2427653670310974, "learning_rate": 2.592691496837667e-05, "loss": 0.0321, "step": 6152 }, { "epoch": 4.323963457484188, "grad_norm": 0.24490854144096375, "learning_rate": 2.5931131412508784e-05, "loss": 0.0268, "step": 6153 }, { "epoch": 4.324666198172874, "grad_norm": 0.16777139902114868, "learning_rate": 2.59353478566409e-05, "loss": 0.024, "step": 6154 }, { "epoch": 4.32536893886156, "grad_norm": 0.1404649168252945, "learning_rate": 2.5939564300773014e-05, "loss": 0.0208, "step": 6155 }, { "epoch": 4.326071679550246, "grad_norm": 0.5499770045280457, "learning_rate": 2.594378074490513e-05, "loss": 0.0319, "step": 6156 }, { "epoch": 4.326774420238932, "grad_norm": 0.516811192035675, "learning_rate": 2.5947997189037243e-05, "loss": 0.0632, "step": 6157 }, { "epoch": 4.327477160927618, "grad_norm": 0.3670041263103485, "learning_rate": 2.595221363316936e-05, "loss": 0.0278, "step": 6158 }, { "epoch": 4.328179901616304, "grad_norm": 0.3672870397567749, "learning_rate": 2.5956430077301477e-05, "loss": 0.0436, "step": 6159 }, { "epoch": 4.32888264230499, "grad_norm": 0.23046116530895233, "learning_rate": 2.5960646521433593e-05, "loss": 0.0267, "step": 6160 }, { "epoch": 4.329585382993676, "grad_norm": 0.43216729164123535, "learning_rate": 2.5964862965565707e-05, "loss": 0.0762, "step": 6161 }, { "epoch": 4.330288123682362, "grad_norm": 0.35184282064437866, "learning_rate": 2.5969079409697823e-05, "loss": 0.0666, "step": 6162 }, { "epoch": 4.3309908643710475, "grad_norm": 0.5067138075828552, "learning_rate": 2.5973295853829937e-05, "loss": 0.0768, "step": 6163 }, { "epoch": 4.3316936050597326, "grad_norm": 1.101170539855957, "learning_rate": 2.5977512297962053e-05, "loss": 0.1473, "step": 6164 }, { "epoch": 4.3323963457484185, "grad_norm": 0.4744367301464081, "learning_rate": 2.5981728742094166e-05, "loss": 0.1707, "step": 6165 }, { "epoch": 4.333099086437104, "grad_norm": 1.326694130897522, "learning_rate": 2.5985945186226283e-05, "loss": 0.2029, "step": 6166 }, { "epoch": 4.33380182712579, "grad_norm": 2.0524587631225586, "learning_rate": 2.59901616303584e-05, "loss": 0.2967, "step": 6167 }, { "epoch": 4.334504567814476, "grad_norm": 0.3958429992198944, "learning_rate": 2.5994378074490513e-05, "loss": 0.0938, "step": 6168 }, { "epoch": 4.335207308503162, "grad_norm": 0.15169194340705872, "learning_rate": 2.599859451862263e-05, "loss": 0.0265, "step": 6169 }, { "epoch": 4.335910049191848, "grad_norm": 0.23575010895729065, "learning_rate": 2.6002810962754743e-05, "loss": 0.0204, "step": 6170 }, { "epoch": 4.336612789880534, "grad_norm": 0.2413557767868042, "learning_rate": 2.600702740688686e-05, "loss": 0.0261, "step": 6171 }, { "epoch": 4.33731553056922, "grad_norm": 0.15300188958644867, "learning_rate": 2.6011243851018973e-05, "loss": 0.0246, "step": 6172 }, { "epoch": 4.338018271257906, "grad_norm": 0.25138476490974426, "learning_rate": 2.6015460295151093e-05, "loss": 0.0193, "step": 6173 }, { "epoch": 4.338721011946592, "grad_norm": 0.19075770676136017, "learning_rate": 2.6019676739283206e-05, "loss": 0.0353, "step": 6174 }, { "epoch": 4.339423752635278, "grad_norm": 0.44510164856910706, "learning_rate": 2.6023893183415323e-05, "loss": 0.0281, "step": 6175 }, { "epoch": 4.340126493323964, "grad_norm": 0.19313347339630127, "learning_rate": 2.6028109627547436e-05, "loss": 0.0203, "step": 6176 }, { "epoch": 4.3408292340126495, "grad_norm": 0.20830024778842926, "learning_rate": 2.6032326071679553e-05, "loss": 0.02, "step": 6177 }, { "epoch": 4.3415319747013355, "grad_norm": 0.18321554362773895, "learning_rate": 2.6036542515811666e-05, "loss": 0.0257, "step": 6178 }, { "epoch": 4.342234715390021, "grad_norm": 0.14055325090885162, "learning_rate": 2.6040758959943782e-05, "loss": 0.0158, "step": 6179 }, { "epoch": 4.342937456078707, "grad_norm": 0.34267890453338623, "learning_rate": 2.6044975404075896e-05, "loss": 0.0366, "step": 6180 }, { "epoch": 4.343640196767393, "grad_norm": 0.39670756459236145, "learning_rate": 2.6049191848208012e-05, "loss": 0.028, "step": 6181 }, { "epoch": 4.344342937456078, "grad_norm": 0.3875651955604553, "learning_rate": 2.6053408292340126e-05, "loss": 0.0391, "step": 6182 }, { "epoch": 4.345045678144764, "grad_norm": 0.35789886116981506, "learning_rate": 2.6057624736472242e-05, "loss": 0.0466, "step": 6183 }, { "epoch": 4.34574841883345, "grad_norm": 0.18715670704841614, "learning_rate": 2.6061841180604355e-05, "loss": 0.0273, "step": 6184 }, { "epoch": 4.346451159522136, "grad_norm": 0.28187477588653564, "learning_rate": 2.6066057624736472e-05, "loss": 0.046, "step": 6185 }, { "epoch": 4.347153900210822, "grad_norm": 0.2604435384273529, "learning_rate": 2.6070274068868585e-05, "loss": 0.0454, "step": 6186 }, { "epoch": 4.347856640899508, "grad_norm": 0.3541487157344818, "learning_rate": 2.6074490513000705e-05, "loss": 0.0608, "step": 6187 }, { "epoch": 4.348559381588194, "grad_norm": 0.6447835564613342, "learning_rate": 2.607870695713282e-05, "loss": 0.143, "step": 6188 }, { "epoch": 4.34926212227688, "grad_norm": 0.6410861611366272, "learning_rate": 2.6082923401264935e-05, "loss": 0.1077, "step": 6189 }, { "epoch": 4.349964862965566, "grad_norm": 0.6993528008460999, "learning_rate": 2.608713984539705e-05, "loss": 0.185, "step": 6190 }, { "epoch": 4.3506676036542515, "grad_norm": 0.9993997812271118, "learning_rate": 2.6091356289529165e-05, "loss": 0.2422, "step": 6191 }, { "epoch": 4.3513703443429375, "grad_norm": 1.893342137336731, "learning_rate": 2.609557273366128e-05, "loss": 0.277, "step": 6192 }, { "epoch": 4.352073085031623, "grad_norm": 0.275725781917572, "learning_rate": 2.6099789177793395e-05, "loss": 0.0724, "step": 6193 }, { "epoch": 4.352775825720309, "grad_norm": 0.1685856133699417, "learning_rate": 2.6104005621925508e-05, "loss": 0.0381, "step": 6194 }, { "epoch": 4.353478566408995, "grad_norm": 0.3043741285800934, "learning_rate": 2.6108222066057625e-05, "loss": 0.0358, "step": 6195 }, { "epoch": 4.354181307097681, "grad_norm": 0.1808660477399826, "learning_rate": 2.611243851018974e-05, "loss": 0.0291, "step": 6196 }, { "epoch": 4.354884047786367, "grad_norm": 0.15172822773456573, "learning_rate": 2.6116654954321855e-05, "loss": 0.021, "step": 6197 }, { "epoch": 4.355586788475053, "grad_norm": 0.12951965630054474, "learning_rate": 2.612087139845397e-05, "loss": 0.0123, "step": 6198 }, { "epoch": 4.356289529163739, "grad_norm": 0.14511491358280182, "learning_rate": 2.6125087842586085e-05, "loss": 0.0143, "step": 6199 }, { "epoch": 4.356992269852425, "grad_norm": 0.2047601193189621, "learning_rate": 2.6129304286718205e-05, "loss": 0.0194, "step": 6200 }, { "epoch": 4.357695010541111, "grad_norm": 0.18350401520729065, "learning_rate": 2.6133520730850318e-05, "loss": 0.0251, "step": 6201 }, { "epoch": 4.358397751229797, "grad_norm": 0.17158405482769012, "learning_rate": 2.6137737174982435e-05, "loss": 0.0147, "step": 6202 }, { "epoch": 4.359100491918482, "grad_norm": 0.21181228756904602, "learning_rate": 2.6141953619114548e-05, "loss": 0.0304, "step": 6203 }, { "epoch": 4.359803232607168, "grad_norm": 0.2315390557050705, "learning_rate": 2.6146170063246664e-05, "loss": 0.0199, "step": 6204 }, { "epoch": 4.3605059732958535, "grad_norm": 0.2442435622215271, "learning_rate": 2.6150386507378778e-05, "loss": 0.0298, "step": 6205 }, { "epoch": 4.3612087139845395, "grad_norm": 0.19161804020404816, "learning_rate": 2.6154602951510894e-05, "loss": 0.0226, "step": 6206 }, { "epoch": 4.361911454673225, "grad_norm": 0.2913595736026764, "learning_rate": 2.6158819395643008e-05, "loss": 0.039, "step": 6207 }, { "epoch": 4.362614195361911, "grad_norm": 0.4369325041770935, "learning_rate": 2.6163035839775124e-05, "loss": 0.0508, "step": 6208 }, { "epoch": 4.363316936050597, "grad_norm": 0.21972137689590454, "learning_rate": 2.6167252283907237e-05, "loss": 0.0313, "step": 6209 }, { "epoch": 4.364019676739283, "grad_norm": 0.45947229862213135, "learning_rate": 2.6171468728039354e-05, "loss": 0.0494, "step": 6210 }, { "epoch": 4.364722417427969, "grad_norm": 0.3375554084777832, "learning_rate": 2.6175685172171467e-05, "loss": 0.0413, "step": 6211 }, { "epoch": 4.365425158116655, "grad_norm": 0.6366312503814697, "learning_rate": 2.6179901616303584e-05, "loss": 0.0733, "step": 6212 }, { "epoch": 4.366127898805341, "grad_norm": 0.4213358759880066, "learning_rate": 2.6184118060435697e-05, "loss": 0.0994, "step": 6213 }, { "epoch": 4.366830639494027, "grad_norm": 0.7462170124053955, "learning_rate": 2.6188334504567817e-05, "loss": 0.1589, "step": 6214 }, { "epoch": 4.367533380182713, "grad_norm": 0.7187256813049316, "learning_rate": 2.619255094869993e-05, "loss": 0.1608, "step": 6215 }, { "epoch": 4.368236120871399, "grad_norm": 1.048661231994629, "learning_rate": 2.6196767392832047e-05, "loss": 0.2975, "step": 6216 }, { "epoch": 4.368938861560085, "grad_norm": 1.3341624736785889, "learning_rate": 2.620098383696416e-05, "loss": 0.2747, "step": 6217 }, { "epoch": 4.3696416022487705, "grad_norm": 0.21095342934131622, "learning_rate": 2.6205200281096277e-05, "loss": 0.0714, "step": 6218 }, { "epoch": 4.370344342937456, "grad_norm": 0.19908985495567322, "learning_rate": 2.620941672522839e-05, "loss": 0.0284, "step": 6219 }, { "epoch": 4.371047083626142, "grad_norm": 0.1964283138513565, "learning_rate": 2.6213633169360507e-05, "loss": 0.0348, "step": 6220 }, { "epoch": 4.371749824314827, "grad_norm": 0.1717553287744522, "learning_rate": 2.621784961349262e-05, "loss": 0.0273, "step": 6221 }, { "epoch": 4.372452565003513, "grad_norm": 0.15662100911140442, "learning_rate": 2.6222066057624737e-05, "loss": 0.0219, "step": 6222 }, { "epoch": 4.373155305692199, "grad_norm": 0.1577717363834381, "learning_rate": 2.622628250175685e-05, "loss": 0.0163, "step": 6223 }, { "epoch": 4.373858046380885, "grad_norm": 0.19696426391601562, "learning_rate": 2.6230498945888967e-05, "loss": 0.0388, "step": 6224 }, { "epoch": 4.374560787069571, "grad_norm": 0.2711816430091858, "learning_rate": 2.6234715390021083e-05, "loss": 0.0297, "step": 6225 }, { "epoch": 4.375263527758257, "grad_norm": 0.21939322352409363, "learning_rate": 2.6238931834153197e-05, "loss": 0.034, "step": 6226 }, { "epoch": 4.375966268446943, "grad_norm": 0.175084188580513, "learning_rate": 2.6243148278285313e-05, "loss": 0.0121, "step": 6227 }, { "epoch": 4.376669009135629, "grad_norm": 0.19311434030532837, "learning_rate": 2.624736472241743e-05, "loss": 0.022, "step": 6228 }, { "epoch": 4.377371749824315, "grad_norm": 0.25751793384552, "learning_rate": 2.6251581166549547e-05, "loss": 0.0188, "step": 6229 }, { "epoch": 4.378074490513001, "grad_norm": 0.26443448662757874, "learning_rate": 2.625579761068166e-05, "loss": 0.0406, "step": 6230 }, { "epoch": 4.378777231201687, "grad_norm": 0.19408515095710754, "learning_rate": 2.6260014054813776e-05, "loss": 0.017, "step": 6231 }, { "epoch": 4.3794799718903725, "grad_norm": 0.2563977539539337, "learning_rate": 2.626423049894589e-05, "loss": 0.0397, "step": 6232 }, { "epoch": 4.3801827125790584, "grad_norm": 0.2234109342098236, "learning_rate": 2.6268446943078006e-05, "loss": 0.0328, "step": 6233 }, { "epoch": 4.380885453267744, "grad_norm": 0.5145342946052551, "learning_rate": 2.627266338721012e-05, "loss": 0.0243, "step": 6234 }, { "epoch": 4.38158819395643, "grad_norm": 0.33424195647239685, "learning_rate": 2.6276879831342236e-05, "loss": 0.0565, "step": 6235 }, { "epoch": 4.382290934645116, "grad_norm": 0.2561005651950836, "learning_rate": 2.628109627547435e-05, "loss": 0.0369, "step": 6236 }, { "epoch": 4.382993675333802, "grad_norm": 0.3781004548072815, "learning_rate": 2.6285312719606466e-05, "loss": 0.067, "step": 6237 }, { "epoch": 4.383696416022488, "grad_norm": 0.5685368180274963, "learning_rate": 2.628952916373858e-05, "loss": 0.0793, "step": 6238 }, { "epoch": 4.384399156711174, "grad_norm": 0.6726898550987244, "learning_rate": 2.6293745607870696e-05, "loss": 0.1386, "step": 6239 }, { "epoch": 4.38510189739986, "grad_norm": 0.7473098039627075, "learning_rate": 2.629796205200281e-05, "loss": 0.1701, "step": 6240 }, { "epoch": 4.385804638088545, "grad_norm": 0.9767914414405823, "learning_rate": 2.630217849613493e-05, "loss": 0.2228, "step": 6241 }, { "epoch": 4.386507378777231, "grad_norm": 1.7048335075378418, "learning_rate": 2.6306394940267042e-05, "loss": 0.2755, "step": 6242 }, { "epoch": 4.387210119465917, "grad_norm": 0.24858719110488892, "learning_rate": 2.631061138439916e-05, "loss": 0.0754, "step": 6243 }, { "epoch": 4.387912860154603, "grad_norm": 0.19834400713443756, "learning_rate": 2.6314827828531272e-05, "loss": 0.0253, "step": 6244 }, { "epoch": 4.388615600843289, "grad_norm": 0.25114378333091736, "learning_rate": 2.631904427266339e-05, "loss": 0.0346, "step": 6245 }, { "epoch": 4.3893183415319745, "grad_norm": 0.18584483861923218, "learning_rate": 2.6323260716795502e-05, "loss": 0.028, "step": 6246 }, { "epoch": 4.3900210822206605, "grad_norm": 0.21032488346099854, "learning_rate": 2.632747716092762e-05, "loss": 0.0385, "step": 6247 }, { "epoch": 4.390723822909346, "grad_norm": 0.21454963088035583, "learning_rate": 2.6331693605059732e-05, "loss": 0.0168, "step": 6248 }, { "epoch": 4.391426563598032, "grad_norm": 0.4432254731655121, "learning_rate": 2.633591004919185e-05, "loss": 0.0287, "step": 6249 }, { "epoch": 4.392129304286718, "grad_norm": 0.13522621989250183, "learning_rate": 2.6340126493323962e-05, "loss": 0.0227, "step": 6250 }, { "epoch": 4.392832044975404, "grad_norm": 0.2567808926105499, "learning_rate": 2.634434293745608e-05, "loss": 0.0292, "step": 6251 }, { "epoch": 4.39353478566409, "grad_norm": 0.1911407858133316, "learning_rate": 2.6348559381588192e-05, "loss": 0.0494, "step": 6252 }, { "epoch": 4.394237526352776, "grad_norm": 0.29765284061431885, "learning_rate": 2.635277582572031e-05, "loss": 0.0437, "step": 6253 }, { "epoch": 4.394940267041462, "grad_norm": 0.17665083706378937, "learning_rate": 2.6356992269852422e-05, "loss": 0.0146, "step": 6254 }, { "epoch": 4.395643007730148, "grad_norm": 0.17225484549999237, "learning_rate": 2.6361208713984542e-05, "loss": 0.0438, "step": 6255 }, { "epoch": 4.396345748418834, "grad_norm": 0.23305658996105194, "learning_rate": 2.636542515811666e-05, "loss": 0.0264, "step": 6256 }, { "epoch": 4.39704848910752, "grad_norm": 0.17773231863975525, "learning_rate": 2.6369641602248772e-05, "loss": 0.0212, "step": 6257 }, { "epoch": 4.397751229796206, "grad_norm": 0.2859762907028198, "learning_rate": 2.637385804638089e-05, "loss": 0.0411, "step": 6258 }, { "epoch": 4.398453970484891, "grad_norm": 0.22748306393623352, "learning_rate": 2.6378074490513e-05, "loss": 0.0239, "step": 6259 }, { "epoch": 4.3991567111735765, "grad_norm": 0.8704110980033875, "learning_rate": 2.6382290934645118e-05, "loss": 0.0397, "step": 6260 }, { "epoch": 4.3998594518622625, "grad_norm": 0.263258695602417, "learning_rate": 2.638650737877723e-05, "loss": 0.0309, "step": 6261 }, { "epoch": 4.400562192550948, "grad_norm": 0.3901433050632477, "learning_rate": 2.6390723822909348e-05, "loss": 0.0504, "step": 6262 }, { "epoch": 4.401264933239634, "grad_norm": 0.3004485070705414, "learning_rate": 2.639494026704146e-05, "loss": 0.0876, "step": 6263 }, { "epoch": 4.40196767392832, "grad_norm": 0.4892844557762146, "learning_rate": 2.6399156711173578e-05, "loss": 0.1234, "step": 6264 }, { "epoch": 4.402670414617006, "grad_norm": 0.6288076043128967, "learning_rate": 2.640337315530569e-05, "loss": 0.2283, "step": 6265 }, { "epoch": 4.403373155305692, "grad_norm": 0.7423327565193176, "learning_rate": 2.6407589599437808e-05, "loss": 0.2051, "step": 6266 }, { "epoch": 4.404075895994378, "grad_norm": 1.3284364938735962, "learning_rate": 2.641180604356992e-05, "loss": 0.3359, "step": 6267 }, { "epoch": 4.404778636683064, "grad_norm": 0.5940676331520081, "learning_rate": 2.641602248770204e-05, "loss": 0.0958, "step": 6268 }, { "epoch": 4.40548137737175, "grad_norm": 0.15911366045475006, "learning_rate": 2.6420238931834154e-05, "loss": 0.0384, "step": 6269 }, { "epoch": 4.406184118060436, "grad_norm": 0.12998302280902863, "learning_rate": 2.642445537596627e-05, "loss": 0.0353, "step": 6270 }, { "epoch": 4.406886858749122, "grad_norm": 0.1619381457567215, "learning_rate": 2.6428671820098384e-05, "loss": 0.0201, "step": 6271 }, { "epoch": 4.407589599437808, "grad_norm": 0.13294540345668793, "learning_rate": 2.64328882642305e-05, "loss": 0.0264, "step": 6272 }, { "epoch": 4.4082923401264935, "grad_norm": 0.16425421833992004, "learning_rate": 2.6437104708362614e-05, "loss": 0.0206, "step": 6273 }, { "epoch": 4.408995080815179, "grad_norm": 0.9322714805603027, "learning_rate": 2.644132115249473e-05, "loss": 0.0267, "step": 6274 }, { "epoch": 4.409697821503865, "grad_norm": 0.14889688789844513, "learning_rate": 2.6445537596626844e-05, "loss": 0.0144, "step": 6275 }, { "epoch": 4.410400562192551, "grad_norm": 0.18086488544940948, "learning_rate": 2.644975404075896e-05, "loss": 0.0282, "step": 6276 }, { "epoch": 4.411103302881237, "grad_norm": 0.20108292996883392, "learning_rate": 2.6453970484891074e-05, "loss": 0.0232, "step": 6277 }, { "epoch": 4.411806043569923, "grad_norm": 0.11176098883152008, "learning_rate": 2.645818692902319e-05, "loss": 0.0156, "step": 6278 }, { "epoch": 4.412508784258609, "grad_norm": 0.21956723928451538, "learning_rate": 2.6462403373155304e-05, "loss": 0.0292, "step": 6279 }, { "epoch": 4.413211524947294, "grad_norm": 0.28387513756752014, "learning_rate": 2.646661981728742e-05, "loss": 0.0428, "step": 6280 }, { "epoch": 4.41391426563598, "grad_norm": 0.2973315417766571, "learning_rate": 2.6470836261419534e-05, "loss": 0.0289, "step": 6281 }, { "epoch": 4.414617006324666, "grad_norm": 0.15899303555488586, "learning_rate": 2.6475052705551654e-05, "loss": 0.0297, "step": 6282 }, { "epoch": 4.415319747013352, "grad_norm": 0.29281488060951233, "learning_rate": 2.6479269149683767e-05, "loss": 0.0429, "step": 6283 }, { "epoch": 4.416022487702038, "grad_norm": 0.2542743682861328, "learning_rate": 2.6483485593815884e-05, "loss": 0.0387, "step": 6284 }, { "epoch": 4.416725228390724, "grad_norm": 0.4260026812553406, "learning_rate": 2.6487702037948e-05, "loss": 0.0452, "step": 6285 }, { "epoch": 4.41742796907941, "grad_norm": 0.43525344133377075, "learning_rate": 2.6491918482080114e-05, "loss": 0.0728, "step": 6286 }, { "epoch": 4.4181307097680955, "grad_norm": 0.30731528997421265, "learning_rate": 2.649613492621223e-05, "loss": 0.0691, "step": 6287 }, { "epoch": 4.418833450456781, "grad_norm": 0.6472182869911194, "learning_rate": 2.6500351370344343e-05, "loss": 0.0864, "step": 6288 }, { "epoch": 4.419536191145467, "grad_norm": 0.7756013870239258, "learning_rate": 2.650456781447646e-05, "loss": 0.1781, "step": 6289 }, { "epoch": 4.420238931834153, "grad_norm": 0.67159503698349, "learning_rate": 2.6508784258608573e-05, "loss": 0.1742, "step": 6290 }, { "epoch": 4.420941672522839, "grad_norm": 0.8215969204902649, "learning_rate": 2.651300070274069e-05, "loss": 0.2214, "step": 6291 }, { "epoch": 4.421644413211525, "grad_norm": 1.5572675466537476, "learning_rate": 2.6517217146872803e-05, "loss": 0.3127, "step": 6292 }, { "epoch": 4.422347153900211, "grad_norm": 0.2532278895378113, "learning_rate": 2.652143359100492e-05, "loss": 0.0866, "step": 6293 }, { "epoch": 4.423049894588897, "grad_norm": 0.19466643035411835, "learning_rate": 2.6525650035137033e-05, "loss": 0.0519, "step": 6294 }, { "epoch": 4.423752635277583, "grad_norm": 0.18857035040855408, "learning_rate": 2.652986647926915e-05, "loss": 0.049, "step": 6295 }, { "epoch": 4.424455375966269, "grad_norm": 0.22530107200145721, "learning_rate": 2.6534082923401266e-05, "loss": 0.0243, "step": 6296 }, { "epoch": 4.425158116654955, "grad_norm": 0.1924951821565628, "learning_rate": 2.6538299367533383e-05, "loss": 0.018, "step": 6297 }, { "epoch": 4.42586085734364, "grad_norm": 0.16532135009765625, "learning_rate": 2.6542515811665496e-05, "loss": 0.0161, "step": 6298 }, { "epoch": 4.426563598032326, "grad_norm": 0.26626282930374146, "learning_rate": 2.6546732255797613e-05, "loss": 0.0191, "step": 6299 }, { "epoch": 4.427266338721012, "grad_norm": 0.19726251065731049, "learning_rate": 2.6550948699929726e-05, "loss": 0.0236, "step": 6300 }, { "epoch": 4.4279690794096975, "grad_norm": 0.2127082347869873, "learning_rate": 2.6555165144061843e-05, "loss": 0.0333, "step": 6301 }, { "epoch": 4.4286718200983834, "grad_norm": 0.2615152597427368, "learning_rate": 2.6559381588193956e-05, "loss": 0.0209, "step": 6302 }, { "epoch": 4.429374560787069, "grad_norm": 2.1137709617614746, "learning_rate": 2.6563598032326073e-05, "loss": 0.0279, "step": 6303 }, { "epoch": 4.430077301475755, "grad_norm": 0.18122050166130066, "learning_rate": 2.6567814476458186e-05, "loss": 0.0159, "step": 6304 }, { "epoch": 4.430780042164441, "grad_norm": 0.27765122056007385, "learning_rate": 2.6572030920590303e-05, "loss": 0.0535, "step": 6305 }, { "epoch": 4.431482782853127, "grad_norm": 0.27842482924461365, "learning_rate": 2.6576247364722416e-05, "loss": 0.0415, "step": 6306 }, { "epoch": 4.432185523541813, "grad_norm": 0.3060230612754822, "learning_rate": 2.6580463808854532e-05, "loss": 0.0292, "step": 6307 }, { "epoch": 4.432888264230499, "grad_norm": 0.1849532425403595, "learning_rate": 2.6584680252986646e-05, "loss": 0.0271, "step": 6308 }, { "epoch": 4.433591004919185, "grad_norm": 0.1932356357574463, "learning_rate": 2.6588896697118766e-05, "loss": 0.0439, "step": 6309 }, { "epoch": 4.434293745607871, "grad_norm": 0.2954069972038269, "learning_rate": 2.659311314125088e-05, "loss": 0.0592, "step": 6310 }, { "epoch": 4.434996486296557, "grad_norm": 0.2887956500053406, "learning_rate": 2.6597329585382996e-05, "loss": 0.0513, "step": 6311 }, { "epoch": 4.435699226985243, "grad_norm": 0.35349059104919434, "learning_rate": 2.660154602951511e-05, "loss": 0.0773, "step": 6312 }, { "epoch": 4.436401967673929, "grad_norm": 0.41624075174331665, "learning_rate": 2.6605762473647226e-05, "loss": 0.0789, "step": 6313 }, { "epoch": 4.4371047083626145, "grad_norm": 0.760941743850708, "learning_rate": 2.660997891777934e-05, "loss": 0.1654, "step": 6314 }, { "epoch": 4.4378074490513, "grad_norm": 0.7845421433448792, "learning_rate": 2.6614195361911455e-05, "loss": 0.1884, "step": 6315 }, { "epoch": 4.438510189739986, "grad_norm": 0.8832811713218689, "learning_rate": 2.6618411806043572e-05, "loss": 0.2515, "step": 6316 }, { "epoch": 4.439212930428672, "grad_norm": 3.0146477222442627, "learning_rate": 2.6622628250175685e-05, "loss": 0.2716, "step": 6317 }, { "epoch": 4.439915671117357, "grad_norm": 1.0802046060562134, "learning_rate": 2.6626844694307802e-05, "loss": 0.109, "step": 6318 }, { "epoch": 4.440618411806043, "grad_norm": 0.20422765612602234, "learning_rate": 2.6631061138439915e-05, "loss": 0.0326, "step": 6319 }, { "epoch": 4.441321152494729, "grad_norm": 0.2587590217590332, "learning_rate": 2.6635277582572032e-05, "loss": 0.0268, "step": 6320 }, { "epoch": 4.442023893183415, "grad_norm": 0.1682039201259613, "learning_rate": 2.6639494026704145e-05, "loss": 0.0217, "step": 6321 }, { "epoch": 4.442726633872101, "grad_norm": 0.12235955893993378, "learning_rate": 2.664371047083626e-05, "loss": 0.0195, "step": 6322 }, { "epoch": 4.443429374560787, "grad_norm": 0.14725498855113983, "learning_rate": 2.664792691496838e-05, "loss": 0.014, "step": 6323 }, { "epoch": 4.444132115249473, "grad_norm": 0.1905423104763031, "learning_rate": 2.6652143359100495e-05, "loss": 0.0159, "step": 6324 }, { "epoch": 4.444834855938159, "grad_norm": 0.414224237203598, "learning_rate": 2.6656359803232608e-05, "loss": 0.0311, "step": 6325 }, { "epoch": 4.445537596626845, "grad_norm": 0.26092520356178284, "learning_rate": 2.6660576247364725e-05, "loss": 0.0308, "step": 6326 }, { "epoch": 4.446240337315531, "grad_norm": 0.7098008394241333, "learning_rate": 2.6664792691496838e-05, "loss": 0.0386, "step": 6327 }, { "epoch": 4.4469430780042165, "grad_norm": 0.19301410019397736, "learning_rate": 2.6669009135628955e-05, "loss": 0.0266, "step": 6328 }, { "epoch": 4.447645818692902, "grad_norm": 0.17936769127845764, "learning_rate": 2.6673225579761068e-05, "loss": 0.0275, "step": 6329 }, { "epoch": 4.448348559381588, "grad_norm": 0.19276918470859528, "learning_rate": 2.6677442023893185e-05, "loss": 0.0294, "step": 6330 }, { "epoch": 4.449051300070274, "grad_norm": 0.26864904165267944, "learning_rate": 2.6681658468025298e-05, "loss": 0.0539, "step": 6331 }, { "epoch": 4.44975404075896, "grad_norm": 0.37388819456100464, "learning_rate": 2.6685874912157415e-05, "loss": 0.0696, "step": 6332 }, { "epoch": 4.450456781447646, "grad_norm": 0.2703861594200134, "learning_rate": 2.6690091356289528e-05, "loss": 0.0249, "step": 6333 }, { "epoch": 4.451159522136332, "grad_norm": 0.388005793094635, "learning_rate": 2.6694307800421644e-05, "loss": 0.0619, "step": 6334 }, { "epoch": 4.451862262825018, "grad_norm": 0.12598618865013123, "learning_rate": 2.6698524244553758e-05, "loss": 0.0262, "step": 6335 }, { "epoch": 4.452565003513703, "grad_norm": 0.34327811002731323, "learning_rate": 2.6702740688685878e-05, "loss": 0.0728, "step": 6336 }, { "epoch": 4.453267744202389, "grad_norm": 0.33255481719970703, "learning_rate": 2.670695713281799e-05, "loss": 0.0653, "step": 6337 }, { "epoch": 4.453970484891075, "grad_norm": 0.5353061556816101, "learning_rate": 2.6711173576950108e-05, "loss": 0.1089, "step": 6338 }, { "epoch": 4.454673225579761, "grad_norm": 0.41574716567993164, "learning_rate": 2.671539002108222e-05, "loss": 0.1262, "step": 6339 }, { "epoch": 4.455375966268447, "grad_norm": 0.977558970451355, "learning_rate": 2.6719606465214337e-05, "loss": 0.2025, "step": 6340 }, { "epoch": 4.456078706957133, "grad_norm": 2.1099772453308105, "learning_rate": 2.672382290934645e-05, "loss": 0.2427, "step": 6341 }, { "epoch": 4.4567814476458185, "grad_norm": 1.6386473178863525, "learning_rate": 2.6728039353478567e-05, "loss": 0.3155, "step": 6342 }, { "epoch": 4.457484188334504, "grad_norm": 0.23995661735534668, "learning_rate": 2.673225579761068e-05, "loss": 0.0817, "step": 6343 }, { "epoch": 4.45818692902319, "grad_norm": 0.12825354933738708, "learning_rate": 2.6736472241742797e-05, "loss": 0.0291, "step": 6344 }, { "epoch": 4.458889669711876, "grad_norm": 0.25176718831062317, "learning_rate": 2.6740688685874914e-05, "loss": 0.0383, "step": 6345 }, { "epoch": 4.459592410400562, "grad_norm": 0.25887331366539, "learning_rate": 2.6744905130007027e-05, "loss": 0.0304, "step": 6346 }, { "epoch": 4.460295151089248, "grad_norm": 0.20030254125595093, "learning_rate": 2.6749121574139144e-05, "loss": 0.0344, "step": 6347 }, { "epoch": 4.460997891777934, "grad_norm": 0.27179643511772156, "learning_rate": 2.6753338018271257e-05, "loss": 0.0259, "step": 6348 }, { "epoch": 4.46170063246662, "grad_norm": 0.2068435102701187, "learning_rate": 2.6757554462403374e-05, "loss": 0.0192, "step": 6349 }, { "epoch": 4.462403373155306, "grad_norm": 0.2532503604888916, "learning_rate": 2.676177090653549e-05, "loss": 0.0232, "step": 6350 }, { "epoch": 4.463106113843992, "grad_norm": 0.2009878009557724, "learning_rate": 2.6765987350667607e-05, "loss": 0.044, "step": 6351 }, { "epoch": 4.463808854532678, "grad_norm": 0.34141817688941956, "learning_rate": 2.677020379479972e-05, "loss": 0.0225, "step": 6352 }, { "epoch": 4.464511595221364, "grad_norm": 0.2310340255498886, "learning_rate": 2.6774420238931837e-05, "loss": 0.0336, "step": 6353 }, { "epoch": 4.46521433591005, "grad_norm": 0.14149099588394165, "learning_rate": 2.677863668306395e-05, "loss": 0.026, "step": 6354 }, { "epoch": 4.4659170765987355, "grad_norm": 0.4242728054523468, "learning_rate": 2.6782853127196067e-05, "loss": 0.0261, "step": 6355 }, { "epoch": 4.466619817287421, "grad_norm": 0.2613811790943146, "learning_rate": 2.678706957132818e-05, "loss": 0.0384, "step": 6356 }, { "epoch": 4.4673225579761064, "grad_norm": 0.2054900825023651, "learning_rate": 2.6791286015460297e-05, "loss": 0.0406, "step": 6357 }, { "epoch": 4.468025298664792, "grad_norm": 0.2784702777862549, "learning_rate": 2.679550245959241e-05, "loss": 0.0583, "step": 6358 }, { "epoch": 4.468728039353478, "grad_norm": 0.17776893079280853, "learning_rate": 2.6799718903724526e-05, "loss": 0.0276, "step": 6359 }, { "epoch": 4.469430780042164, "grad_norm": 0.3318902254104614, "learning_rate": 2.680393534785664e-05, "loss": 0.0773, "step": 6360 }, { "epoch": 4.47013352073085, "grad_norm": 0.26291364431381226, "learning_rate": 2.6808151791988756e-05, "loss": 0.0334, "step": 6361 }, { "epoch": 4.470836261419536, "grad_norm": 0.4211001396179199, "learning_rate": 2.681236823612087e-05, "loss": 0.0825, "step": 6362 }, { "epoch": 4.471539002108222, "grad_norm": 0.5891136527061462, "learning_rate": 2.6816584680252986e-05, "loss": 0.0743, "step": 6363 }, { "epoch": 4.472241742796908, "grad_norm": 0.5030520558357239, "learning_rate": 2.6820801124385103e-05, "loss": 0.1597, "step": 6364 }, { "epoch": 4.472944483485594, "grad_norm": 0.7228496074676514, "learning_rate": 2.682501756851722e-05, "loss": 0.1961, "step": 6365 }, { "epoch": 4.47364722417428, "grad_norm": 0.9854082465171814, "learning_rate": 2.6829234012649333e-05, "loss": 0.217, "step": 6366 }, { "epoch": 4.474349964862966, "grad_norm": 1.7496024370193481, "learning_rate": 2.683345045678145e-05, "loss": 0.297, "step": 6367 }, { "epoch": 4.475052705551652, "grad_norm": 0.32406923174858093, "learning_rate": 2.6837666900913563e-05, "loss": 0.0836, "step": 6368 }, { "epoch": 4.4757554462403375, "grad_norm": 0.23365014791488647, "learning_rate": 2.684188334504568e-05, "loss": 0.0442, "step": 6369 }, { "epoch": 4.476458186929023, "grad_norm": 0.22642019391059875, "learning_rate": 2.6846099789177793e-05, "loss": 0.0458, "step": 6370 }, { "epoch": 4.477160927617709, "grad_norm": 0.20930291712284088, "learning_rate": 2.685031623330991e-05, "loss": 0.0206, "step": 6371 }, { "epoch": 4.477863668306395, "grad_norm": 0.21858441829681396, "learning_rate": 2.6854532677442022e-05, "loss": 0.0222, "step": 6372 }, { "epoch": 4.478566408995081, "grad_norm": 0.1655815988779068, "learning_rate": 2.685874912157414e-05, "loss": 0.0281, "step": 6373 }, { "epoch": 4.479269149683767, "grad_norm": 0.41372236609458923, "learning_rate": 2.6862965565706256e-05, "loss": 0.0384, "step": 6374 }, { "epoch": 4.479971890372452, "grad_norm": 0.21914879977703094, "learning_rate": 2.686718200983837e-05, "loss": 0.039, "step": 6375 }, { "epoch": 4.480674631061138, "grad_norm": 0.13800182938575745, "learning_rate": 2.6871398453970486e-05, "loss": 0.0267, "step": 6376 }, { "epoch": 4.481377371749824, "grad_norm": 0.15210869908332825, "learning_rate": 2.6875614898102602e-05, "loss": 0.0165, "step": 6377 }, { "epoch": 4.48208011243851, "grad_norm": 0.3598565459251404, "learning_rate": 2.687983134223472e-05, "loss": 0.0361, "step": 6378 }, { "epoch": 4.482782853127196, "grad_norm": 0.23583249747753143, "learning_rate": 2.6884047786366832e-05, "loss": 0.0277, "step": 6379 }, { "epoch": 4.483485593815882, "grad_norm": 0.2876974940299988, "learning_rate": 2.688826423049895e-05, "loss": 0.0405, "step": 6380 }, { "epoch": 4.484188334504568, "grad_norm": 0.19904173910617828, "learning_rate": 2.6892480674631062e-05, "loss": 0.0234, "step": 6381 }, { "epoch": 4.484891075193254, "grad_norm": 0.5021383762359619, "learning_rate": 2.689669711876318e-05, "loss": 0.0377, "step": 6382 }, { "epoch": 4.4855938158819395, "grad_norm": 0.24528902769088745, "learning_rate": 2.6900913562895292e-05, "loss": 0.0421, "step": 6383 }, { "epoch": 4.486296556570625, "grad_norm": 0.23796282708644867, "learning_rate": 2.690513000702741e-05, "loss": 0.0252, "step": 6384 }, { "epoch": 4.486999297259311, "grad_norm": 0.2705674469470978, "learning_rate": 2.6909346451159522e-05, "loss": 0.0456, "step": 6385 }, { "epoch": 4.487702037947997, "grad_norm": 0.2979729473590851, "learning_rate": 2.691356289529164e-05, "loss": 0.049, "step": 6386 }, { "epoch": 4.488404778636683, "grad_norm": 0.3342866897583008, "learning_rate": 2.691777933942375e-05, "loss": 0.0632, "step": 6387 }, { "epoch": 4.489107519325369, "grad_norm": 0.8999193906784058, "learning_rate": 2.6921995783555868e-05, "loss": 0.0725, "step": 6388 }, { "epoch": 4.489810260014055, "grad_norm": 0.6978486776351929, "learning_rate": 2.692621222768798e-05, "loss": 0.1468, "step": 6389 }, { "epoch": 4.490513000702741, "grad_norm": 5.052612781524658, "learning_rate": 2.6930428671820098e-05, "loss": 0.2025, "step": 6390 }, { "epoch": 4.491215741391427, "grad_norm": 1.047018051147461, "learning_rate": 2.6934645115952215e-05, "loss": 0.2698, "step": 6391 }, { "epoch": 4.491918482080113, "grad_norm": 1.397857666015625, "learning_rate": 2.693886156008433e-05, "loss": 0.2913, "step": 6392 }, { "epoch": 4.492621222768799, "grad_norm": 0.25898104906082153, "learning_rate": 2.6943078004216445e-05, "loss": 0.0646, "step": 6393 }, { "epoch": 4.493323963457485, "grad_norm": 0.22272361814975739, "learning_rate": 2.694729444834856e-05, "loss": 0.0364, "step": 6394 }, { "epoch": 4.49402670414617, "grad_norm": 0.3499932885169983, "learning_rate": 2.6951510892480675e-05, "loss": 0.0403, "step": 6395 }, { "epoch": 4.494729444834856, "grad_norm": 0.17736509442329407, "learning_rate": 2.695572733661279e-05, "loss": 0.0254, "step": 6396 }, { "epoch": 4.4954321855235415, "grad_norm": 0.1957833468914032, "learning_rate": 2.6959943780744904e-05, "loss": 0.0234, "step": 6397 }, { "epoch": 4.496134926212227, "grad_norm": 0.17073148488998413, "learning_rate": 2.696416022487702e-05, "loss": 0.0188, "step": 6398 }, { "epoch": 4.496837666900913, "grad_norm": 0.27238699793815613, "learning_rate": 2.6968376669009134e-05, "loss": 0.0248, "step": 6399 }, { "epoch": 4.497540407589599, "grad_norm": 0.22141489386558533, "learning_rate": 2.697259311314125e-05, "loss": 0.0307, "step": 6400 }, { "epoch": 4.498243148278285, "grad_norm": 0.26561543345451355, "learning_rate": 2.6976809557273364e-05, "loss": 0.0343, "step": 6401 }, { "epoch": 4.498945888966971, "grad_norm": 0.3577340543270111, "learning_rate": 2.698102600140548e-05, "loss": 0.0164, "step": 6402 }, { "epoch": 4.499648629655657, "grad_norm": 0.3066326677799225, "learning_rate": 2.6985242445537594e-05, "loss": 0.0464, "step": 6403 }, { "epoch": 4.500351370344343, "grad_norm": 0.28493526577949524, "learning_rate": 2.6989458889669714e-05, "loss": 0.0166, "step": 6404 }, { "epoch": 4.501054111033029, "grad_norm": 0.21338914334774017, "learning_rate": 2.699367533380183e-05, "loss": 0.0417, "step": 6405 }, { "epoch": 4.501756851721715, "grad_norm": 0.22559216618537903, "learning_rate": 2.6997891777933944e-05, "loss": 0.0214, "step": 6406 }, { "epoch": 4.502459592410401, "grad_norm": 0.3460957407951355, "learning_rate": 2.700210822206606e-05, "loss": 0.0582, "step": 6407 }, { "epoch": 4.503162333099087, "grad_norm": 0.397222638130188, "learning_rate": 2.7006324666198174e-05, "loss": 0.0599, "step": 6408 }, { "epoch": 4.503865073787773, "grad_norm": 0.16646535694599152, "learning_rate": 2.701054111033029e-05, "loss": 0.0217, "step": 6409 }, { "epoch": 4.5045678144764585, "grad_norm": 0.3096378743648529, "learning_rate": 2.7014757554462404e-05, "loss": 0.0474, "step": 6410 }, { "epoch": 4.505270555165144, "grad_norm": 0.2818789780139923, "learning_rate": 2.701897399859452e-05, "loss": 0.0509, "step": 6411 }, { "epoch": 4.505973295853829, "grad_norm": 0.3938984274864197, "learning_rate": 2.7023190442726634e-05, "loss": 0.063, "step": 6412 }, { "epoch": 4.506676036542515, "grad_norm": 0.374308317899704, "learning_rate": 2.702740688685875e-05, "loss": 0.0895, "step": 6413 }, { "epoch": 4.507378777231201, "grad_norm": 0.41450732946395874, "learning_rate": 2.7031623330990864e-05, "loss": 0.099, "step": 6414 }, { "epoch": 4.508081517919887, "grad_norm": 0.5774344205856323, "learning_rate": 2.703583977512298e-05, "loss": 0.1529, "step": 6415 }, { "epoch": 4.508784258608573, "grad_norm": 1.321242094039917, "learning_rate": 2.7040056219255093e-05, "loss": 0.2554, "step": 6416 }, { "epoch": 4.509486999297259, "grad_norm": 1.4166388511657715, "learning_rate": 2.704427266338721e-05, "loss": 0.2935, "step": 6417 }, { "epoch": 4.510189739985945, "grad_norm": 0.29790425300598145, "learning_rate": 2.7048489107519327e-05, "loss": 0.0731, "step": 6418 }, { "epoch": 4.510892480674631, "grad_norm": 0.1685221940279007, "learning_rate": 2.7052705551651443e-05, "loss": 0.0268, "step": 6419 }, { "epoch": 4.511595221363317, "grad_norm": 0.22333140671253204, "learning_rate": 2.7056921995783557e-05, "loss": 0.0245, "step": 6420 }, { "epoch": 4.512297962052003, "grad_norm": 0.14885854721069336, "learning_rate": 2.7061138439915673e-05, "loss": 0.0231, "step": 6421 }, { "epoch": 4.513000702740689, "grad_norm": 0.25812995433807373, "learning_rate": 2.7065354884047787e-05, "loss": 0.0273, "step": 6422 }, { "epoch": 4.513703443429375, "grad_norm": 0.15139225125312805, "learning_rate": 2.7069571328179903e-05, "loss": 0.0164, "step": 6423 }, { "epoch": 4.5144061841180605, "grad_norm": 0.4990427792072296, "learning_rate": 2.7073787772312016e-05, "loss": 0.0432, "step": 6424 }, { "epoch": 4.515108924806746, "grad_norm": 0.1565137356519699, "learning_rate": 2.7078004216444133e-05, "loss": 0.0261, "step": 6425 }, { "epoch": 4.515811665495432, "grad_norm": 0.3284175395965576, "learning_rate": 2.7082220660576246e-05, "loss": 0.0283, "step": 6426 }, { "epoch": 4.516514406184118, "grad_norm": 0.2961803376674652, "learning_rate": 2.7086437104708363e-05, "loss": 0.0179, "step": 6427 }, { "epoch": 4.517217146872804, "grad_norm": 0.5000996589660645, "learning_rate": 2.7090653548840476e-05, "loss": 0.0323, "step": 6428 }, { "epoch": 4.51791988756149, "grad_norm": 0.2096172571182251, "learning_rate": 2.7094869992972593e-05, "loss": 0.0218, "step": 6429 }, { "epoch": 4.518622628250176, "grad_norm": 0.5631721019744873, "learning_rate": 2.7099086437104706e-05, "loss": 0.0429, "step": 6430 }, { "epoch": 4.519325368938862, "grad_norm": 0.2489878088235855, "learning_rate": 2.7103302881236826e-05, "loss": 0.0228, "step": 6431 }, { "epoch": 4.520028109627548, "grad_norm": 0.4175037741661072, "learning_rate": 2.710751932536894e-05, "loss": 0.035, "step": 6432 }, { "epoch": 4.520730850316234, "grad_norm": 0.332663893699646, "learning_rate": 2.7111735769501056e-05, "loss": 0.0439, "step": 6433 }, { "epoch": 4.521433591004919, "grad_norm": 0.3271087110042572, "learning_rate": 2.7115952213633173e-05, "loss": 0.0293, "step": 6434 }, { "epoch": 4.522136331693605, "grad_norm": 0.4690577983856201, "learning_rate": 2.7120168657765286e-05, "loss": 0.0401, "step": 6435 }, { "epoch": 4.522839072382291, "grad_norm": 0.3940000534057617, "learning_rate": 2.7124385101897403e-05, "loss": 0.0452, "step": 6436 }, { "epoch": 4.523541813070977, "grad_norm": 0.3596409857273102, "learning_rate": 2.7128601546029516e-05, "loss": 0.0659, "step": 6437 }, { "epoch": 4.5242445537596625, "grad_norm": 0.4656267464160919, "learning_rate": 2.7132817990161632e-05, "loss": 0.1001, "step": 6438 }, { "epoch": 4.524947294448348, "grad_norm": 0.8656374216079712, "learning_rate": 2.7137034434293746e-05, "loss": 0.1629, "step": 6439 }, { "epoch": 4.525650035137034, "grad_norm": 0.7040641903877258, "learning_rate": 2.7141250878425862e-05, "loss": 0.1871, "step": 6440 }, { "epoch": 4.52635277582572, "grad_norm": 1.7870997190475464, "learning_rate": 2.7145467322557976e-05, "loss": 0.2445, "step": 6441 }, { "epoch": 4.527055516514406, "grad_norm": 1.4679051637649536, "learning_rate": 2.7149683766690092e-05, "loss": 0.321, "step": 6442 }, { "epoch": 4.527758257203092, "grad_norm": 0.2557228207588196, "learning_rate": 2.7153900210822205e-05, "loss": 0.0849, "step": 6443 }, { "epoch": 4.528460997891778, "grad_norm": 0.21338623762130737, "learning_rate": 2.7158116654954322e-05, "loss": 0.0374, "step": 6444 }, { "epoch": 4.529163738580464, "grad_norm": 0.17466895282268524, "learning_rate": 2.716233309908644e-05, "loss": 0.0184, "step": 6445 }, { "epoch": 4.52986647926915, "grad_norm": 0.1516409069299698, "learning_rate": 2.7166549543218555e-05, "loss": 0.0262, "step": 6446 }, { "epoch": 4.530569219957836, "grad_norm": 0.13219879567623138, "learning_rate": 2.717076598735067e-05, "loss": 0.0134, "step": 6447 }, { "epoch": 4.531271960646522, "grad_norm": 0.20892737805843353, "learning_rate": 2.7174982431482785e-05, "loss": 0.0194, "step": 6448 }, { "epoch": 4.531974701335208, "grad_norm": 0.20950064063072205, "learning_rate": 2.71791988756149e-05, "loss": 0.0202, "step": 6449 }, { "epoch": 4.5326774420238936, "grad_norm": 0.23892256617546082, "learning_rate": 2.7183415319747015e-05, "loss": 0.0304, "step": 6450 }, { "epoch": 4.533380182712579, "grad_norm": 0.24873027205467224, "learning_rate": 2.718763176387913e-05, "loss": 0.0286, "step": 6451 }, { "epoch": 4.5340829234012645, "grad_norm": 0.17526990175247192, "learning_rate": 2.7191848208011245e-05, "loss": 0.0257, "step": 6452 }, { "epoch": 4.53478566408995, "grad_norm": 0.35313940048217773, "learning_rate": 2.7196064652143358e-05, "loss": 0.0346, "step": 6453 }, { "epoch": 4.535488404778636, "grad_norm": 0.18224886059761047, "learning_rate": 2.7200281096275475e-05, "loss": 0.0126, "step": 6454 }, { "epoch": 4.536191145467322, "grad_norm": 0.2733128070831299, "learning_rate": 2.7204497540407588e-05, "loss": 0.036, "step": 6455 }, { "epoch": 4.536893886156008, "grad_norm": 0.2339337170124054, "learning_rate": 2.7208713984539705e-05, "loss": 0.0254, "step": 6456 }, { "epoch": 4.537596626844694, "grad_norm": 0.7451375722885132, "learning_rate": 2.7212930428671818e-05, "loss": 0.042, "step": 6457 }, { "epoch": 4.53829936753338, "grad_norm": 0.38473713397979736, "learning_rate": 2.7217146872803935e-05, "loss": 0.049, "step": 6458 }, { "epoch": 4.539002108222066, "grad_norm": 0.22077427804470062, "learning_rate": 2.722136331693605e-05, "loss": 0.0258, "step": 6459 }, { "epoch": 4.539704848910752, "grad_norm": 0.3523794114589691, "learning_rate": 2.7225579761068168e-05, "loss": 0.041, "step": 6460 }, { "epoch": 4.540407589599438, "grad_norm": 0.6260344386100769, "learning_rate": 2.722979620520028e-05, "loss": 0.057, "step": 6461 }, { "epoch": 4.541110330288124, "grad_norm": 0.517398476600647, "learning_rate": 2.7234012649332398e-05, "loss": 0.0671, "step": 6462 }, { "epoch": 4.54181307097681, "grad_norm": 0.9087995886802673, "learning_rate": 2.723822909346451e-05, "loss": 0.1056, "step": 6463 }, { "epoch": 4.542515811665496, "grad_norm": 3.2402660846710205, "learning_rate": 2.7242445537596628e-05, "loss": 0.1386, "step": 6464 }, { "epoch": 4.5432185523541815, "grad_norm": 0.7338157296180725, "learning_rate": 2.7246661981728744e-05, "loss": 0.1762, "step": 6465 }, { "epoch": 4.543921293042867, "grad_norm": 1.2170361280441284, "learning_rate": 2.7250878425860858e-05, "loss": 0.221, "step": 6466 }, { "epoch": 4.544624033731553, "grad_norm": 1.5508148670196533, "learning_rate": 2.7255094869992974e-05, "loss": 0.3241, "step": 6467 }, { "epoch": 4.545326774420239, "grad_norm": 0.29437196254730225, "learning_rate": 2.7259311314125088e-05, "loss": 0.085, "step": 6468 }, { "epoch": 4.546029515108925, "grad_norm": 0.2036522775888443, "learning_rate": 2.7263527758257204e-05, "loss": 0.0325, "step": 6469 }, { "epoch": 4.546732255797611, "grad_norm": 0.2268998771905899, "learning_rate": 2.7267744202389317e-05, "loss": 0.0328, "step": 6470 }, { "epoch": 4.547434996486297, "grad_norm": 0.19170860946178436, "learning_rate": 2.7271960646521434e-05, "loss": 0.0204, "step": 6471 }, { "epoch": 4.548137737174983, "grad_norm": 0.2868161201477051, "learning_rate": 2.727617709065355e-05, "loss": 0.045, "step": 6472 }, { "epoch": 4.548840477863668, "grad_norm": 0.2335374504327774, "learning_rate": 2.7280393534785667e-05, "loss": 0.0188, "step": 6473 }, { "epoch": 4.549543218552354, "grad_norm": 0.29910528659820557, "learning_rate": 2.728460997891778e-05, "loss": 0.0201, "step": 6474 }, { "epoch": 4.55024595924104, "grad_norm": 0.18286186456680298, "learning_rate": 2.7288826423049897e-05, "loss": 0.0352, "step": 6475 }, { "epoch": 4.550948699929726, "grad_norm": 0.24747075140476227, "learning_rate": 2.729304286718201e-05, "loss": 0.0293, "step": 6476 }, { "epoch": 4.551651440618412, "grad_norm": 0.26448097825050354, "learning_rate": 2.7297259311314127e-05, "loss": 0.0206, "step": 6477 }, { "epoch": 4.552354181307098, "grad_norm": 0.3022823631763458, "learning_rate": 2.730147575544624e-05, "loss": 0.0226, "step": 6478 }, { "epoch": 4.5530569219957835, "grad_norm": 0.3772543966770172, "learning_rate": 2.7305692199578357e-05, "loss": 0.0187, "step": 6479 }, { "epoch": 4.553759662684469, "grad_norm": 0.17781999707221985, "learning_rate": 2.730990864371047e-05, "loss": 0.0223, "step": 6480 }, { "epoch": 4.554462403373155, "grad_norm": 0.20053283870220184, "learning_rate": 2.7314125087842587e-05, "loss": 0.0187, "step": 6481 }, { "epoch": 4.555165144061841, "grad_norm": 0.432113379240036, "learning_rate": 2.73183415319747e-05, "loss": 0.0407, "step": 6482 }, { "epoch": 4.555867884750527, "grad_norm": 0.5303354263305664, "learning_rate": 2.7322557976106817e-05, "loss": 0.0465, "step": 6483 }, { "epoch": 4.556570625439213, "grad_norm": 0.42569079995155334, "learning_rate": 2.732677442023893e-05, "loss": 0.0293, "step": 6484 }, { "epoch": 4.557273366127899, "grad_norm": 0.2566137909889221, "learning_rate": 2.7330990864371047e-05, "loss": 0.0437, "step": 6485 }, { "epoch": 4.557976106816585, "grad_norm": 0.3491691052913666, "learning_rate": 2.7335207308503163e-05, "loss": 0.0548, "step": 6486 }, { "epoch": 4.558678847505271, "grad_norm": 0.4484145939350128, "learning_rate": 2.733942375263528e-05, "loss": 0.0615, "step": 6487 }, { "epoch": 4.559381588193957, "grad_norm": 0.38940173387527466, "learning_rate": 2.7343640196767393e-05, "loss": 0.109, "step": 6488 }, { "epoch": 4.560084328882642, "grad_norm": 0.6510281562805176, "learning_rate": 2.734785664089951e-05, "loss": 0.1331, "step": 6489 }, { "epoch": 4.560787069571328, "grad_norm": 1.8826426267623901, "learning_rate": 2.7352073085031623e-05, "loss": 0.2337, "step": 6490 }, { "epoch": 4.561489810260014, "grad_norm": 1.1238433122634888, "learning_rate": 2.735628952916374e-05, "loss": 0.2672, "step": 6491 }, { "epoch": 4.5621925509487, "grad_norm": 1.3898109197616577, "learning_rate": 2.7360505973295853e-05, "loss": 0.2711, "step": 6492 }, { "epoch": 4.5628952916373855, "grad_norm": 0.23211967945098877, "learning_rate": 2.736472241742797e-05, "loss": 0.0938, "step": 6493 }, { "epoch": 4.563598032326071, "grad_norm": 0.1314411759376526, "learning_rate": 2.7368938861560086e-05, "loss": 0.0249, "step": 6494 }, { "epoch": 4.564300773014757, "grad_norm": 0.17863082885742188, "learning_rate": 2.73731553056922e-05, "loss": 0.0356, "step": 6495 }, { "epoch": 4.565003513703443, "grad_norm": 0.29103153944015503, "learning_rate": 2.7377371749824316e-05, "loss": 0.029, "step": 6496 }, { "epoch": 4.565706254392129, "grad_norm": 0.1702120304107666, "learning_rate": 2.738158819395643e-05, "loss": 0.0206, "step": 6497 }, { "epoch": 4.566408995080815, "grad_norm": 0.20251326262950897, "learning_rate": 2.7385804638088546e-05, "loss": 0.0201, "step": 6498 }, { "epoch": 4.567111735769501, "grad_norm": 0.16325359046459198, "learning_rate": 2.7390021082220663e-05, "loss": 0.0134, "step": 6499 }, { "epoch": 4.567814476458187, "grad_norm": 0.24074417352676392, "learning_rate": 2.739423752635278e-05, "loss": 0.0416, "step": 6500 }, { "epoch": 4.568517217146873, "grad_norm": 0.20463360846042633, "learning_rate": 2.7398453970484893e-05, "loss": 0.047, "step": 6501 }, { "epoch": 4.569219957835559, "grad_norm": 0.2879696488380432, "learning_rate": 2.740267041461701e-05, "loss": 0.0353, "step": 6502 }, { "epoch": 4.569922698524245, "grad_norm": 0.29096531867980957, "learning_rate": 2.7406886858749122e-05, "loss": 0.0305, "step": 6503 }, { "epoch": 4.570625439212931, "grad_norm": 0.16977857053279877, "learning_rate": 2.741110330288124e-05, "loss": 0.0258, "step": 6504 }, { "epoch": 4.5713281799016166, "grad_norm": 0.2765266001224518, "learning_rate": 2.7415319747013352e-05, "loss": 0.0311, "step": 6505 }, { "epoch": 4.5720309205903025, "grad_norm": 0.23480896651744843, "learning_rate": 2.741953619114547e-05, "loss": 0.0294, "step": 6506 }, { "epoch": 4.572733661278988, "grad_norm": 0.2855510413646698, "learning_rate": 2.7423752635277582e-05, "loss": 0.0321, "step": 6507 }, { "epoch": 4.573436401967674, "grad_norm": 0.4024507701396942, "learning_rate": 2.74279690794097e-05, "loss": 0.0576, "step": 6508 }, { "epoch": 4.57413914265636, "grad_norm": 0.2597343623638153, "learning_rate": 2.7432185523541812e-05, "loss": 0.0392, "step": 6509 }, { "epoch": 4.574841883345046, "grad_norm": 0.2434176206588745, "learning_rate": 2.743640196767393e-05, "loss": 0.0613, "step": 6510 }, { "epoch": 4.575544624033731, "grad_norm": 0.3513893783092499, "learning_rate": 2.7440618411806042e-05, "loss": 0.0581, "step": 6511 }, { "epoch": 4.576247364722417, "grad_norm": 0.2892964482307434, "learning_rate": 2.744483485593816e-05, "loss": 0.0654, "step": 6512 }, { "epoch": 4.576950105411103, "grad_norm": 0.5678826570510864, "learning_rate": 2.7449051300070275e-05, "loss": 0.068, "step": 6513 }, { "epoch": 4.577652846099789, "grad_norm": 0.47879642248153687, "learning_rate": 2.7453267744202392e-05, "loss": 0.1407, "step": 6514 }, { "epoch": 4.578355586788475, "grad_norm": 0.8087746500968933, "learning_rate": 2.7457484188334505e-05, "loss": 0.172, "step": 6515 }, { "epoch": 4.579058327477161, "grad_norm": 1.0978260040283203, "learning_rate": 2.7461700632466622e-05, "loss": 0.2599, "step": 6516 }, { "epoch": 4.579761068165847, "grad_norm": 2.0371408462524414, "learning_rate": 2.7465917076598735e-05, "loss": 0.3146, "step": 6517 }, { "epoch": 4.580463808854533, "grad_norm": 0.3097948729991913, "learning_rate": 2.747013352073085e-05, "loss": 0.075, "step": 6518 }, { "epoch": 4.581166549543219, "grad_norm": 0.17020151019096375, "learning_rate": 2.7474349964862965e-05, "loss": 0.0308, "step": 6519 }, { "epoch": 4.5818692902319045, "grad_norm": 0.2869734764099121, "learning_rate": 2.747856640899508e-05, "loss": 0.0281, "step": 6520 }, { "epoch": 4.58257203092059, "grad_norm": 0.1488194763660431, "learning_rate": 2.7482782853127195e-05, "loss": 0.0274, "step": 6521 }, { "epoch": 4.583274771609276, "grad_norm": 0.19605223834514618, "learning_rate": 2.748699929725931e-05, "loss": 0.0206, "step": 6522 }, { "epoch": 4.583977512297962, "grad_norm": 0.254097044467926, "learning_rate": 2.7491215741391428e-05, "loss": 0.0204, "step": 6523 }, { "epoch": 4.584680252986648, "grad_norm": 0.24480274319648743, "learning_rate": 2.749543218552354e-05, "loss": 0.0209, "step": 6524 }, { "epoch": 4.585382993675334, "grad_norm": 0.20196527242660522, "learning_rate": 2.7499648629655658e-05, "loss": 0.0314, "step": 6525 }, { "epoch": 4.58608573436402, "grad_norm": 0.19578219950199127, "learning_rate": 2.750386507378777e-05, "loss": 0.0259, "step": 6526 }, { "epoch": 4.586788475052706, "grad_norm": 0.21977148950099945, "learning_rate": 2.750808151791989e-05, "loss": 0.0244, "step": 6527 }, { "epoch": 4.587491215741391, "grad_norm": 0.2811819612979889, "learning_rate": 2.7512297962052004e-05, "loss": 0.032, "step": 6528 }, { "epoch": 4.588193956430077, "grad_norm": 0.20384742319583893, "learning_rate": 2.751651440618412e-05, "loss": 0.0247, "step": 6529 }, { "epoch": 4.588896697118763, "grad_norm": 0.2594483494758606, "learning_rate": 2.7520730850316234e-05, "loss": 0.0285, "step": 6530 }, { "epoch": 4.589599437807449, "grad_norm": 0.1641080379486084, "learning_rate": 2.752494729444835e-05, "loss": 0.0196, "step": 6531 }, { "epoch": 4.590302178496135, "grad_norm": 0.28939783573150635, "learning_rate": 2.7529163738580464e-05, "loss": 0.034, "step": 6532 }, { "epoch": 4.591004919184821, "grad_norm": 0.30520400404930115, "learning_rate": 2.753338018271258e-05, "loss": 0.0591, "step": 6533 }, { "epoch": 4.5917076598735065, "grad_norm": 0.22896572947502136, "learning_rate": 2.7537596626844694e-05, "loss": 0.0489, "step": 6534 }, { "epoch": 4.592410400562192, "grad_norm": 0.4940264821052551, "learning_rate": 2.754181307097681e-05, "loss": 0.0272, "step": 6535 }, { "epoch": 4.593113141250878, "grad_norm": 0.3506864309310913, "learning_rate": 2.7546029515108924e-05, "loss": 0.1022, "step": 6536 }, { "epoch": 4.593815881939564, "grad_norm": 0.39014720916748047, "learning_rate": 2.755024595924104e-05, "loss": 0.0872, "step": 6537 }, { "epoch": 4.59451862262825, "grad_norm": 0.5694363117218018, "learning_rate": 2.7554462403373154e-05, "loss": 0.0727, "step": 6538 }, { "epoch": 4.595221363316936, "grad_norm": 0.700164258480072, "learning_rate": 2.755867884750527e-05, "loss": 0.1483, "step": 6539 }, { "epoch": 4.595924104005622, "grad_norm": 0.8651959896087646, "learning_rate": 2.7562895291637387e-05, "loss": 0.2195, "step": 6540 }, { "epoch": 4.596626844694308, "grad_norm": 0.8497485518455505, "learning_rate": 2.7567111735769504e-05, "loss": 0.2421, "step": 6541 }, { "epoch": 4.597329585382994, "grad_norm": 2.718747615814209, "learning_rate": 2.7571328179901617e-05, "loss": 0.2861, "step": 6542 }, { "epoch": 4.59803232607168, "grad_norm": 0.4316946268081665, "learning_rate": 2.7575544624033734e-05, "loss": 0.1028, "step": 6543 }, { "epoch": 4.598735066760366, "grad_norm": 0.19555173814296722, "learning_rate": 2.7579761068165847e-05, "loss": 0.0601, "step": 6544 }, { "epoch": 4.599437807449052, "grad_norm": 0.13892437517642975, "learning_rate": 2.7583977512297964e-05, "loss": 0.0239, "step": 6545 }, { "epoch": 4.6001405481377375, "grad_norm": 0.21173515915870667, "learning_rate": 2.7588193956430077e-05, "loss": 0.0278, "step": 6546 }, { "epoch": 4.6008432888264235, "grad_norm": 0.2622445821762085, "learning_rate": 2.7592410400562193e-05, "loss": 0.0303, "step": 6547 }, { "epoch": 4.601546029515109, "grad_norm": 0.17083577811717987, "learning_rate": 2.7596626844694307e-05, "loss": 0.0193, "step": 6548 }, { "epoch": 4.602248770203794, "grad_norm": 0.1609336882829666, "learning_rate": 2.7600843288826423e-05, "loss": 0.0205, "step": 6549 }, { "epoch": 4.60295151089248, "grad_norm": 0.21337200701236725, "learning_rate": 2.7605059732958537e-05, "loss": 0.0427, "step": 6550 }, { "epoch": 4.603654251581166, "grad_norm": 0.12339304387569427, "learning_rate": 2.7609276177090653e-05, "loss": 0.0177, "step": 6551 }, { "epoch": 4.604356992269852, "grad_norm": 0.20858731865882874, "learning_rate": 2.7613492621222766e-05, "loss": 0.0181, "step": 6552 }, { "epoch": 4.605059732958538, "grad_norm": 0.2540615200996399, "learning_rate": 2.7617709065354883e-05, "loss": 0.0318, "step": 6553 }, { "epoch": 4.605762473647224, "grad_norm": 0.13216379284858704, "learning_rate": 2.7621925509487003e-05, "loss": 0.018, "step": 6554 }, { "epoch": 4.60646521433591, "grad_norm": 0.19628074765205383, "learning_rate": 2.7626141953619116e-05, "loss": 0.0264, "step": 6555 }, { "epoch": 4.607167955024596, "grad_norm": 0.1897033005952835, "learning_rate": 2.7630358397751233e-05, "loss": 0.0136, "step": 6556 }, { "epoch": 4.607870695713282, "grad_norm": 0.25732192397117615, "learning_rate": 2.7634574841883346e-05, "loss": 0.0427, "step": 6557 }, { "epoch": 4.608573436401968, "grad_norm": 0.6130675077438354, "learning_rate": 2.7638791286015463e-05, "loss": 0.0629, "step": 6558 }, { "epoch": 4.609276177090654, "grad_norm": 0.28568077087402344, "learning_rate": 2.7643007730147576e-05, "loss": 0.054, "step": 6559 }, { "epoch": 4.6099789177793395, "grad_norm": 0.24784715473651886, "learning_rate": 2.7647224174279693e-05, "loss": 0.0424, "step": 6560 }, { "epoch": 4.6106816584680255, "grad_norm": 0.6561524271965027, "learning_rate": 2.7651440618411806e-05, "loss": 0.0926, "step": 6561 }, { "epoch": 4.611384399156711, "grad_norm": 0.33771929144859314, "learning_rate": 2.7655657062543923e-05, "loss": 0.0713, "step": 6562 }, { "epoch": 4.612087139845397, "grad_norm": 0.510964035987854, "learning_rate": 2.7659873506676036e-05, "loss": 0.1156, "step": 6563 }, { "epoch": 4.612789880534083, "grad_norm": 0.4762673079967499, "learning_rate": 2.7664089950808153e-05, "loss": 0.1324, "step": 6564 }, { "epoch": 4.613492621222769, "grad_norm": 1.1119686365127563, "learning_rate": 2.7668306394940266e-05, "loss": 0.2058, "step": 6565 }, { "epoch": 4.614195361911454, "grad_norm": 0.8835681676864624, "learning_rate": 2.7672522839072382e-05, "loss": 0.2107, "step": 6566 }, { "epoch": 4.61489810260014, "grad_norm": 1.8615167140960693, "learning_rate": 2.76767392832045e-05, "loss": 0.3137, "step": 6567 }, { "epoch": 4.615600843288826, "grad_norm": 0.5608939528465271, "learning_rate": 2.7680955727336616e-05, "loss": 0.1124, "step": 6568 }, { "epoch": 4.616303583977512, "grad_norm": 0.23037420213222504, "learning_rate": 2.768517217146873e-05, "loss": 0.0317, "step": 6569 }, { "epoch": 4.617006324666198, "grad_norm": 0.2095218300819397, "learning_rate": 2.7689388615600846e-05, "loss": 0.0419, "step": 6570 }, { "epoch": 4.617709065354884, "grad_norm": 0.24480201303958893, "learning_rate": 2.769360505973296e-05, "loss": 0.025, "step": 6571 }, { "epoch": 4.61841180604357, "grad_norm": 0.2663764953613281, "learning_rate": 2.7697821503865076e-05, "loss": 0.0223, "step": 6572 }, { "epoch": 4.619114546732256, "grad_norm": 0.6323765516281128, "learning_rate": 2.770203794799719e-05, "loss": 0.0362, "step": 6573 }, { "epoch": 4.6198172874209416, "grad_norm": 0.3373676538467407, "learning_rate": 2.7706254392129305e-05, "loss": 0.0331, "step": 6574 }, { "epoch": 4.6205200281096275, "grad_norm": 0.15594185888767242, "learning_rate": 2.771047083626142e-05, "loss": 0.0275, "step": 6575 }, { "epoch": 4.621222768798313, "grad_norm": 0.27093538641929626, "learning_rate": 2.7714687280393535e-05, "loss": 0.0354, "step": 6576 }, { "epoch": 4.621925509486999, "grad_norm": 0.2124369591474533, "learning_rate": 2.771890372452565e-05, "loss": 0.021, "step": 6577 }, { "epoch": 4.622628250175685, "grad_norm": 0.2796577513217926, "learning_rate": 2.7723120168657765e-05, "loss": 0.0275, "step": 6578 }, { "epoch": 4.623330990864371, "grad_norm": 0.2172469049692154, "learning_rate": 2.772733661278988e-05, "loss": 0.0214, "step": 6579 }, { "epoch": 4.624033731553057, "grad_norm": 0.30595332384109497, "learning_rate": 2.7731553056921995e-05, "loss": 0.0342, "step": 6580 }, { "epoch": 4.624736472241743, "grad_norm": 0.2844182252883911, "learning_rate": 2.7735769501054112e-05, "loss": 0.0219, "step": 6581 }, { "epoch": 4.625439212930429, "grad_norm": 0.1973377913236618, "learning_rate": 2.773998594518623e-05, "loss": 0.0336, "step": 6582 }, { "epoch": 4.626141953619115, "grad_norm": 0.6224938631057739, "learning_rate": 2.7744202389318345e-05, "loss": 0.0378, "step": 6583 }, { "epoch": 4.626844694307801, "grad_norm": 0.18876373767852783, "learning_rate": 2.7748418833450458e-05, "loss": 0.0323, "step": 6584 }, { "epoch": 4.627547434996487, "grad_norm": 0.7053959369659424, "learning_rate": 2.7752635277582575e-05, "loss": 0.0407, "step": 6585 }, { "epoch": 4.628250175685173, "grad_norm": 0.3435909152030945, "learning_rate": 2.7756851721714688e-05, "loss": 0.0501, "step": 6586 }, { "epoch": 4.6289529163738585, "grad_norm": 0.3456464111804962, "learning_rate": 2.7761068165846805e-05, "loss": 0.0893, "step": 6587 }, { "epoch": 4.629655657062544, "grad_norm": 0.3659335672855377, "learning_rate": 2.7765284609978918e-05, "loss": 0.0762, "step": 6588 }, { "epoch": 4.6303583977512295, "grad_norm": 0.7080227136611938, "learning_rate": 2.7769501054111035e-05, "loss": 0.1529, "step": 6589 }, { "epoch": 4.631061138439915, "grad_norm": 0.690557599067688, "learning_rate": 2.7773717498243148e-05, "loss": 0.2178, "step": 6590 }, { "epoch": 4.631763879128601, "grad_norm": 0.8011117577552795, "learning_rate": 2.7777933942375265e-05, "loss": 0.2791, "step": 6591 }, { "epoch": 4.632466619817287, "grad_norm": 1.3776880502700806, "learning_rate": 2.7782150386507378e-05, "loss": 0.2696, "step": 6592 }, { "epoch": 4.633169360505973, "grad_norm": 0.2720726728439331, "learning_rate": 2.7786366830639494e-05, "loss": 0.0789, "step": 6593 }, { "epoch": 4.633872101194659, "grad_norm": 0.27950775623321533, "learning_rate": 2.7790583274771608e-05, "loss": 0.0353, "step": 6594 }, { "epoch": 4.634574841883345, "grad_norm": 0.20341700315475464, "learning_rate": 2.7794799718903728e-05, "loss": 0.0296, "step": 6595 }, { "epoch": 4.635277582572031, "grad_norm": 0.20573724806308746, "learning_rate": 2.779901616303584e-05, "loss": 0.025, "step": 6596 }, { "epoch": 4.635980323260717, "grad_norm": 0.6936368346214294, "learning_rate": 2.7803232607167958e-05, "loss": 0.0235, "step": 6597 }, { "epoch": 4.636683063949403, "grad_norm": 0.11477043479681015, "learning_rate": 2.780744905130007e-05, "loss": 0.0151, "step": 6598 }, { "epoch": 4.637385804638089, "grad_norm": 0.19636762142181396, "learning_rate": 2.7811665495432187e-05, "loss": 0.0313, "step": 6599 }, { "epoch": 4.638088545326775, "grad_norm": 0.27795591950416565, "learning_rate": 2.78158819395643e-05, "loss": 0.0428, "step": 6600 }, { "epoch": 4.6387912860154605, "grad_norm": 0.15923896431922913, "learning_rate": 2.7820098383696417e-05, "loss": 0.0412, "step": 6601 }, { "epoch": 4.6394940267041465, "grad_norm": 0.22100374102592468, "learning_rate": 2.782431482782853e-05, "loss": 0.0339, "step": 6602 }, { "epoch": 4.640196767392832, "grad_norm": 0.23986904323101044, "learning_rate": 2.7828531271960647e-05, "loss": 0.0274, "step": 6603 }, { "epoch": 4.640899508081518, "grad_norm": 0.16237850487232208, "learning_rate": 2.783274771609276e-05, "loss": 0.0155, "step": 6604 }, { "epoch": 4.641602248770203, "grad_norm": 0.20927630364894867, "learning_rate": 2.7836964160224877e-05, "loss": 0.0363, "step": 6605 }, { "epoch": 4.642304989458889, "grad_norm": 0.17017985880374908, "learning_rate": 2.784118060435699e-05, "loss": 0.0225, "step": 6606 }, { "epoch": 4.643007730147575, "grad_norm": 0.24474823474884033, "learning_rate": 2.7845397048489107e-05, "loss": 0.0396, "step": 6607 }, { "epoch": 4.643710470836261, "grad_norm": 0.24839484691619873, "learning_rate": 2.7849613492621224e-05, "loss": 0.0436, "step": 6608 }, { "epoch": 4.644413211524947, "grad_norm": 0.3446909785270691, "learning_rate": 2.785382993675334e-05, "loss": 0.0396, "step": 6609 }, { "epoch": 4.645115952213633, "grad_norm": 0.22704902291297913, "learning_rate": 2.7858046380885454e-05, "loss": 0.0483, "step": 6610 }, { "epoch": 4.645818692902319, "grad_norm": 0.6127332448959351, "learning_rate": 2.786226282501757e-05, "loss": 0.0528, "step": 6611 }, { "epoch": 4.646521433591005, "grad_norm": 0.514729380607605, "learning_rate": 2.7866479269149687e-05, "loss": 0.0928, "step": 6612 }, { "epoch": 4.647224174279691, "grad_norm": 0.34227773547172546, "learning_rate": 2.78706957132818e-05, "loss": 0.075, "step": 6613 }, { "epoch": 4.647926914968377, "grad_norm": 0.5595040321350098, "learning_rate": 2.7874912157413917e-05, "loss": 0.1358, "step": 6614 }, { "epoch": 4.6486296556570625, "grad_norm": 0.6592699885368347, "learning_rate": 2.787912860154603e-05, "loss": 0.2187, "step": 6615 }, { "epoch": 4.6493323963457485, "grad_norm": 1.0957716703414917, "learning_rate": 2.7883345045678147e-05, "loss": 0.2217, "step": 6616 }, { "epoch": 4.650035137034434, "grad_norm": 2.269227981567383, "learning_rate": 2.788756148981026e-05, "loss": 0.2577, "step": 6617 }, { "epoch": 4.65073787772312, "grad_norm": 0.29485073685646057, "learning_rate": 2.7891777933942376e-05, "loss": 0.0798, "step": 6618 }, { "epoch": 4.651440618411806, "grad_norm": 0.2362973839044571, "learning_rate": 2.789599437807449e-05, "loss": 0.0291, "step": 6619 }, { "epoch": 4.652143359100492, "grad_norm": 0.2629019618034363, "learning_rate": 2.7900210822206606e-05, "loss": 0.0399, "step": 6620 }, { "epoch": 4.652846099789178, "grad_norm": 0.18745701014995575, "learning_rate": 2.790442726633872e-05, "loss": 0.0217, "step": 6621 }, { "epoch": 4.653548840477864, "grad_norm": 0.2034825086593628, "learning_rate": 2.790864371047084e-05, "loss": 0.0298, "step": 6622 }, { "epoch": 4.65425158116655, "grad_norm": 0.3092630207538605, "learning_rate": 2.7912860154602953e-05, "loss": 0.0232, "step": 6623 }, { "epoch": 4.654954321855236, "grad_norm": 0.16706451773643494, "learning_rate": 2.791707659873507e-05, "loss": 0.0188, "step": 6624 }, { "epoch": 4.655657062543922, "grad_norm": 0.22873082756996155, "learning_rate": 2.7921293042867183e-05, "loss": 0.027, "step": 6625 }, { "epoch": 4.656359803232607, "grad_norm": 0.23273850977420807, "learning_rate": 2.79255094869993e-05, "loss": 0.0329, "step": 6626 }, { "epoch": 4.657062543921293, "grad_norm": 0.168072909116745, "learning_rate": 2.7929725931131413e-05, "loss": 0.015, "step": 6627 }, { "epoch": 4.657765284609979, "grad_norm": 0.2744898796081543, "learning_rate": 2.793394237526353e-05, "loss": 0.0368, "step": 6628 }, { "epoch": 4.6584680252986645, "grad_norm": 0.25163257122039795, "learning_rate": 2.7938158819395643e-05, "loss": 0.0163, "step": 6629 }, { "epoch": 4.6591707659873505, "grad_norm": 0.25787895917892456, "learning_rate": 2.794237526352776e-05, "loss": 0.0326, "step": 6630 }, { "epoch": 4.659873506676036, "grad_norm": 0.33731600642204285, "learning_rate": 2.7946591707659872e-05, "loss": 0.0418, "step": 6631 }, { "epoch": 4.660576247364722, "grad_norm": 0.3218785226345062, "learning_rate": 2.795080815179199e-05, "loss": 0.0404, "step": 6632 }, { "epoch": 4.661278988053408, "grad_norm": 0.2340392917394638, "learning_rate": 2.7955024595924102e-05, "loss": 0.0453, "step": 6633 }, { "epoch": 4.661981728742094, "grad_norm": 0.2666214406490326, "learning_rate": 2.795924104005622e-05, "loss": 0.0264, "step": 6634 }, { "epoch": 4.66268446943078, "grad_norm": 0.32238805294036865, "learning_rate": 2.7963457484188336e-05, "loss": 0.0476, "step": 6635 }, { "epoch": 4.663387210119466, "grad_norm": 0.3804260790348053, "learning_rate": 2.7967673928320452e-05, "loss": 0.0582, "step": 6636 }, { "epoch": 4.664089950808152, "grad_norm": 0.3571678400039673, "learning_rate": 2.7971890372452565e-05, "loss": 0.0602, "step": 6637 }, { "epoch": 4.664792691496838, "grad_norm": 0.34851396083831787, "learning_rate": 2.7976106816584682e-05, "loss": 0.0955, "step": 6638 }, { "epoch": 4.665495432185524, "grad_norm": 0.6039999723434448, "learning_rate": 2.7980323260716795e-05, "loss": 0.1495, "step": 6639 }, { "epoch": 4.66619817287421, "grad_norm": 0.7299416065216064, "learning_rate": 2.7984539704848912e-05, "loss": 0.1877, "step": 6640 }, { "epoch": 4.666900913562896, "grad_norm": 1.2660610675811768, "learning_rate": 2.7988756148981025e-05, "loss": 0.2147, "step": 6641 }, { "epoch": 4.6676036542515815, "grad_norm": 1.8966383934020996, "learning_rate": 2.7992972593113142e-05, "loss": 0.2842, "step": 6642 }, { "epoch": 4.668306394940267, "grad_norm": 0.601138710975647, "learning_rate": 2.799718903724526e-05, "loss": 0.0965, "step": 6643 }, { "epoch": 4.6690091356289525, "grad_norm": 0.18305642902851105, "learning_rate": 2.8001405481377372e-05, "loss": 0.0288, "step": 6644 }, { "epoch": 4.669711876317638, "grad_norm": 0.16233819723129272, "learning_rate": 2.800562192550949e-05, "loss": 0.0214, "step": 6645 }, { "epoch": 4.670414617006324, "grad_norm": 0.19931884109973907, "learning_rate": 2.80098383696416e-05, "loss": 0.0293, "step": 6646 }, { "epoch": 4.67111735769501, "grad_norm": 0.2708229422569275, "learning_rate": 2.801405481377372e-05, "loss": 0.0295, "step": 6647 }, { "epoch": 4.671820098383696, "grad_norm": 0.23289038240909576, "learning_rate": 2.801827125790583e-05, "loss": 0.0261, "step": 6648 }, { "epoch": 4.672522839072382, "grad_norm": 0.20825111865997314, "learning_rate": 2.802248770203795e-05, "loss": 0.0327, "step": 6649 }, { "epoch": 4.673225579761068, "grad_norm": 0.18929411470890045, "learning_rate": 2.8026704146170065e-05, "loss": 0.0233, "step": 6650 }, { "epoch": 4.673928320449754, "grad_norm": 0.23388011753559113, "learning_rate": 2.803092059030218e-05, "loss": 0.0318, "step": 6651 }, { "epoch": 4.67463106113844, "grad_norm": 0.18976463377475739, "learning_rate": 2.8035137034434295e-05, "loss": 0.0438, "step": 6652 }, { "epoch": 4.675333801827126, "grad_norm": 0.3817099332809448, "learning_rate": 2.803935347856641e-05, "loss": 0.0403, "step": 6653 }, { "epoch": 4.676036542515812, "grad_norm": 0.15015222132205963, "learning_rate": 2.8043569922698525e-05, "loss": 0.0116, "step": 6654 }, { "epoch": 4.676739283204498, "grad_norm": 0.29624924063682556, "learning_rate": 2.804778636683064e-05, "loss": 0.0335, "step": 6655 }, { "epoch": 4.6774420238931835, "grad_norm": 0.2500048577785492, "learning_rate": 2.8052002810962755e-05, "loss": 0.0213, "step": 6656 }, { "epoch": 4.6781447645818695, "grad_norm": 0.25730690360069275, "learning_rate": 2.805621925509487e-05, "loss": 0.0351, "step": 6657 }, { "epoch": 4.678847505270555, "grad_norm": 0.3993737995624542, "learning_rate": 2.8060435699226984e-05, "loss": 0.0455, "step": 6658 }, { "epoch": 4.679550245959241, "grad_norm": 0.330846905708313, "learning_rate": 2.80646521433591e-05, "loss": 0.0349, "step": 6659 }, { "epoch": 4.680252986647927, "grad_norm": 0.36633339524269104, "learning_rate": 2.8068868587491214e-05, "loss": 0.0543, "step": 6660 }, { "epoch": 4.680955727336613, "grad_norm": 0.32634368538856506, "learning_rate": 2.807308503162333e-05, "loss": 0.0517, "step": 6661 }, { "epoch": 4.681658468025299, "grad_norm": 0.4409547448158264, "learning_rate": 2.8077301475755444e-05, "loss": 0.0852, "step": 6662 }, { "epoch": 4.682361208713985, "grad_norm": 1.0423880815505981, "learning_rate": 2.8081517919887564e-05, "loss": 0.1093, "step": 6663 }, { "epoch": 4.683063949402671, "grad_norm": 0.9407961964607239, "learning_rate": 2.8085734364019677e-05, "loss": 0.1537, "step": 6664 }, { "epoch": 4.683766690091356, "grad_norm": 1.1707955598831177, "learning_rate": 2.8089950808151794e-05, "loss": 0.1973, "step": 6665 }, { "epoch": 4.684469430780042, "grad_norm": 3.3602991104125977, "learning_rate": 2.8094167252283907e-05, "loss": 0.2524, "step": 6666 }, { "epoch": 4.685172171468728, "grad_norm": Infinity, "learning_rate": 2.8094167252283907e-05, "loss": 0.2965, "step": 6667 }, { "epoch": 4.685874912157414, "grad_norm": 0.2355576753616333, "learning_rate": 2.8098383696416024e-05, "loss": 0.0702, "step": 6668 }, { "epoch": 4.6865776528461, "grad_norm": 0.2697128653526306, "learning_rate": 2.8102600140548137e-05, "loss": 0.0316, "step": 6669 }, { "epoch": 4.6872803935347855, "grad_norm": 0.18134087324142456, "learning_rate": 2.8106816584680254e-05, "loss": 0.0276, "step": 6670 }, { "epoch": 4.6879831342234715, "grad_norm": 0.1535300463438034, "learning_rate": 2.8111033028812367e-05, "loss": 0.0297, "step": 6671 }, { "epoch": 4.688685874912157, "grad_norm": 0.17790289223194122, "learning_rate": 2.8115249472944484e-05, "loss": 0.0276, "step": 6672 }, { "epoch": 4.689388615600843, "grad_norm": 0.2748766541481018, "learning_rate": 2.81194659170766e-05, "loss": 0.0183, "step": 6673 }, { "epoch": 4.690091356289529, "grad_norm": 0.14235804975032806, "learning_rate": 2.8123682361208714e-05, "loss": 0.0245, "step": 6674 }, { "epoch": 4.690794096978215, "grad_norm": 0.49709823727607727, "learning_rate": 2.812789880534083e-05, "loss": 0.0271, "step": 6675 }, { "epoch": 4.691496837666901, "grad_norm": 0.4937898516654968, "learning_rate": 2.8132115249472944e-05, "loss": 0.0364, "step": 6676 }, { "epoch": 4.692199578355587, "grad_norm": 0.32397201657295227, "learning_rate": 2.8136331693605064e-05, "loss": 0.0259, "step": 6677 }, { "epoch": 4.692902319044273, "grad_norm": 0.17063720524311066, "learning_rate": 2.8140548137737177e-05, "loss": 0.0231, "step": 6678 }, { "epoch": 4.693605059732959, "grad_norm": 0.29688358306884766, "learning_rate": 2.8144764581869293e-05, "loss": 0.0331, "step": 6679 }, { "epoch": 4.694307800421645, "grad_norm": 0.22427138686180115, "learning_rate": 2.8148981026001407e-05, "loss": 0.0303, "step": 6680 }, { "epoch": 4.695010541110331, "grad_norm": 0.2400653213262558, "learning_rate": 2.8153197470133523e-05, "loss": 0.0218, "step": 6681 }, { "epoch": 4.695713281799016, "grad_norm": 0.2181304395198822, "learning_rate": 2.8157413914265637e-05, "loss": 0.05, "step": 6682 }, { "epoch": 4.696416022487702, "grad_norm": 0.20786529779434204, "learning_rate": 2.8161630358397753e-05, "loss": 0.0423, "step": 6683 }, { "epoch": 4.6971187631763875, "grad_norm": 0.32653409242630005, "learning_rate": 2.8165846802529866e-05, "loss": 0.0276, "step": 6684 }, { "epoch": 4.6978215038650735, "grad_norm": 0.37944966554641724, "learning_rate": 2.8170063246661983e-05, "loss": 0.0488, "step": 6685 }, { "epoch": 4.698524244553759, "grad_norm": 0.2478720098733902, "learning_rate": 2.8174279690794096e-05, "loss": 0.0341, "step": 6686 }, { "epoch": 4.699226985242445, "grad_norm": 0.36066731810569763, "learning_rate": 2.8178496134926213e-05, "loss": 0.0939, "step": 6687 }, { "epoch": 4.699929725931131, "grad_norm": 0.672423779964447, "learning_rate": 2.8182712579058326e-05, "loss": 0.1013, "step": 6688 }, { "epoch": 4.700632466619817, "grad_norm": 0.57358717918396, "learning_rate": 2.8186929023190443e-05, "loss": 0.1261, "step": 6689 }, { "epoch": 4.701335207308503, "grad_norm": 0.7211507558822632, "learning_rate": 2.8191145467322556e-05, "loss": 0.1746, "step": 6690 }, { "epoch": 4.702037947997189, "grad_norm": 0.9773074388504028, "learning_rate": 2.8195361911454676e-05, "loss": 0.2465, "step": 6691 }, { "epoch": 4.702740688685875, "grad_norm": 1.2058258056640625, "learning_rate": 2.819957835558679e-05, "loss": 0.3171, "step": 6692 }, { "epoch": 4.703443429374561, "grad_norm": 0.2893076539039612, "learning_rate": 2.8203794799718906e-05, "loss": 0.087, "step": 6693 }, { "epoch": 4.704146170063247, "grad_norm": 0.18998245894908905, "learning_rate": 2.820801124385102e-05, "loss": 0.0267, "step": 6694 }, { "epoch": 4.704848910751933, "grad_norm": 0.20479415357112885, "learning_rate": 2.8212227687983136e-05, "loss": 0.028, "step": 6695 }, { "epoch": 4.705551651440619, "grad_norm": 0.1664658486843109, "learning_rate": 2.821644413211525e-05, "loss": 0.0204, "step": 6696 }, { "epoch": 4.7062543921293045, "grad_norm": 0.1493387073278427, "learning_rate": 2.8220660576247366e-05, "loss": 0.0311, "step": 6697 }, { "epoch": 4.70695713281799, "grad_norm": 0.15754954516887665, "learning_rate": 2.822487702037948e-05, "loss": 0.0135, "step": 6698 }, { "epoch": 4.707659873506676, "grad_norm": 0.12547092139720917, "learning_rate": 2.8229093464511596e-05, "loss": 0.0174, "step": 6699 }, { "epoch": 4.708362614195362, "grad_norm": 0.27127787470817566, "learning_rate": 2.823330990864371e-05, "loss": 0.0327, "step": 6700 }, { "epoch": 4.709065354884048, "grad_norm": 0.14893195033073425, "learning_rate": 2.8237526352775826e-05, "loss": 0.0253, "step": 6701 }, { "epoch": 4.709768095572734, "grad_norm": 0.2199811041355133, "learning_rate": 2.824174279690794e-05, "loss": 0.0188, "step": 6702 }, { "epoch": 4.710470836261419, "grad_norm": 0.21494919061660767, "learning_rate": 2.8245959241040055e-05, "loss": 0.039, "step": 6703 }, { "epoch": 4.711173576950105, "grad_norm": 0.1878211945295334, "learning_rate": 2.8250175685172172e-05, "loss": 0.0315, "step": 6704 }, { "epoch": 4.711876317638791, "grad_norm": 0.26167958974838257, "learning_rate": 2.825439212930429e-05, "loss": 0.0469, "step": 6705 }, { "epoch": 4.712579058327477, "grad_norm": 0.18565967679023743, "learning_rate": 2.8258608573436405e-05, "loss": 0.0177, "step": 6706 }, { "epoch": 4.713281799016163, "grad_norm": 0.3152605891227722, "learning_rate": 2.826282501756852e-05, "loss": 0.0255, "step": 6707 }, { "epoch": 4.713984539704849, "grad_norm": 0.22209997475147247, "learning_rate": 2.8267041461700635e-05, "loss": 0.039, "step": 6708 }, { "epoch": 4.714687280393535, "grad_norm": 0.22821871936321259, "learning_rate": 2.827125790583275e-05, "loss": 0.0189, "step": 6709 }, { "epoch": 4.715390021082221, "grad_norm": 0.2154996246099472, "learning_rate": 2.8275474349964865e-05, "loss": 0.0331, "step": 6710 }, { "epoch": 4.7160927617709065, "grad_norm": 0.34539860486984253, "learning_rate": 2.827969079409698e-05, "loss": 0.0672, "step": 6711 }, { "epoch": 4.7167955024595924, "grad_norm": 0.4025905430316925, "learning_rate": 2.8283907238229095e-05, "loss": 0.0691, "step": 6712 }, { "epoch": 4.717498243148278, "grad_norm": 0.6019862294197083, "learning_rate": 2.8288123682361208e-05, "loss": 0.1314, "step": 6713 }, { "epoch": 4.718200983836964, "grad_norm": 0.5238941311836243, "learning_rate": 2.8292340126493325e-05, "loss": 0.1209, "step": 6714 }, { "epoch": 4.71890372452565, "grad_norm": 1.2511656284332275, "learning_rate": 2.8296556570625438e-05, "loss": 0.1999, "step": 6715 }, { "epoch": 4.719606465214336, "grad_norm": 1.6473616361618042, "learning_rate": 2.8300773014757555e-05, "loss": 0.2354, "step": 6716 }, { "epoch": 4.720309205903022, "grad_norm": 1.6410027742385864, "learning_rate": 2.8304989458889668e-05, "loss": 0.2816, "step": 6717 }, { "epoch": 4.721011946591708, "grad_norm": 0.32106372714042664, "learning_rate": 2.8309205903021788e-05, "loss": 0.0775, "step": 6718 }, { "epoch": 4.721714687280394, "grad_norm": 0.19667145609855652, "learning_rate": 2.83134223471539e-05, "loss": 0.0302, "step": 6719 }, { "epoch": 4.722417427969079, "grad_norm": 0.19410869479179382, "learning_rate": 2.8317638791286018e-05, "loss": 0.0327, "step": 6720 }, { "epoch": 4.723120168657765, "grad_norm": 0.18951855599880219, "learning_rate": 2.832185523541813e-05, "loss": 0.034, "step": 6721 }, { "epoch": 4.723822909346451, "grad_norm": 0.21897968649864197, "learning_rate": 2.8326071679550248e-05, "loss": 0.0227, "step": 6722 }, { "epoch": 4.724525650035137, "grad_norm": 0.09815362840890884, "learning_rate": 2.833028812368236e-05, "loss": 0.0102, "step": 6723 }, { "epoch": 4.725228390723823, "grad_norm": 0.2977122366428375, "learning_rate": 2.8334504567814478e-05, "loss": 0.0268, "step": 6724 }, { "epoch": 4.7259311314125085, "grad_norm": 0.34087979793548584, "learning_rate": 2.833872101194659e-05, "loss": 0.0584, "step": 6725 }, { "epoch": 4.7266338721011945, "grad_norm": 0.1987573802471161, "learning_rate": 2.8342937456078708e-05, "loss": 0.0203, "step": 6726 }, { "epoch": 4.72733661278988, "grad_norm": 0.2055371105670929, "learning_rate": 2.834715390021082e-05, "loss": 0.028, "step": 6727 }, { "epoch": 4.728039353478566, "grad_norm": 0.27448153495788574, "learning_rate": 2.8351370344342938e-05, "loss": 0.0497, "step": 6728 }, { "epoch": 4.728742094167252, "grad_norm": 0.14343225955963135, "learning_rate": 2.835558678847505e-05, "loss": 0.0183, "step": 6729 }, { "epoch": 4.729444834855938, "grad_norm": 0.1946418732404709, "learning_rate": 2.8359803232607167e-05, "loss": 0.0438, "step": 6730 }, { "epoch": 4.730147575544624, "grad_norm": 0.20742899179458618, "learning_rate": 2.836401967673928e-05, "loss": 0.0208, "step": 6731 }, { "epoch": 4.73085031623331, "grad_norm": 0.20751453936100006, "learning_rate": 2.83682361208714e-05, "loss": 0.0341, "step": 6732 }, { "epoch": 4.731553056921996, "grad_norm": 0.17325489223003387, "learning_rate": 2.8372452565003517e-05, "loss": 0.0224, "step": 6733 }, { "epoch": 4.732255797610682, "grad_norm": 0.2325342893600464, "learning_rate": 2.837666900913563e-05, "loss": 0.0468, "step": 6734 }, { "epoch": 4.732958538299368, "grad_norm": 0.3199363350868225, "learning_rate": 2.8380885453267747e-05, "loss": 0.0673, "step": 6735 }, { "epoch": 4.733661278988054, "grad_norm": 0.2798765003681183, "learning_rate": 2.838510189739986e-05, "loss": 0.0535, "step": 6736 }, { "epoch": 4.73436401967674, "grad_norm": 0.43723979592323303, "learning_rate": 2.8389318341531977e-05, "loss": 0.1084, "step": 6737 }, { "epoch": 4.7350667603654255, "grad_norm": 0.790664792060852, "learning_rate": 2.839353478566409e-05, "loss": 0.1075, "step": 6738 }, { "epoch": 4.735769501054111, "grad_norm": 0.5124601721763611, "learning_rate": 2.8397751229796207e-05, "loss": 0.1411, "step": 6739 }, { "epoch": 4.736472241742797, "grad_norm": 0.6973966956138611, "learning_rate": 2.840196767392832e-05, "loss": 0.1852, "step": 6740 }, { "epoch": 4.737174982431483, "grad_norm": 0.9976853728294373, "learning_rate": 2.8406184118060437e-05, "loss": 0.2061, "step": 6741 }, { "epoch": 4.737877723120168, "grad_norm": 1.4500571489334106, "learning_rate": 2.841040056219255e-05, "loss": 0.265, "step": 6742 }, { "epoch": 4.738580463808854, "grad_norm": 0.32987451553344727, "learning_rate": 2.8414617006324667e-05, "loss": 0.0858, "step": 6743 }, { "epoch": 4.73928320449754, "grad_norm": 0.31976374983787537, "learning_rate": 2.841883345045678e-05, "loss": 0.0473, "step": 6744 }, { "epoch": 4.739985945186226, "grad_norm": 0.16325978934764862, "learning_rate": 2.84230498945889e-05, "loss": 0.0282, "step": 6745 }, { "epoch": 4.740688685874912, "grad_norm": 0.210079625248909, "learning_rate": 2.8427266338721013e-05, "loss": 0.0293, "step": 6746 }, { "epoch": 4.741391426563598, "grad_norm": 0.24667322635650635, "learning_rate": 2.843148278285313e-05, "loss": 0.0306, "step": 6747 }, { "epoch": 4.742094167252284, "grad_norm": 0.20920132100582123, "learning_rate": 2.8435699226985243e-05, "loss": 0.019, "step": 6748 }, { "epoch": 4.74279690794097, "grad_norm": 0.18052969872951508, "learning_rate": 2.843991567111736e-05, "loss": 0.0272, "step": 6749 }, { "epoch": 4.743499648629656, "grad_norm": 0.20408812165260315, "learning_rate": 2.8444132115249473e-05, "loss": 0.0312, "step": 6750 }, { "epoch": 4.744202389318342, "grad_norm": 0.33264681696891785, "learning_rate": 2.844834855938159e-05, "loss": 0.0237, "step": 6751 }, { "epoch": 4.7449051300070275, "grad_norm": 0.27682626247406006, "learning_rate": 2.8452565003513703e-05, "loss": 0.0246, "step": 6752 }, { "epoch": 4.745607870695713, "grad_norm": 0.22281061112880707, "learning_rate": 2.845678144764582e-05, "loss": 0.0419, "step": 6753 }, { "epoch": 4.746310611384399, "grad_norm": 0.2180820107460022, "learning_rate": 2.8460997891777933e-05, "loss": 0.0177, "step": 6754 }, { "epoch": 4.747013352073085, "grad_norm": 0.2379680871963501, "learning_rate": 2.846521433591005e-05, "loss": 0.0389, "step": 6755 }, { "epoch": 4.747716092761771, "grad_norm": 0.21829526126384735, "learning_rate": 2.8469430780042163e-05, "loss": 0.025, "step": 6756 }, { "epoch": 4.748418833450457, "grad_norm": 0.2755739688873291, "learning_rate": 2.847364722417428e-05, "loss": 0.0528, "step": 6757 }, { "epoch": 4.749121574139143, "grad_norm": 0.3521100580692291, "learning_rate": 2.8477863668306393e-05, "loss": 0.0456, "step": 6758 }, { "epoch": 4.749824314827828, "grad_norm": 0.2125186324119568, "learning_rate": 2.8482080112438513e-05, "loss": 0.0313, "step": 6759 }, { "epoch": 4.750527055516514, "grad_norm": 0.20480255782604218, "learning_rate": 2.8486296556570626e-05, "loss": 0.0358, "step": 6760 }, { "epoch": 4.7512297962052, "grad_norm": 0.3129578232765198, "learning_rate": 2.8490513000702743e-05, "loss": 0.0697, "step": 6761 }, { "epoch": 4.751932536893886, "grad_norm": 0.3689694404602051, "learning_rate": 2.849472944483486e-05, "loss": 0.0549, "step": 6762 }, { "epoch": 4.752635277582572, "grad_norm": 0.46421948075294495, "learning_rate": 2.8498945888966972e-05, "loss": 0.1104, "step": 6763 }, { "epoch": 4.753338018271258, "grad_norm": 0.5771089196205139, "learning_rate": 2.850316233309909e-05, "loss": 0.1225, "step": 6764 }, { "epoch": 4.754040758959944, "grad_norm": 0.6581290364265442, "learning_rate": 2.8507378777231202e-05, "loss": 0.1615, "step": 6765 }, { "epoch": 4.7547434996486295, "grad_norm": 1.1335201263427734, "learning_rate": 2.851159522136332e-05, "loss": 0.2359, "step": 6766 }, { "epoch": 4.7554462403373154, "grad_norm": 1.7213774919509888, "learning_rate": 2.8515811665495432e-05, "loss": 0.3211, "step": 6767 }, { "epoch": 4.756148981026001, "grad_norm": 0.39085790514945984, "learning_rate": 2.852002810962755e-05, "loss": 0.0701, "step": 6768 }, { "epoch": 4.756851721714687, "grad_norm": 0.22389189898967743, "learning_rate": 2.8524244553759662e-05, "loss": 0.0583, "step": 6769 }, { "epoch": 4.757554462403373, "grad_norm": 0.15530793368816376, "learning_rate": 2.852846099789178e-05, "loss": 0.0264, "step": 6770 }, { "epoch": 4.758257203092059, "grad_norm": 0.18227994441986084, "learning_rate": 2.8532677442023892e-05, "loss": 0.0212, "step": 6771 }, { "epoch": 4.758959943780745, "grad_norm": 0.1881146878004074, "learning_rate": 2.853689388615601e-05, "loss": 0.021, "step": 6772 }, { "epoch": 4.759662684469431, "grad_norm": 0.14057517051696777, "learning_rate": 2.8541110330288125e-05, "loss": 0.0151, "step": 6773 }, { "epoch": 4.760365425158117, "grad_norm": 0.1819869428873062, "learning_rate": 2.8545326774420242e-05, "loss": 0.0259, "step": 6774 }, { "epoch": 4.761068165846803, "grad_norm": 0.1997978240251541, "learning_rate": 2.8549543218552355e-05, "loss": 0.038, "step": 6775 }, { "epoch": 4.761770906535489, "grad_norm": 0.46883609890937805, "learning_rate": 2.8553759662684472e-05, "loss": 0.024, "step": 6776 }, { "epoch": 4.762473647224175, "grad_norm": 0.16631104052066803, "learning_rate": 2.8557976106816585e-05, "loss": 0.0163, "step": 6777 }, { "epoch": 4.763176387912861, "grad_norm": 0.30406907200813293, "learning_rate": 2.85621925509487e-05, "loss": 0.0293, "step": 6778 }, { "epoch": 4.7638791286015465, "grad_norm": 0.14858770370483398, "learning_rate": 2.8566408995080815e-05, "loss": 0.0214, "step": 6779 }, { "epoch": 4.7645818692902315, "grad_norm": 0.20769213140010834, "learning_rate": 2.857062543921293e-05, "loss": 0.0243, "step": 6780 }, { "epoch": 4.7652846099789175, "grad_norm": 0.3220365345478058, "learning_rate": 2.8574841883345045e-05, "loss": 0.0281, "step": 6781 }, { "epoch": 4.765987350667603, "grad_norm": 0.3284478485584259, "learning_rate": 2.857905832747716e-05, "loss": 0.0443, "step": 6782 }, { "epoch": 4.766690091356289, "grad_norm": 0.3355002999305725, "learning_rate": 2.8583274771609275e-05, "loss": 0.0526, "step": 6783 }, { "epoch": 4.767392832044975, "grad_norm": 0.18607303500175476, "learning_rate": 2.858749121574139e-05, "loss": 0.0395, "step": 6784 }, { "epoch": 4.768095572733661, "grad_norm": 0.5623014569282532, "learning_rate": 2.8591707659873505e-05, "loss": 0.0355, "step": 6785 }, { "epoch": 4.768798313422347, "grad_norm": 0.4560624063014984, "learning_rate": 2.8595924104005625e-05, "loss": 0.0573, "step": 6786 }, { "epoch": 4.769501054111033, "grad_norm": 0.2969096601009369, "learning_rate": 2.8600140548137738e-05, "loss": 0.0762, "step": 6787 }, { "epoch": 4.770203794799719, "grad_norm": 0.6257981657981873, "learning_rate": 2.8604356992269854e-05, "loss": 0.1077, "step": 6788 }, { "epoch": 4.770906535488405, "grad_norm": 0.7855430841445923, "learning_rate": 2.8608573436401968e-05, "loss": 0.1092, "step": 6789 }, { "epoch": 4.771609276177091, "grad_norm": 0.7996420860290527, "learning_rate": 2.8612789880534084e-05, "loss": 0.1824, "step": 6790 }, { "epoch": 4.772312016865777, "grad_norm": 0.7550327777862549, "learning_rate": 2.8617006324666198e-05, "loss": 0.217, "step": 6791 }, { "epoch": 4.773014757554463, "grad_norm": 1.8338490724563599, "learning_rate": 2.8621222768798314e-05, "loss": 0.3528, "step": 6792 }, { "epoch": 4.7737174982431485, "grad_norm": 0.23981158435344696, "learning_rate": 2.862543921293043e-05, "loss": 0.0638, "step": 6793 }, { "epoch": 4.774420238931834, "grad_norm": 0.14189857244491577, "learning_rate": 2.8629655657062544e-05, "loss": 0.0255, "step": 6794 }, { "epoch": 4.77512297962052, "grad_norm": 0.1728201061487198, "learning_rate": 2.863387210119466e-05, "loss": 0.032, "step": 6795 }, { "epoch": 4.775825720309206, "grad_norm": 0.150954008102417, "learning_rate": 2.8638088545326774e-05, "loss": 0.0198, "step": 6796 }, { "epoch": 4.776528460997891, "grad_norm": 0.1772432178258896, "learning_rate": 2.864230498945889e-05, "loss": 0.0211, "step": 6797 }, { "epoch": 4.777231201686577, "grad_norm": 0.17136485874652863, "learning_rate": 2.8646521433591004e-05, "loss": 0.0196, "step": 6798 }, { "epoch": 4.777933942375263, "grad_norm": 0.18785357475280762, "learning_rate": 2.865073787772312e-05, "loss": 0.0248, "step": 6799 }, { "epoch": 4.778636683063949, "grad_norm": 0.1907835751771927, "learning_rate": 2.8654954321855237e-05, "loss": 0.0264, "step": 6800 }, { "epoch": 4.779339423752635, "grad_norm": 0.17820154130458832, "learning_rate": 2.8659170765987354e-05, "loss": 0.029, "step": 6801 }, { "epoch": 4.780042164441321, "grad_norm": 0.14953546226024628, "learning_rate": 2.8663387210119467e-05, "loss": 0.0241, "step": 6802 }, { "epoch": 4.780744905130007, "grad_norm": 0.16754932701587677, "learning_rate": 2.8667603654251584e-05, "loss": 0.0222, "step": 6803 }, { "epoch": 4.781447645818693, "grad_norm": 0.1549072563648224, "learning_rate": 2.8671820098383697e-05, "loss": 0.02, "step": 6804 }, { "epoch": 4.782150386507379, "grad_norm": 0.24307881295681, "learning_rate": 2.8676036542515814e-05, "loss": 0.0302, "step": 6805 }, { "epoch": 4.782853127196065, "grad_norm": 0.3530001938343048, "learning_rate": 2.8680252986647927e-05, "loss": 0.042, "step": 6806 }, { "epoch": 4.7835558678847505, "grad_norm": 0.1900126039981842, "learning_rate": 2.8684469430780043e-05, "loss": 0.0216, "step": 6807 }, { "epoch": 4.784258608573436, "grad_norm": 0.2953585088253021, "learning_rate": 2.8688685874912157e-05, "loss": 0.0509, "step": 6808 }, { "epoch": 4.784961349262122, "grad_norm": 0.3624804615974426, "learning_rate": 2.8692902319044273e-05, "loss": 0.0702, "step": 6809 }, { "epoch": 4.785664089950808, "grad_norm": 0.29264283180236816, "learning_rate": 2.8697118763176387e-05, "loss": 0.0532, "step": 6810 }, { "epoch": 4.786366830639494, "grad_norm": 0.2476656436920166, "learning_rate": 2.8701335207308503e-05, "loss": 0.0373, "step": 6811 }, { "epoch": 4.78706957132818, "grad_norm": 0.31168216466903687, "learning_rate": 2.8705551651440617e-05, "loss": 0.0622, "step": 6812 }, { "epoch": 4.787772312016866, "grad_norm": 1.7123922109603882, "learning_rate": 2.8709768095572737e-05, "loss": 0.0821, "step": 6813 }, { "epoch": 4.788475052705552, "grad_norm": 1.2914836406707764, "learning_rate": 2.871398453970485e-05, "loss": 0.1702, "step": 6814 }, { "epoch": 4.789177793394238, "grad_norm": 0.7054232358932495, "learning_rate": 2.8718200983836966e-05, "loss": 0.1445, "step": 6815 }, { "epoch": 4.789880534082924, "grad_norm": 0.987497091293335, "learning_rate": 2.872241742796908e-05, "loss": 0.2452, "step": 6816 }, { "epoch": 4.79058327477161, "grad_norm": 1.3993886709213257, "learning_rate": 2.8726633872101196e-05, "loss": 0.2731, "step": 6817 }, { "epoch": 4.791286015460296, "grad_norm": 0.3028234839439392, "learning_rate": 2.873085031623331e-05, "loss": 0.1003, "step": 6818 }, { "epoch": 4.791988756148981, "grad_norm": 0.3177639842033386, "learning_rate": 2.8735066760365426e-05, "loss": 0.0286, "step": 6819 }, { "epoch": 4.792691496837667, "grad_norm": 0.13729587197303772, "learning_rate": 2.873928320449754e-05, "loss": 0.0239, "step": 6820 }, { "epoch": 4.7933942375263525, "grad_norm": 0.24955801665782928, "learning_rate": 2.8743499648629656e-05, "loss": 0.0306, "step": 6821 }, { "epoch": 4.794096978215038, "grad_norm": 0.1928303986787796, "learning_rate": 2.8747716092761773e-05, "loss": 0.0245, "step": 6822 }, { "epoch": 4.794799718903724, "grad_norm": 0.17446881532669067, "learning_rate": 2.8751932536893886e-05, "loss": 0.0148, "step": 6823 }, { "epoch": 4.79550245959241, "grad_norm": 0.18900622427463531, "learning_rate": 2.8756148981026003e-05, "loss": 0.0258, "step": 6824 }, { "epoch": 4.796205200281096, "grad_norm": 0.16285662353038788, "learning_rate": 2.8760365425158116e-05, "loss": 0.023, "step": 6825 }, { "epoch": 4.796907940969782, "grad_norm": 0.17241977155208588, "learning_rate": 2.8764581869290232e-05, "loss": 0.0263, "step": 6826 }, { "epoch": 4.797610681658468, "grad_norm": 0.13585513830184937, "learning_rate": 2.876879831342235e-05, "loss": 0.0178, "step": 6827 }, { "epoch": 4.798313422347154, "grad_norm": 0.32747286558151245, "learning_rate": 2.8773014757554466e-05, "loss": 0.0382, "step": 6828 }, { "epoch": 4.79901616303584, "grad_norm": 0.48934194445610046, "learning_rate": 2.877723120168658e-05, "loss": 0.0196, "step": 6829 }, { "epoch": 4.799718903724526, "grad_norm": 0.24359029531478882, "learning_rate": 2.8781447645818696e-05, "loss": 0.0324, "step": 6830 }, { "epoch": 4.800421644413212, "grad_norm": 0.24352535605430603, "learning_rate": 2.878566408995081e-05, "loss": 0.0313, "step": 6831 }, { "epoch": 4.801124385101898, "grad_norm": 0.2560611963272095, "learning_rate": 2.8789880534082926e-05, "loss": 0.0438, "step": 6832 }, { "epoch": 4.801827125790584, "grad_norm": 0.283952921628952, "learning_rate": 2.879409697821504e-05, "loss": 0.0306, "step": 6833 }, { "epoch": 4.8025298664792695, "grad_norm": 0.27865928411483765, "learning_rate": 2.8798313422347155e-05, "loss": 0.0217, "step": 6834 }, { "epoch": 4.8032326071679545, "grad_norm": 2.549318552017212, "learning_rate": 2.880252986647927e-05, "loss": 0.046, "step": 6835 }, { "epoch": 4.8039353478566404, "grad_norm": 0.4290410280227661, "learning_rate": 2.8806746310611385e-05, "loss": 0.0598, "step": 6836 }, { "epoch": 4.804638088545326, "grad_norm": 0.641441285610199, "learning_rate": 2.88109627547435e-05, "loss": 0.0518, "step": 6837 }, { "epoch": 4.805340829234012, "grad_norm": 0.4846619665622711, "learning_rate": 2.8815179198875615e-05, "loss": 0.0797, "step": 6838 }, { "epoch": 4.806043569922698, "grad_norm": 0.5315062999725342, "learning_rate": 2.881939564300773e-05, "loss": 0.1726, "step": 6839 }, { "epoch": 4.806746310611384, "grad_norm": 0.6979954838752747, "learning_rate": 2.8823612087139845e-05, "loss": 0.203, "step": 6840 }, { "epoch": 4.80744905130007, "grad_norm": 0.6804975271224976, "learning_rate": 2.8827828531271962e-05, "loss": 0.2274, "step": 6841 }, { "epoch": 4.808151791988756, "grad_norm": 1.097564935684204, "learning_rate": 2.883204497540408e-05, "loss": 0.2557, "step": 6842 }, { "epoch": 4.808854532677442, "grad_norm": 0.33238092064857483, "learning_rate": 2.883626141953619e-05, "loss": 0.0789, "step": 6843 }, { "epoch": 4.809557273366128, "grad_norm": 0.3044651746749878, "learning_rate": 2.8840477863668308e-05, "loss": 0.0328, "step": 6844 }, { "epoch": 4.810260014054814, "grad_norm": 0.16497811675071716, "learning_rate": 2.884469430780042e-05, "loss": 0.0175, "step": 6845 }, { "epoch": 4.8109627547435, "grad_norm": 0.1711985468864441, "learning_rate": 2.8848910751932538e-05, "loss": 0.024, "step": 6846 }, { "epoch": 4.811665495432186, "grad_norm": 0.2711280882358551, "learning_rate": 2.885312719606465e-05, "loss": 0.0246, "step": 6847 }, { "epoch": 4.8123682361208715, "grad_norm": 0.22549085319042206, "learning_rate": 2.8857343640196768e-05, "loss": 0.0286, "step": 6848 }, { "epoch": 4.813070976809557, "grad_norm": 0.17106454074382782, "learning_rate": 2.886156008432888e-05, "loss": 0.02, "step": 6849 }, { "epoch": 4.813773717498243, "grad_norm": 0.20583860576152802, "learning_rate": 2.8865776528460998e-05, "loss": 0.029, "step": 6850 }, { "epoch": 4.814476458186929, "grad_norm": 0.29240351915359497, "learning_rate": 2.886999297259311e-05, "loss": 0.0287, "step": 6851 }, { "epoch": 4.815179198875615, "grad_norm": 0.20275084674358368, "learning_rate": 2.8874209416725228e-05, "loss": 0.0222, "step": 6852 }, { "epoch": 4.815881939564301, "grad_norm": 0.1788255125284195, "learning_rate": 2.8878425860857344e-05, "loss": 0.0203, "step": 6853 }, { "epoch": 4.816584680252987, "grad_norm": 0.18253861367702484, "learning_rate": 2.888264230498946e-05, "loss": 0.0168, "step": 6854 }, { "epoch": 4.817287420941673, "grad_norm": 0.35907283425331116, "learning_rate": 2.8886858749121578e-05, "loss": 0.0403, "step": 6855 }, { "epoch": 4.817990161630359, "grad_norm": 0.13020050525665283, "learning_rate": 2.889107519325369e-05, "loss": 0.0205, "step": 6856 }, { "epoch": 4.818692902319044, "grad_norm": 0.30471453070640564, "learning_rate": 2.8895291637385808e-05, "loss": 0.0398, "step": 6857 }, { "epoch": 4.81939564300773, "grad_norm": 0.3571326732635498, "learning_rate": 2.889950808151792e-05, "loss": 0.0491, "step": 6858 }, { "epoch": 4.820098383696416, "grad_norm": 1.2635549306869507, "learning_rate": 2.8903724525650037e-05, "loss": 0.0518, "step": 6859 }, { "epoch": 4.820801124385102, "grad_norm": 0.23485760390758514, "learning_rate": 2.890794096978215e-05, "loss": 0.0328, "step": 6860 }, { "epoch": 4.821503865073788, "grad_norm": 0.37461772561073303, "learning_rate": 2.8912157413914267e-05, "loss": 0.0698, "step": 6861 }, { "epoch": 4.8222066057624735, "grad_norm": 0.482597291469574, "learning_rate": 2.891637385804638e-05, "loss": 0.0632, "step": 6862 }, { "epoch": 4.822909346451159, "grad_norm": 0.9679247736930847, "learning_rate": 2.8920590302178497e-05, "loss": 0.1172, "step": 6863 }, { "epoch": 4.823612087139845, "grad_norm": 0.6235312223434448, "learning_rate": 2.892480674631061e-05, "loss": 0.1417, "step": 6864 }, { "epoch": 4.824314827828531, "grad_norm": 0.7637677788734436, "learning_rate": 2.8929023190442727e-05, "loss": 0.1996, "step": 6865 }, { "epoch": 4.825017568517217, "grad_norm": 0.9992109537124634, "learning_rate": 2.893323963457484e-05, "loss": 0.2046, "step": 6866 }, { "epoch": 4.825720309205903, "grad_norm": 4.205570220947266, "learning_rate": 2.8937456078706957e-05, "loss": 0.2992, "step": 6867 }, { "epoch": 4.826423049894589, "grad_norm": 0.28292447328567505, "learning_rate": 2.8941672522839074e-05, "loss": 0.1073, "step": 6868 }, { "epoch": 4.827125790583275, "grad_norm": 0.14876142144203186, "learning_rate": 2.894588896697119e-05, "loss": 0.0294, "step": 6869 }, { "epoch": 4.827828531271961, "grad_norm": 0.3037256896495819, "learning_rate": 2.8950105411103304e-05, "loss": 0.0299, "step": 6870 }, { "epoch": 4.828531271960647, "grad_norm": 0.18293873965740204, "learning_rate": 2.895432185523542e-05, "loss": 0.0291, "step": 6871 }, { "epoch": 4.829234012649333, "grad_norm": 0.16166894137859344, "learning_rate": 2.8958538299367533e-05, "loss": 0.0198, "step": 6872 }, { "epoch": 4.829936753338019, "grad_norm": 0.21301138401031494, "learning_rate": 2.896275474349965e-05, "loss": 0.0239, "step": 6873 }, { "epoch": 4.830639494026704, "grad_norm": 0.19691181182861328, "learning_rate": 2.8966971187631763e-05, "loss": 0.0291, "step": 6874 }, { "epoch": 4.83134223471539, "grad_norm": 0.422681987285614, "learning_rate": 2.897118763176388e-05, "loss": 0.025, "step": 6875 }, { "epoch": 4.8320449754040755, "grad_norm": 0.1895064264535904, "learning_rate": 2.8975404075895993e-05, "loss": 0.0229, "step": 6876 }, { "epoch": 4.832747716092761, "grad_norm": 0.20255155861377716, "learning_rate": 2.897962052002811e-05, "loss": 0.0239, "step": 6877 }, { "epoch": 4.833450456781447, "grad_norm": 0.39959537982940674, "learning_rate": 2.8983836964160223e-05, "loss": 0.0336, "step": 6878 }, { "epoch": 4.834153197470133, "grad_norm": 0.17890919744968414, "learning_rate": 2.898805340829234e-05, "loss": 0.0177, "step": 6879 }, { "epoch": 4.834855938158819, "grad_norm": 0.20770014822483063, "learning_rate": 2.8992269852424453e-05, "loss": 0.036, "step": 6880 }, { "epoch": 4.835558678847505, "grad_norm": 0.3782055079936981, "learning_rate": 2.8996486296556573e-05, "loss": 0.0185, "step": 6881 }, { "epoch": 4.836261419536191, "grad_norm": 0.17933014035224915, "learning_rate": 2.900070274068869e-05, "loss": 0.0341, "step": 6882 }, { "epoch": 4.836964160224877, "grad_norm": 0.26547786593437195, "learning_rate": 2.9004919184820803e-05, "loss": 0.0367, "step": 6883 }, { "epoch": 4.837666900913563, "grad_norm": 0.2714449167251587, "learning_rate": 2.900913562895292e-05, "loss": 0.0315, "step": 6884 }, { "epoch": 4.838369641602249, "grad_norm": 0.26947376132011414, "learning_rate": 2.9013352073085033e-05, "loss": 0.0511, "step": 6885 }, { "epoch": 4.839072382290935, "grad_norm": 0.5123331546783447, "learning_rate": 2.901756851721715e-05, "loss": 0.0511, "step": 6886 }, { "epoch": 4.839775122979621, "grad_norm": 0.38102149963378906, "learning_rate": 2.9021784961349263e-05, "loss": 0.0488, "step": 6887 }, { "epoch": 4.840477863668307, "grad_norm": 0.8104220628738403, "learning_rate": 2.902600140548138e-05, "loss": 0.0867, "step": 6888 }, { "epoch": 4.8411806043569925, "grad_norm": 0.7936144471168518, "learning_rate": 2.9030217849613493e-05, "loss": 0.1765, "step": 6889 }, { "epoch": 4.841883345045678, "grad_norm": 0.8276838660240173, "learning_rate": 2.903443429374561e-05, "loss": 0.2015, "step": 6890 }, { "epoch": 4.842586085734364, "grad_norm": 1.1499652862548828, "learning_rate": 2.9038650737877722e-05, "loss": 0.2345, "step": 6891 }, { "epoch": 4.84328882642305, "grad_norm": 1.4926310777664185, "learning_rate": 2.904286718200984e-05, "loss": 0.3171, "step": 6892 }, { "epoch": 4.843991567111736, "grad_norm": 0.2236521989107132, "learning_rate": 2.9047083626141952e-05, "loss": 0.0764, "step": 6893 }, { "epoch": 4.844694307800422, "grad_norm": 0.14519983530044556, "learning_rate": 2.905130007027407e-05, "loss": 0.0389, "step": 6894 }, { "epoch": 4.845397048489108, "grad_norm": 0.2674698233604431, "learning_rate": 2.9055516514406186e-05, "loss": 0.0269, "step": 6895 }, { "epoch": 4.846099789177793, "grad_norm": 0.1471986025571823, "learning_rate": 2.9059732958538302e-05, "loss": 0.0226, "step": 6896 }, { "epoch": 4.846802529866479, "grad_norm": 0.17864537239074707, "learning_rate": 2.9063949402670416e-05, "loss": 0.0305, "step": 6897 }, { "epoch": 4.847505270555165, "grad_norm": 0.17827832698822021, "learning_rate": 2.9068165846802532e-05, "loss": 0.026, "step": 6898 }, { "epoch": 4.848208011243851, "grad_norm": 0.21556271612644196, "learning_rate": 2.9072382290934645e-05, "loss": 0.0292, "step": 6899 }, { "epoch": 4.848910751932537, "grad_norm": 0.20447179675102234, "learning_rate": 2.9076598735066762e-05, "loss": 0.017, "step": 6900 }, { "epoch": 4.849613492621223, "grad_norm": 0.1271839141845703, "learning_rate": 2.9080815179198875e-05, "loss": 0.0191, "step": 6901 }, { "epoch": 4.850316233309909, "grad_norm": 0.21518242359161377, "learning_rate": 2.9085031623330992e-05, "loss": 0.0513, "step": 6902 }, { "epoch": 4.8510189739985945, "grad_norm": 0.3077070713043213, "learning_rate": 2.9089248067463105e-05, "loss": 0.0427, "step": 6903 }, { "epoch": 4.85172171468728, "grad_norm": 0.20432834327220917, "learning_rate": 2.9093464511595222e-05, "loss": 0.0288, "step": 6904 }, { "epoch": 4.852424455375966, "grad_norm": 0.1667805165052414, "learning_rate": 2.9097680955727335e-05, "loss": 0.038, "step": 6905 }, { "epoch": 4.853127196064652, "grad_norm": 0.3312290906906128, "learning_rate": 2.9101897399859452e-05, "loss": 0.0511, "step": 6906 }, { "epoch": 4.853829936753338, "grad_norm": 0.15261106193065643, "learning_rate": 2.9106113843991565e-05, "loss": 0.025, "step": 6907 }, { "epoch": 4.854532677442024, "grad_norm": 0.26438093185424805, "learning_rate": 2.911033028812368e-05, "loss": 0.0405, "step": 6908 }, { "epoch": 4.85523541813071, "grad_norm": 0.2381300926208496, "learning_rate": 2.9114546732255798e-05, "loss": 0.0416, "step": 6909 }, { "epoch": 4.855938158819396, "grad_norm": 0.2611582279205322, "learning_rate": 2.9118763176387915e-05, "loss": 0.0551, "step": 6910 }, { "epoch": 4.856640899508082, "grad_norm": 0.27585235238075256, "learning_rate": 2.912297962052003e-05, "loss": 0.0534, "step": 6911 }, { "epoch": 4.857343640196767, "grad_norm": 0.45446309447288513, "learning_rate": 2.9127196064652145e-05, "loss": 0.0773, "step": 6912 }, { "epoch": 4.858046380885453, "grad_norm": 0.501116931438446, "learning_rate": 2.913141250878426e-05, "loss": 0.0943, "step": 6913 }, { "epoch": 4.858749121574139, "grad_norm": 0.49357953667640686, "learning_rate": 2.9135628952916375e-05, "loss": 0.1327, "step": 6914 }, { "epoch": 4.859451862262825, "grad_norm": 2.4444596767425537, "learning_rate": 2.913984539704849e-05, "loss": 0.1826, "step": 6915 }, { "epoch": 4.860154602951511, "grad_norm": 1.1698918342590332, "learning_rate": 2.9144061841180605e-05, "loss": 0.2659, "step": 6916 }, { "epoch": 4.8608573436401965, "grad_norm": 1.5820225477218628, "learning_rate": 2.914827828531272e-05, "loss": 0.2815, "step": 6917 }, { "epoch": 4.861560084328882, "grad_norm": 0.3372035026550293, "learning_rate": 2.9152494729444834e-05, "loss": 0.0841, "step": 6918 }, { "epoch": 4.862262825017568, "grad_norm": 0.18627995252609253, "learning_rate": 2.915671117357695e-05, "loss": 0.038, "step": 6919 }, { "epoch": 4.862965565706254, "grad_norm": 0.21120917797088623, "learning_rate": 2.9160927617709064e-05, "loss": 0.0296, "step": 6920 }, { "epoch": 4.86366830639494, "grad_norm": 0.3155261278152466, "learning_rate": 2.916514406184118e-05, "loss": 0.0527, "step": 6921 }, { "epoch": 4.864371047083626, "grad_norm": 0.16023027896881104, "learning_rate": 2.9169360505973298e-05, "loss": 0.0187, "step": 6922 }, { "epoch": 4.865073787772312, "grad_norm": 0.192024365067482, "learning_rate": 2.9173576950105414e-05, "loss": 0.0203, "step": 6923 }, { "epoch": 4.865776528460998, "grad_norm": 0.16219554841518402, "learning_rate": 2.9177793394237527e-05, "loss": 0.0201, "step": 6924 }, { "epoch": 4.866479269149684, "grad_norm": 0.21254029870033264, "learning_rate": 2.9182009838369644e-05, "loss": 0.0281, "step": 6925 }, { "epoch": 4.86718200983837, "grad_norm": 0.18791188299655914, "learning_rate": 2.9186226282501757e-05, "loss": 0.0331, "step": 6926 }, { "epoch": 4.867884750527056, "grad_norm": 0.16937045753002167, "learning_rate": 2.9190442726633874e-05, "loss": 0.0139, "step": 6927 }, { "epoch": 4.868587491215742, "grad_norm": 0.34945181012153625, "learning_rate": 2.9194659170765987e-05, "loss": 0.0556, "step": 6928 }, { "epoch": 4.869290231904428, "grad_norm": 0.1608734130859375, "learning_rate": 2.9198875614898104e-05, "loss": 0.0224, "step": 6929 }, { "epoch": 4.8699929725931135, "grad_norm": 0.2114756554365158, "learning_rate": 2.9203092059030217e-05, "loss": 0.0283, "step": 6930 }, { "epoch": 4.870695713281799, "grad_norm": 0.2063433676958084, "learning_rate": 2.9207308503162334e-05, "loss": 0.0276, "step": 6931 }, { "epoch": 4.871398453970485, "grad_norm": 0.297557532787323, "learning_rate": 2.9211524947294447e-05, "loss": 0.0296, "step": 6932 }, { "epoch": 4.872101194659171, "grad_norm": 0.3083944022655487, "learning_rate": 2.9215741391426564e-05, "loss": 0.0454, "step": 6933 }, { "epoch": 4.872803935347856, "grad_norm": 0.626340925693512, "learning_rate": 2.9219957835558677e-05, "loss": 0.0356, "step": 6934 }, { "epoch": 4.873506676036542, "grad_norm": 1.2770497798919678, "learning_rate": 2.9224174279690794e-05, "loss": 0.0476, "step": 6935 }, { "epoch": 4.874209416725228, "grad_norm": 0.2817930579185486, "learning_rate": 2.922839072382291e-05, "loss": 0.0513, "step": 6936 }, { "epoch": 4.874912157413914, "grad_norm": 0.36874207854270935, "learning_rate": 2.9232607167955027e-05, "loss": 0.0957, "step": 6937 }, { "epoch": 4.8756148981026, "grad_norm": 0.5841784477233887, "learning_rate": 2.923682361208714e-05, "loss": 0.1054, "step": 6938 }, { "epoch": 4.876317638791286, "grad_norm": 0.71095210313797, "learning_rate": 2.9241040056219257e-05, "loss": 0.134, "step": 6939 }, { "epoch": 4.877020379479972, "grad_norm": 0.9951574206352234, "learning_rate": 2.924525650035137e-05, "loss": 0.1527, "step": 6940 }, { "epoch": 4.877723120168658, "grad_norm": 1.7225862741470337, "learning_rate": 2.9249472944483487e-05, "loss": 0.2516, "step": 6941 }, { "epoch": 4.878425860857344, "grad_norm": 1.084078311920166, "learning_rate": 2.9253689388615603e-05, "loss": 0.286, "step": 6942 }, { "epoch": 4.87912860154603, "grad_norm": 0.2595394253730774, "learning_rate": 2.9257905832747716e-05, "loss": 0.0718, "step": 6943 }, { "epoch": 4.8798313422347155, "grad_norm": 0.15934720635414124, "learning_rate": 2.9262122276879833e-05, "loss": 0.0251, "step": 6944 }, { "epoch": 4.880534082923401, "grad_norm": 0.3015764653682709, "learning_rate": 2.9266338721011946e-05, "loss": 0.0392, "step": 6945 }, { "epoch": 4.881236823612087, "grad_norm": 0.209547221660614, "learning_rate": 2.9270555165144063e-05, "loss": 0.0379, "step": 6946 }, { "epoch": 4.881939564300773, "grad_norm": 0.187783345580101, "learning_rate": 2.9274771609276176e-05, "loss": 0.0337, "step": 6947 }, { "epoch": 4.882642304989459, "grad_norm": 0.1269521564245224, "learning_rate": 2.9278988053408293e-05, "loss": 0.0129, "step": 6948 }, { "epoch": 4.883345045678145, "grad_norm": 0.22414317727088928, "learning_rate": 2.928320449754041e-05, "loss": 0.0189, "step": 6949 }, { "epoch": 4.884047786366831, "grad_norm": 0.24703866243362427, "learning_rate": 2.9287420941672526e-05, "loss": 0.0464, "step": 6950 }, { "epoch": 4.884750527055516, "grad_norm": 0.15834183990955353, "learning_rate": 2.929163738580464e-05, "loss": 0.0207, "step": 6951 }, { "epoch": 4.885453267744202, "grad_norm": 0.20964045822620392, "learning_rate": 2.9295853829936756e-05, "loss": 0.0252, "step": 6952 }, { "epoch": 4.886156008432888, "grad_norm": 0.40909647941589355, "learning_rate": 2.930007027406887e-05, "loss": 0.0617, "step": 6953 }, { "epoch": 4.886858749121574, "grad_norm": 0.1952565461397171, "learning_rate": 2.9304286718200986e-05, "loss": 0.0165, "step": 6954 }, { "epoch": 4.88756148981026, "grad_norm": 0.6691052317619324, "learning_rate": 2.93085031623331e-05, "loss": 0.0525, "step": 6955 }, { "epoch": 4.888264230498946, "grad_norm": 0.176228865981102, "learning_rate": 2.9312719606465216e-05, "loss": 0.0271, "step": 6956 }, { "epoch": 4.888966971187632, "grad_norm": 0.22871236503124237, "learning_rate": 2.931693605059733e-05, "loss": 0.0378, "step": 6957 }, { "epoch": 4.8896697118763175, "grad_norm": 0.1986338496208191, "learning_rate": 2.9321152494729446e-05, "loss": 0.0546, "step": 6958 }, { "epoch": 4.890372452565003, "grad_norm": 0.2332354187965393, "learning_rate": 2.932536893886156e-05, "loss": 0.0364, "step": 6959 }, { "epoch": 4.891075193253689, "grad_norm": 0.24852906167507172, "learning_rate": 2.9329585382993676e-05, "loss": 0.0392, "step": 6960 }, { "epoch": 4.891777933942375, "grad_norm": 0.2454293966293335, "learning_rate": 2.933380182712579e-05, "loss": 0.0434, "step": 6961 }, { "epoch": 4.892480674631061, "grad_norm": 0.24178634583950043, "learning_rate": 2.9338018271257905e-05, "loss": 0.0384, "step": 6962 }, { "epoch": 4.893183415319747, "grad_norm": 0.5026727318763733, "learning_rate": 2.9342234715390022e-05, "loss": 0.1154, "step": 6963 }, { "epoch": 4.893886156008433, "grad_norm": 0.42054322361946106, "learning_rate": 2.934645115952214e-05, "loss": 0.1286, "step": 6964 }, { "epoch": 4.894588896697119, "grad_norm": 0.7172042727470398, "learning_rate": 2.9350667603654252e-05, "loss": 0.1819, "step": 6965 }, { "epoch": 4.895291637385805, "grad_norm": 0.9981964826583862, "learning_rate": 2.935488404778637e-05, "loss": 0.2298, "step": 6966 }, { "epoch": 4.895994378074491, "grad_norm": 1.3558273315429688, "learning_rate": 2.9359100491918482e-05, "loss": 0.309, "step": 6967 }, { "epoch": 4.896697118763177, "grad_norm": 0.27323266863822937, "learning_rate": 2.93633169360506e-05, "loss": 0.0806, "step": 6968 }, { "epoch": 4.897399859451863, "grad_norm": 0.15660595893859863, "learning_rate": 2.9367533380182712e-05, "loss": 0.0259, "step": 6969 }, { "epoch": 4.8981026001405485, "grad_norm": 0.5755212903022766, "learning_rate": 2.937174982431483e-05, "loss": 0.0359, "step": 6970 }, { "epoch": 4.8988053408292345, "grad_norm": 0.14615565538406372, "learning_rate": 2.9375966268446945e-05, "loss": 0.0123, "step": 6971 }, { "epoch": 4.8995080815179195, "grad_norm": 0.16496390104293823, "learning_rate": 2.938018271257906e-05, "loss": 0.0188, "step": 6972 }, { "epoch": 4.900210822206605, "grad_norm": 0.2180212289094925, "learning_rate": 2.9384399156711175e-05, "loss": 0.0266, "step": 6973 }, { "epoch": 4.900913562895291, "grad_norm": 0.21661996841430664, "learning_rate": 2.9388615600843288e-05, "loss": 0.0333, "step": 6974 }, { "epoch": 4.901616303583977, "grad_norm": 0.20248740911483765, "learning_rate": 2.9392832044975405e-05, "loss": 0.0206, "step": 6975 }, { "epoch": 4.902319044272663, "grad_norm": 0.2851742208003998, "learning_rate": 2.9397048489107518e-05, "loss": 0.0505, "step": 6976 }, { "epoch": 4.903021784961349, "grad_norm": 0.1310613453388214, "learning_rate": 2.9401264933239638e-05, "loss": 0.014, "step": 6977 }, { "epoch": 4.903724525650035, "grad_norm": 0.2680917978286743, "learning_rate": 2.940548137737175e-05, "loss": 0.0397, "step": 6978 }, { "epoch": 4.904427266338721, "grad_norm": 0.32188236713409424, "learning_rate": 2.9409697821503868e-05, "loss": 0.0165, "step": 6979 }, { "epoch": 4.905130007027407, "grad_norm": 0.24778875708580017, "learning_rate": 2.941391426563598e-05, "loss": 0.0246, "step": 6980 }, { "epoch": 4.905832747716093, "grad_norm": 0.3008274435997009, "learning_rate": 2.9418130709768098e-05, "loss": 0.0265, "step": 6981 }, { "epoch": 4.906535488404779, "grad_norm": 0.2434152513742447, "learning_rate": 2.942234715390021e-05, "loss": 0.0346, "step": 6982 }, { "epoch": 4.907238229093465, "grad_norm": 0.22710579633712769, "learning_rate": 2.9426563598032328e-05, "loss": 0.0436, "step": 6983 }, { "epoch": 4.9079409697821506, "grad_norm": 0.22789223492145538, "learning_rate": 2.943078004216444e-05, "loss": 0.0591, "step": 6984 }, { "epoch": 4.9086437104708365, "grad_norm": 0.20706133544445038, "learning_rate": 2.9434996486296558e-05, "loss": 0.0304, "step": 6985 }, { "epoch": 4.909346451159522, "grad_norm": 0.3718039393424988, "learning_rate": 2.943921293042867e-05, "loss": 0.0711, "step": 6986 }, { "epoch": 4.910049191848208, "grad_norm": 0.4962078332901001, "learning_rate": 2.9443429374560788e-05, "loss": 0.0792, "step": 6987 }, { "epoch": 4.910751932536894, "grad_norm": 1.3199156522750854, "learning_rate": 2.94476458186929e-05, "loss": 0.1249, "step": 6988 }, { "epoch": 4.911454673225579, "grad_norm": 0.5693628787994385, "learning_rate": 2.9451862262825017e-05, "loss": 0.1581, "step": 6989 }, { "epoch": 4.912157413914265, "grad_norm": 0.7793270349502563, "learning_rate": 2.9456078706957134e-05, "loss": 0.1885, "step": 6990 }, { "epoch": 4.912860154602951, "grad_norm": 1.1669625043869019, "learning_rate": 2.946029515108925e-05, "loss": 0.1996, "step": 6991 }, { "epoch": 4.913562895291637, "grad_norm": 1.1007376909255981, "learning_rate": 2.9464511595221364e-05, "loss": 0.2725, "step": 6992 }, { "epoch": 4.914265635980323, "grad_norm": 0.25706809759140015, "learning_rate": 2.946872803935348e-05, "loss": 0.1046, "step": 6993 }, { "epoch": 4.914968376669009, "grad_norm": 0.22960016131401062, "learning_rate": 2.9472944483485594e-05, "loss": 0.0307, "step": 6994 }, { "epoch": 4.915671117357695, "grad_norm": 0.32804688811302185, "learning_rate": 2.947716092761771e-05, "loss": 0.0392, "step": 6995 }, { "epoch": 4.916373858046381, "grad_norm": 0.20377708971500397, "learning_rate": 2.9481377371749824e-05, "loss": 0.022, "step": 6996 }, { "epoch": 4.917076598735067, "grad_norm": 0.24827797710895538, "learning_rate": 2.948559381588194e-05, "loss": 0.0243, "step": 6997 }, { "epoch": 4.917779339423753, "grad_norm": 0.17768152058124542, "learning_rate": 2.9489810260014054e-05, "loss": 0.0275, "step": 6998 }, { "epoch": 4.9184820801124385, "grad_norm": 0.21098729968070984, "learning_rate": 2.949402670414617e-05, "loss": 0.0274, "step": 6999 }, { "epoch": 4.919184820801124, "grad_norm": 0.2830641269683838, "learning_rate": 2.9498243148278287e-05, "loss": 0.0292, "step": 7000 }, { "epoch": 4.919184820801124, "eval_cer": 0.19871689961048739, "eval_loss": 0.32148003578186035, "eval_runtime": 86.5783, "eval_samples_per_second": 52.415, "eval_steps_per_second": 0.173, "eval_wer": 0.36836959861456386, "step": 7000 }, { "epoch": 4.91988756148981, "grad_norm": 0.1934104710817337, "learning_rate": 2.95024595924104e-05, "loss": 0.0458, "step": 7001 }, { "epoch": 4.920590302178496, "grad_norm": 0.15732508897781372, "learning_rate": 2.9506676036542517e-05, "loss": 0.032, "step": 7002 }, { "epoch": 4.921293042867182, "grad_norm": 0.21131230890750885, "learning_rate": 2.951089248067463e-05, "loss": 0.0321, "step": 7003 }, { "epoch": 4.921995783555868, "grad_norm": 0.1517535001039505, "learning_rate": 2.951510892480675e-05, "loss": 0.0244, "step": 7004 }, { "epoch": 4.922698524244554, "grad_norm": 0.1502402126789093, "learning_rate": 2.9519325368938863e-05, "loss": 0.0392, "step": 7005 }, { "epoch": 4.92340126493324, "grad_norm": 0.15266305208206177, "learning_rate": 2.952354181307098e-05, "loss": 0.0196, "step": 7006 }, { "epoch": 4.924104005621926, "grad_norm": 0.3676675260066986, "learning_rate": 2.9527758257203093e-05, "loss": 0.0613, "step": 7007 }, { "epoch": 4.924806746310612, "grad_norm": 0.23784828186035156, "learning_rate": 2.953197470133521e-05, "loss": 0.0438, "step": 7008 }, { "epoch": 4.925509486999298, "grad_norm": 0.2126443088054657, "learning_rate": 2.9536191145467323e-05, "loss": 0.0349, "step": 7009 }, { "epoch": 4.926212227687984, "grad_norm": 0.24710683524608612, "learning_rate": 2.954040758959944e-05, "loss": 0.0451, "step": 7010 }, { "epoch": 4.926914968376669, "grad_norm": 0.24717353284358978, "learning_rate": 2.9544624033731553e-05, "loss": 0.0771, "step": 7011 }, { "epoch": 4.927617709065355, "grad_norm": 0.5098283886909485, "learning_rate": 2.954884047786367e-05, "loss": 0.0639, "step": 7012 }, { "epoch": 4.9283204497540405, "grad_norm": 0.3661726117134094, "learning_rate": 2.9553056921995783e-05, "loss": 0.0844, "step": 7013 }, { "epoch": 4.929023190442726, "grad_norm": 0.6199425458908081, "learning_rate": 2.95572733661279e-05, "loss": 0.166, "step": 7014 }, { "epoch": 4.929725931131412, "grad_norm": 0.6454159021377563, "learning_rate": 2.9561489810260013e-05, "loss": 0.1896, "step": 7015 }, { "epoch": 4.930428671820098, "grad_norm": 1.15126633644104, "learning_rate": 2.956570625439213e-05, "loss": 0.2185, "step": 7016 }, { "epoch": 4.931131412508784, "grad_norm": 1.0673351287841797, "learning_rate": 2.9569922698524246e-05, "loss": 0.3107, "step": 7017 }, { "epoch": 4.93183415319747, "grad_norm": 0.28389307856559753, "learning_rate": 2.9574139142656363e-05, "loss": 0.0855, "step": 7018 }, { "epoch": 4.932536893886156, "grad_norm": 0.3051954209804535, "learning_rate": 2.9578355586788476e-05, "loss": 0.0622, "step": 7019 }, { "epoch": 4.933239634574842, "grad_norm": 0.23503293097019196, "learning_rate": 2.9582572030920593e-05, "loss": 0.0346, "step": 7020 }, { "epoch": 4.933942375263528, "grad_norm": 0.17108944058418274, "learning_rate": 2.9586788475052706e-05, "loss": 0.0284, "step": 7021 }, { "epoch": 4.934645115952214, "grad_norm": 0.2180021107196808, "learning_rate": 2.9591004919184822e-05, "loss": 0.0222, "step": 7022 }, { "epoch": 4.9353478566409, "grad_norm": 0.14400063455104828, "learning_rate": 2.9595221363316936e-05, "loss": 0.0269, "step": 7023 }, { "epoch": 4.936050597329586, "grad_norm": 0.1984318047761917, "learning_rate": 2.9599437807449052e-05, "loss": 0.0323, "step": 7024 }, { "epoch": 4.9367533380182715, "grad_norm": 0.18908843398094177, "learning_rate": 2.9603654251581166e-05, "loss": 0.0271, "step": 7025 }, { "epoch": 4.9374560787069575, "grad_norm": 0.1787995845079422, "learning_rate": 2.9607870695713282e-05, "loss": 0.0313, "step": 7026 }, { "epoch": 4.938158819395643, "grad_norm": 0.22617456316947937, "learning_rate": 2.9612087139845395e-05, "loss": 0.0222, "step": 7027 }, { "epoch": 4.938861560084328, "grad_norm": 0.18090905249118805, "learning_rate": 2.9616303583977512e-05, "loss": 0.0475, "step": 7028 }, { "epoch": 4.939564300773014, "grad_norm": 0.12345860153436661, "learning_rate": 2.9620520028109625e-05, "loss": 0.0175, "step": 7029 }, { "epoch": 4.9402670414617, "grad_norm": 0.22297878563404083, "learning_rate": 2.9624736472241742e-05, "loss": 0.0398, "step": 7030 }, { "epoch": 4.940969782150386, "grad_norm": 0.2582424581050873, "learning_rate": 2.9628952916373862e-05, "loss": 0.0225, "step": 7031 }, { "epoch": 4.941672522839072, "grad_norm": 0.2088623195886612, "learning_rate": 2.9633169360505975e-05, "loss": 0.0323, "step": 7032 }, { "epoch": 4.942375263527758, "grad_norm": 0.36535829305648804, "learning_rate": 2.9637385804638092e-05, "loss": 0.0389, "step": 7033 }, { "epoch": 4.943078004216444, "grad_norm": 0.1627335250377655, "learning_rate": 2.9641602248770205e-05, "loss": 0.0168, "step": 7034 }, { "epoch": 4.94378074490513, "grad_norm": 0.22402098774909973, "learning_rate": 2.9645818692902322e-05, "loss": 0.0446, "step": 7035 }, { "epoch": 4.944483485593816, "grad_norm": 0.3930913209915161, "learning_rate": 2.9650035137034435e-05, "loss": 0.0741, "step": 7036 }, { "epoch": 4.945186226282502, "grad_norm": 0.3425208032131195, "learning_rate": 2.965425158116655e-05, "loss": 0.1056, "step": 7037 }, { "epoch": 4.945888966971188, "grad_norm": 0.3913267254829407, "learning_rate": 2.9658468025298665e-05, "loss": 0.0943, "step": 7038 }, { "epoch": 4.9465917076598735, "grad_norm": 0.3840882480144501, "learning_rate": 2.966268446943078e-05, "loss": 0.1431, "step": 7039 }, { "epoch": 4.9472944483485595, "grad_norm": 0.6056797504425049, "learning_rate": 2.9666900913562895e-05, "loss": 0.1798, "step": 7040 }, { "epoch": 4.947997189037245, "grad_norm": 0.8603782057762146, "learning_rate": 2.967111735769501e-05, "loss": 0.227, "step": 7041 }, { "epoch": 4.948699929725931, "grad_norm": 1.1256802082061768, "learning_rate": 2.9675333801827125e-05, "loss": 0.3125, "step": 7042 }, { "epoch": 4.949402670414617, "grad_norm": 0.2565560042858124, "learning_rate": 2.967955024595924e-05, "loss": 0.0781, "step": 7043 }, { "epoch": 4.950105411103303, "grad_norm": 0.1746237576007843, "learning_rate": 2.9683766690091355e-05, "loss": 0.0404, "step": 7044 }, { "epoch": 4.950808151791989, "grad_norm": 0.25778162479400635, "learning_rate": 2.9687983134223475e-05, "loss": 0.0341, "step": 7045 }, { "epoch": 4.951510892480675, "grad_norm": 0.20574626326560974, "learning_rate": 2.9692199578355588e-05, "loss": 0.0281, "step": 7046 }, { "epoch": 4.952213633169361, "grad_norm": 0.2489202916622162, "learning_rate": 2.9696416022487704e-05, "loss": 0.0232, "step": 7047 }, { "epoch": 4.952916373858047, "grad_norm": 0.2827945947647095, "learning_rate": 2.9700632466619818e-05, "loss": 0.0151, "step": 7048 }, { "epoch": 4.953619114546732, "grad_norm": 0.2684025466442108, "learning_rate": 2.9704848910751934e-05, "loss": 0.0241, "step": 7049 }, { "epoch": 4.954321855235418, "grad_norm": 0.19585761427879333, "learning_rate": 2.9709065354884048e-05, "loss": 0.0209, "step": 7050 }, { "epoch": 4.955024595924104, "grad_norm": 0.20889484882354736, "learning_rate": 2.9713281799016164e-05, "loss": 0.0209, "step": 7051 }, { "epoch": 4.95572733661279, "grad_norm": 0.25585630536079407, "learning_rate": 2.9717498243148278e-05, "loss": 0.0186, "step": 7052 }, { "epoch": 4.956430077301476, "grad_norm": 0.2450440675020218, "learning_rate": 2.9721714687280394e-05, "loss": 0.0404, "step": 7053 }, { "epoch": 4.9571328179901615, "grad_norm": 0.2569912374019623, "learning_rate": 2.9725931131412507e-05, "loss": 0.0258, "step": 7054 }, { "epoch": 4.957835558678847, "grad_norm": 0.2862156927585602, "learning_rate": 2.9730147575544624e-05, "loss": 0.041, "step": 7055 }, { "epoch": 4.958538299367533, "grad_norm": 0.3169426918029785, "learning_rate": 2.9734364019676737e-05, "loss": 0.0401, "step": 7056 }, { "epoch": 4.959241040056219, "grad_norm": 0.3118356466293335, "learning_rate": 2.9738580463808854e-05, "loss": 0.0348, "step": 7057 }, { "epoch": 4.959943780744905, "grad_norm": 0.3852181136608124, "learning_rate": 2.974279690794097e-05, "loss": 0.0431, "step": 7058 }, { "epoch": 4.960646521433591, "grad_norm": 0.45186376571655273, "learning_rate": 2.9747013352073087e-05, "loss": 0.027, "step": 7059 }, { "epoch": 4.961349262122277, "grad_norm": 0.3396901786327362, "learning_rate": 2.9751229796205204e-05, "loss": 0.0619, "step": 7060 }, { "epoch": 4.962052002810963, "grad_norm": 0.35193949937820435, "learning_rate": 2.9755446240337317e-05, "loss": 0.0954, "step": 7061 }, { "epoch": 4.962754743499649, "grad_norm": 0.4138906002044678, "learning_rate": 2.9759662684469434e-05, "loss": 0.071, "step": 7062 }, { "epoch": 4.963457484188335, "grad_norm": 0.4670344293117523, "learning_rate": 2.9763879128601547e-05, "loss": 0.108, "step": 7063 }, { "epoch": 4.964160224877021, "grad_norm": 0.46326926350593567, "learning_rate": 2.9768095572733664e-05, "loss": 0.1482, "step": 7064 }, { "epoch": 4.964862965565707, "grad_norm": 0.6463460922241211, "learning_rate": 2.9772312016865777e-05, "loss": 0.1964, "step": 7065 }, { "epoch": 4.965565706254392, "grad_norm": 3.1431074142456055, "learning_rate": 2.9776528460997894e-05, "loss": 0.2431, "step": 7066 }, { "epoch": 4.966268446943078, "grad_norm": 2.657606840133667, "learning_rate": 2.9780744905130007e-05, "loss": 0.2582, "step": 7067 }, { "epoch": 4.9669711876317635, "grad_norm": 0.24428193271160126, "learning_rate": 2.9784961349262123e-05, "loss": 0.0812, "step": 7068 }, { "epoch": 4.967673928320449, "grad_norm": 0.23332546651363373, "learning_rate": 2.9789177793394237e-05, "loss": 0.051, "step": 7069 }, { "epoch": 4.968376669009135, "grad_norm": 0.213247150182724, "learning_rate": 2.9793394237526353e-05, "loss": 0.0367, "step": 7070 }, { "epoch": 4.969079409697821, "grad_norm": 0.2728801369667053, "learning_rate": 2.9797610681658467e-05, "loss": 0.042, "step": 7071 }, { "epoch": 4.969782150386507, "grad_norm": 0.16710004210472107, "learning_rate": 2.9801827125790587e-05, "loss": 0.0238, "step": 7072 }, { "epoch": 4.970484891075193, "grad_norm": 0.18476097285747528, "learning_rate": 2.98060435699227e-05, "loss": 0.0294, "step": 7073 }, { "epoch": 4.971187631763879, "grad_norm": 0.17096176743507385, "learning_rate": 2.9810260014054816e-05, "loss": 0.0188, "step": 7074 }, { "epoch": 4.971890372452565, "grad_norm": 0.2705177366733551, "learning_rate": 2.981447645818693e-05, "loss": 0.0282, "step": 7075 }, { "epoch": 4.972593113141251, "grad_norm": 0.15823279321193695, "learning_rate": 2.9818692902319046e-05, "loss": 0.0283, "step": 7076 }, { "epoch": 4.973295853829937, "grad_norm": 0.15756286680698395, "learning_rate": 2.982290934645116e-05, "loss": 0.0138, "step": 7077 }, { "epoch": 4.973998594518623, "grad_norm": 0.27731066942214966, "learning_rate": 2.9827125790583276e-05, "loss": 0.0467, "step": 7078 }, { "epoch": 4.974701335207309, "grad_norm": 0.1923743635416031, "learning_rate": 2.983134223471539e-05, "loss": 0.0232, "step": 7079 }, { "epoch": 4.9754040758959945, "grad_norm": 0.2872786223888397, "learning_rate": 2.9835558678847506e-05, "loss": 0.0375, "step": 7080 }, { "epoch": 4.9761068165846805, "grad_norm": 0.2315407693386078, "learning_rate": 2.983977512297962e-05, "loss": 0.0213, "step": 7081 }, { "epoch": 4.976809557273366, "grad_norm": 0.39478567242622375, "learning_rate": 2.9843991567111736e-05, "loss": 0.0337, "step": 7082 }, { "epoch": 4.977512297962052, "grad_norm": 0.29035457968711853, "learning_rate": 2.984820801124385e-05, "loss": 0.0394, "step": 7083 }, { "epoch": 4.978215038650738, "grad_norm": 0.22828422486782074, "learning_rate": 2.9852424455375966e-05, "loss": 0.0225, "step": 7084 }, { "epoch": 4.978917779339424, "grad_norm": 0.3942031264305115, "learning_rate": 2.9856640899508083e-05, "loss": 0.0416, "step": 7085 }, { "epoch": 4.97962052002811, "grad_norm": 0.2887871563434601, "learning_rate": 2.98608573436402e-05, "loss": 0.0421, "step": 7086 }, { "epoch": 4.980323260716796, "grad_norm": 0.3664858043193817, "learning_rate": 2.9865073787772312e-05, "loss": 0.0546, "step": 7087 }, { "epoch": 4.981026001405481, "grad_norm": 0.6026016473770142, "learning_rate": 2.986929023190443e-05, "loss": 0.1024, "step": 7088 }, { "epoch": 4.981728742094167, "grad_norm": 0.6680776476860046, "learning_rate": 2.9873506676036542e-05, "loss": 0.139, "step": 7089 }, { "epoch": 4.982431482782853, "grad_norm": 0.8726975917816162, "learning_rate": 2.987772312016866e-05, "loss": 0.1819, "step": 7090 }, { "epoch": 4.983134223471539, "grad_norm": 1.0058499574661255, "learning_rate": 2.9881939564300776e-05, "loss": 0.2215, "step": 7091 }, { "epoch": 4.983836964160225, "grad_norm": 1.36276376247406, "learning_rate": 2.988615600843289e-05, "loss": 0.3034, "step": 7092 }, { "epoch": 4.984539704848911, "grad_norm": 0.2999083697795868, "learning_rate": 2.9890372452565005e-05, "loss": 0.0941, "step": 7093 }, { "epoch": 4.9852424455375965, "grad_norm": 0.19339698553085327, "learning_rate": 2.989458889669712e-05, "loss": 0.0337, "step": 7094 }, { "epoch": 4.9859451862262825, "grad_norm": 0.15516836941242218, "learning_rate": 2.9898805340829235e-05, "loss": 0.0336, "step": 7095 }, { "epoch": 4.986647926914968, "grad_norm": 0.21580497920513153, "learning_rate": 2.990302178496135e-05, "loss": 0.0441, "step": 7096 }, { "epoch": 4.987350667603654, "grad_norm": 0.34303149580955505, "learning_rate": 2.9907238229093465e-05, "loss": 0.0133, "step": 7097 }, { "epoch": 4.98805340829234, "grad_norm": 0.16153249144554138, "learning_rate": 2.991145467322558e-05, "loss": 0.0256, "step": 7098 }, { "epoch": 4.988756148981026, "grad_norm": 0.759032666683197, "learning_rate": 2.99156711173577e-05, "loss": 0.0177, "step": 7099 }, { "epoch": 4.989458889669712, "grad_norm": 0.2092856466770172, "learning_rate": 2.9919887561489812e-05, "loss": 0.0291, "step": 7100 }, { "epoch": 4.990161630358398, "grad_norm": 0.1210612952709198, "learning_rate": 2.992410400562193e-05, "loss": 0.0147, "step": 7101 }, { "epoch": 4.990864371047084, "grad_norm": 0.18891707062721252, "learning_rate": 2.992832044975404e-05, "loss": 0.0218, "step": 7102 }, { "epoch": 4.99156711173577, "grad_norm": 0.35556694865226746, "learning_rate": 2.9932536893886158e-05, "loss": 0.0204, "step": 7103 }, { "epoch": 4.992269852424456, "grad_norm": 0.22388169169425964, "learning_rate": 2.993675333801827e-05, "loss": 0.0357, "step": 7104 }, { "epoch": 4.992972593113141, "grad_norm": 0.29951798915863037, "learning_rate": 2.9940969782150388e-05, "loss": 0.0212, "step": 7105 }, { "epoch": 4.993675333801827, "grad_norm": 0.8212118148803711, "learning_rate": 2.99451862262825e-05, "loss": 0.0602, "step": 7106 }, { "epoch": 4.994378074490513, "grad_norm": 0.30835017561912537, "learning_rate": 2.9949402670414618e-05, "loss": 0.0244, "step": 7107 }, { "epoch": 4.9950808151791986, "grad_norm": 0.39083683490753174, "learning_rate": 2.995361911454673e-05, "loss": 0.0645, "step": 7108 }, { "epoch": 4.9957835558678845, "grad_norm": 0.25347474217414856, "learning_rate": 2.9957835558678848e-05, "loss": 0.0469, "step": 7109 }, { "epoch": 4.99648629655657, "grad_norm": 0.5363970994949341, "learning_rate": 2.996205200281096e-05, "loss": 0.0622, "step": 7110 }, { "epoch": 4.997189037245256, "grad_norm": 0.43508002161979675, "learning_rate": 2.9966268446943078e-05, "loss": 0.1204, "step": 7111 }, { "epoch": 4.997891777933942, "grad_norm": 1.4812309741973877, "learning_rate": 2.997048489107519e-05, "loss": 0.1573, "step": 7112 }, { "epoch": 4.998594518622628, "grad_norm": 0.7836503982543945, "learning_rate": 2.997470133520731e-05, "loss": 0.2323, "step": 7113 }, { "epoch": 4.999297259311314, "grad_norm": 1.461397409439087, "learning_rate": 2.9978917779339424e-05, "loss": 0.2957, "step": 7114 }, { "epoch": 5.0, "grad_norm": 0.8232632279396057, "learning_rate": 2.998313422347154e-05, "loss": 0.2063, "step": 7115 }, { "epoch": 5.000702740688686, "grad_norm": 0.3503764867782593, "learning_rate": 2.9987350667603654e-05, "loss": 0.1042, "step": 7116 }, { "epoch": 5.001405481377372, "grad_norm": 0.15472997725009918, "learning_rate": 2.999156711173577e-05, "loss": 0.0287, "step": 7117 }, { "epoch": 5.002108222066058, "grad_norm": 0.1404210329055786, "learning_rate": 2.9995783555867884e-05, "loss": 0.0339, "step": 7118 }, { "epoch": 5.002810962754744, "grad_norm": 0.12290169298648834, "learning_rate": 3e-05, "loss": 0.0152, "step": 7119 }, { "epoch": 5.00351370344343, "grad_norm": 0.24206362664699554, "learning_rate": 2.9999531506207544e-05, "loss": 0.0211, "step": 7120 }, { "epoch": 5.0042164441321155, "grad_norm": 0.21703851222991943, "learning_rate": 2.9999063012415088e-05, "loss": 0.0237, "step": 7121 }, { "epoch": 5.0049191848208014, "grad_norm": 0.30720213055610657, "learning_rate": 2.999859451862263e-05, "loss": 0.0358, "step": 7122 }, { "epoch": 5.005621925509487, "grad_norm": 0.2501491904258728, "learning_rate": 2.9998126024830172e-05, "loss": 0.0267, "step": 7123 }, { "epoch": 5.006324666198173, "grad_norm": 0.9376144409179688, "learning_rate": 2.9997657531037716e-05, "loss": 0.0352, "step": 7124 }, { "epoch": 5.007027406886858, "grad_norm": 0.17793017625808716, "learning_rate": 2.999718903724526e-05, "loss": 0.0132, "step": 7125 }, { "epoch": 5.007730147575544, "grad_norm": 0.21429257094860077, "learning_rate": 2.99967205434528e-05, "loss": 0.0304, "step": 7126 }, { "epoch": 5.00843288826423, "grad_norm": 0.18100792169570923, "learning_rate": 2.9996252049660344e-05, "loss": 0.0271, "step": 7127 }, { "epoch": 5.009135628952916, "grad_norm": 0.19019973278045654, "learning_rate": 2.9995783555867884e-05, "loss": 0.0448, "step": 7128 }, { "epoch": 5.009838369641602, "grad_norm": 0.1713096648454666, "learning_rate": 2.9995315062075428e-05, "loss": 0.0213, "step": 7129 }, { "epoch": 5.010541110330288, "grad_norm": 0.26619258522987366, "learning_rate": 2.9994846568282968e-05, "loss": 0.0289, "step": 7130 }, { "epoch": 5.011243851018974, "grad_norm": 0.25103670358657837, "learning_rate": 2.9994378074490512e-05, "loss": 0.0457, "step": 7131 }, { "epoch": 5.01194659170766, "grad_norm": 0.177617609500885, "learning_rate": 2.9993909580698056e-05, "loss": 0.0353, "step": 7132 }, { "epoch": 5.012649332396346, "grad_norm": 0.3054504692554474, "learning_rate": 2.99934410869056e-05, "loss": 0.0468, "step": 7133 }, { "epoch": 5.013352073085032, "grad_norm": 0.2935950458049774, "learning_rate": 2.9992972593113143e-05, "loss": 0.042, "step": 7134 }, { "epoch": 5.0140548137737175, "grad_norm": 0.8026213645935059, "learning_rate": 2.9992504099320683e-05, "loss": 0.0458, "step": 7135 }, { "epoch": 5.0147575544624035, "grad_norm": 0.5522305369377136, "learning_rate": 2.9992035605528227e-05, "loss": 0.0851, "step": 7136 }, { "epoch": 5.015460295151089, "grad_norm": 0.8687443733215332, "learning_rate": 2.999156711173577e-05, "loss": 0.1458, "step": 7137 }, { "epoch": 5.016163035839775, "grad_norm": 1.7470459938049316, "learning_rate": 2.9991098617943315e-05, "loss": 0.1652, "step": 7138 }, { "epoch": 5.016865776528461, "grad_norm": 0.877411425113678, "learning_rate": 2.9990630124150855e-05, "loss": 0.208, "step": 7139 }, { "epoch": 5.017568517217147, "grad_norm": 1.2951736450195312, "learning_rate": 2.99901616303584e-05, "loss": 0.2311, "step": 7140 }, { "epoch": 5.018271257905833, "grad_norm": 0.5611412525177002, "learning_rate": 2.9989693136565942e-05, "loss": 0.0777, "step": 7141 }, { "epoch": 5.018973998594519, "grad_norm": 0.14648588001728058, "learning_rate": 2.9989224642773486e-05, "loss": 0.0253, "step": 7142 }, { "epoch": 5.019676739283205, "grad_norm": 0.19279369711875916, "learning_rate": 2.9988756148981026e-05, "loss": 0.0287, "step": 7143 }, { "epoch": 5.02037947997189, "grad_norm": 0.2217462956905365, "learning_rate": 2.998828765518857e-05, "loss": 0.0365, "step": 7144 }, { "epoch": 5.021082220660576, "grad_norm": 0.47407621145248413, "learning_rate": 2.9987819161396114e-05, "loss": 0.0191, "step": 7145 }, { "epoch": 5.021784961349262, "grad_norm": 0.17421479523181915, "learning_rate": 2.9987350667603654e-05, "loss": 0.0177, "step": 7146 }, { "epoch": 5.022487702037948, "grad_norm": 0.1497834026813507, "learning_rate": 2.9986882173811198e-05, "loss": 0.0157, "step": 7147 }, { "epoch": 5.023190442726634, "grad_norm": 0.3747480511665344, "learning_rate": 2.9986413680018738e-05, "loss": 0.0456, "step": 7148 }, { "epoch": 5.0238931834153195, "grad_norm": 0.22619567811489105, "learning_rate": 2.9985945186226282e-05, "loss": 0.0269, "step": 7149 }, { "epoch": 5.0245959241040055, "grad_norm": 0.15768253803253174, "learning_rate": 2.9985476692433826e-05, "loss": 0.0141, "step": 7150 }, { "epoch": 5.025298664792691, "grad_norm": 0.13680483400821686, "learning_rate": 2.998500819864137e-05, "loss": 0.0227, "step": 7151 }, { "epoch": 5.026001405481377, "grad_norm": 0.20326875150203705, "learning_rate": 2.998453970484891e-05, "loss": 0.0233, "step": 7152 }, { "epoch": 5.026704146170063, "grad_norm": 0.21064545214176178, "learning_rate": 2.9984071211056454e-05, "loss": 0.0316, "step": 7153 }, { "epoch": 5.027406886858749, "grad_norm": 0.29694247245788574, "learning_rate": 2.9983602717263997e-05, "loss": 0.0259, "step": 7154 }, { "epoch": 5.028109627547435, "grad_norm": 0.251529723405838, "learning_rate": 2.998313422347154e-05, "loss": 0.0332, "step": 7155 }, { "epoch": 5.028812368236121, "grad_norm": 0.4021762013435364, "learning_rate": 2.998266572967908e-05, "loss": 0.0362, "step": 7156 }, { "epoch": 5.029515108924807, "grad_norm": 0.44713979959487915, "learning_rate": 2.9982197235886625e-05, "loss": 0.0268, "step": 7157 }, { "epoch": 5.030217849613493, "grad_norm": 0.22594526410102844, "learning_rate": 2.998172874209417e-05, "loss": 0.0363, "step": 7158 }, { "epoch": 5.030920590302179, "grad_norm": 3.120939016342163, "learning_rate": 2.9981260248301712e-05, "loss": 0.0393, "step": 7159 }, { "epoch": 5.031623330990865, "grad_norm": 0.29808294773101807, "learning_rate": 2.9980791754509256e-05, "loss": 0.0763, "step": 7160 }, { "epoch": 5.032326071679551, "grad_norm": 0.5117647647857666, "learning_rate": 2.9980323260716797e-05, "loss": 0.1181, "step": 7161 }, { "epoch": 5.0330288123682365, "grad_norm": 0.5002261400222778, "learning_rate": 2.997985476692434e-05, "loss": 0.1213, "step": 7162 }, { "epoch": 5.033731553056922, "grad_norm": 0.6340384483337402, "learning_rate": 2.997938627313188e-05, "loss": 0.1699, "step": 7163 }, { "epoch": 5.0344342937456075, "grad_norm": 1.291917324066162, "learning_rate": 2.9978917779339424e-05, "loss": 0.2601, "step": 7164 }, { "epoch": 5.035137034434293, "grad_norm": 1.2539689540863037, "learning_rate": 2.9978449285546965e-05, "loss": 0.2614, "step": 7165 }, { "epoch": 5.035839775122979, "grad_norm": 0.27466699481010437, "learning_rate": 2.997798079175451e-05, "loss": 0.086, "step": 7166 }, { "epoch": 5.036542515811665, "grad_norm": 0.2934017777442932, "learning_rate": 2.9977512297962052e-05, "loss": 0.0299, "step": 7167 }, { "epoch": 5.037245256500351, "grad_norm": 0.11896409094333649, "learning_rate": 2.9977043804169596e-05, "loss": 0.0194, "step": 7168 }, { "epoch": 5.037947997189037, "grad_norm": 0.22877304255962372, "learning_rate": 2.9976575310377136e-05, "loss": 0.0331, "step": 7169 }, { "epoch": 5.038650737877723, "grad_norm": 0.11219342797994614, "learning_rate": 2.997610681658468e-05, "loss": 0.0177, "step": 7170 }, { "epoch": 5.039353478566409, "grad_norm": 0.1764705628156662, "learning_rate": 2.9975638322792224e-05, "loss": 0.0225, "step": 7171 }, { "epoch": 5.040056219255095, "grad_norm": 0.15317639708518982, "learning_rate": 2.9975169828999767e-05, "loss": 0.0218, "step": 7172 }, { "epoch": 5.040758959943781, "grad_norm": 0.17415547370910645, "learning_rate": 2.997470133520731e-05, "loss": 0.0152, "step": 7173 }, { "epoch": 5.041461700632467, "grad_norm": 0.23271310329437256, "learning_rate": 2.997423284141485e-05, "loss": 0.0332, "step": 7174 }, { "epoch": 5.042164441321153, "grad_norm": 0.18693704903125763, "learning_rate": 2.9973764347622395e-05, "loss": 0.0189, "step": 7175 }, { "epoch": 5.0428671820098385, "grad_norm": 0.17971393465995789, "learning_rate": 2.997329585382994e-05, "loss": 0.0303, "step": 7176 }, { "epoch": 5.043569922698524, "grad_norm": 0.2844192683696747, "learning_rate": 2.9972827360037483e-05, "loss": 0.0254, "step": 7177 }, { "epoch": 5.04427266338721, "grad_norm": 0.15818315744400024, "learning_rate": 2.9972358866245023e-05, "loss": 0.0315, "step": 7178 }, { "epoch": 5.044975404075896, "grad_norm": 0.44242915511131287, "learning_rate": 2.9971890372452567e-05, "loss": 0.0396, "step": 7179 }, { "epoch": 5.045678144764582, "grad_norm": 0.23529136180877686, "learning_rate": 2.9971421878660107e-05, "loss": 0.0408, "step": 7180 }, { "epoch": 5.046380885453268, "grad_norm": 0.3119789659976959, "learning_rate": 2.997095338486765e-05, "loss": 0.0662, "step": 7181 }, { "epoch": 5.047083626141954, "grad_norm": 0.34162411093711853, "learning_rate": 2.997048489107519e-05, "loss": 0.0328, "step": 7182 }, { "epoch": 5.047786366830639, "grad_norm": 0.5342933535575867, "learning_rate": 2.9970016397282735e-05, "loss": 0.0394, "step": 7183 }, { "epoch": 5.048489107519325, "grad_norm": 0.26571202278137207, "learning_rate": 2.996954790349028e-05, "loss": 0.0399, "step": 7184 }, { "epoch": 5.049191848208011, "grad_norm": 0.5337769389152527, "learning_rate": 2.9969079409697822e-05, "loss": 0.0764, "step": 7185 }, { "epoch": 5.049894588896697, "grad_norm": 0.522738516330719, "learning_rate": 2.9968610915905366e-05, "loss": 0.0852, "step": 7186 }, { "epoch": 5.050597329585383, "grad_norm": 0.8656845688819885, "learning_rate": 2.9968142422112906e-05, "loss": 0.1152, "step": 7187 }, { "epoch": 5.051300070274069, "grad_norm": 1.3242770433425903, "learning_rate": 2.996767392832045e-05, "loss": 0.2063, "step": 7188 }, { "epoch": 5.052002810962755, "grad_norm": 1.0802770853042603, "learning_rate": 2.9967205434527994e-05, "loss": 0.2371, "step": 7189 }, { "epoch": 5.0527055516514405, "grad_norm": 1.7838101387023926, "learning_rate": 2.9966736940735537e-05, "loss": 0.2683, "step": 7190 }, { "epoch": 5.0534082923401265, "grad_norm": 0.30801236629486084, "learning_rate": 2.9966268446943078e-05, "loss": 0.0795, "step": 7191 }, { "epoch": 5.054111033028812, "grad_norm": 0.3035440742969513, "learning_rate": 2.996579995315062e-05, "loss": 0.0362, "step": 7192 }, { "epoch": 5.054813773717498, "grad_norm": 0.22993867099285126, "learning_rate": 2.9965331459358165e-05, "loss": 0.0404, "step": 7193 }, { "epoch": 5.055516514406184, "grad_norm": 0.19400526583194733, "learning_rate": 2.996486296556571e-05, "loss": 0.0235, "step": 7194 }, { "epoch": 5.05621925509487, "grad_norm": 0.19566622376441956, "learning_rate": 2.996439447177325e-05, "loss": 0.0241, "step": 7195 }, { "epoch": 5.056921995783556, "grad_norm": 0.21945297718048096, "learning_rate": 2.9963925977980793e-05, "loss": 0.0292, "step": 7196 }, { "epoch": 5.057624736472242, "grad_norm": 0.17654789984226227, "learning_rate": 2.9963457484188337e-05, "loss": 0.0204, "step": 7197 }, { "epoch": 5.058327477160928, "grad_norm": 0.17854194343090057, "learning_rate": 2.9962988990395877e-05, "loss": 0.0298, "step": 7198 }, { "epoch": 5.059030217849614, "grad_norm": 0.1911163032054901, "learning_rate": 2.996252049660342e-05, "loss": 0.03, "step": 7199 }, { "epoch": 5.0597329585383, "grad_norm": 0.720078706741333, "learning_rate": 2.996205200281096e-05, "loss": 0.0225, "step": 7200 }, { "epoch": 5.060435699226986, "grad_norm": 0.24159865081310272, "learning_rate": 2.9961583509018505e-05, "loss": 0.0297, "step": 7201 }, { "epoch": 5.061138439915671, "grad_norm": 0.24862898886203766, "learning_rate": 2.996111501522605e-05, "loss": 0.021, "step": 7202 }, { "epoch": 5.061841180604357, "grad_norm": 0.21326051652431488, "learning_rate": 2.9960646521433592e-05, "loss": 0.0249, "step": 7203 }, { "epoch": 5.0625439212930425, "grad_norm": 0.22662872076034546, "learning_rate": 2.9960178027641133e-05, "loss": 0.0281, "step": 7204 }, { "epoch": 5.0632466619817285, "grad_norm": 0.1875026375055313, "learning_rate": 2.9959709533848676e-05, "loss": 0.0542, "step": 7205 }, { "epoch": 5.063949402670414, "grad_norm": 0.7829228639602661, "learning_rate": 2.995924104005622e-05, "loss": 0.0398, "step": 7206 }, { "epoch": 5.0646521433591, "grad_norm": 0.3718477189540863, "learning_rate": 2.9958772546263764e-05, "loss": 0.0633, "step": 7207 }, { "epoch": 5.065354884047786, "grad_norm": 0.2227337658405304, "learning_rate": 2.9958304052471304e-05, "loss": 0.0511, "step": 7208 }, { "epoch": 5.066057624736472, "grad_norm": 0.4785904586315155, "learning_rate": 2.9957835558678848e-05, "loss": 0.061, "step": 7209 }, { "epoch": 5.066760365425158, "grad_norm": 0.3715161085128784, "learning_rate": 2.995736706488639e-05, "loss": 0.0644, "step": 7210 }, { "epoch": 5.067463106113844, "grad_norm": 0.5349627733230591, "learning_rate": 2.9956898571093935e-05, "loss": 0.1228, "step": 7211 }, { "epoch": 5.06816584680253, "grad_norm": 0.5650061368942261, "learning_rate": 2.995643007730148e-05, "loss": 0.1267, "step": 7212 }, { "epoch": 5.068868587491216, "grad_norm": 0.7487751841545105, "learning_rate": 2.995596158350902e-05, "loss": 0.1751, "step": 7213 }, { "epoch": 5.069571328179902, "grad_norm": 1.4194613695144653, "learning_rate": 2.9955493089716563e-05, "loss": 0.226, "step": 7214 }, { "epoch": 5.070274068868588, "grad_norm": 1.6331067085266113, "learning_rate": 2.9955024595924104e-05, "loss": 0.2869, "step": 7215 }, { "epoch": 5.070976809557274, "grad_norm": 0.38547655940055847, "learning_rate": 2.9954556102131647e-05, "loss": 0.0987, "step": 7216 }, { "epoch": 5.0716795502459595, "grad_norm": 0.34600210189819336, "learning_rate": 2.9954087608339188e-05, "loss": 0.0324, "step": 7217 }, { "epoch": 5.072382290934645, "grad_norm": 0.17336703836917877, "learning_rate": 2.995361911454673e-05, "loss": 0.0275, "step": 7218 }, { "epoch": 5.073085031623331, "grad_norm": 0.21274247765541077, "learning_rate": 2.9953150620754275e-05, "loss": 0.0282, "step": 7219 }, { "epoch": 5.073787772312017, "grad_norm": 0.2201123833656311, "learning_rate": 2.995268212696182e-05, "loss": 0.0211, "step": 7220 }, { "epoch": 5.074490513000702, "grad_norm": 0.20007435977458954, "learning_rate": 2.9952213633169362e-05, "loss": 0.0164, "step": 7221 }, { "epoch": 5.075193253689388, "grad_norm": 0.576537549495697, "learning_rate": 2.9951745139376903e-05, "loss": 0.0269, "step": 7222 }, { "epoch": 5.075895994378074, "grad_norm": 0.2746924161911011, "learning_rate": 2.9951276645584447e-05, "loss": 0.0405, "step": 7223 }, { "epoch": 5.07659873506676, "grad_norm": 0.23982630670070648, "learning_rate": 2.995080815179199e-05, "loss": 0.0247, "step": 7224 }, { "epoch": 5.077301475755446, "grad_norm": 0.13691186904907227, "learning_rate": 2.9950339657999534e-05, "loss": 0.0116, "step": 7225 }, { "epoch": 5.078004216444132, "grad_norm": 0.13399283587932587, "learning_rate": 2.9949871164207074e-05, "loss": 0.0412, "step": 7226 }, { "epoch": 5.078706957132818, "grad_norm": 0.1560508906841278, "learning_rate": 2.9949402670414618e-05, "loss": 0.0268, "step": 7227 }, { "epoch": 5.079409697821504, "grad_norm": 0.14120405912399292, "learning_rate": 2.9948934176622162e-05, "loss": 0.029, "step": 7228 }, { "epoch": 5.08011243851019, "grad_norm": 0.12577702105045319, "learning_rate": 2.9948465682829705e-05, "loss": 0.0199, "step": 7229 }, { "epoch": 5.080815179198876, "grad_norm": 0.17400392889976501, "learning_rate": 2.9947997189037246e-05, "loss": 0.0542, "step": 7230 }, { "epoch": 5.0815179198875615, "grad_norm": 0.20650549232959747, "learning_rate": 2.994752869524479e-05, "loss": 0.0402, "step": 7231 }, { "epoch": 5.082220660576247, "grad_norm": 0.1671803742647171, "learning_rate": 2.9947060201452333e-05, "loss": 0.023, "step": 7232 }, { "epoch": 5.082923401264933, "grad_norm": 0.35133761167526245, "learning_rate": 2.9946591707659874e-05, "loss": 0.0545, "step": 7233 }, { "epoch": 5.083626141953619, "grad_norm": 0.25658300518989563, "learning_rate": 2.9946123213867417e-05, "loss": 0.0454, "step": 7234 }, { "epoch": 5.084328882642305, "grad_norm": 0.3533341884613037, "learning_rate": 2.9945654720074958e-05, "loss": 0.0529, "step": 7235 }, { "epoch": 5.085031623330991, "grad_norm": 0.32659876346588135, "learning_rate": 2.99451862262825e-05, "loss": 0.0767, "step": 7236 }, { "epoch": 5.085734364019677, "grad_norm": 0.6594269871711731, "learning_rate": 2.9944717732490045e-05, "loss": 0.1566, "step": 7237 }, { "epoch": 5.086437104708363, "grad_norm": 0.6330965161323547, "learning_rate": 2.994424923869759e-05, "loss": 0.1536, "step": 7238 }, { "epoch": 5.087139845397049, "grad_norm": 4.084334373474121, "learning_rate": 2.994378074490513e-05, "loss": 0.2278, "step": 7239 }, { "epoch": 5.087842586085735, "grad_norm": 1.3776886463165283, "learning_rate": 2.9943312251112673e-05, "loss": 0.3092, "step": 7240 }, { "epoch": 5.08854532677442, "grad_norm": 0.27242138981819153, "learning_rate": 2.9942843757320217e-05, "loss": 0.0917, "step": 7241 }, { "epoch": 5.089248067463106, "grad_norm": 0.2096564918756485, "learning_rate": 2.994237526352776e-05, "loss": 0.0312, "step": 7242 }, { "epoch": 5.089950808151792, "grad_norm": 0.2166728377342224, "learning_rate": 2.99419067697353e-05, "loss": 0.0311, "step": 7243 }, { "epoch": 5.090653548840478, "grad_norm": 0.1486634463071823, "learning_rate": 2.9941438275942844e-05, "loss": 0.0219, "step": 7244 }, { "epoch": 5.0913562895291635, "grad_norm": 0.2402956336736679, "learning_rate": 2.9940969782150388e-05, "loss": 0.0232, "step": 7245 }, { "epoch": 5.0920590302178494, "grad_norm": 0.16227537393569946, "learning_rate": 2.9940501288357932e-05, "loss": 0.0186, "step": 7246 }, { "epoch": 5.092761770906535, "grad_norm": 0.15748785436153412, "learning_rate": 2.9940032794565476e-05, "loss": 0.0161, "step": 7247 }, { "epoch": 5.093464511595221, "grad_norm": 0.17581750452518463, "learning_rate": 2.9939564300773016e-05, "loss": 0.0174, "step": 7248 }, { "epoch": 5.094167252283907, "grad_norm": 0.18515917658805847, "learning_rate": 2.993909580698056e-05, "loss": 0.0255, "step": 7249 }, { "epoch": 5.094869992972593, "grad_norm": 0.1663016378879547, "learning_rate": 2.99386273131881e-05, "loss": 0.0197, "step": 7250 }, { "epoch": 5.095572733661279, "grad_norm": 0.21807482838630676, "learning_rate": 2.9938158819395644e-05, "loss": 0.0288, "step": 7251 }, { "epoch": 5.096275474349965, "grad_norm": 0.22173097729682922, "learning_rate": 2.9937690325603184e-05, "loss": 0.019, "step": 7252 }, { "epoch": 5.096978215038651, "grad_norm": 0.29986342787742615, "learning_rate": 2.9937221831810728e-05, "loss": 0.0412, "step": 7253 }, { "epoch": 5.097680955727337, "grad_norm": 0.30509206652641296, "learning_rate": 2.993675333801827e-05, "loss": 0.0244, "step": 7254 }, { "epoch": 5.098383696416023, "grad_norm": 0.2732740342617035, "learning_rate": 2.9936284844225815e-05, "loss": 0.0293, "step": 7255 }, { "epoch": 5.099086437104709, "grad_norm": 0.23693178594112396, "learning_rate": 2.9935816350433356e-05, "loss": 0.0541, "step": 7256 }, { "epoch": 5.099789177793395, "grad_norm": 0.322490930557251, "learning_rate": 2.99353478566409e-05, "loss": 0.0275, "step": 7257 }, { "epoch": 5.1004919184820805, "grad_norm": 0.27575117349624634, "learning_rate": 2.9934879362848443e-05, "loss": 0.05, "step": 7258 }, { "epoch": 5.101194659170766, "grad_norm": 0.272152841091156, "learning_rate": 2.9934410869055987e-05, "loss": 0.0596, "step": 7259 }, { "epoch": 5.1018973998594515, "grad_norm": 0.789758563041687, "learning_rate": 2.993394237526353e-05, "loss": 0.056, "step": 7260 }, { "epoch": 5.102600140548137, "grad_norm": 0.8816899061203003, "learning_rate": 2.993347388147107e-05, "loss": 0.1387, "step": 7261 }, { "epoch": 5.103302881236823, "grad_norm": 0.7497177720069885, "learning_rate": 2.9933005387678615e-05, "loss": 0.127, "step": 7262 }, { "epoch": 5.104005621925509, "grad_norm": 2.888493537902832, "learning_rate": 2.9932536893886158e-05, "loss": 0.2113, "step": 7263 }, { "epoch": 5.104708362614195, "grad_norm": 4.932112216949463, "learning_rate": 2.9932068400093702e-05, "loss": 0.2046, "step": 7264 }, { "epoch": 5.105411103302881, "grad_norm": 1.274932622909546, "learning_rate": 2.9931599906301242e-05, "loss": 0.2589, "step": 7265 }, { "epoch": 5.106113843991567, "grad_norm": 0.2701186537742615, "learning_rate": 2.9931131412508786e-05, "loss": 0.0694, "step": 7266 }, { "epoch": 5.106816584680253, "grad_norm": 0.18773798644542694, "learning_rate": 2.993066291871633e-05, "loss": 0.0419, "step": 7267 }, { "epoch": 5.107519325368939, "grad_norm": 0.2240193486213684, "learning_rate": 2.993019442492387e-05, "loss": 0.043, "step": 7268 }, { "epoch": 5.108222066057625, "grad_norm": 0.21164491772651672, "learning_rate": 2.992972593113141e-05, "loss": 0.03, "step": 7269 }, { "epoch": 5.108924806746311, "grad_norm": 0.15264521539211273, "learning_rate": 2.9929257437338954e-05, "loss": 0.0248, "step": 7270 }, { "epoch": 5.109627547434997, "grad_norm": 0.4835006594657898, "learning_rate": 2.9928788943546498e-05, "loss": 0.016, "step": 7271 }, { "epoch": 5.1103302881236825, "grad_norm": 0.25879865884780884, "learning_rate": 2.992832044975404e-05, "loss": 0.0329, "step": 7272 }, { "epoch": 5.111033028812368, "grad_norm": 0.21329531073570251, "learning_rate": 2.9927851955961585e-05, "loss": 0.031, "step": 7273 }, { "epoch": 5.111735769501054, "grad_norm": 0.15548087656497955, "learning_rate": 2.9927383462169126e-05, "loss": 0.0212, "step": 7274 }, { "epoch": 5.11243851018974, "grad_norm": 0.13886256515979767, "learning_rate": 2.992691496837667e-05, "loss": 0.0134, "step": 7275 }, { "epoch": 5.113141250878426, "grad_norm": 0.1439008116722107, "learning_rate": 2.9926446474584213e-05, "loss": 0.0217, "step": 7276 }, { "epoch": 5.113843991567112, "grad_norm": 0.1760713905096054, "learning_rate": 2.9925977980791757e-05, "loss": 0.0114, "step": 7277 }, { "epoch": 5.114546732255798, "grad_norm": 0.1849372684955597, "learning_rate": 2.9925509486999297e-05, "loss": 0.0222, "step": 7278 }, { "epoch": 5.115249472944483, "grad_norm": 0.2678564190864563, "learning_rate": 2.992504099320684e-05, "loss": 0.0202, "step": 7279 }, { "epoch": 5.115952213633169, "grad_norm": 0.2789084315299988, "learning_rate": 2.9924572499414385e-05, "loss": 0.0439, "step": 7280 }, { "epoch": 5.116654954321855, "grad_norm": 0.28838226199150085, "learning_rate": 2.992410400562193e-05, "loss": 0.0414, "step": 7281 }, { "epoch": 5.117357695010541, "grad_norm": 0.3182353079319, "learning_rate": 2.992363551182947e-05, "loss": 0.0421, "step": 7282 }, { "epoch": 5.118060435699227, "grad_norm": 0.8502454161643982, "learning_rate": 2.9923167018037012e-05, "loss": 0.0294, "step": 7283 }, { "epoch": 5.118763176387913, "grad_norm": 0.27121928334236145, "learning_rate": 2.9922698524244556e-05, "loss": 0.0604, "step": 7284 }, { "epoch": 5.119465917076599, "grad_norm": 0.37877801060676575, "learning_rate": 2.9922230030452097e-05, "loss": 0.0803, "step": 7285 }, { "epoch": 5.1201686577652845, "grad_norm": 0.5427423119544983, "learning_rate": 2.992176153665964e-05, "loss": 0.1124, "step": 7286 }, { "epoch": 5.12087139845397, "grad_norm": 1.0841526985168457, "learning_rate": 2.992129304286718e-05, "loss": 0.1336, "step": 7287 }, { "epoch": 5.121574139142656, "grad_norm": 0.6385695934295654, "learning_rate": 2.9920824549074724e-05, "loss": 0.1325, "step": 7288 }, { "epoch": 5.122276879831342, "grad_norm": 1.0008200407028198, "learning_rate": 2.9920356055282268e-05, "loss": 0.2397, "step": 7289 }, { "epoch": 5.122979620520028, "grad_norm": 1.5784703493118286, "learning_rate": 2.9919887561489812e-05, "loss": 0.2565, "step": 7290 }, { "epoch": 5.123682361208714, "grad_norm": 0.20375841856002808, "learning_rate": 2.9919419067697352e-05, "loss": 0.0726, "step": 7291 }, { "epoch": 5.1243851018974, "grad_norm": 0.2613120675086975, "learning_rate": 2.9918950573904896e-05, "loss": 0.0248, "step": 7292 }, { "epoch": 5.125087842586086, "grad_norm": 0.12198465317487717, "learning_rate": 2.991848208011244e-05, "loss": 0.021, "step": 7293 }, { "epoch": 5.125790583274772, "grad_norm": 0.21812492609024048, "learning_rate": 2.9918013586319983e-05, "loss": 0.0228, "step": 7294 }, { "epoch": 5.126493323963458, "grad_norm": 0.20520180463790894, "learning_rate": 2.9917545092527524e-05, "loss": 0.0205, "step": 7295 }, { "epoch": 5.127196064652144, "grad_norm": 0.1555146723985672, "learning_rate": 2.9917076598735067e-05, "loss": 0.0189, "step": 7296 }, { "epoch": 5.12789880534083, "grad_norm": 0.20072270929813385, "learning_rate": 2.991660810494261e-05, "loss": 0.0257, "step": 7297 }, { "epoch": 5.128601546029515, "grad_norm": 0.46580255031585693, "learning_rate": 2.9916139611150155e-05, "loss": 0.027, "step": 7298 }, { "epoch": 5.129304286718201, "grad_norm": 0.16070930659770966, "learning_rate": 2.99156711173577e-05, "loss": 0.0188, "step": 7299 }, { "epoch": 5.1300070274068865, "grad_norm": 0.18558748066425323, "learning_rate": 2.991520262356524e-05, "loss": 0.0159, "step": 7300 }, { "epoch": 5.130709768095572, "grad_norm": 0.22959814965724945, "learning_rate": 2.9914734129772783e-05, "loss": 0.0315, "step": 7301 }, { "epoch": 5.131412508784258, "grad_norm": 0.527684211730957, "learning_rate": 2.9914265635980323e-05, "loss": 0.0158, "step": 7302 }, { "epoch": 5.132115249472944, "grad_norm": 0.2748446464538574, "learning_rate": 2.9913797142187867e-05, "loss": 0.0355, "step": 7303 }, { "epoch": 5.13281799016163, "grad_norm": 0.23953069746494293, "learning_rate": 2.9913328648395407e-05, "loss": 0.0213, "step": 7304 }, { "epoch": 5.133520730850316, "grad_norm": 0.2704862058162689, "learning_rate": 2.991286015460295e-05, "loss": 0.0462, "step": 7305 }, { "epoch": 5.134223471539002, "grad_norm": 0.3342060446739197, "learning_rate": 2.9912391660810494e-05, "loss": 0.044, "step": 7306 }, { "epoch": 5.134926212227688, "grad_norm": 0.13498404622077942, "learning_rate": 2.9911923167018038e-05, "loss": 0.0244, "step": 7307 }, { "epoch": 5.135628952916374, "grad_norm": 0.4269661009311676, "learning_rate": 2.991145467322558e-05, "loss": 0.0405, "step": 7308 }, { "epoch": 5.13633169360506, "grad_norm": 0.45038557052612305, "learning_rate": 2.9910986179433122e-05, "loss": 0.0532, "step": 7309 }, { "epoch": 5.137034434293746, "grad_norm": 0.3330816626548767, "learning_rate": 2.9910517685640666e-05, "loss": 0.0676, "step": 7310 }, { "epoch": 5.137737174982432, "grad_norm": 0.4218854606151581, "learning_rate": 2.991004919184821e-05, "loss": 0.0707, "step": 7311 }, { "epoch": 5.138439915671118, "grad_norm": 0.5750783085823059, "learning_rate": 2.9909580698055753e-05, "loss": 0.1279, "step": 7312 }, { "epoch": 5.1391426563598035, "grad_norm": 0.8244590759277344, "learning_rate": 2.9909112204263294e-05, "loss": 0.1892, "step": 7313 }, { "epoch": 5.139845397048489, "grad_norm": 1.3126966953277588, "learning_rate": 2.9908643710470837e-05, "loss": 0.2172, "step": 7314 }, { "epoch": 5.140548137737175, "grad_norm": 1.3619844913482666, "learning_rate": 2.990817521667838e-05, "loss": 0.32, "step": 7315 }, { "epoch": 5.141250878425861, "grad_norm": 0.3627021014690399, "learning_rate": 2.9907706722885925e-05, "loss": 0.0818, "step": 7316 }, { "epoch": 5.141953619114547, "grad_norm": 0.18559126555919647, "learning_rate": 2.9907238229093465e-05, "loss": 0.0259, "step": 7317 }, { "epoch": 5.142656359803232, "grad_norm": 0.3187492787837982, "learning_rate": 2.990676973530101e-05, "loss": 0.0341, "step": 7318 }, { "epoch": 5.143359100491918, "grad_norm": 0.18124602735042572, "learning_rate": 2.9906301241508553e-05, "loss": 0.013, "step": 7319 }, { "epoch": 5.144061841180604, "grad_norm": 0.16518132388591766, "learning_rate": 2.9905832747716093e-05, "loss": 0.0257, "step": 7320 }, { "epoch": 5.14476458186929, "grad_norm": 0.15622727572917938, "learning_rate": 2.9905364253923633e-05, "loss": 0.0371, "step": 7321 }, { "epoch": 5.145467322557976, "grad_norm": 0.11168255656957626, "learning_rate": 2.9904895760131177e-05, "loss": 0.0154, "step": 7322 }, { "epoch": 5.146170063246662, "grad_norm": 0.15876922011375427, "learning_rate": 2.990442726633872e-05, "loss": 0.0268, "step": 7323 }, { "epoch": 5.146872803935348, "grad_norm": 0.22932447493076324, "learning_rate": 2.9903958772546265e-05, "loss": 0.0384, "step": 7324 }, { "epoch": 5.147575544624034, "grad_norm": 0.25071191787719727, "learning_rate": 2.9903490278753808e-05, "loss": 0.0228, "step": 7325 }, { "epoch": 5.14827828531272, "grad_norm": 0.2878361940383911, "learning_rate": 2.990302178496135e-05, "loss": 0.025, "step": 7326 }, { "epoch": 5.1489810260014055, "grad_norm": 0.16367031633853912, "learning_rate": 2.9902553291168892e-05, "loss": 0.0238, "step": 7327 }, { "epoch": 5.149683766690091, "grad_norm": 0.29045364260673523, "learning_rate": 2.9902084797376436e-05, "loss": 0.0548, "step": 7328 }, { "epoch": 5.150386507378777, "grad_norm": 0.16828496754169464, "learning_rate": 2.990161630358398e-05, "loss": 0.0165, "step": 7329 }, { "epoch": 5.151089248067463, "grad_norm": 0.3625718355178833, "learning_rate": 2.990114780979152e-05, "loss": 0.0367, "step": 7330 }, { "epoch": 5.151791988756149, "grad_norm": 0.2249990552663803, "learning_rate": 2.9900679315999064e-05, "loss": 0.0315, "step": 7331 }, { "epoch": 5.152494729444835, "grad_norm": 0.22428880631923676, "learning_rate": 2.9900210822206608e-05, "loss": 0.0194, "step": 7332 }, { "epoch": 5.153197470133521, "grad_norm": 0.20253139734268188, "learning_rate": 2.989974232841415e-05, "loss": 0.05, "step": 7333 }, { "epoch": 5.153900210822207, "grad_norm": 0.2883929908275604, "learning_rate": 2.989927383462169e-05, "loss": 0.0416, "step": 7334 }, { "epoch": 5.154602951510893, "grad_norm": 0.3195645213127136, "learning_rate": 2.9898805340829235e-05, "loss": 0.0414, "step": 7335 }, { "epoch": 5.155305692199578, "grad_norm": 0.46615591645240784, "learning_rate": 2.989833684703678e-05, "loss": 0.0817, "step": 7336 }, { "epoch": 5.156008432888264, "grad_norm": 1.4185609817504883, "learning_rate": 2.989786835324432e-05, "loss": 0.1545, "step": 7337 }, { "epoch": 5.15671117357695, "grad_norm": 1.236722469329834, "learning_rate": 2.9897399859451863e-05, "loss": 0.1967, "step": 7338 }, { "epoch": 5.157413914265636, "grad_norm": 1.5894438028335571, "learning_rate": 2.9896931365659403e-05, "loss": 0.2689, "step": 7339 }, { "epoch": 5.158116654954322, "grad_norm": 1.2172435522079468, "learning_rate": 2.9896462871866947e-05, "loss": 0.2487, "step": 7340 }, { "epoch": 5.1588193956430075, "grad_norm": 0.32621896266937256, "learning_rate": 2.989599437807449e-05, "loss": 0.0704, "step": 7341 }, { "epoch": 5.159522136331693, "grad_norm": 0.19584569334983826, "learning_rate": 2.9895525884282035e-05, "loss": 0.0279, "step": 7342 }, { "epoch": 5.160224877020379, "grad_norm": 0.19377660751342773, "learning_rate": 2.9895057390489575e-05, "loss": 0.0269, "step": 7343 }, { "epoch": 5.160927617709065, "grad_norm": 0.15153591334819794, "learning_rate": 2.989458889669712e-05, "loss": 0.0263, "step": 7344 }, { "epoch": 5.161630358397751, "grad_norm": 0.23425835371017456, "learning_rate": 2.9894120402904662e-05, "loss": 0.0284, "step": 7345 }, { "epoch": 5.162333099086437, "grad_norm": 0.3285210430622101, "learning_rate": 2.9893651909112206e-05, "loss": 0.019, "step": 7346 }, { "epoch": 5.163035839775123, "grad_norm": 0.12046031653881073, "learning_rate": 2.9893183415319746e-05, "loss": 0.0172, "step": 7347 }, { "epoch": 5.163738580463809, "grad_norm": 0.20099419355392456, "learning_rate": 2.989271492152729e-05, "loss": 0.0474, "step": 7348 }, { "epoch": 5.164441321152495, "grad_norm": 0.25305822491645813, "learning_rate": 2.9892246427734834e-05, "loss": 0.0289, "step": 7349 }, { "epoch": 5.165144061841181, "grad_norm": 0.24056029319763184, "learning_rate": 2.9891777933942378e-05, "loss": 0.0192, "step": 7350 }, { "epoch": 5.165846802529867, "grad_norm": 0.38425979018211365, "learning_rate": 2.989130944014992e-05, "loss": 0.0389, "step": 7351 }, { "epoch": 5.166549543218553, "grad_norm": 0.16442546248435974, "learning_rate": 2.9890840946357462e-05, "loss": 0.0145, "step": 7352 }, { "epoch": 5.167252283907239, "grad_norm": 0.395033061504364, "learning_rate": 2.9890372452565005e-05, "loss": 0.0349, "step": 7353 }, { "epoch": 5.1679550245959245, "grad_norm": 0.24778397381305695, "learning_rate": 2.988990395877255e-05, "loss": 0.0232, "step": 7354 }, { "epoch": 5.16865776528461, "grad_norm": 0.2125503271818161, "learning_rate": 2.988943546498009e-05, "loss": 0.0345, "step": 7355 }, { "epoch": 5.169360505973295, "grad_norm": 0.23822064697742462, "learning_rate": 2.988896697118763e-05, "loss": 0.0463, "step": 7356 }, { "epoch": 5.170063246661981, "grad_norm": 0.19581487774848938, "learning_rate": 2.9888498477395174e-05, "loss": 0.025, "step": 7357 }, { "epoch": 5.170765987350667, "grad_norm": 0.24563512206077576, "learning_rate": 2.9888029983602717e-05, "loss": 0.0519, "step": 7358 }, { "epoch": 5.171468728039353, "grad_norm": 0.2953232228755951, "learning_rate": 2.988756148981026e-05, "loss": 0.0834, "step": 7359 }, { "epoch": 5.172171468728039, "grad_norm": 0.29864588379859924, "learning_rate": 2.98870929960178e-05, "loss": 0.0412, "step": 7360 }, { "epoch": 5.172874209416725, "grad_norm": 0.4409042000770569, "learning_rate": 2.9886624502225345e-05, "loss": 0.0824, "step": 7361 }, { "epoch": 5.173576950105411, "grad_norm": 0.5776435732841492, "learning_rate": 2.988615600843289e-05, "loss": 0.1604, "step": 7362 }, { "epoch": 5.174279690794097, "grad_norm": 1.6399388313293457, "learning_rate": 2.9885687514640433e-05, "loss": 0.1846, "step": 7363 }, { "epoch": 5.174982431482783, "grad_norm": 1.010785460472107, "learning_rate": 2.9885219020847976e-05, "loss": 0.275, "step": 7364 }, { "epoch": 5.175685172171469, "grad_norm": 1.4798086881637573, "learning_rate": 2.9884750527055517e-05, "loss": 0.241, "step": 7365 }, { "epoch": 5.176387912860155, "grad_norm": 0.3902947008609772, "learning_rate": 2.988428203326306e-05, "loss": 0.1356, "step": 7366 }, { "epoch": 5.177090653548841, "grad_norm": 0.27417561411857605, "learning_rate": 2.9883813539470604e-05, "loss": 0.0617, "step": 7367 }, { "epoch": 5.1777933942375265, "grad_norm": 0.11433660238981247, "learning_rate": 2.9883345045678148e-05, "loss": 0.0197, "step": 7368 }, { "epoch": 5.178496134926212, "grad_norm": 0.1392306089401245, "learning_rate": 2.9882876551885688e-05, "loss": 0.0219, "step": 7369 }, { "epoch": 5.179198875614898, "grad_norm": 0.14415773749351501, "learning_rate": 2.9882408058093232e-05, "loss": 0.0168, "step": 7370 }, { "epoch": 5.179901616303584, "grad_norm": 0.24529623985290527, "learning_rate": 2.9881939564300776e-05, "loss": 0.0254, "step": 7371 }, { "epoch": 5.18060435699227, "grad_norm": 0.22616459429264069, "learning_rate": 2.9881471070508316e-05, "loss": 0.0262, "step": 7372 }, { "epoch": 5.181307097680956, "grad_norm": 0.1644178330898285, "learning_rate": 2.9881002576715856e-05, "loss": 0.0382, "step": 7373 }, { "epoch": 5.182009838369642, "grad_norm": 0.1905241459608078, "learning_rate": 2.98805340829234e-05, "loss": 0.0251, "step": 7374 }, { "epoch": 5.182712579058327, "grad_norm": 0.18924696743488312, "learning_rate": 2.9880065589130944e-05, "loss": 0.0266, "step": 7375 }, { "epoch": 5.183415319747013, "grad_norm": 0.174935445189476, "learning_rate": 2.9879597095338487e-05, "loss": 0.0352, "step": 7376 }, { "epoch": 5.184118060435699, "grad_norm": 0.2593996226787567, "learning_rate": 2.987912860154603e-05, "loss": 0.0201, "step": 7377 }, { "epoch": 5.184820801124385, "grad_norm": 0.27202266454696655, "learning_rate": 2.987866010775357e-05, "loss": 0.0382, "step": 7378 }, { "epoch": 5.185523541813071, "grad_norm": 0.1960482895374298, "learning_rate": 2.9878191613961115e-05, "loss": 0.0287, "step": 7379 }, { "epoch": 5.186226282501757, "grad_norm": 0.2642674744129181, "learning_rate": 2.987772312016866e-05, "loss": 0.0493, "step": 7380 }, { "epoch": 5.186929023190443, "grad_norm": 0.19964948296546936, "learning_rate": 2.9877254626376203e-05, "loss": 0.0516, "step": 7381 }, { "epoch": 5.1876317638791285, "grad_norm": 0.23431845009326935, "learning_rate": 2.9876786132583743e-05, "loss": 0.0235, "step": 7382 }, { "epoch": 5.188334504567814, "grad_norm": 0.2617785930633545, "learning_rate": 2.9876317638791287e-05, "loss": 0.0309, "step": 7383 }, { "epoch": 5.1890372452565, "grad_norm": 0.27577510476112366, "learning_rate": 2.987584914499883e-05, "loss": 0.0481, "step": 7384 }, { "epoch": 5.189739985945186, "grad_norm": 0.6450905203819275, "learning_rate": 2.9875380651206374e-05, "loss": 0.0788, "step": 7385 }, { "epoch": 5.190442726633872, "grad_norm": 0.36274200677871704, "learning_rate": 2.9874912157413914e-05, "loss": 0.0955, "step": 7386 }, { "epoch": 5.191145467322558, "grad_norm": 0.7850545048713684, "learning_rate": 2.9874443663621458e-05, "loss": 0.1128, "step": 7387 }, { "epoch": 5.191848208011244, "grad_norm": 0.7765017747879028, "learning_rate": 2.9873975169829002e-05, "loss": 0.1903, "step": 7388 }, { "epoch": 5.19255094869993, "grad_norm": 1.912069320678711, "learning_rate": 2.9873506676036542e-05, "loss": 0.2332, "step": 7389 }, { "epoch": 5.193253689388616, "grad_norm": 1.3673151731491089, "learning_rate": 2.9873038182244086e-05, "loss": 0.2792, "step": 7390 }, { "epoch": 5.193956430077302, "grad_norm": 0.21248027682304382, "learning_rate": 2.9872569688451626e-05, "loss": 0.0819, "step": 7391 }, { "epoch": 5.194659170765988, "grad_norm": 0.17456333339214325, "learning_rate": 2.987210119465917e-05, "loss": 0.0327, "step": 7392 }, { "epoch": 5.195361911454674, "grad_norm": 0.15857647359371185, "learning_rate": 2.9871632700866714e-05, "loss": 0.0278, "step": 7393 }, { "epoch": 5.1960646521433596, "grad_norm": 0.1730838268995285, "learning_rate": 2.9871164207074258e-05, "loss": 0.0334, "step": 7394 }, { "epoch": 5.196767392832045, "grad_norm": 0.15534718334674835, "learning_rate": 2.9870695713281798e-05, "loss": 0.0296, "step": 7395 }, { "epoch": 5.1974701335207305, "grad_norm": 0.5441619157791138, "learning_rate": 2.987022721948934e-05, "loss": 0.0151, "step": 7396 }, { "epoch": 5.198172874209416, "grad_norm": 0.10341664403676987, "learning_rate": 2.9869758725696885e-05, "loss": 0.0133, "step": 7397 }, { "epoch": 5.198875614898102, "grad_norm": 0.24136371910572052, "learning_rate": 2.986929023190443e-05, "loss": 0.0342, "step": 7398 }, { "epoch": 5.199578355586788, "grad_norm": 0.15627521276474, "learning_rate": 2.986882173811197e-05, "loss": 0.0164, "step": 7399 }, { "epoch": 5.200281096275474, "grad_norm": 0.1850413978099823, "learning_rate": 2.9868353244319513e-05, "loss": 0.0152, "step": 7400 }, { "epoch": 5.20098383696416, "grad_norm": 0.22993138432502747, "learning_rate": 2.9867884750527057e-05, "loss": 0.03, "step": 7401 }, { "epoch": 5.201686577652846, "grad_norm": 0.23272369801998138, "learning_rate": 2.98674162567346e-05, "loss": 0.0199, "step": 7402 }, { "epoch": 5.202389318341532, "grad_norm": 0.19568713009357452, "learning_rate": 2.9866947762942144e-05, "loss": 0.0296, "step": 7403 }, { "epoch": 5.203092059030218, "grad_norm": 0.30285170674324036, "learning_rate": 2.9866479269149685e-05, "loss": 0.0238, "step": 7404 }, { "epoch": 5.203794799718904, "grad_norm": 0.33412355184555054, "learning_rate": 2.986601077535723e-05, "loss": 0.0344, "step": 7405 }, { "epoch": 5.20449754040759, "grad_norm": 0.2368408441543579, "learning_rate": 2.9865542281564772e-05, "loss": 0.0213, "step": 7406 }, { "epoch": 5.205200281096276, "grad_norm": 0.27374085783958435, "learning_rate": 2.9865073787772312e-05, "loss": 0.0381, "step": 7407 }, { "epoch": 5.205903021784962, "grad_norm": 0.3312644064426422, "learning_rate": 2.9864605293979853e-05, "loss": 0.0434, "step": 7408 }, { "epoch": 5.2066057624736475, "grad_norm": 0.49683189392089844, "learning_rate": 2.9864136800187396e-05, "loss": 0.0592, "step": 7409 }, { "epoch": 5.207308503162333, "grad_norm": 0.5407440662384033, "learning_rate": 2.986366830639494e-05, "loss": 0.0986, "step": 7410 }, { "epoch": 5.208011243851019, "grad_norm": 0.4326605796813965, "learning_rate": 2.9863199812602484e-05, "loss": 0.0889, "step": 7411 }, { "epoch": 5.208713984539705, "grad_norm": 0.780116617679596, "learning_rate": 2.9862731318810024e-05, "loss": 0.1245, "step": 7412 }, { "epoch": 5.20941672522839, "grad_norm": 1.3815876245498657, "learning_rate": 2.9862262825017568e-05, "loss": 0.1788, "step": 7413 }, { "epoch": 5.210119465917076, "grad_norm": 0.9622195363044739, "learning_rate": 2.9861794331225112e-05, "loss": 0.2293, "step": 7414 }, { "epoch": 5.210822206605762, "grad_norm": 1.493632435798645, "learning_rate": 2.9861325837432655e-05, "loss": 0.2901, "step": 7415 }, { "epoch": 5.211524947294448, "grad_norm": 0.8396345973014832, "learning_rate": 2.98608573436402e-05, "loss": 0.0674, "step": 7416 }, { "epoch": 5.212227687983134, "grad_norm": 0.2908991873264313, "learning_rate": 2.986038884984774e-05, "loss": 0.0233, "step": 7417 }, { "epoch": 5.21293042867182, "grad_norm": 0.32140490412712097, "learning_rate": 2.9859920356055283e-05, "loss": 0.0327, "step": 7418 }, { "epoch": 5.213633169360506, "grad_norm": 0.26488062739372253, "learning_rate": 2.9859451862262827e-05, "loss": 0.0386, "step": 7419 }, { "epoch": 5.214335910049192, "grad_norm": 0.21077467501163483, "learning_rate": 2.985898336847037e-05, "loss": 0.0239, "step": 7420 }, { "epoch": 5.215038650737878, "grad_norm": 0.1883024126291275, "learning_rate": 2.985851487467791e-05, "loss": 0.0312, "step": 7421 }, { "epoch": 5.215741391426564, "grad_norm": 0.14292864501476288, "learning_rate": 2.9858046380885455e-05, "loss": 0.0171, "step": 7422 }, { "epoch": 5.2164441321152495, "grad_norm": 0.2523246109485626, "learning_rate": 2.9857577887093e-05, "loss": 0.0284, "step": 7423 }, { "epoch": 5.217146872803935, "grad_norm": 0.27413100004196167, "learning_rate": 2.985710939330054e-05, "loss": 0.017, "step": 7424 }, { "epoch": 5.217849613492621, "grad_norm": 0.09919514507055283, "learning_rate": 2.9856640899508083e-05, "loss": 0.013, "step": 7425 }, { "epoch": 5.218552354181307, "grad_norm": 0.4908366799354553, "learning_rate": 2.9856172405715623e-05, "loss": 0.0236, "step": 7426 }, { "epoch": 5.219255094869993, "grad_norm": 0.1840611845254898, "learning_rate": 2.9855703911923167e-05, "loss": 0.0316, "step": 7427 }, { "epoch": 5.219957835558679, "grad_norm": 0.3883676528930664, "learning_rate": 2.985523541813071e-05, "loss": 0.0634, "step": 7428 }, { "epoch": 5.220660576247365, "grad_norm": 0.3670042157173157, "learning_rate": 2.9854766924338254e-05, "loss": 0.0198, "step": 7429 }, { "epoch": 5.221363316936051, "grad_norm": 0.2555113732814789, "learning_rate": 2.9854298430545794e-05, "loss": 0.0374, "step": 7430 }, { "epoch": 5.222066057624737, "grad_norm": 0.7090804576873779, "learning_rate": 2.9853829936753338e-05, "loss": 0.0647, "step": 7431 }, { "epoch": 5.222768798313423, "grad_norm": 0.3040018081665039, "learning_rate": 2.9853361442960882e-05, "loss": 0.0277, "step": 7432 }, { "epoch": 5.223471539002108, "grad_norm": 0.5332582592964172, "learning_rate": 2.9852892949168426e-05, "loss": 0.0511, "step": 7433 }, { "epoch": 5.224174279690794, "grad_norm": 0.38035985827445984, "learning_rate": 2.9852424455375966e-05, "loss": 0.0741, "step": 7434 }, { "epoch": 5.22487702037948, "grad_norm": 0.29842323064804077, "learning_rate": 2.985195596158351e-05, "loss": 0.0517, "step": 7435 }, { "epoch": 5.225579761068166, "grad_norm": 0.6554241180419922, "learning_rate": 2.9851487467791053e-05, "loss": 0.0998, "step": 7436 }, { "epoch": 5.2262825017568515, "grad_norm": 0.5371680855751038, "learning_rate": 2.9851018973998597e-05, "loss": 0.13, "step": 7437 }, { "epoch": 5.226985242445537, "grad_norm": 0.7424364686012268, "learning_rate": 2.985055048020614e-05, "loss": 0.2033, "step": 7438 }, { "epoch": 5.227687983134223, "grad_norm": 1.4166607856750488, "learning_rate": 2.985008198641368e-05, "loss": 0.2407, "step": 7439 }, { "epoch": 5.228390723822909, "grad_norm": 1.5839236974716187, "learning_rate": 2.9849613492621225e-05, "loss": 0.3057, "step": 7440 }, { "epoch": 5.229093464511595, "grad_norm": 0.3198295831680298, "learning_rate": 2.984914499882877e-05, "loss": 0.0861, "step": 7441 }, { "epoch": 5.229796205200281, "grad_norm": 0.21139654517173767, "learning_rate": 2.984867650503631e-05, "loss": 0.0415, "step": 7442 }, { "epoch": 5.230498945888967, "grad_norm": 0.16481660306453705, "learning_rate": 2.984820801124385e-05, "loss": 0.022, "step": 7443 }, { "epoch": 5.231201686577653, "grad_norm": 0.1974223405122757, "learning_rate": 2.9847739517451393e-05, "loss": 0.0282, "step": 7444 }, { "epoch": 5.231904427266339, "grad_norm": 0.2524951696395874, "learning_rate": 2.9847271023658937e-05, "loss": 0.0446, "step": 7445 }, { "epoch": 5.232607167955025, "grad_norm": 0.30802416801452637, "learning_rate": 2.984680252986648e-05, "loss": 0.0148, "step": 7446 }, { "epoch": 5.233309908643711, "grad_norm": 0.2349034696817398, "learning_rate": 2.984633403607402e-05, "loss": 0.0226, "step": 7447 }, { "epoch": 5.234012649332397, "grad_norm": 0.1886502504348755, "learning_rate": 2.9845865542281564e-05, "loss": 0.03, "step": 7448 }, { "epoch": 5.2347153900210825, "grad_norm": 0.16745615005493164, "learning_rate": 2.9845397048489108e-05, "loss": 0.0414, "step": 7449 }, { "epoch": 5.2354181307097685, "grad_norm": 0.16431504487991333, "learning_rate": 2.9844928554696652e-05, "loss": 0.0213, "step": 7450 }, { "epoch": 5.236120871398454, "grad_norm": 0.18204431235790253, "learning_rate": 2.9844460060904196e-05, "loss": 0.0254, "step": 7451 }, { "epoch": 5.236823612087139, "grad_norm": 0.21029552817344666, "learning_rate": 2.9843991567111736e-05, "loss": 0.0203, "step": 7452 }, { "epoch": 5.237526352775825, "grad_norm": 0.6584470272064209, "learning_rate": 2.984352307331928e-05, "loss": 0.0324, "step": 7453 }, { "epoch": 5.238229093464511, "grad_norm": 0.22050204873085022, "learning_rate": 2.9843054579526823e-05, "loss": 0.016, "step": 7454 }, { "epoch": 5.238931834153197, "grad_norm": 0.3206985592842102, "learning_rate": 2.9842586085734367e-05, "loss": 0.0424, "step": 7455 }, { "epoch": 5.239634574841883, "grad_norm": 0.2937818765640259, "learning_rate": 2.9842117591941908e-05, "loss": 0.0534, "step": 7456 }, { "epoch": 5.240337315530569, "grad_norm": 0.7221601009368896, "learning_rate": 2.984164909814945e-05, "loss": 0.0259, "step": 7457 }, { "epoch": 5.241040056219255, "grad_norm": 0.21450865268707275, "learning_rate": 2.9841180604356995e-05, "loss": 0.0471, "step": 7458 }, { "epoch": 5.241742796907941, "grad_norm": 0.5027645826339722, "learning_rate": 2.9840712110564535e-05, "loss": 0.0519, "step": 7459 }, { "epoch": 5.242445537596627, "grad_norm": 0.36311668157577515, "learning_rate": 2.9840243616772076e-05, "loss": 0.0802, "step": 7460 }, { "epoch": 5.243148278285313, "grad_norm": 0.65028315782547, "learning_rate": 2.983977512297962e-05, "loss": 0.0742, "step": 7461 }, { "epoch": 5.243851018973999, "grad_norm": 1.0343900918960571, "learning_rate": 2.9839306629187163e-05, "loss": 0.1089, "step": 7462 }, { "epoch": 5.2445537596626846, "grad_norm": 0.8248065710067749, "learning_rate": 2.9838838135394707e-05, "loss": 0.2077, "step": 7463 }, { "epoch": 5.2452565003513705, "grad_norm": 1.3172008991241455, "learning_rate": 2.983836964160225e-05, "loss": 0.2117, "step": 7464 }, { "epoch": 5.245959241040056, "grad_norm": 3.083893299102783, "learning_rate": 2.983790114780979e-05, "loss": 0.311, "step": 7465 }, { "epoch": 5.246661981728742, "grad_norm": 0.4097607135772705, "learning_rate": 2.9837432654017335e-05, "loss": 0.077, "step": 7466 }, { "epoch": 5.247364722417428, "grad_norm": 0.26144370436668396, "learning_rate": 2.983696416022488e-05, "loss": 0.026, "step": 7467 }, { "epoch": 5.248067463106114, "grad_norm": 0.15250864624977112, "learning_rate": 2.9836495666432422e-05, "loss": 0.0266, "step": 7468 }, { "epoch": 5.2487702037948, "grad_norm": 0.17078901827335358, "learning_rate": 2.9836027172639962e-05, "loss": 0.0286, "step": 7469 }, { "epoch": 5.249472944483486, "grad_norm": 0.29059073328971863, "learning_rate": 2.9835558678847506e-05, "loss": 0.0289, "step": 7470 }, { "epoch": 5.250175685172172, "grad_norm": 0.17056411504745483, "learning_rate": 2.983509018505505e-05, "loss": 0.0226, "step": 7471 }, { "epoch": 5.250878425860857, "grad_norm": 0.13830021023750305, "learning_rate": 2.9834621691262594e-05, "loss": 0.0133, "step": 7472 }, { "epoch": 5.251581166549543, "grad_norm": 0.16548006236553192, "learning_rate": 2.9834153197470134e-05, "loss": 0.0352, "step": 7473 }, { "epoch": 5.252283907238229, "grad_norm": 0.19057969748973846, "learning_rate": 2.9833684703677678e-05, "loss": 0.0234, "step": 7474 }, { "epoch": 5.252986647926915, "grad_norm": 0.2833617031574249, "learning_rate": 2.983321620988522e-05, "loss": 0.0205, "step": 7475 }, { "epoch": 5.253689388615601, "grad_norm": 0.29001384973526, "learning_rate": 2.9832747716092765e-05, "loss": 0.0305, "step": 7476 }, { "epoch": 5.254392129304287, "grad_norm": 0.15155018866062164, "learning_rate": 2.9832279222300305e-05, "loss": 0.0158, "step": 7477 }, { "epoch": 5.2550948699929725, "grad_norm": 0.3729470372200012, "learning_rate": 2.9831810728507846e-05, "loss": 0.0405, "step": 7478 }, { "epoch": 5.255797610681658, "grad_norm": 0.20860718190670013, "learning_rate": 2.983134223471539e-05, "loss": 0.0243, "step": 7479 }, { "epoch": 5.256500351370344, "grad_norm": 0.33953988552093506, "learning_rate": 2.9830873740922933e-05, "loss": 0.0455, "step": 7480 }, { "epoch": 5.25720309205903, "grad_norm": 0.2850889563560486, "learning_rate": 2.9830405247130477e-05, "loss": 0.0476, "step": 7481 }, { "epoch": 5.257905832747716, "grad_norm": 0.23904980719089508, "learning_rate": 2.9829936753338017e-05, "loss": 0.0261, "step": 7482 }, { "epoch": 5.258608573436402, "grad_norm": 0.2081947773694992, "learning_rate": 2.982946825954556e-05, "loss": 0.039, "step": 7483 }, { "epoch": 5.259311314125088, "grad_norm": 0.5495910048484802, "learning_rate": 2.9828999765753105e-05, "loss": 0.064, "step": 7484 }, { "epoch": 5.260014054813774, "grad_norm": 0.2222878634929657, "learning_rate": 2.982853127196065e-05, "loss": 0.039, "step": 7485 }, { "epoch": 5.26071679550246, "grad_norm": 0.5822715163230896, "learning_rate": 2.982806277816819e-05, "loss": 0.0936, "step": 7486 }, { "epoch": 5.261419536191146, "grad_norm": 0.5206252336502075, "learning_rate": 2.9827594284375732e-05, "loss": 0.1449, "step": 7487 }, { "epoch": 5.262122276879832, "grad_norm": 0.5408294796943665, "learning_rate": 2.9827125790583276e-05, "loss": 0.1616, "step": 7488 }, { "epoch": 5.262825017568518, "grad_norm": 0.811492383480072, "learning_rate": 2.982665729679082e-05, "loss": 0.2776, "step": 7489 }, { "epoch": 5.263527758257203, "grad_norm": 1.39350163936615, "learning_rate": 2.9826188802998364e-05, "loss": 0.3114, "step": 7490 }, { "epoch": 5.264230498945889, "grad_norm": 0.2623130977153778, "learning_rate": 2.9825720309205904e-05, "loss": 0.0805, "step": 7491 }, { "epoch": 5.2649332396345745, "grad_norm": 0.31953373551368713, "learning_rate": 2.9825251815413448e-05, "loss": 0.0393, "step": 7492 }, { "epoch": 5.26563598032326, "grad_norm": 0.1205609068274498, "learning_rate": 2.982478332162099e-05, "loss": 0.0231, "step": 7493 }, { "epoch": 5.266338721011946, "grad_norm": 0.18068453669548035, "learning_rate": 2.9824314827828532e-05, "loss": 0.0217, "step": 7494 }, { "epoch": 5.267041461700632, "grad_norm": 0.15174274146556854, "learning_rate": 2.9823846334036072e-05, "loss": 0.0261, "step": 7495 }, { "epoch": 5.267744202389318, "grad_norm": 0.13270209729671478, "learning_rate": 2.9823377840243616e-05, "loss": 0.024, "step": 7496 }, { "epoch": 5.268446943078004, "grad_norm": 0.1895974576473236, "learning_rate": 2.982290934645116e-05, "loss": 0.033, "step": 7497 }, { "epoch": 5.26914968376669, "grad_norm": 0.2009132355451584, "learning_rate": 2.9822440852658703e-05, "loss": 0.0385, "step": 7498 }, { "epoch": 5.269852424455376, "grad_norm": 0.1168844997882843, "learning_rate": 2.9821972358866244e-05, "loss": 0.0199, "step": 7499 }, { "epoch": 5.270555165144062, "grad_norm": 0.1443035751581192, "learning_rate": 2.9821503865073787e-05, "loss": 0.0198, "step": 7500 }, { "epoch": 5.271257905832748, "grad_norm": 0.22485437989234924, "learning_rate": 2.982103537128133e-05, "loss": 0.0281, "step": 7501 }, { "epoch": 5.271960646521434, "grad_norm": 0.18827736377716064, "learning_rate": 2.9820566877488875e-05, "loss": 0.0268, "step": 7502 }, { "epoch": 5.27266338721012, "grad_norm": 0.42196860909461975, "learning_rate": 2.982009838369642e-05, "loss": 0.0338, "step": 7503 }, { "epoch": 5.2733661278988055, "grad_norm": 0.20253562927246094, "learning_rate": 2.981962988990396e-05, "loss": 0.0191, "step": 7504 }, { "epoch": 5.2740688685874915, "grad_norm": 0.29579731822013855, "learning_rate": 2.9819161396111503e-05, "loss": 0.0463, "step": 7505 }, { "epoch": 5.274771609276177, "grad_norm": 0.5156646370887756, "learning_rate": 2.9818692902319046e-05, "loss": 0.0383, "step": 7506 }, { "epoch": 5.275474349964863, "grad_norm": 0.24251390993595123, "learning_rate": 2.981822440852659e-05, "loss": 0.036, "step": 7507 }, { "epoch": 5.276177090653549, "grad_norm": 0.37189796566963196, "learning_rate": 2.981775591473413e-05, "loss": 0.0487, "step": 7508 }, { "epoch": 5.276879831342235, "grad_norm": 0.4946422278881073, "learning_rate": 2.9817287420941674e-05, "loss": 0.0406, "step": 7509 }, { "epoch": 5.27758257203092, "grad_norm": 0.3179152011871338, "learning_rate": 2.9816818927149218e-05, "loss": 0.0549, "step": 7510 }, { "epoch": 5.278285312719606, "grad_norm": 0.3324033319950104, "learning_rate": 2.9816350433356758e-05, "loss": 0.0773, "step": 7511 }, { "epoch": 5.278988053408292, "grad_norm": 0.4633297920227051, "learning_rate": 2.98158819395643e-05, "loss": 0.111, "step": 7512 }, { "epoch": 5.279690794096978, "grad_norm": 0.679871141910553, "learning_rate": 2.9815413445771842e-05, "loss": 0.1527, "step": 7513 }, { "epoch": 5.280393534785664, "grad_norm": 1.2895021438598633, "learning_rate": 2.9814944951979386e-05, "loss": 0.2559, "step": 7514 }, { "epoch": 5.28109627547435, "grad_norm": 1.7777997255325317, "learning_rate": 2.981447645818693e-05, "loss": 0.2914, "step": 7515 }, { "epoch": 5.281799016163036, "grad_norm": 0.6290530562400818, "learning_rate": 2.9814007964394473e-05, "loss": 0.0903, "step": 7516 }, { "epoch": 5.282501756851722, "grad_norm": 0.19401656091213226, "learning_rate": 2.9813539470602014e-05, "loss": 0.0282, "step": 7517 }, { "epoch": 5.2832044975404076, "grad_norm": 0.1997075080871582, "learning_rate": 2.9813070976809557e-05, "loss": 0.0299, "step": 7518 }, { "epoch": 5.2839072382290935, "grad_norm": 0.1338367462158203, "learning_rate": 2.98126024830171e-05, "loss": 0.0214, "step": 7519 }, { "epoch": 5.284609978917779, "grad_norm": 0.15314781665802002, "learning_rate": 2.9812133989224645e-05, "loss": 0.0323, "step": 7520 }, { "epoch": 5.285312719606465, "grad_norm": 0.14865009486675262, "learning_rate": 2.9811665495432185e-05, "loss": 0.0152, "step": 7521 }, { "epoch": 5.286015460295151, "grad_norm": 0.10401035100221634, "learning_rate": 2.981119700163973e-05, "loss": 0.0119, "step": 7522 }, { "epoch": 5.286718200983837, "grad_norm": 0.20845696330070496, "learning_rate": 2.9810728507847273e-05, "loss": 0.0267, "step": 7523 }, { "epoch": 5.287420941672523, "grad_norm": 0.30612415075302124, "learning_rate": 2.9810260014054816e-05, "loss": 0.0272, "step": 7524 }, { "epoch": 5.288123682361209, "grad_norm": 0.18337199091911316, "learning_rate": 2.9809791520262357e-05, "loss": 0.0159, "step": 7525 }, { "epoch": 5.288826423049895, "grad_norm": 0.3305620551109314, "learning_rate": 2.98093230264699e-05, "loss": 0.0346, "step": 7526 }, { "epoch": 5.289529163738581, "grad_norm": 0.22610752284526825, "learning_rate": 2.9808854532677444e-05, "loss": 0.0233, "step": 7527 }, { "epoch": 5.290231904427266, "grad_norm": 0.1998172402381897, "learning_rate": 2.9808386038884988e-05, "loss": 0.0226, "step": 7528 }, { "epoch": 5.290934645115952, "grad_norm": 0.21324846148490906, "learning_rate": 2.9807917545092528e-05, "loss": 0.0394, "step": 7529 }, { "epoch": 5.291637385804638, "grad_norm": 0.3277212679386139, "learning_rate": 2.980744905130007e-05, "loss": 0.0363, "step": 7530 }, { "epoch": 5.292340126493324, "grad_norm": 0.22639964520931244, "learning_rate": 2.9806980557507612e-05, "loss": 0.0456, "step": 7531 }, { "epoch": 5.29304286718201, "grad_norm": 0.2545869052410126, "learning_rate": 2.9806512063715156e-05, "loss": 0.0441, "step": 7532 }, { "epoch": 5.2937456078706955, "grad_norm": 0.15932367742061615, "learning_rate": 2.98060435699227e-05, "loss": 0.0271, "step": 7533 }, { "epoch": 5.294448348559381, "grad_norm": 0.39255744218826294, "learning_rate": 2.980557507613024e-05, "loss": 0.0613, "step": 7534 }, { "epoch": 5.295151089248067, "grad_norm": 0.3236428499221802, "learning_rate": 2.9805106582337784e-05, "loss": 0.0602, "step": 7535 }, { "epoch": 5.295853829936753, "grad_norm": 1.5499695539474487, "learning_rate": 2.9804638088545328e-05, "loss": 0.0982, "step": 7536 }, { "epoch": 5.296556570625439, "grad_norm": 0.8935629725456238, "learning_rate": 2.980416959475287e-05, "loss": 0.1522, "step": 7537 }, { "epoch": 5.297259311314125, "grad_norm": 0.570812463760376, "learning_rate": 2.980370110096041e-05, "loss": 0.1681, "step": 7538 }, { "epoch": 5.297962052002811, "grad_norm": 0.694123387336731, "learning_rate": 2.9803232607167955e-05, "loss": 0.2345, "step": 7539 }, { "epoch": 5.298664792691497, "grad_norm": 1.3863072395324707, "learning_rate": 2.98027641133755e-05, "loss": 0.2713, "step": 7540 }, { "epoch": 5.299367533380183, "grad_norm": 0.48900678753852844, "learning_rate": 2.9802295619583043e-05, "loss": 0.0996, "step": 7541 }, { "epoch": 5.300070274068869, "grad_norm": 0.1753379851579666, "learning_rate": 2.9801827125790587e-05, "loss": 0.0325, "step": 7542 }, { "epoch": 5.300773014757555, "grad_norm": 0.18960033357143402, "learning_rate": 2.9801358631998127e-05, "loss": 0.0361, "step": 7543 }, { "epoch": 5.301475755446241, "grad_norm": 0.28149673342704773, "learning_rate": 2.980089013820567e-05, "loss": 0.0238, "step": 7544 }, { "epoch": 5.3021784961349265, "grad_norm": 0.12142794579267502, "learning_rate": 2.9800421644413214e-05, "loss": 0.0146, "step": 7545 }, { "epoch": 5.3028812368236125, "grad_norm": 0.18524357676506042, "learning_rate": 2.9799953150620755e-05, "loss": 0.0178, "step": 7546 }, { "epoch": 5.303583977512298, "grad_norm": 0.4276416003704071, "learning_rate": 2.9799484656828295e-05, "loss": 0.0267, "step": 7547 }, { "epoch": 5.304286718200984, "grad_norm": 0.1989080011844635, "learning_rate": 2.979901616303584e-05, "loss": 0.0263, "step": 7548 }, { "epoch": 5.304989458889669, "grad_norm": 0.18972717225551605, "learning_rate": 2.9798547669243382e-05, "loss": 0.0217, "step": 7549 }, { "epoch": 5.305692199578355, "grad_norm": 0.262751966714859, "learning_rate": 2.9798079175450926e-05, "loss": 0.0168, "step": 7550 }, { "epoch": 5.306394940267041, "grad_norm": 0.26609402894973755, "learning_rate": 2.9797610681658467e-05, "loss": 0.0478, "step": 7551 }, { "epoch": 5.307097680955727, "grad_norm": 0.20307716727256775, "learning_rate": 2.979714218786601e-05, "loss": 0.0226, "step": 7552 }, { "epoch": 5.307800421644413, "grad_norm": 0.2926856279373169, "learning_rate": 2.9796673694073554e-05, "loss": 0.0462, "step": 7553 }, { "epoch": 5.308503162333099, "grad_norm": 0.18351182341575623, "learning_rate": 2.9796205200281098e-05, "loss": 0.0222, "step": 7554 }, { "epoch": 5.309205903021785, "grad_norm": 0.2031424194574356, "learning_rate": 2.979573670648864e-05, "loss": 0.0317, "step": 7555 }, { "epoch": 5.309908643710471, "grad_norm": 0.3008790612220764, "learning_rate": 2.9795268212696182e-05, "loss": 0.04, "step": 7556 }, { "epoch": 5.310611384399157, "grad_norm": 0.178042933344841, "learning_rate": 2.9794799718903725e-05, "loss": 0.0433, "step": 7557 }, { "epoch": 5.311314125087843, "grad_norm": 0.2570413053035736, "learning_rate": 2.979433122511127e-05, "loss": 0.0452, "step": 7558 }, { "epoch": 5.3120168657765285, "grad_norm": 0.8952637314796448, "learning_rate": 2.9793862731318813e-05, "loss": 0.0616, "step": 7559 }, { "epoch": 5.3127196064652145, "grad_norm": 0.5505931973457336, "learning_rate": 2.9793394237526353e-05, "loss": 0.0616, "step": 7560 }, { "epoch": 5.3134223471539, "grad_norm": 0.356049120426178, "learning_rate": 2.9792925743733897e-05, "loss": 0.0906, "step": 7561 }, { "epoch": 5.314125087842586, "grad_norm": 0.46670371294021606, "learning_rate": 2.979245724994144e-05, "loss": 0.166, "step": 7562 }, { "epoch": 5.314827828531272, "grad_norm": 0.6106122732162476, "learning_rate": 2.9791988756148984e-05, "loss": 0.2189, "step": 7563 }, { "epoch": 5.315530569219958, "grad_norm": 0.7991392612457275, "learning_rate": 2.979152026235652e-05, "loss": 0.1912, "step": 7564 }, { "epoch": 5.316233309908644, "grad_norm": 3.281324863433838, "learning_rate": 2.9791051768564065e-05, "loss": 0.3326, "step": 7565 }, { "epoch": 5.31693605059733, "grad_norm": 0.27078139781951904, "learning_rate": 2.979058327477161e-05, "loss": 0.0653, "step": 7566 }, { "epoch": 5.317638791286015, "grad_norm": 0.18690985441207886, "learning_rate": 2.9790114780979153e-05, "loss": 0.0274, "step": 7567 }, { "epoch": 5.318341531974701, "grad_norm": 0.24148081243038177, "learning_rate": 2.9789646287186696e-05, "loss": 0.0537, "step": 7568 }, { "epoch": 5.319044272663387, "grad_norm": 0.2032971978187561, "learning_rate": 2.9789177793394237e-05, "loss": 0.022, "step": 7569 }, { "epoch": 5.319747013352073, "grad_norm": 0.15415038168430328, "learning_rate": 2.978870929960178e-05, "loss": 0.0231, "step": 7570 }, { "epoch": 5.320449754040759, "grad_norm": 0.34992101788520813, "learning_rate": 2.9788240805809324e-05, "loss": 0.0152, "step": 7571 }, { "epoch": 5.321152494729445, "grad_norm": 0.18565300107002258, "learning_rate": 2.9787772312016868e-05, "loss": 0.0208, "step": 7572 }, { "epoch": 5.3218552354181305, "grad_norm": 0.2284514456987381, "learning_rate": 2.9787303818224408e-05, "loss": 0.0363, "step": 7573 }, { "epoch": 5.3225579761068165, "grad_norm": 0.15079663693904877, "learning_rate": 2.9786835324431952e-05, "loss": 0.0352, "step": 7574 }, { "epoch": 5.323260716795502, "grad_norm": 0.15776817500591278, "learning_rate": 2.9786366830639496e-05, "loss": 0.0127, "step": 7575 }, { "epoch": 5.323963457484188, "grad_norm": 0.27727195620536804, "learning_rate": 2.978589833684704e-05, "loss": 0.0269, "step": 7576 }, { "epoch": 5.324666198172874, "grad_norm": 0.16578932106494904, "learning_rate": 2.978542984305458e-05, "loss": 0.0214, "step": 7577 }, { "epoch": 5.32536893886156, "grad_norm": 0.23209069669246674, "learning_rate": 2.9784961349262123e-05, "loss": 0.0355, "step": 7578 }, { "epoch": 5.326071679550246, "grad_norm": 0.3242781162261963, "learning_rate": 2.9784492855469667e-05, "loss": 0.0219, "step": 7579 }, { "epoch": 5.326774420238932, "grad_norm": 0.48729467391967773, "learning_rate": 2.978402436167721e-05, "loss": 0.023, "step": 7580 }, { "epoch": 5.327477160927618, "grad_norm": 0.18482555449008942, "learning_rate": 2.978355586788475e-05, "loss": 0.0457, "step": 7581 }, { "epoch": 5.328179901616304, "grad_norm": 0.2772617042064667, "learning_rate": 2.978308737409229e-05, "loss": 0.0257, "step": 7582 }, { "epoch": 5.32888264230499, "grad_norm": 0.21778681874275208, "learning_rate": 2.9782618880299835e-05, "loss": 0.042, "step": 7583 }, { "epoch": 5.329585382993676, "grad_norm": 0.28527313470840454, "learning_rate": 2.978215038650738e-05, "loss": 0.0486, "step": 7584 }, { "epoch": 5.330288123682362, "grad_norm": 0.5741040706634521, "learning_rate": 2.9781681892714923e-05, "loss": 0.0863, "step": 7585 }, { "epoch": 5.3309908643710475, "grad_norm": 0.6160460114479065, "learning_rate": 2.9781213398922463e-05, "loss": 0.1148, "step": 7586 }, { "epoch": 5.3316936050597326, "grad_norm": 0.8254830837249756, "learning_rate": 2.9780744905130007e-05, "loss": 0.1144, "step": 7587 }, { "epoch": 5.3323963457484185, "grad_norm": 0.7625781297683716, "learning_rate": 2.978027641133755e-05, "loss": 0.172, "step": 7588 }, { "epoch": 5.333099086437104, "grad_norm": 0.7866466045379639, "learning_rate": 2.9779807917545094e-05, "loss": 0.2056, "step": 7589 }, { "epoch": 5.33380182712579, "grad_norm": 0.9342631101608276, "learning_rate": 2.9779339423752635e-05, "loss": 0.2504, "step": 7590 }, { "epoch": 5.334504567814476, "grad_norm": 0.2904582619667053, "learning_rate": 2.9778870929960178e-05, "loss": 0.0661, "step": 7591 }, { "epoch": 5.335207308503162, "grad_norm": 0.14356257021427155, "learning_rate": 2.9778402436167722e-05, "loss": 0.0264, "step": 7592 }, { "epoch": 5.335910049191848, "grad_norm": 0.1528477966785431, "learning_rate": 2.9777933942375266e-05, "loss": 0.023, "step": 7593 }, { "epoch": 5.336612789880534, "grad_norm": 0.201907679438591, "learning_rate": 2.977746544858281e-05, "loss": 0.0195, "step": 7594 }, { "epoch": 5.33731553056922, "grad_norm": 0.10798216611146927, "learning_rate": 2.977699695479035e-05, "loss": 0.0192, "step": 7595 }, { "epoch": 5.338018271257906, "grad_norm": 0.15975558757781982, "learning_rate": 2.9776528460997894e-05, "loss": 0.0177, "step": 7596 }, { "epoch": 5.338721011946592, "grad_norm": 0.23903599381446838, "learning_rate": 2.9776059967205437e-05, "loss": 0.0169, "step": 7597 }, { "epoch": 5.339423752635278, "grad_norm": 0.21848657727241516, "learning_rate": 2.9775591473412978e-05, "loss": 0.0271, "step": 7598 }, { "epoch": 5.340126493323964, "grad_norm": 0.18074922263622284, "learning_rate": 2.9775122979620518e-05, "loss": 0.023, "step": 7599 }, { "epoch": 5.3408292340126495, "grad_norm": 0.22732087969779968, "learning_rate": 2.977465448582806e-05, "loss": 0.0272, "step": 7600 }, { "epoch": 5.3415319747013355, "grad_norm": 0.39863064885139465, "learning_rate": 2.9774185992035605e-05, "loss": 0.0408, "step": 7601 }, { "epoch": 5.342234715390021, "grad_norm": 0.12500986456871033, "learning_rate": 2.977371749824315e-05, "loss": 0.0135, "step": 7602 }, { "epoch": 5.342937456078707, "grad_norm": 0.18189845979213715, "learning_rate": 2.977324900445069e-05, "loss": 0.0263, "step": 7603 }, { "epoch": 5.343640196767393, "grad_norm": 0.2812473177909851, "learning_rate": 2.9772780510658233e-05, "loss": 0.0242, "step": 7604 }, { "epoch": 5.344342937456078, "grad_norm": 0.33083978295326233, "learning_rate": 2.9772312016865777e-05, "loss": 0.0578, "step": 7605 }, { "epoch": 5.345045678144764, "grad_norm": 0.17560453712940216, "learning_rate": 2.977184352307332e-05, "loss": 0.0322, "step": 7606 }, { "epoch": 5.34574841883345, "grad_norm": 0.2269139289855957, "learning_rate": 2.9771375029280864e-05, "loss": 0.0363, "step": 7607 }, { "epoch": 5.346451159522136, "grad_norm": 0.7913895845413208, "learning_rate": 2.9770906535488405e-05, "loss": 0.0515, "step": 7608 }, { "epoch": 5.347153900210822, "grad_norm": 0.25206461548805237, "learning_rate": 2.977043804169595e-05, "loss": 0.045, "step": 7609 }, { "epoch": 5.347856640899508, "grad_norm": 0.6111821532249451, "learning_rate": 2.9769969547903492e-05, "loss": 0.0517, "step": 7610 }, { "epoch": 5.348559381588194, "grad_norm": 0.3678880035877228, "learning_rate": 2.9769501054111036e-05, "loss": 0.1161, "step": 7611 }, { "epoch": 5.34926212227688, "grad_norm": 0.4021284878253937, "learning_rate": 2.9769032560318576e-05, "loss": 0.1184, "step": 7612 }, { "epoch": 5.349964862965566, "grad_norm": 0.6776660680770874, "learning_rate": 2.976856406652612e-05, "loss": 0.1862, "step": 7613 }, { "epoch": 5.3506676036542515, "grad_norm": 1.248519778251648, "learning_rate": 2.9768095572733664e-05, "loss": 0.2253, "step": 7614 }, { "epoch": 5.3513703443429375, "grad_norm": 0.8814519643783569, "learning_rate": 2.9767627078941207e-05, "loss": 0.2431, "step": 7615 }, { "epoch": 5.352073085031623, "grad_norm": 0.19891846179962158, "learning_rate": 2.9767158585148748e-05, "loss": 0.0698, "step": 7616 }, { "epoch": 5.352775825720309, "grad_norm": 0.24616344273090363, "learning_rate": 2.9766690091356288e-05, "loss": 0.0366, "step": 7617 }, { "epoch": 5.353478566408995, "grad_norm": 0.2951551377773285, "learning_rate": 2.9766221597563832e-05, "loss": 0.0334, "step": 7618 }, { "epoch": 5.354181307097681, "grad_norm": 0.14652325212955475, "learning_rate": 2.9765753103771375e-05, "loss": 0.0223, "step": 7619 }, { "epoch": 5.354884047786367, "grad_norm": 0.17359335720539093, "learning_rate": 2.976528460997892e-05, "loss": 0.0277, "step": 7620 }, { "epoch": 5.355586788475053, "grad_norm": 0.20852093398571014, "learning_rate": 2.976481611618646e-05, "loss": 0.0191, "step": 7621 }, { "epoch": 5.356289529163739, "grad_norm": 0.22195923328399658, "learning_rate": 2.9764347622394003e-05, "loss": 0.0217, "step": 7622 }, { "epoch": 5.356992269852425, "grad_norm": 0.20321622490882874, "learning_rate": 2.9763879128601547e-05, "loss": 0.0253, "step": 7623 }, { "epoch": 5.357695010541111, "grad_norm": 0.303876131772995, "learning_rate": 2.976341063480909e-05, "loss": 0.0381, "step": 7624 }, { "epoch": 5.358397751229797, "grad_norm": 0.1320468783378601, "learning_rate": 2.976294214101663e-05, "loss": 0.0234, "step": 7625 }, { "epoch": 5.359100491918482, "grad_norm": 0.19911883771419525, "learning_rate": 2.9762473647224175e-05, "loss": 0.025, "step": 7626 }, { "epoch": 5.359803232607168, "grad_norm": 0.19107191264629364, "learning_rate": 2.976200515343172e-05, "loss": 0.0218, "step": 7627 }, { "epoch": 5.3605059732958535, "grad_norm": 0.6765775680541992, "learning_rate": 2.9761536659639262e-05, "loss": 0.0234, "step": 7628 }, { "epoch": 5.3612087139845395, "grad_norm": 0.23183566331863403, "learning_rate": 2.9761068165846803e-05, "loss": 0.0277, "step": 7629 }, { "epoch": 5.361911454673225, "grad_norm": 0.2067132592201233, "learning_rate": 2.9760599672054346e-05, "loss": 0.0382, "step": 7630 }, { "epoch": 5.362614195361911, "grad_norm": 0.24001330137252808, "learning_rate": 2.976013117826189e-05, "loss": 0.0593, "step": 7631 }, { "epoch": 5.363316936050597, "grad_norm": 0.24829579889774323, "learning_rate": 2.9759662684469434e-05, "loss": 0.0283, "step": 7632 }, { "epoch": 5.364019676739283, "grad_norm": 0.22817789018154144, "learning_rate": 2.9759194190676974e-05, "loss": 0.0438, "step": 7633 }, { "epoch": 5.364722417427969, "grad_norm": 0.32996997237205505, "learning_rate": 2.9758725696884514e-05, "loss": 0.0847, "step": 7634 }, { "epoch": 5.365425158116655, "grad_norm": 0.36312416195869446, "learning_rate": 2.9758257203092058e-05, "loss": 0.0528, "step": 7635 }, { "epoch": 5.366127898805341, "grad_norm": 0.32324448227882385, "learning_rate": 2.9757788709299602e-05, "loss": 0.0928, "step": 7636 }, { "epoch": 5.366830639494027, "grad_norm": 1.4107753038406372, "learning_rate": 2.9757320215507146e-05, "loss": 0.1235, "step": 7637 }, { "epoch": 5.367533380182713, "grad_norm": 0.5544338226318359, "learning_rate": 2.9756851721714686e-05, "loss": 0.1766, "step": 7638 }, { "epoch": 5.368236120871399, "grad_norm": 1.2678459882736206, "learning_rate": 2.975638322792223e-05, "loss": 0.2379, "step": 7639 }, { "epoch": 5.368938861560085, "grad_norm": 1.560505986213684, "learning_rate": 2.9755914734129773e-05, "loss": 0.283, "step": 7640 }, { "epoch": 5.3696416022487705, "grad_norm": 0.36570748686790466, "learning_rate": 2.9755446240337317e-05, "loss": 0.0551, "step": 7641 }, { "epoch": 5.370344342937456, "grad_norm": 0.17202630639076233, "learning_rate": 2.975497774654486e-05, "loss": 0.0394, "step": 7642 }, { "epoch": 5.371047083626142, "grad_norm": 0.21311336755752563, "learning_rate": 2.97545092527524e-05, "loss": 0.0277, "step": 7643 }, { "epoch": 5.371749824314827, "grad_norm": 0.11689529567956924, "learning_rate": 2.9754040758959945e-05, "loss": 0.0233, "step": 7644 }, { "epoch": 5.372452565003513, "grad_norm": 0.14304383099079132, "learning_rate": 2.975357226516749e-05, "loss": 0.0211, "step": 7645 }, { "epoch": 5.373155305692199, "grad_norm": 0.22355082631111145, "learning_rate": 2.9753103771375032e-05, "loss": 0.0206, "step": 7646 }, { "epoch": 5.373858046380885, "grad_norm": 0.13975757360458374, "learning_rate": 2.9752635277582573e-05, "loss": 0.0171, "step": 7647 }, { "epoch": 5.374560787069571, "grad_norm": 0.13452604413032532, "learning_rate": 2.9752166783790116e-05, "loss": 0.0147, "step": 7648 }, { "epoch": 5.375263527758257, "grad_norm": 0.2975134253501892, "learning_rate": 2.975169828999766e-05, "loss": 0.0453, "step": 7649 }, { "epoch": 5.375966268446943, "grad_norm": 0.1525174081325531, "learning_rate": 2.9751229796205204e-05, "loss": 0.0158, "step": 7650 }, { "epoch": 5.376669009135629, "grad_norm": 0.19895760715007782, "learning_rate": 2.975076130241274e-05, "loss": 0.0336, "step": 7651 }, { "epoch": 5.377371749824315, "grad_norm": 0.16904397308826447, "learning_rate": 2.9750292808620285e-05, "loss": 0.0178, "step": 7652 }, { "epoch": 5.378074490513001, "grad_norm": 0.23223263025283813, "learning_rate": 2.9749824314827828e-05, "loss": 0.0273, "step": 7653 }, { "epoch": 5.378777231201687, "grad_norm": 0.16357341408729553, "learning_rate": 2.9749355821035372e-05, "loss": 0.0183, "step": 7654 }, { "epoch": 5.3794799718903725, "grad_norm": 0.18014125525951385, "learning_rate": 2.9748887327242916e-05, "loss": 0.0299, "step": 7655 }, { "epoch": 5.3801827125790584, "grad_norm": 0.21008925139904022, "learning_rate": 2.9748418833450456e-05, "loss": 0.0314, "step": 7656 }, { "epoch": 5.380885453267744, "grad_norm": 0.3834001421928406, "learning_rate": 2.9747950339658e-05, "loss": 0.0148, "step": 7657 }, { "epoch": 5.38158819395643, "grad_norm": 0.29169753193855286, "learning_rate": 2.9747481845865543e-05, "loss": 0.0458, "step": 7658 }, { "epoch": 5.382290934645116, "grad_norm": 0.5673812627792358, "learning_rate": 2.9747013352073087e-05, "loss": 0.0424, "step": 7659 }, { "epoch": 5.382993675333802, "grad_norm": 0.46846288442611694, "learning_rate": 2.9746544858280628e-05, "loss": 0.0832, "step": 7660 }, { "epoch": 5.383696416022488, "grad_norm": 0.3724325895309448, "learning_rate": 2.974607636448817e-05, "loss": 0.0848, "step": 7661 }, { "epoch": 5.384399156711174, "grad_norm": 0.47316670417785645, "learning_rate": 2.9745607870695715e-05, "loss": 0.1319, "step": 7662 }, { "epoch": 5.38510189739986, "grad_norm": 0.9042780995368958, "learning_rate": 2.974513937690326e-05, "loss": 0.2127, "step": 7663 }, { "epoch": 5.385804638088545, "grad_norm": 0.8808352947235107, "learning_rate": 2.97446708831108e-05, "loss": 0.2212, "step": 7664 }, { "epoch": 5.386507378777231, "grad_norm": 1.9386335611343384, "learning_rate": 2.9744202389318343e-05, "loss": 0.2828, "step": 7665 }, { "epoch": 5.387210119465917, "grad_norm": 0.2213788628578186, "learning_rate": 2.9743733895525887e-05, "loss": 0.0611, "step": 7666 }, { "epoch": 5.387912860154603, "grad_norm": 0.12989884614944458, "learning_rate": 2.974326540173343e-05, "loss": 0.021, "step": 7667 }, { "epoch": 5.388615600843289, "grad_norm": 0.3535607159137726, "learning_rate": 2.974279690794097e-05, "loss": 0.0505, "step": 7668 }, { "epoch": 5.3893183415319745, "grad_norm": 0.32725998759269714, "learning_rate": 2.974232841414851e-05, "loss": 0.0234, "step": 7669 }, { "epoch": 5.3900210822206605, "grad_norm": 0.15564529597759247, "learning_rate": 2.9741859920356055e-05, "loss": 0.0207, "step": 7670 }, { "epoch": 5.390723822909346, "grad_norm": 0.12823538482189178, "learning_rate": 2.97413914265636e-05, "loss": 0.0127, "step": 7671 }, { "epoch": 5.391426563598032, "grad_norm": 0.25487810373306274, "learning_rate": 2.9740922932771142e-05, "loss": 0.0289, "step": 7672 }, { "epoch": 5.392129304286718, "grad_norm": 0.30611157417297363, "learning_rate": 2.9740454438978682e-05, "loss": 0.0321, "step": 7673 }, { "epoch": 5.392832044975404, "grad_norm": 0.18846377730369568, "learning_rate": 2.9739985945186226e-05, "loss": 0.0323, "step": 7674 }, { "epoch": 5.39353478566409, "grad_norm": 0.13847748935222626, "learning_rate": 2.973951745139377e-05, "loss": 0.013, "step": 7675 }, { "epoch": 5.394237526352776, "grad_norm": 0.12461090832948685, "learning_rate": 2.9739048957601314e-05, "loss": 0.0197, "step": 7676 }, { "epoch": 5.394940267041462, "grad_norm": 0.19845890998840332, "learning_rate": 2.9738580463808854e-05, "loss": 0.0303, "step": 7677 }, { "epoch": 5.395643007730148, "grad_norm": 0.38931015133857727, "learning_rate": 2.9738111970016398e-05, "loss": 0.0345, "step": 7678 }, { "epoch": 5.396345748418834, "grad_norm": 0.16086110472679138, "learning_rate": 2.973764347622394e-05, "loss": 0.0207, "step": 7679 }, { "epoch": 5.39704848910752, "grad_norm": 0.1677362322807312, "learning_rate": 2.9737174982431485e-05, "loss": 0.0267, "step": 7680 }, { "epoch": 5.397751229796206, "grad_norm": 0.3164353668689728, "learning_rate": 2.973670648863903e-05, "loss": 0.0393, "step": 7681 }, { "epoch": 5.398453970484891, "grad_norm": 0.2622019052505493, "learning_rate": 2.973623799484657e-05, "loss": 0.0346, "step": 7682 }, { "epoch": 5.3991567111735765, "grad_norm": 0.223236545920372, "learning_rate": 2.9735769501054113e-05, "loss": 0.0422, "step": 7683 }, { "epoch": 5.3998594518622625, "grad_norm": 0.2764679193496704, "learning_rate": 2.9735301007261657e-05, "loss": 0.0412, "step": 7684 }, { "epoch": 5.400562192550948, "grad_norm": 0.2934773564338684, "learning_rate": 2.9734832513469197e-05, "loss": 0.0447, "step": 7685 }, { "epoch": 5.401264933239634, "grad_norm": 0.5623652338981628, "learning_rate": 2.9734364019676737e-05, "loss": 0.0972, "step": 7686 }, { "epoch": 5.40196767392832, "grad_norm": 0.612143337726593, "learning_rate": 2.973389552588428e-05, "loss": 0.1186, "step": 7687 }, { "epoch": 5.402670414617006, "grad_norm": 0.9175444841384888, "learning_rate": 2.9733427032091825e-05, "loss": 0.1682, "step": 7688 }, { "epoch": 5.403373155305692, "grad_norm": 0.8939987421035767, "learning_rate": 2.973295853829937e-05, "loss": 0.1913, "step": 7689 }, { "epoch": 5.404075895994378, "grad_norm": 1.3128650188446045, "learning_rate": 2.973249004450691e-05, "loss": 0.2672, "step": 7690 }, { "epoch": 5.404778636683064, "grad_norm": 0.45563969016075134, "learning_rate": 2.9732021550714453e-05, "loss": 0.0868, "step": 7691 }, { "epoch": 5.40548137737175, "grad_norm": 0.2659050226211548, "learning_rate": 2.9731553056921996e-05, "loss": 0.0319, "step": 7692 }, { "epoch": 5.406184118060436, "grad_norm": 0.3247660994529724, "learning_rate": 2.973108456312954e-05, "loss": 0.0248, "step": 7693 }, { "epoch": 5.406886858749122, "grad_norm": 0.19235216081142426, "learning_rate": 2.9730616069337084e-05, "loss": 0.0266, "step": 7694 }, { "epoch": 5.407589599437808, "grad_norm": 0.11511465162038803, "learning_rate": 2.9730147575544624e-05, "loss": 0.0229, "step": 7695 }, { "epoch": 5.4082923401264935, "grad_norm": 0.6514063477516174, "learning_rate": 2.9729679081752168e-05, "loss": 0.0183, "step": 7696 }, { "epoch": 5.408995080815179, "grad_norm": 0.14734035730361938, "learning_rate": 2.972921058795971e-05, "loss": 0.0253, "step": 7697 }, { "epoch": 5.409697821503865, "grad_norm": 0.1651179939508438, "learning_rate": 2.9728742094167255e-05, "loss": 0.0154, "step": 7698 }, { "epoch": 5.410400562192551, "grad_norm": 0.14170295000076294, "learning_rate": 2.9728273600374796e-05, "loss": 0.0182, "step": 7699 }, { "epoch": 5.411103302881237, "grad_norm": 0.1488049477338791, "learning_rate": 2.972780510658234e-05, "loss": 0.0178, "step": 7700 }, { "epoch": 5.411806043569923, "grad_norm": 0.7307521104812622, "learning_rate": 2.9727336612789883e-05, "loss": 0.058, "step": 7701 }, { "epoch": 5.412508784258609, "grad_norm": 0.20193251967430115, "learning_rate": 2.9726868118997427e-05, "loss": 0.0209, "step": 7702 }, { "epoch": 5.413211524947294, "grad_norm": 0.23633012175559998, "learning_rate": 2.9726399625204964e-05, "loss": 0.0238, "step": 7703 }, { "epoch": 5.41391426563598, "grad_norm": 0.2149834930896759, "learning_rate": 2.9725931131412507e-05, "loss": 0.037, "step": 7704 }, { "epoch": 5.414617006324666, "grad_norm": 0.2905685007572174, "learning_rate": 2.972546263762005e-05, "loss": 0.0387, "step": 7705 }, { "epoch": 5.415319747013352, "grad_norm": 0.4669819176197052, "learning_rate": 2.9724994143827595e-05, "loss": 0.0462, "step": 7706 }, { "epoch": 5.416022487702038, "grad_norm": 0.2232101410627365, "learning_rate": 2.972452565003514e-05, "loss": 0.0359, "step": 7707 }, { "epoch": 5.416725228390724, "grad_norm": 0.3203928470611572, "learning_rate": 2.972405715624268e-05, "loss": 0.0395, "step": 7708 }, { "epoch": 5.41742796907941, "grad_norm": 0.3633910119533539, "learning_rate": 2.9723588662450223e-05, "loss": 0.0545, "step": 7709 }, { "epoch": 5.4181307097680955, "grad_norm": 0.43806788325309753, "learning_rate": 2.9723120168657766e-05, "loss": 0.06, "step": 7710 }, { "epoch": 5.418833450456781, "grad_norm": 0.40478718280792236, "learning_rate": 2.972265167486531e-05, "loss": 0.0867, "step": 7711 }, { "epoch": 5.419536191145467, "grad_norm": 0.8038181662559509, "learning_rate": 2.972218318107285e-05, "loss": 0.1538, "step": 7712 }, { "epoch": 5.420238931834153, "grad_norm": 0.6487658023834229, "learning_rate": 2.9721714687280394e-05, "loss": 0.1929, "step": 7713 }, { "epoch": 5.420941672522839, "grad_norm": 1.148665189743042, "learning_rate": 2.9721246193487938e-05, "loss": 0.2315, "step": 7714 }, { "epoch": 5.421644413211525, "grad_norm": 1.337494134902954, "learning_rate": 2.972077769969548e-05, "loss": 0.284, "step": 7715 }, { "epoch": 5.422347153900211, "grad_norm": 0.2332848459482193, "learning_rate": 2.9720309205903022e-05, "loss": 0.0865, "step": 7716 }, { "epoch": 5.423049894588897, "grad_norm": 0.15665791928768158, "learning_rate": 2.9719840712110566e-05, "loss": 0.0312, "step": 7717 }, { "epoch": 5.423752635277583, "grad_norm": 0.16431084275245667, "learning_rate": 2.971937221831811e-05, "loss": 0.0229, "step": 7718 }, { "epoch": 5.424455375966269, "grad_norm": 0.21676287055015564, "learning_rate": 2.9718903724525653e-05, "loss": 0.0293, "step": 7719 }, { "epoch": 5.425158116654955, "grad_norm": 0.20313894748687744, "learning_rate": 2.9718435230733193e-05, "loss": 0.0291, "step": 7720 }, { "epoch": 5.42586085734364, "grad_norm": 0.11157665401697159, "learning_rate": 2.9717966736940734e-05, "loss": 0.0121, "step": 7721 }, { "epoch": 5.426563598032326, "grad_norm": 0.23936210572719574, "learning_rate": 2.9717498243148278e-05, "loss": 0.0177, "step": 7722 }, { "epoch": 5.427266338721012, "grad_norm": 0.2593331038951874, "learning_rate": 2.971702974935582e-05, "loss": 0.0299, "step": 7723 }, { "epoch": 5.4279690794096975, "grad_norm": 0.6193409562110901, "learning_rate": 2.9716561255563365e-05, "loss": 0.0223, "step": 7724 }, { "epoch": 5.4286718200983834, "grad_norm": 0.18462646007537842, "learning_rate": 2.9716092761770905e-05, "loss": 0.0319, "step": 7725 }, { "epoch": 5.429374560787069, "grad_norm": 0.18978077173233032, "learning_rate": 2.971562426797845e-05, "loss": 0.0227, "step": 7726 }, { "epoch": 5.430077301475755, "grad_norm": 0.7372552752494812, "learning_rate": 2.9715155774185993e-05, "loss": 0.0535, "step": 7727 }, { "epoch": 5.430780042164441, "grad_norm": 0.24539029598236084, "learning_rate": 2.9714687280393536e-05, "loss": 0.0448, "step": 7728 }, { "epoch": 5.431482782853127, "grad_norm": 0.2100801020860672, "learning_rate": 2.9714218786601077e-05, "loss": 0.0261, "step": 7729 }, { "epoch": 5.432185523541813, "grad_norm": 0.16368679702281952, "learning_rate": 2.971375029280862e-05, "loss": 0.0291, "step": 7730 }, { "epoch": 5.432888264230499, "grad_norm": 0.15861822664737701, "learning_rate": 2.9713281799016164e-05, "loss": 0.031, "step": 7731 }, { "epoch": 5.433591004919185, "grad_norm": 0.2778562009334564, "learning_rate": 2.9712813305223708e-05, "loss": 0.0331, "step": 7732 }, { "epoch": 5.434293745607871, "grad_norm": 0.21100300550460815, "learning_rate": 2.9712344811431252e-05, "loss": 0.0418, "step": 7733 }, { "epoch": 5.434996486296557, "grad_norm": 0.22934012115001678, "learning_rate": 2.9711876317638792e-05, "loss": 0.0316, "step": 7734 }, { "epoch": 5.435699226985243, "grad_norm": 0.2748725116252899, "learning_rate": 2.9711407823846336e-05, "loss": 0.0807, "step": 7735 }, { "epoch": 5.436401967673929, "grad_norm": 0.38165992498397827, "learning_rate": 2.971093933005388e-05, "loss": 0.0962, "step": 7736 }, { "epoch": 5.4371047083626145, "grad_norm": 0.6606733798980713, "learning_rate": 2.9710470836261423e-05, "loss": 0.1544, "step": 7737 }, { "epoch": 5.4378074490513, "grad_norm": 1.103392481803894, "learning_rate": 2.971000234246896e-05, "loss": 0.1543, "step": 7738 }, { "epoch": 5.438510189739986, "grad_norm": 1.113646388053894, "learning_rate": 2.9709533848676504e-05, "loss": 0.1949, "step": 7739 }, { "epoch": 5.439212930428672, "grad_norm": 1.891357421875, "learning_rate": 2.9709065354884048e-05, "loss": 0.2635, "step": 7740 }, { "epoch": 5.439915671117357, "grad_norm": 0.25144705176353455, "learning_rate": 2.970859686109159e-05, "loss": 0.0869, "step": 7741 }, { "epoch": 5.440618411806043, "grad_norm": 0.22742661833763123, "learning_rate": 2.970812836729913e-05, "loss": 0.0561, "step": 7742 }, { "epoch": 5.441321152494729, "grad_norm": 0.15619641542434692, "learning_rate": 2.9707659873506675e-05, "loss": 0.0356, "step": 7743 }, { "epoch": 5.442023893183415, "grad_norm": 0.12329357862472534, "learning_rate": 2.970719137971422e-05, "loss": 0.0197, "step": 7744 }, { "epoch": 5.442726633872101, "grad_norm": 0.21647578477859497, "learning_rate": 2.9706722885921763e-05, "loss": 0.0253, "step": 7745 }, { "epoch": 5.443429374560787, "grad_norm": 0.21146488189697266, "learning_rate": 2.9706254392129307e-05, "loss": 0.0197, "step": 7746 }, { "epoch": 5.444132115249473, "grad_norm": 0.16269062459468842, "learning_rate": 2.9705785898336847e-05, "loss": 0.0125, "step": 7747 }, { "epoch": 5.444834855938159, "grad_norm": 0.12938182055950165, "learning_rate": 2.970531740454439e-05, "loss": 0.0209, "step": 7748 }, { "epoch": 5.445537596626845, "grad_norm": 0.22093313932418823, "learning_rate": 2.9704848910751934e-05, "loss": 0.0277, "step": 7749 }, { "epoch": 5.446240337315531, "grad_norm": 0.22739998996257782, "learning_rate": 2.9704380416959478e-05, "loss": 0.0264, "step": 7750 }, { "epoch": 5.4469430780042165, "grad_norm": 0.6041073799133301, "learning_rate": 2.970391192316702e-05, "loss": 0.0392, "step": 7751 }, { "epoch": 5.447645818692902, "grad_norm": 0.18163050711154938, "learning_rate": 2.9703443429374562e-05, "loss": 0.0191, "step": 7752 }, { "epoch": 5.448348559381588, "grad_norm": 0.21989063918590546, "learning_rate": 2.9702974935582106e-05, "loss": 0.0338, "step": 7753 }, { "epoch": 5.449051300070274, "grad_norm": 0.2560272812843323, "learning_rate": 2.970250644178965e-05, "loss": 0.0237, "step": 7754 }, { "epoch": 5.44975404075896, "grad_norm": 0.19021731615066528, "learning_rate": 2.9702037947997187e-05, "loss": 0.0435, "step": 7755 }, { "epoch": 5.450456781447646, "grad_norm": 0.17124328017234802, "learning_rate": 2.970156945420473e-05, "loss": 0.0343, "step": 7756 }, { "epoch": 5.451159522136332, "grad_norm": 0.29615578055381775, "learning_rate": 2.9701100960412274e-05, "loss": 0.0426, "step": 7757 }, { "epoch": 5.451862262825018, "grad_norm": 0.21898315846920013, "learning_rate": 2.9700632466619818e-05, "loss": 0.0329, "step": 7758 }, { "epoch": 5.452565003513703, "grad_norm": 0.30981749296188354, "learning_rate": 2.970016397282736e-05, "loss": 0.0623, "step": 7759 }, { "epoch": 5.453267744202389, "grad_norm": 0.48012083768844604, "learning_rate": 2.9699695479034902e-05, "loss": 0.0862, "step": 7760 }, { "epoch": 5.453970484891075, "grad_norm": 0.39880529046058655, "learning_rate": 2.9699226985242446e-05, "loss": 0.0888, "step": 7761 }, { "epoch": 5.454673225579761, "grad_norm": 0.6023617386817932, "learning_rate": 2.969875849144999e-05, "loss": 0.138, "step": 7762 }, { "epoch": 5.455375966268447, "grad_norm": 0.5678142309188843, "learning_rate": 2.9698289997657533e-05, "loss": 0.1601, "step": 7763 }, { "epoch": 5.456078706957133, "grad_norm": 1.0363060235977173, "learning_rate": 2.9697821503865073e-05, "loss": 0.2203, "step": 7764 }, { "epoch": 5.4567814476458185, "grad_norm": 1.1850290298461914, "learning_rate": 2.9697353010072617e-05, "loss": 0.2729, "step": 7765 }, { "epoch": 5.457484188334504, "grad_norm": 0.3538316488265991, "learning_rate": 2.969688451628016e-05, "loss": 0.1082, "step": 7766 }, { "epoch": 5.45818692902319, "grad_norm": 0.2049713432788849, "learning_rate": 2.9696416022487704e-05, "loss": 0.0372, "step": 7767 }, { "epoch": 5.458889669711876, "grad_norm": 0.15808232128620148, "learning_rate": 2.9695947528695245e-05, "loss": 0.0333, "step": 7768 }, { "epoch": 5.459592410400562, "grad_norm": 0.3129546642303467, "learning_rate": 2.969547903490279e-05, "loss": 0.04, "step": 7769 }, { "epoch": 5.460295151089248, "grad_norm": 0.20502294600009918, "learning_rate": 2.9695010541110332e-05, "loss": 0.0289, "step": 7770 }, { "epoch": 5.460997891777934, "grad_norm": 0.16609446704387665, "learning_rate": 2.9694542047317876e-05, "loss": 0.0204, "step": 7771 }, { "epoch": 5.46170063246662, "grad_norm": 0.1932562291622162, "learning_rate": 2.969407355352542e-05, "loss": 0.0175, "step": 7772 }, { "epoch": 5.462403373155306, "grad_norm": 0.18704023957252502, "learning_rate": 2.9693605059732957e-05, "loss": 0.0254, "step": 7773 }, { "epoch": 5.463106113843992, "grad_norm": 0.16206081211566925, "learning_rate": 2.96931365659405e-05, "loss": 0.0331, "step": 7774 }, { "epoch": 5.463808854532678, "grad_norm": 0.17731282114982605, "learning_rate": 2.9692668072148044e-05, "loss": 0.0281, "step": 7775 }, { "epoch": 5.464511595221364, "grad_norm": 0.255511999130249, "learning_rate": 2.9692199578355588e-05, "loss": 0.0512, "step": 7776 }, { "epoch": 5.46521433591005, "grad_norm": 0.37165194749832153, "learning_rate": 2.9691731084563128e-05, "loss": 0.0302, "step": 7777 }, { "epoch": 5.4659170765987355, "grad_norm": 0.18181933462619781, "learning_rate": 2.9691262590770672e-05, "loss": 0.0244, "step": 7778 }, { "epoch": 5.466619817287421, "grad_norm": 0.39423778653144836, "learning_rate": 2.9690794096978216e-05, "loss": 0.0201, "step": 7779 }, { "epoch": 5.4673225579761064, "grad_norm": 0.1984872967004776, "learning_rate": 2.969032560318576e-05, "loss": 0.0472, "step": 7780 }, { "epoch": 5.468025298664792, "grad_norm": 0.21726983785629272, "learning_rate": 2.96898571093933e-05, "loss": 0.0398, "step": 7781 }, { "epoch": 5.468728039353478, "grad_norm": 0.16648058593273163, "learning_rate": 2.9689388615600843e-05, "loss": 0.0249, "step": 7782 }, { "epoch": 5.469430780042164, "grad_norm": 0.3304961323738098, "learning_rate": 2.9688920121808387e-05, "loss": 0.0528, "step": 7783 }, { "epoch": 5.47013352073085, "grad_norm": 0.29957425594329834, "learning_rate": 2.968845162801593e-05, "loss": 0.0743, "step": 7784 }, { "epoch": 5.470836261419536, "grad_norm": 0.18716290593147278, "learning_rate": 2.9687983134223475e-05, "loss": 0.0353, "step": 7785 }, { "epoch": 5.471539002108222, "grad_norm": 0.36283013224601746, "learning_rate": 2.9687514640431015e-05, "loss": 0.0891, "step": 7786 }, { "epoch": 5.472241742796908, "grad_norm": 0.49757441878318787, "learning_rate": 2.968704614663856e-05, "loss": 0.1351, "step": 7787 }, { "epoch": 5.472944483485594, "grad_norm": 0.7987452149391174, "learning_rate": 2.9686577652846102e-05, "loss": 0.2062, "step": 7788 }, { "epoch": 5.47364722417428, "grad_norm": 0.735885500907898, "learning_rate": 2.9686109159053646e-05, "loss": 0.2171, "step": 7789 }, { "epoch": 5.474349964862966, "grad_norm": 1.1343321800231934, "learning_rate": 2.9685640665261183e-05, "loss": 0.257, "step": 7790 }, { "epoch": 5.475052705551652, "grad_norm": 0.2677462697029114, "learning_rate": 2.9685172171468727e-05, "loss": 0.0863, "step": 7791 }, { "epoch": 5.4757554462403375, "grad_norm": 0.16063588857650757, "learning_rate": 2.968470367767627e-05, "loss": 0.0332, "step": 7792 }, { "epoch": 5.476458186929023, "grad_norm": 0.16148048639297485, "learning_rate": 2.9684235183883814e-05, "loss": 0.037, "step": 7793 }, { "epoch": 5.477160927617709, "grad_norm": 0.1979285180568695, "learning_rate": 2.9683766690091355e-05, "loss": 0.0236, "step": 7794 }, { "epoch": 5.477863668306395, "grad_norm": 0.12901975214481354, "learning_rate": 2.9683298196298898e-05, "loss": 0.0243, "step": 7795 }, { "epoch": 5.478566408995081, "grad_norm": 0.22281964123249054, "learning_rate": 2.9682829702506442e-05, "loss": 0.0198, "step": 7796 }, { "epoch": 5.479269149683767, "grad_norm": 0.17414280772209167, "learning_rate": 2.9682361208713986e-05, "loss": 0.019, "step": 7797 }, { "epoch": 5.479971890372452, "grad_norm": 0.2041338086128235, "learning_rate": 2.968189271492153e-05, "loss": 0.0329, "step": 7798 }, { "epoch": 5.480674631061138, "grad_norm": 0.26805680990219116, "learning_rate": 2.968142422112907e-05, "loss": 0.0238, "step": 7799 }, { "epoch": 5.481377371749824, "grad_norm": 0.22005143761634827, "learning_rate": 2.9680955727336614e-05, "loss": 0.026, "step": 7800 }, { "epoch": 5.48208011243851, "grad_norm": 0.28386175632476807, "learning_rate": 2.9680487233544157e-05, "loss": 0.0347, "step": 7801 }, { "epoch": 5.482782853127196, "grad_norm": 0.21484802663326263, "learning_rate": 2.96800187397517e-05, "loss": 0.0138, "step": 7802 }, { "epoch": 5.483485593815882, "grad_norm": 0.16350658237934113, "learning_rate": 2.967955024595924e-05, "loss": 0.0248, "step": 7803 }, { "epoch": 5.484188334504568, "grad_norm": 0.18338999152183533, "learning_rate": 2.9679081752166785e-05, "loss": 0.0185, "step": 7804 }, { "epoch": 5.484891075193254, "grad_norm": 0.23244373500347137, "learning_rate": 2.967861325837433e-05, "loss": 0.028, "step": 7805 }, { "epoch": 5.4855938158819395, "grad_norm": 0.2477308213710785, "learning_rate": 2.9678144764581873e-05, "loss": 0.0426, "step": 7806 }, { "epoch": 5.486296556570625, "grad_norm": 0.19732420146465302, "learning_rate": 2.9677676270789413e-05, "loss": 0.0252, "step": 7807 }, { "epoch": 5.486999297259311, "grad_norm": 0.22958345711231232, "learning_rate": 2.9677207776996953e-05, "loss": 0.0362, "step": 7808 }, { "epoch": 5.487702037947997, "grad_norm": 0.3298903703689575, "learning_rate": 2.9676739283204497e-05, "loss": 0.0418, "step": 7809 }, { "epoch": 5.488404778636683, "grad_norm": 0.2914637327194214, "learning_rate": 2.967627078941204e-05, "loss": 0.0594, "step": 7810 }, { "epoch": 5.489107519325369, "grad_norm": 0.48423561453819275, "learning_rate": 2.9675802295619584e-05, "loss": 0.1177, "step": 7811 }, { "epoch": 5.489810260014055, "grad_norm": 0.452495276927948, "learning_rate": 2.9675333801827125e-05, "loss": 0.1226, "step": 7812 }, { "epoch": 5.490513000702741, "grad_norm": 0.7815123200416565, "learning_rate": 2.967486530803467e-05, "loss": 0.184, "step": 7813 }, { "epoch": 5.491215741391427, "grad_norm": 1.0633544921875, "learning_rate": 2.9674396814242212e-05, "loss": 0.2027, "step": 7814 }, { "epoch": 5.491918482080113, "grad_norm": 1.8542596101760864, "learning_rate": 2.9673928320449756e-05, "loss": 0.2727, "step": 7815 }, { "epoch": 5.492621222768799, "grad_norm": 0.28542324900627136, "learning_rate": 2.9673459826657296e-05, "loss": 0.0869, "step": 7816 }, { "epoch": 5.493323963457485, "grad_norm": 0.25611501932144165, "learning_rate": 2.967299133286484e-05, "loss": 0.0371, "step": 7817 }, { "epoch": 5.49402670414617, "grad_norm": 0.15527158975601196, "learning_rate": 2.9672522839072384e-05, "loss": 0.0342, "step": 7818 }, { "epoch": 5.494729444834856, "grad_norm": 0.26707956194877625, "learning_rate": 2.9672054345279927e-05, "loss": 0.0239, "step": 7819 }, { "epoch": 5.4954321855235415, "grad_norm": 0.14550158381462097, "learning_rate": 2.9671585851487468e-05, "loss": 0.0261, "step": 7820 }, { "epoch": 5.496134926212227, "grad_norm": 0.13209326565265656, "learning_rate": 2.967111735769501e-05, "loss": 0.0203, "step": 7821 }, { "epoch": 5.496837666900913, "grad_norm": 0.11199918389320374, "learning_rate": 2.9670648863902555e-05, "loss": 0.0184, "step": 7822 }, { "epoch": 5.497540407589599, "grad_norm": 0.14537280797958374, "learning_rate": 2.96701803701101e-05, "loss": 0.0258, "step": 7823 }, { "epoch": 5.498243148278285, "grad_norm": 0.14711368083953857, "learning_rate": 2.9669711876317643e-05, "loss": 0.0171, "step": 7824 }, { "epoch": 5.498945888966971, "grad_norm": 0.16551417112350464, "learning_rate": 2.966924338252518e-05, "loss": 0.0406, "step": 7825 }, { "epoch": 5.499648629655657, "grad_norm": 0.20898400247097015, "learning_rate": 2.9668774888732723e-05, "loss": 0.0175, "step": 7826 }, { "epoch": 5.500351370344343, "grad_norm": 0.16364161670207977, "learning_rate": 2.9668306394940267e-05, "loss": 0.0298, "step": 7827 }, { "epoch": 5.501054111033029, "grad_norm": 0.25528427958488464, "learning_rate": 2.966783790114781e-05, "loss": 0.019, "step": 7828 }, { "epoch": 5.501756851721715, "grad_norm": 0.2734260857105255, "learning_rate": 2.966736940735535e-05, "loss": 0.0323, "step": 7829 }, { "epoch": 5.502459592410401, "grad_norm": 0.3947972357273102, "learning_rate": 2.9666900913562895e-05, "loss": 0.0544, "step": 7830 }, { "epoch": 5.503162333099087, "grad_norm": 0.24060939252376556, "learning_rate": 2.966643241977044e-05, "loss": 0.0361, "step": 7831 }, { "epoch": 5.503865073787773, "grad_norm": 0.27159208059310913, "learning_rate": 2.9665963925977982e-05, "loss": 0.0311, "step": 7832 }, { "epoch": 5.5045678144764585, "grad_norm": 0.19550208747386932, "learning_rate": 2.9665495432185526e-05, "loss": 0.0357, "step": 7833 }, { "epoch": 5.505270555165144, "grad_norm": 0.2032126486301422, "learning_rate": 2.9665026938393066e-05, "loss": 0.0455, "step": 7834 }, { "epoch": 5.505973295853829, "grad_norm": 0.2746981382369995, "learning_rate": 2.966455844460061e-05, "loss": 0.0659, "step": 7835 }, { "epoch": 5.506676036542515, "grad_norm": 0.39839881658554077, "learning_rate": 2.9664089950808154e-05, "loss": 0.1008, "step": 7836 }, { "epoch": 5.507378777231201, "grad_norm": 0.5581663846969604, "learning_rate": 2.9663621457015697e-05, "loss": 0.1378, "step": 7837 }, { "epoch": 5.508081517919887, "grad_norm": 0.6595010757446289, "learning_rate": 2.9663152963223238e-05, "loss": 0.195, "step": 7838 }, { "epoch": 5.508784258608573, "grad_norm": 0.9260771870613098, "learning_rate": 2.966268446943078e-05, "loss": 0.2145, "step": 7839 }, { "epoch": 5.509486999297259, "grad_norm": 1.1041473150253296, "learning_rate": 2.9662215975638325e-05, "loss": 0.2721, "step": 7840 }, { "epoch": 5.510189739985945, "grad_norm": 0.2121284008026123, "learning_rate": 2.966174748184587e-05, "loss": 0.0718, "step": 7841 }, { "epoch": 5.510892480674631, "grad_norm": 0.13083533942699432, "learning_rate": 2.9661278988053406e-05, "loss": 0.0297, "step": 7842 }, { "epoch": 5.511595221363317, "grad_norm": 0.3306998312473297, "learning_rate": 2.966081049426095e-05, "loss": 0.0239, "step": 7843 }, { "epoch": 5.512297962052003, "grad_norm": 0.1584322601556778, "learning_rate": 2.9660342000468493e-05, "loss": 0.0147, "step": 7844 }, { "epoch": 5.513000702740689, "grad_norm": 0.18018805980682373, "learning_rate": 2.9659873506676037e-05, "loss": 0.0321, "step": 7845 }, { "epoch": 5.513703443429375, "grad_norm": 0.19629333913326263, "learning_rate": 2.965940501288358e-05, "loss": 0.0255, "step": 7846 }, { "epoch": 5.5144061841180605, "grad_norm": 0.19617462158203125, "learning_rate": 2.965893651909112e-05, "loss": 0.0221, "step": 7847 }, { "epoch": 5.515108924806746, "grad_norm": 0.23385682702064514, "learning_rate": 2.9658468025298665e-05, "loss": 0.0236, "step": 7848 }, { "epoch": 5.515811665495432, "grad_norm": 0.15930695831775665, "learning_rate": 2.965799953150621e-05, "loss": 0.027, "step": 7849 }, { "epoch": 5.516514406184118, "grad_norm": 0.17092125117778778, "learning_rate": 2.9657531037713752e-05, "loss": 0.015, "step": 7850 }, { "epoch": 5.517217146872804, "grad_norm": 0.25676068663597107, "learning_rate": 2.9657062543921293e-05, "loss": 0.0341, "step": 7851 }, { "epoch": 5.51791988756149, "grad_norm": 0.15639141201972961, "learning_rate": 2.9656594050128836e-05, "loss": 0.0126, "step": 7852 }, { "epoch": 5.518622628250176, "grad_norm": 0.291781485080719, "learning_rate": 2.965612555633638e-05, "loss": 0.0322, "step": 7853 }, { "epoch": 5.519325368938862, "grad_norm": 0.2090199589729309, "learning_rate": 2.9655657062543924e-05, "loss": 0.0218, "step": 7854 }, { "epoch": 5.520028109627548, "grad_norm": 0.24712258577346802, "learning_rate": 2.9655188568751464e-05, "loss": 0.0364, "step": 7855 }, { "epoch": 5.520730850316234, "grad_norm": 0.20602791011333466, "learning_rate": 2.9654720074959008e-05, "loss": 0.0286, "step": 7856 }, { "epoch": 5.521433591004919, "grad_norm": 0.3044671416282654, "learning_rate": 2.965425158116655e-05, "loss": 0.0259, "step": 7857 }, { "epoch": 5.522136331693605, "grad_norm": 0.40798792243003845, "learning_rate": 2.9653783087374095e-05, "loss": 0.0513, "step": 7858 }, { "epoch": 5.522839072382291, "grad_norm": 0.40153607726097107, "learning_rate": 2.965331459358164e-05, "loss": 0.0431, "step": 7859 }, { "epoch": 5.523541813070977, "grad_norm": 0.27657291293144226, "learning_rate": 2.9652846099789176e-05, "loss": 0.0586, "step": 7860 }, { "epoch": 5.5242445537596625, "grad_norm": 0.35955896973609924, "learning_rate": 2.965237760599672e-05, "loss": 0.0735, "step": 7861 }, { "epoch": 5.524947294448348, "grad_norm": 0.9675022959709167, "learning_rate": 2.9651909112204264e-05, "loss": 0.1554, "step": 7862 }, { "epoch": 5.525650035137034, "grad_norm": 0.735092282295227, "learning_rate": 2.9651440618411807e-05, "loss": 0.1664, "step": 7863 }, { "epoch": 5.52635277582572, "grad_norm": 1.0087003707885742, "learning_rate": 2.9650972124619348e-05, "loss": 0.2328, "step": 7864 }, { "epoch": 5.527055516514406, "grad_norm": 1.2335745096206665, "learning_rate": 2.965050363082689e-05, "loss": 0.2144, "step": 7865 }, { "epoch": 5.527758257203092, "grad_norm": 0.23975619673728943, "learning_rate": 2.9650035137034435e-05, "loss": 0.0695, "step": 7866 }, { "epoch": 5.528460997891778, "grad_norm": 0.22299040853977203, "learning_rate": 2.964956664324198e-05, "loss": 0.0557, "step": 7867 }, { "epoch": 5.529163738580464, "grad_norm": 0.19838851690292358, "learning_rate": 2.964909814944952e-05, "loss": 0.0215, "step": 7868 }, { "epoch": 5.52986647926915, "grad_norm": 0.1592581868171692, "learning_rate": 2.9648629655657063e-05, "loss": 0.0325, "step": 7869 }, { "epoch": 5.530569219957836, "grad_norm": 0.2404683232307434, "learning_rate": 2.9648161161864607e-05, "loss": 0.0233, "step": 7870 }, { "epoch": 5.531271960646522, "grad_norm": 0.20040272176265717, "learning_rate": 2.964769266807215e-05, "loss": 0.0215, "step": 7871 }, { "epoch": 5.531974701335208, "grad_norm": 0.20630009472370148, "learning_rate": 2.9647224174279694e-05, "loss": 0.0146, "step": 7872 }, { "epoch": 5.5326774420238936, "grad_norm": 0.1727074533700943, "learning_rate": 2.9646755680487234e-05, "loss": 0.0284, "step": 7873 }, { "epoch": 5.533380182712579, "grad_norm": 0.19815580546855927, "learning_rate": 2.9646287186694778e-05, "loss": 0.0246, "step": 7874 }, { "epoch": 5.5340829234012645, "grad_norm": 0.17283251881599426, "learning_rate": 2.9645818692902322e-05, "loss": 0.0269, "step": 7875 }, { "epoch": 5.53478566408995, "grad_norm": 0.19821658730506897, "learning_rate": 2.9645350199109866e-05, "loss": 0.0249, "step": 7876 }, { "epoch": 5.535488404778636, "grad_norm": 0.24200624227523804, "learning_rate": 2.9644881705317402e-05, "loss": 0.0269, "step": 7877 }, { "epoch": 5.536191145467322, "grad_norm": 0.2747345268726349, "learning_rate": 2.9644413211524946e-05, "loss": 0.0306, "step": 7878 }, { "epoch": 5.536893886156008, "grad_norm": 0.18268156051635742, "learning_rate": 2.964394471773249e-05, "loss": 0.0274, "step": 7879 }, { "epoch": 5.537596626844694, "grad_norm": 0.3078385293483734, "learning_rate": 2.9643476223940034e-05, "loss": 0.048, "step": 7880 }, { "epoch": 5.53829936753338, "grad_norm": 0.26662734150886536, "learning_rate": 2.9643007730147574e-05, "loss": 0.0471, "step": 7881 }, { "epoch": 5.539002108222066, "grad_norm": 0.32579463720321655, "learning_rate": 2.9642539236355118e-05, "loss": 0.0278, "step": 7882 }, { "epoch": 5.539704848910752, "grad_norm": 0.22368381917476654, "learning_rate": 2.964207074256266e-05, "loss": 0.0399, "step": 7883 }, { "epoch": 5.540407589599438, "grad_norm": 0.3027983009815216, "learning_rate": 2.9641602248770205e-05, "loss": 0.0666, "step": 7884 }, { "epoch": 5.541110330288124, "grad_norm": 0.2968800663948059, "learning_rate": 2.964113375497775e-05, "loss": 0.0463, "step": 7885 }, { "epoch": 5.54181307097681, "grad_norm": 0.6791741847991943, "learning_rate": 2.964066526118529e-05, "loss": 0.0875, "step": 7886 }, { "epoch": 5.542515811665496, "grad_norm": 0.45148470997810364, "learning_rate": 2.9640196767392833e-05, "loss": 0.1673, "step": 7887 }, { "epoch": 5.5432185523541815, "grad_norm": 0.7137035131454468, "learning_rate": 2.9639728273600377e-05, "loss": 0.1882, "step": 7888 }, { "epoch": 5.543921293042867, "grad_norm": 0.9725578427314758, "learning_rate": 2.963925977980792e-05, "loss": 0.2208, "step": 7889 }, { "epoch": 5.544624033731553, "grad_norm": 1.1752102375030518, "learning_rate": 2.963879128601546e-05, "loss": 0.3247, "step": 7890 }, { "epoch": 5.545326774420239, "grad_norm": 0.24233561754226685, "learning_rate": 2.9638322792223004e-05, "loss": 0.0872, "step": 7891 }, { "epoch": 5.546029515108925, "grad_norm": 0.15563875436782837, "learning_rate": 2.9637854298430548e-05, "loss": 0.0287, "step": 7892 }, { "epoch": 5.546732255797611, "grad_norm": 0.20359446108341217, "learning_rate": 2.9637385804638092e-05, "loss": 0.0195, "step": 7893 }, { "epoch": 5.547434996486297, "grad_norm": 0.13133607804775238, "learning_rate": 2.963691731084563e-05, "loss": 0.0214, "step": 7894 }, { "epoch": 5.548137737174983, "grad_norm": 0.1400357484817505, "learning_rate": 2.9636448817053173e-05, "loss": 0.0272, "step": 7895 }, { "epoch": 5.548840477863668, "grad_norm": 0.13315968215465546, "learning_rate": 2.9635980323260716e-05, "loss": 0.0243, "step": 7896 }, { "epoch": 5.549543218552354, "grad_norm": 0.17561344802379608, "learning_rate": 2.963551182946826e-05, "loss": 0.0167, "step": 7897 }, { "epoch": 5.55024595924104, "grad_norm": 0.30881085991859436, "learning_rate": 2.9635043335675804e-05, "loss": 0.0407, "step": 7898 }, { "epoch": 5.550948699929726, "grad_norm": 0.21636101603507996, "learning_rate": 2.9634574841883344e-05, "loss": 0.0256, "step": 7899 }, { "epoch": 5.551651440618412, "grad_norm": 0.19201169908046722, "learning_rate": 2.9634106348090888e-05, "loss": 0.0225, "step": 7900 }, { "epoch": 5.552354181307098, "grad_norm": 0.20568622648715973, "learning_rate": 2.963363785429843e-05, "loss": 0.0351, "step": 7901 }, { "epoch": 5.5530569219957835, "grad_norm": 0.16609472036361694, "learning_rate": 2.9633169360505975e-05, "loss": 0.0202, "step": 7902 }, { "epoch": 5.553759662684469, "grad_norm": 0.2519652843475342, "learning_rate": 2.9632700866713516e-05, "loss": 0.0354, "step": 7903 }, { "epoch": 5.554462403373155, "grad_norm": 0.2719908058643341, "learning_rate": 2.963223237292106e-05, "loss": 0.017, "step": 7904 }, { "epoch": 5.555165144061841, "grad_norm": 0.18911895155906677, "learning_rate": 2.9631763879128603e-05, "loss": 0.0297, "step": 7905 }, { "epoch": 5.555867884750527, "grad_norm": 0.162324458360672, "learning_rate": 2.9631295385336147e-05, "loss": 0.0414, "step": 7906 }, { "epoch": 5.556570625439213, "grad_norm": 0.1525326818227768, "learning_rate": 2.9630826891543687e-05, "loss": 0.0147, "step": 7907 }, { "epoch": 5.557273366127899, "grad_norm": 0.2645239233970642, "learning_rate": 2.963035839775123e-05, "loss": 0.0366, "step": 7908 }, { "epoch": 5.557976106816585, "grad_norm": 0.3418998420238495, "learning_rate": 2.9629889903958775e-05, "loss": 0.0526, "step": 7909 }, { "epoch": 5.558678847505271, "grad_norm": 0.29159751534461975, "learning_rate": 2.9629421410166318e-05, "loss": 0.0568, "step": 7910 }, { "epoch": 5.559381588193957, "grad_norm": 0.4604395031929016, "learning_rate": 2.9628952916373862e-05, "loss": 0.0839, "step": 7911 }, { "epoch": 5.560084328882642, "grad_norm": 0.5286097526550293, "learning_rate": 2.96284844225814e-05, "loss": 0.1456, "step": 7912 }, { "epoch": 5.560787069571328, "grad_norm": 0.6842241883277893, "learning_rate": 2.9628015928788943e-05, "loss": 0.1954, "step": 7913 }, { "epoch": 5.561489810260014, "grad_norm": 0.7532913684844971, "learning_rate": 2.9627547434996486e-05, "loss": 0.2544, "step": 7914 }, { "epoch": 5.5621925509487, "grad_norm": 1.3109833002090454, "learning_rate": 2.962707894120403e-05, "loss": 0.2822, "step": 7915 }, { "epoch": 5.5628952916373855, "grad_norm": 0.5676263570785522, "learning_rate": 2.962661044741157e-05, "loss": 0.0745, "step": 7916 }, { "epoch": 5.563598032326071, "grad_norm": 0.17257462441921234, "learning_rate": 2.9626141953619114e-05, "loss": 0.0429, "step": 7917 }, { "epoch": 5.564300773014757, "grad_norm": 0.2253546267747879, "learning_rate": 2.9625673459826658e-05, "loss": 0.0262, "step": 7918 }, { "epoch": 5.565003513703443, "grad_norm": 0.169489324092865, "learning_rate": 2.96252049660342e-05, "loss": 0.0295, "step": 7919 }, { "epoch": 5.565706254392129, "grad_norm": 0.2517862617969513, "learning_rate": 2.9624736472241742e-05, "loss": 0.0308, "step": 7920 }, { "epoch": 5.566408995080815, "grad_norm": 0.21697667241096497, "learning_rate": 2.9624267978449286e-05, "loss": 0.0198, "step": 7921 }, { "epoch": 5.567111735769501, "grad_norm": 0.14932163059711456, "learning_rate": 2.962379948465683e-05, "loss": 0.0328, "step": 7922 }, { "epoch": 5.567814476458187, "grad_norm": 0.13505691289901733, "learning_rate": 2.9623330990864373e-05, "loss": 0.0265, "step": 7923 }, { "epoch": 5.568517217146873, "grad_norm": 0.16657213866710663, "learning_rate": 2.9622862497071917e-05, "loss": 0.0192, "step": 7924 }, { "epoch": 5.569219957835559, "grad_norm": 0.1369016170501709, "learning_rate": 2.9622394003279457e-05, "loss": 0.0118, "step": 7925 }, { "epoch": 5.569922698524245, "grad_norm": 0.15460285544395447, "learning_rate": 2.9621925509487e-05, "loss": 0.0263, "step": 7926 }, { "epoch": 5.570625439212931, "grad_norm": 0.1350378841161728, "learning_rate": 2.9621457015694545e-05, "loss": 0.0142, "step": 7927 }, { "epoch": 5.5713281799016166, "grad_norm": 0.2919127345085144, "learning_rate": 2.962098852190209e-05, "loss": 0.026, "step": 7928 }, { "epoch": 5.5720309205903025, "grad_norm": 0.14654874801635742, "learning_rate": 2.9620520028109625e-05, "loss": 0.0147, "step": 7929 }, { "epoch": 5.572733661278988, "grad_norm": 0.20118042826652527, "learning_rate": 2.962005153431717e-05, "loss": 0.033, "step": 7930 }, { "epoch": 5.573436401967674, "grad_norm": 0.5178455114364624, "learning_rate": 2.9619583040524713e-05, "loss": 0.0472, "step": 7931 }, { "epoch": 5.57413914265636, "grad_norm": 0.2724420428276062, "learning_rate": 2.9619114546732257e-05, "loss": 0.0214, "step": 7932 }, { "epoch": 5.574841883345046, "grad_norm": 0.23443393409252167, "learning_rate": 2.9618646052939797e-05, "loss": 0.0394, "step": 7933 }, { "epoch": 5.575544624033731, "grad_norm": 0.32197704911231995, "learning_rate": 2.961817755914734e-05, "loss": 0.0385, "step": 7934 }, { "epoch": 5.576247364722417, "grad_norm": 0.5158899426460266, "learning_rate": 2.9617709065354884e-05, "loss": 0.0874, "step": 7935 }, { "epoch": 5.576950105411103, "grad_norm": 0.4989846646785736, "learning_rate": 2.9617240571562428e-05, "loss": 0.0831, "step": 7936 }, { "epoch": 5.577652846099789, "grad_norm": 0.6932191252708435, "learning_rate": 2.9616772077769972e-05, "loss": 0.1684, "step": 7937 }, { "epoch": 5.578355586788475, "grad_norm": 0.8594722151756287, "learning_rate": 2.9616303583977512e-05, "loss": 0.1874, "step": 7938 }, { "epoch": 5.579058327477161, "grad_norm": 1.044046401977539, "learning_rate": 2.9615835090185056e-05, "loss": 0.2581, "step": 7939 }, { "epoch": 5.579761068165847, "grad_norm": 1.1823772192001343, "learning_rate": 2.96153665963926e-05, "loss": 0.2341, "step": 7940 }, { "epoch": 5.580463808854533, "grad_norm": 0.25434109568595886, "learning_rate": 2.9614898102600143e-05, "loss": 0.0677, "step": 7941 }, { "epoch": 5.581166549543219, "grad_norm": 0.3454763889312744, "learning_rate": 2.9614429608807684e-05, "loss": 0.0316, "step": 7942 }, { "epoch": 5.5818692902319045, "grad_norm": 0.24270276725292206, "learning_rate": 2.9613961115015227e-05, "loss": 0.0601, "step": 7943 }, { "epoch": 5.58257203092059, "grad_norm": 0.16820397973060608, "learning_rate": 2.961349262122277e-05, "loss": 0.0229, "step": 7944 }, { "epoch": 5.583274771609276, "grad_norm": 0.12912166118621826, "learning_rate": 2.9613024127430315e-05, "loss": 0.0236, "step": 7945 }, { "epoch": 5.583977512297962, "grad_norm": 0.14843179285526276, "learning_rate": 2.9612555633637855e-05, "loss": 0.0205, "step": 7946 }, { "epoch": 5.584680252986648, "grad_norm": 0.576906144618988, "learning_rate": 2.9612087139845395e-05, "loss": 0.0143, "step": 7947 }, { "epoch": 5.585382993675334, "grad_norm": 0.11587905138731003, "learning_rate": 2.961161864605294e-05, "loss": 0.0214, "step": 7948 }, { "epoch": 5.58608573436402, "grad_norm": 0.15434570610523224, "learning_rate": 2.9611150152260483e-05, "loss": 0.016, "step": 7949 }, { "epoch": 5.586788475052706, "grad_norm": 0.12222988158464432, "learning_rate": 2.9610681658468027e-05, "loss": 0.0155, "step": 7950 }, { "epoch": 5.587491215741391, "grad_norm": 0.18500268459320068, "learning_rate": 2.9610213164675567e-05, "loss": 0.0263, "step": 7951 }, { "epoch": 5.588193956430077, "grad_norm": 0.1722947359085083, "learning_rate": 2.960974467088311e-05, "loss": 0.0174, "step": 7952 }, { "epoch": 5.588896697118763, "grad_norm": 0.20981411635875702, "learning_rate": 2.9609276177090654e-05, "loss": 0.0345, "step": 7953 }, { "epoch": 5.589599437807449, "grad_norm": 0.1872740387916565, "learning_rate": 2.9608807683298198e-05, "loss": 0.0185, "step": 7954 }, { "epoch": 5.590302178496135, "grad_norm": 0.2893826961517334, "learning_rate": 2.960833918950574e-05, "loss": 0.0379, "step": 7955 }, { "epoch": 5.591004919184821, "grad_norm": 0.24819637835025787, "learning_rate": 2.9607870695713282e-05, "loss": 0.0277, "step": 7956 }, { "epoch": 5.5917076598735065, "grad_norm": 0.23059597611427307, "learning_rate": 2.9607402201920826e-05, "loss": 0.0257, "step": 7957 }, { "epoch": 5.592410400562192, "grad_norm": 0.21984800696372986, "learning_rate": 2.960693370812837e-05, "loss": 0.0444, "step": 7958 }, { "epoch": 5.593113141250878, "grad_norm": 0.2772398889064789, "learning_rate": 2.960646521433591e-05, "loss": 0.0595, "step": 7959 }, { "epoch": 5.593815881939564, "grad_norm": 0.5622789859771729, "learning_rate": 2.9605996720543454e-05, "loss": 0.0519, "step": 7960 }, { "epoch": 5.59451862262825, "grad_norm": 0.43749865889549255, "learning_rate": 2.9605528226750997e-05, "loss": 0.0945, "step": 7961 }, { "epoch": 5.595221363316936, "grad_norm": 0.7805104851722717, "learning_rate": 2.960505973295854e-05, "loss": 0.1296, "step": 7962 }, { "epoch": 5.595924104005622, "grad_norm": 0.5856647491455078, "learning_rate": 2.9604591239166085e-05, "loss": 0.1843, "step": 7963 }, { "epoch": 5.596626844694308, "grad_norm": 1.063559651374817, "learning_rate": 2.9604122745373622e-05, "loss": 0.1855, "step": 7964 }, { "epoch": 5.597329585382994, "grad_norm": 1.0207018852233887, "learning_rate": 2.9603654251581166e-05, "loss": 0.2667, "step": 7965 }, { "epoch": 5.59803232607168, "grad_norm": 0.256561815738678, "learning_rate": 2.960318575778871e-05, "loss": 0.0924, "step": 7966 }, { "epoch": 5.598735066760366, "grad_norm": 0.2010391354560852, "learning_rate": 2.9602717263996253e-05, "loss": 0.0433, "step": 7967 }, { "epoch": 5.599437807449052, "grad_norm": 0.1400989592075348, "learning_rate": 2.9602248770203793e-05, "loss": 0.0329, "step": 7968 }, { "epoch": 5.6001405481377375, "grad_norm": 0.14144718647003174, "learning_rate": 2.9601780276411337e-05, "loss": 0.0204, "step": 7969 }, { "epoch": 5.6008432888264235, "grad_norm": 0.119900181889534, "learning_rate": 2.960131178261888e-05, "loss": 0.0143, "step": 7970 }, { "epoch": 5.601546029515109, "grad_norm": 0.12583595514297485, "learning_rate": 2.9600843288826425e-05, "loss": 0.0189, "step": 7971 }, { "epoch": 5.602248770203794, "grad_norm": 0.22378826141357422, "learning_rate": 2.9600374795033965e-05, "loss": 0.031, "step": 7972 }, { "epoch": 5.60295151089248, "grad_norm": 0.15727518498897552, "learning_rate": 2.959990630124151e-05, "loss": 0.0241, "step": 7973 }, { "epoch": 5.603654251581166, "grad_norm": 0.21488863229751587, "learning_rate": 2.9599437807449052e-05, "loss": 0.0244, "step": 7974 }, { "epoch": 5.604356992269852, "grad_norm": 0.10928952693939209, "learning_rate": 2.9598969313656596e-05, "loss": 0.0156, "step": 7975 }, { "epoch": 5.605059732958538, "grad_norm": 0.23621505498886108, "learning_rate": 2.959850081986414e-05, "loss": 0.0408, "step": 7976 }, { "epoch": 5.605762473647224, "grad_norm": 0.24193181097507477, "learning_rate": 2.959803232607168e-05, "loss": 0.0225, "step": 7977 }, { "epoch": 5.60646521433591, "grad_norm": 0.23681803047657013, "learning_rate": 2.9597563832279224e-05, "loss": 0.0463, "step": 7978 }, { "epoch": 5.607167955024596, "grad_norm": 0.2628649175167084, "learning_rate": 2.9597095338486768e-05, "loss": 0.0163, "step": 7979 }, { "epoch": 5.607870695713282, "grad_norm": 0.48481738567352295, "learning_rate": 2.959662684469431e-05, "loss": 0.0484, "step": 7980 }, { "epoch": 5.608573436401968, "grad_norm": 0.22919195890426636, "learning_rate": 2.959615835090185e-05, "loss": 0.0529, "step": 7981 }, { "epoch": 5.609276177090654, "grad_norm": 0.25941407680511475, "learning_rate": 2.9595689857109392e-05, "loss": 0.0381, "step": 7982 }, { "epoch": 5.6099789177793395, "grad_norm": 0.2544838786125183, "learning_rate": 2.9595221363316936e-05, "loss": 0.0353, "step": 7983 }, { "epoch": 5.6106816584680255, "grad_norm": 0.23173633217811584, "learning_rate": 2.959475286952448e-05, "loss": 0.047, "step": 7984 }, { "epoch": 5.611384399156711, "grad_norm": 0.23968535661697388, "learning_rate": 2.959428437573202e-05, "loss": 0.0525, "step": 7985 }, { "epoch": 5.612087139845397, "grad_norm": 0.32631605863571167, "learning_rate": 2.9593815881939563e-05, "loss": 0.0764, "step": 7986 }, { "epoch": 5.612789880534083, "grad_norm": 0.4372870624065399, "learning_rate": 2.9593347388147107e-05, "loss": 0.1001, "step": 7987 }, { "epoch": 5.613492621222769, "grad_norm": 0.719863772392273, "learning_rate": 2.959287889435465e-05, "loss": 0.1819, "step": 7988 }, { "epoch": 5.614195361911454, "grad_norm": 1.079507827758789, "learning_rate": 2.9592410400562195e-05, "loss": 0.2224, "step": 7989 }, { "epoch": 5.61489810260014, "grad_norm": 1.0609683990478516, "learning_rate": 2.9591941906769735e-05, "loss": 0.2784, "step": 7990 }, { "epoch": 5.615600843288826, "grad_norm": 0.3574804961681366, "learning_rate": 2.959147341297728e-05, "loss": 0.0819, "step": 7991 }, { "epoch": 5.616303583977512, "grad_norm": 0.2273426651954651, "learning_rate": 2.9591004919184822e-05, "loss": 0.0377, "step": 7992 }, { "epoch": 5.617006324666198, "grad_norm": 0.13573457300662994, "learning_rate": 2.9590536425392366e-05, "loss": 0.0205, "step": 7993 }, { "epoch": 5.617709065354884, "grad_norm": 0.2781001925468445, "learning_rate": 2.9590067931599906e-05, "loss": 0.0194, "step": 7994 }, { "epoch": 5.61841180604357, "grad_norm": 0.15616737306118011, "learning_rate": 2.958959943780745e-05, "loss": 0.0228, "step": 7995 }, { "epoch": 5.619114546732256, "grad_norm": 0.1766369640827179, "learning_rate": 2.9589130944014994e-05, "loss": 0.0364, "step": 7996 }, { "epoch": 5.6198172874209416, "grad_norm": 0.13978375494480133, "learning_rate": 2.9588662450222538e-05, "loss": 0.0191, "step": 7997 }, { "epoch": 5.6205200281096275, "grad_norm": 0.2970595359802246, "learning_rate": 2.9588193956430078e-05, "loss": 0.0371, "step": 7998 }, { "epoch": 5.621222768798313, "grad_norm": 0.18108896911144257, "learning_rate": 2.958772546263762e-05, "loss": 0.045, "step": 7999 }, { "epoch": 5.621925509486999, "grad_norm": 0.09335355460643768, "learning_rate": 2.9587256968845162e-05, "loss": 0.0147, "step": 8000 }, { "epoch": 5.621925509486999, "eval_cer": 0.19799679224902622, "eval_loss": 0.2950074076652527, "eval_runtime": 18.4369, "eval_samples_per_second": 246.137, "eval_steps_per_second": 0.814, "eval_wer": 0.3583140135750398, "step": 8000 }, { "epoch": 5.622628250175685, "grad_norm": 0.15903989970684052, "learning_rate": 2.9586788475052706e-05, "loss": 0.031, "step": 8001 }, { "epoch": 5.623330990864371, "grad_norm": 0.15615227818489075, "learning_rate": 2.958631998126025e-05, "loss": 0.0251, "step": 8002 }, { "epoch": 5.624033731553057, "grad_norm": 0.29189637303352356, "learning_rate": 2.958585148746779e-05, "loss": 0.0655, "step": 8003 }, { "epoch": 5.624736472241743, "grad_norm": 0.13444024324417114, "learning_rate": 2.9585382993675334e-05, "loss": 0.0164, "step": 8004 }, { "epoch": 5.625439212930429, "grad_norm": 0.3409136235713959, "learning_rate": 2.9584914499882877e-05, "loss": 0.0377, "step": 8005 }, { "epoch": 5.626141953619115, "grad_norm": 0.19200579822063446, "learning_rate": 2.958444600609042e-05, "loss": 0.0285, "step": 8006 }, { "epoch": 5.626844694307801, "grad_norm": 0.13372589647769928, "learning_rate": 2.958397751229796e-05, "loss": 0.0193, "step": 8007 }, { "epoch": 5.627547434996487, "grad_norm": 0.3104481101036072, "learning_rate": 2.9583509018505505e-05, "loss": 0.0589, "step": 8008 }, { "epoch": 5.628250175685173, "grad_norm": 0.33266982436180115, "learning_rate": 2.958304052471305e-05, "loss": 0.0509, "step": 8009 }, { "epoch": 5.6289529163738585, "grad_norm": 0.30694442987442017, "learning_rate": 2.9582572030920593e-05, "loss": 0.0667, "step": 8010 }, { "epoch": 5.629655657062544, "grad_norm": 0.40677034854888916, "learning_rate": 2.9582103537128133e-05, "loss": 0.0949, "step": 8011 }, { "epoch": 5.6303583977512295, "grad_norm": 0.5985527634620667, "learning_rate": 2.9581635043335677e-05, "loss": 0.1217, "step": 8012 }, { "epoch": 5.631061138439915, "grad_norm": 0.7343234419822693, "learning_rate": 2.958116654954322e-05, "loss": 0.1588, "step": 8013 }, { "epoch": 5.631763879128601, "grad_norm": 0.75759357213974, "learning_rate": 2.9580698055750764e-05, "loss": 0.1901, "step": 8014 }, { "epoch": 5.632466619817287, "grad_norm": 1.8446760177612305, "learning_rate": 2.9580229561958308e-05, "loss": 0.2744, "step": 8015 }, { "epoch": 5.633169360505973, "grad_norm": 0.2777179479598999, "learning_rate": 2.9579761068165845e-05, "loss": 0.1001, "step": 8016 }, { "epoch": 5.633872101194659, "grad_norm": 0.2695471942424774, "learning_rate": 2.957929257437339e-05, "loss": 0.0263, "step": 8017 }, { "epoch": 5.634574841883345, "grad_norm": 0.1863054484128952, "learning_rate": 2.9578824080580932e-05, "loss": 0.0252, "step": 8018 }, { "epoch": 5.635277582572031, "grad_norm": 0.14709734916687012, "learning_rate": 2.9578355586788476e-05, "loss": 0.0228, "step": 8019 }, { "epoch": 5.635980323260717, "grad_norm": 0.1995955854654312, "learning_rate": 2.9577887092996016e-05, "loss": 0.0211, "step": 8020 }, { "epoch": 5.636683063949403, "grad_norm": 0.14364571869373322, "learning_rate": 2.957741859920356e-05, "loss": 0.0174, "step": 8021 }, { "epoch": 5.637385804638089, "grad_norm": 0.2469266951084137, "learning_rate": 2.9576950105411104e-05, "loss": 0.0316, "step": 8022 }, { "epoch": 5.638088545326775, "grad_norm": 0.2539346218109131, "learning_rate": 2.9576481611618647e-05, "loss": 0.0206, "step": 8023 }, { "epoch": 5.6387912860154605, "grad_norm": 0.2152240127325058, "learning_rate": 2.9576013117826188e-05, "loss": 0.027, "step": 8024 }, { "epoch": 5.6394940267041465, "grad_norm": 0.5564555525779724, "learning_rate": 2.957554462403373e-05, "loss": 0.018, "step": 8025 }, { "epoch": 5.640196767392832, "grad_norm": 0.146180659532547, "learning_rate": 2.9575076130241275e-05, "loss": 0.0214, "step": 8026 }, { "epoch": 5.640899508081518, "grad_norm": 0.1402578055858612, "learning_rate": 2.957460763644882e-05, "loss": 0.0175, "step": 8027 }, { "epoch": 5.641602248770203, "grad_norm": 0.26651740074157715, "learning_rate": 2.9574139142656363e-05, "loss": 0.0343, "step": 8028 }, { "epoch": 5.642304989458889, "grad_norm": 0.23884476721286774, "learning_rate": 2.9573670648863903e-05, "loss": 0.0238, "step": 8029 }, { "epoch": 5.643007730147575, "grad_norm": 0.266403466463089, "learning_rate": 2.9573202155071447e-05, "loss": 0.0351, "step": 8030 }, { "epoch": 5.643710470836261, "grad_norm": 0.2196836620569229, "learning_rate": 2.957273366127899e-05, "loss": 0.0578, "step": 8031 }, { "epoch": 5.644413211524947, "grad_norm": 0.1669638752937317, "learning_rate": 2.9572265167486534e-05, "loss": 0.0225, "step": 8032 }, { "epoch": 5.645115952213633, "grad_norm": 0.2658984661102295, "learning_rate": 2.9571796673694075e-05, "loss": 0.029, "step": 8033 }, { "epoch": 5.645818692902319, "grad_norm": 0.27101460099220276, "learning_rate": 2.9571328179901615e-05, "loss": 0.0464, "step": 8034 }, { "epoch": 5.646521433591005, "grad_norm": 0.5104789733886719, "learning_rate": 2.957085968610916e-05, "loss": 0.1097, "step": 8035 }, { "epoch": 5.647224174279691, "grad_norm": 0.5697578191757202, "learning_rate": 2.9570391192316702e-05, "loss": 0.0835, "step": 8036 }, { "epoch": 5.647926914968377, "grad_norm": 0.6359078288078308, "learning_rate": 2.9569922698524246e-05, "loss": 0.1345, "step": 8037 }, { "epoch": 5.6486296556570625, "grad_norm": 0.7937038540840149, "learning_rate": 2.9569454204731786e-05, "loss": 0.1945, "step": 8038 }, { "epoch": 5.6493323963457485, "grad_norm": 2.0311739444732666, "learning_rate": 2.956898571093933e-05, "loss": 0.2179, "step": 8039 }, { "epoch": 5.650035137034434, "grad_norm": 0.9397835731506348, "learning_rate": 2.9568517217146874e-05, "loss": 0.243, "step": 8040 }, { "epoch": 5.65073787772312, "grad_norm": 0.30515795946121216, "learning_rate": 2.9568048723354418e-05, "loss": 0.0618, "step": 8041 }, { "epoch": 5.651440618411806, "grad_norm": 0.1745205521583557, "learning_rate": 2.9567580229561958e-05, "loss": 0.0302, "step": 8042 }, { "epoch": 5.652143359100492, "grad_norm": 0.14761027693748474, "learning_rate": 2.95671117357695e-05, "loss": 0.0232, "step": 8043 }, { "epoch": 5.652846099789178, "grad_norm": 0.18702098727226257, "learning_rate": 2.9566643241977045e-05, "loss": 0.0283, "step": 8044 }, { "epoch": 5.653548840477864, "grad_norm": 0.173322856426239, "learning_rate": 2.956617474818459e-05, "loss": 0.0243, "step": 8045 }, { "epoch": 5.65425158116655, "grad_norm": 0.18043813109397888, "learning_rate": 2.956570625439213e-05, "loss": 0.0176, "step": 8046 }, { "epoch": 5.654954321855236, "grad_norm": 1.2886204719543457, "learning_rate": 2.9565237760599673e-05, "loss": 0.0302, "step": 8047 }, { "epoch": 5.655657062543922, "grad_norm": 0.1452690064907074, "learning_rate": 2.9564769266807217e-05, "loss": 0.0215, "step": 8048 }, { "epoch": 5.656359803232607, "grad_norm": 0.2655898928642273, "learning_rate": 2.956430077301476e-05, "loss": 0.0158, "step": 8049 }, { "epoch": 5.657062543921293, "grad_norm": 0.11323008686304092, "learning_rate": 2.95638322792223e-05, "loss": 0.0206, "step": 8050 }, { "epoch": 5.657765284609979, "grad_norm": 0.18328210711479187, "learning_rate": 2.956336378542984e-05, "loss": 0.0494, "step": 8051 }, { "epoch": 5.6584680252986645, "grad_norm": 0.14511533081531525, "learning_rate": 2.9562895291637385e-05, "loss": 0.0252, "step": 8052 }, { "epoch": 5.6591707659873505, "grad_norm": 0.3641771674156189, "learning_rate": 2.956242679784493e-05, "loss": 0.0296, "step": 8053 }, { "epoch": 5.659873506676036, "grad_norm": 0.1493275761604309, "learning_rate": 2.9561958304052472e-05, "loss": 0.0168, "step": 8054 }, { "epoch": 5.660576247364722, "grad_norm": 0.2368163764476776, "learning_rate": 2.9561489810260013e-05, "loss": 0.0299, "step": 8055 }, { "epoch": 5.661278988053408, "grad_norm": 0.2182588428258896, "learning_rate": 2.9561021316467556e-05, "loss": 0.03, "step": 8056 }, { "epoch": 5.661981728742094, "grad_norm": 0.2065858691930771, "learning_rate": 2.95605528226751e-05, "loss": 0.0321, "step": 8057 }, { "epoch": 5.66268446943078, "grad_norm": 0.1955631971359253, "learning_rate": 2.9560084328882644e-05, "loss": 0.0396, "step": 8058 }, { "epoch": 5.663387210119466, "grad_norm": 1.0602778196334839, "learning_rate": 2.9559615835090184e-05, "loss": 0.0354, "step": 8059 }, { "epoch": 5.664089950808152, "grad_norm": 0.39521706104278564, "learning_rate": 2.9559147341297728e-05, "loss": 0.1, "step": 8060 }, { "epoch": 5.664792691496838, "grad_norm": 0.2676352858543396, "learning_rate": 2.9558678847505272e-05, "loss": 0.0754, "step": 8061 }, { "epoch": 5.665495432185524, "grad_norm": 0.36492303013801575, "learning_rate": 2.9558210353712815e-05, "loss": 0.097, "step": 8062 }, { "epoch": 5.66619817287421, "grad_norm": 0.9942603707313538, "learning_rate": 2.955774185992036e-05, "loss": 0.1953, "step": 8063 }, { "epoch": 5.666900913562896, "grad_norm": 1.1020389795303345, "learning_rate": 2.95572733661279e-05, "loss": 0.2409, "step": 8064 }, { "epoch": 5.6676036542515815, "grad_norm": 1.5038108825683594, "learning_rate": 2.9556804872335443e-05, "loss": 0.2427, "step": 8065 }, { "epoch": 5.668306394940267, "grad_norm": 0.273624062538147, "learning_rate": 2.9556336378542987e-05, "loss": 0.0561, "step": 8066 }, { "epoch": 5.6690091356289525, "grad_norm": 0.14635814726352692, "learning_rate": 2.955586788475053e-05, "loss": 0.0302, "step": 8067 }, { "epoch": 5.669711876317638, "grad_norm": 0.4408627450466156, "learning_rate": 2.955539939095807e-05, "loss": 0.0316, "step": 8068 }, { "epoch": 5.670414617006324, "grad_norm": 0.1790924221277237, "learning_rate": 2.955493089716561e-05, "loss": 0.0168, "step": 8069 }, { "epoch": 5.67111735769501, "grad_norm": 0.12498051673173904, "learning_rate": 2.9554462403373155e-05, "loss": 0.0175, "step": 8070 }, { "epoch": 5.671820098383696, "grad_norm": 0.1662159264087677, "learning_rate": 2.95539939095807e-05, "loss": 0.0128, "step": 8071 }, { "epoch": 5.672522839072382, "grad_norm": 0.1897500455379486, "learning_rate": 2.955352541578824e-05, "loss": 0.0187, "step": 8072 }, { "epoch": 5.673225579761068, "grad_norm": 0.18515664339065552, "learning_rate": 2.9553056921995783e-05, "loss": 0.0269, "step": 8073 }, { "epoch": 5.673928320449754, "grad_norm": 0.31458863615989685, "learning_rate": 2.9552588428203327e-05, "loss": 0.0444, "step": 8074 }, { "epoch": 5.67463106113844, "grad_norm": 0.13183088600635529, "learning_rate": 2.955211993441087e-05, "loss": 0.0128, "step": 8075 }, { "epoch": 5.675333801827126, "grad_norm": 0.3533637523651123, "learning_rate": 2.9551651440618414e-05, "loss": 0.0334, "step": 8076 }, { "epoch": 5.676036542515812, "grad_norm": 0.17968769371509552, "learning_rate": 2.9551182946825954e-05, "loss": 0.0215, "step": 8077 }, { "epoch": 5.676739283204498, "grad_norm": 0.3160805106163025, "learning_rate": 2.9550714453033498e-05, "loss": 0.0316, "step": 8078 }, { "epoch": 5.6774420238931835, "grad_norm": 0.1981748640537262, "learning_rate": 2.9550245959241042e-05, "loss": 0.0361, "step": 8079 }, { "epoch": 5.6781447645818695, "grad_norm": 0.29529714584350586, "learning_rate": 2.9549777465448586e-05, "loss": 0.0347, "step": 8080 }, { "epoch": 5.678847505270555, "grad_norm": 0.20472432672977448, "learning_rate": 2.9549308971656126e-05, "loss": 0.035, "step": 8081 }, { "epoch": 5.679550245959241, "grad_norm": 0.1369422972202301, "learning_rate": 2.954884047786367e-05, "loss": 0.0207, "step": 8082 }, { "epoch": 5.680252986647927, "grad_norm": 0.34324946999549866, "learning_rate": 2.9548371984071213e-05, "loss": 0.0534, "step": 8083 }, { "epoch": 5.680955727336613, "grad_norm": 0.3788290023803711, "learning_rate": 2.9547903490278757e-05, "loss": 0.0588, "step": 8084 }, { "epoch": 5.681658468025299, "grad_norm": 0.4293175935745239, "learning_rate": 2.9547434996486297e-05, "loss": 0.0849, "step": 8085 }, { "epoch": 5.682361208713985, "grad_norm": 0.3544159233570099, "learning_rate": 2.9546966502693838e-05, "loss": 0.0676, "step": 8086 }, { "epoch": 5.683063949402671, "grad_norm": 0.6720353364944458, "learning_rate": 2.954649800890138e-05, "loss": 0.1337, "step": 8087 }, { "epoch": 5.683766690091356, "grad_norm": 0.8122348189353943, "learning_rate": 2.9546029515108925e-05, "loss": 0.2134, "step": 8088 }, { "epoch": 5.684469430780042, "grad_norm": 1.6019657850265503, "learning_rate": 2.954556102131647e-05, "loss": 0.1928, "step": 8089 }, { "epoch": 5.685172171468728, "grad_norm": 1.2662230730056763, "learning_rate": 2.954509252752401e-05, "loss": 0.2637, "step": 8090 }, { "epoch": 5.685874912157414, "grad_norm": 0.27961546182632446, "learning_rate": 2.9544624033731553e-05, "loss": 0.0981, "step": 8091 }, { "epoch": 5.6865776528461, "grad_norm": 0.23123827576637268, "learning_rate": 2.9544155539939097e-05, "loss": 0.0348, "step": 8092 }, { "epoch": 5.6872803935347855, "grad_norm": 0.12997792661190033, "learning_rate": 2.954368704614664e-05, "loss": 0.027, "step": 8093 }, { "epoch": 5.6879831342234715, "grad_norm": 0.2687389850616455, "learning_rate": 2.954321855235418e-05, "loss": 0.0182, "step": 8094 }, { "epoch": 5.688685874912157, "grad_norm": 0.18591365218162537, "learning_rate": 2.9542750058561724e-05, "loss": 0.0214, "step": 8095 }, { "epoch": 5.689388615600843, "grad_norm": 0.155782550573349, "learning_rate": 2.9542281564769268e-05, "loss": 0.0189, "step": 8096 }, { "epoch": 5.690091356289529, "grad_norm": 0.4363710880279541, "learning_rate": 2.9541813070976812e-05, "loss": 0.0303, "step": 8097 }, { "epoch": 5.690794096978215, "grad_norm": 0.13677732646465302, "learning_rate": 2.9541344577184352e-05, "loss": 0.0217, "step": 8098 }, { "epoch": 5.691496837666901, "grad_norm": 0.2798653841018677, "learning_rate": 2.9540876083391896e-05, "loss": 0.0349, "step": 8099 }, { "epoch": 5.692199578355587, "grad_norm": 0.18300184607505798, "learning_rate": 2.954040758959944e-05, "loss": 0.0264, "step": 8100 }, { "epoch": 5.692902319044273, "grad_norm": 0.19262175261974335, "learning_rate": 2.9539939095806983e-05, "loss": 0.0362, "step": 8101 }, { "epoch": 5.693605059732959, "grad_norm": 0.1484457105398178, "learning_rate": 2.9539470602014527e-05, "loss": 0.0126, "step": 8102 }, { "epoch": 5.694307800421645, "grad_norm": 0.1104019284248352, "learning_rate": 2.9539002108222064e-05, "loss": 0.0175, "step": 8103 }, { "epoch": 5.695010541110331, "grad_norm": 0.18646904826164246, "learning_rate": 2.9538533614429608e-05, "loss": 0.0224, "step": 8104 }, { "epoch": 5.695713281799016, "grad_norm": 0.2086886167526245, "learning_rate": 2.953806512063715e-05, "loss": 0.0506, "step": 8105 }, { "epoch": 5.696416022487702, "grad_norm": 0.2475224882364273, "learning_rate": 2.9537596626844695e-05, "loss": 0.0332, "step": 8106 }, { "epoch": 5.6971187631763875, "grad_norm": 0.2663404643535614, "learning_rate": 2.9537128133052236e-05, "loss": 0.0227, "step": 8107 }, { "epoch": 5.6978215038650735, "grad_norm": 0.2564612030982971, "learning_rate": 2.953665963925978e-05, "loss": 0.0438, "step": 8108 }, { "epoch": 5.698524244553759, "grad_norm": 0.27662909030914307, "learning_rate": 2.9536191145467323e-05, "loss": 0.041, "step": 8109 }, { "epoch": 5.699226985242445, "grad_norm": 0.33676275610923767, "learning_rate": 2.9535722651674867e-05, "loss": 0.0922, "step": 8110 }, { "epoch": 5.699929725931131, "grad_norm": 0.3663691282272339, "learning_rate": 2.9535254157882407e-05, "loss": 0.0769, "step": 8111 }, { "epoch": 5.700632466619817, "grad_norm": 1.0098975896835327, "learning_rate": 2.953478566408995e-05, "loss": 0.1577, "step": 8112 }, { "epoch": 5.701335207308503, "grad_norm": 0.5957418084144592, "learning_rate": 2.9534317170297495e-05, "loss": 0.1455, "step": 8113 }, { "epoch": 5.702037947997189, "grad_norm": 2.3300390243530273, "learning_rate": 2.953384867650504e-05, "loss": 0.2471, "step": 8114 }, { "epoch": 5.702740688685875, "grad_norm": 1.6386076211929321, "learning_rate": 2.9533380182712582e-05, "loss": 0.267, "step": 8115 }, { "epoch": 5.703443429374561, "grad_norm": 0.22603581845760345, "learning_rate": 2.9532911688920122e-05, "loss": 0.0734, "step": 8116 }, { "epoch": 5.704146170063247, "grad_norm": 0.3209744691848755, "learning_rate": 2.9532443195127666e-05, "loss": 0.0365, "step": 8117 }, { "epoch": 5.704848910751933, "grad_norm": 0.13302886486053467, "learning_rate": 2.953197470133521e-05, "loss": 0.0346, "step": 8118 }, { "epoch": 5.705551651440619, "grad_norm": 0.25359776616096497, "learning_rate": 2.9531506207542754e-05, "loss": 0.0313, "step": 8119 }, { "epoch": 5.7062543921293045, "grad_norm": 0.14203760027885437, "learning_rate": 2.9531037713750294e-05, "loss": 0.022, "step": 8120 }, { "epoch": 5.70695713281799, "grad_norm": 0.2901654541492462, "learning_rate": 2.9530569219957834e-05, "loss": 0.0221, "step": 8121 }, { "epoch": 5.707659873506676, "grad_norm": 0.2789877951145172, "learning_rate": 2.9530100726165378e-05, "loss": 0.0286, "step": 8122 }, { "epoch": 5.708362614195362, "grad_norm": 0.15501874685287476, "learning_rate": 2.952963223237292e-05, "loss": 0.0227, "step": 8123 }, { "epoch": 5.709065354884048, "grad_norm": 0.1911403387784958, "learning_rate": 2.9529163738580462e-05, "loss": 0.0304, "step": 8124 }, { "epoch": 5.709768095572734, "grad_norm": 0.18237023055553436, "learning_rate": 2.9528695244788006e-05, "loss": 0.0142, "step": 8125 }, { "epoch": 5.710470836261419, "grad_norm": 0.1174076572060585, "learning_rate": 2.952822675099555e-05, "loss": 0.0168, "step": 8126 }, { "epoch": 5.711173576950105, "grad_norm": 0.1991131752729416, "learning_rate": 2.9527758257203093e-05, "loss": 0.0195, "step": 8127 }, { "epoch": 5.711876317638791, "grad_norm": 0.18480761349201202, "learning_rate": 2.9527289763410637e-05, "loss": 0.0324, "step": 8128 }, { "epoch": 5.712579058327477, "grad_norm": 0.24769343435764313, "learning_rate": 2.9526821269618177e-05, "loss": 0.028, "step": 8129 }, { "epoch": 5.713281799016163, "grad_norm": 0.2771133780479431, "learning_rate": 2.952635277582572e-05, "loss": 0.0307, "step": 8130 }, { "epoch": 5.713984539704849, "grad_norm": 0.19222834706306458, "learning_rate": 2.9525884282033265e-05, "loss": 0.0444, "step": 8131 }, { "epoch": 5.714687280393535, "grad_norm": 0.22555623948574066, "learning_rate": 2.952541578824081e-05, "loss": 0.0308, "step": 8132 }, { "epoch": 5.715390021082221, "grad_norm": 0.24018751084804535, "learning_rate": 2.952494729444835e-05, "loss": 0.0549, "step": 8133 }, { "epoch": 5.7160927617709065, "grad_norm": 0.31018564105033875, "learning_rate": 2.9524478800655892e-05, "loss": 0.0585, "step": 8134 }, { "epoch": 5.7167955024595924, "grad_norm": 0.26556190848350525, "learning_rate": 2.9524010306863436e-05, "loss": 0.0594, "step": 8135 }, { "epoch": 5.717498243148278, "grad_norm": 0.4069218337535858, "learning_rate": 2.952354181307098e-05, "loss": 0.0927, "step": 8136 }, { "epoch": 5.718200983836964, "grad_norm": 0.61546391248703, "learning_rate": 2.952307331927852e-05, "loss": 0.1226, "step": 8137 }, { "epoch": 5.71890372452565, "grad_norm": 1.2009824514389038, "learning_rate": 2.952260482548606e-05, "loss": 0.1693, "step": 8138 }, { "epoch": 5.719606465214336, "grad_norm": 0.8819776773452759, "learning_rate": 2.9522136331693604e-05, "loss": 0.2169, "step": 8139 }, { "epoch": 5.720309205903022, "grad_norm": 2.255311965942383, "learning_rate": 2.9521667837901148e-05, "loss": 0.2547, "step": 8140 }, { "epoch": 5.721011946591708, "grad_norm": 0.6974627375602722, "learning_rate": 2.9521199344108692e-05, "loss": 0.095, "step": 8141 }, { "epoch": 5.721714687280394, "grad_norm": 0.16770103573799133, "learning_rate": 2.9520730850316232e-05, "loss": 0.0308, "step": 8142 }, { "epoch": 5.722417427969079, "grad_norm": 0.1331307291984558, "learning_rate": 2.9520262356523776e-05, "loss": 0.026, "step": 8143 }, { "epoch": 5.723120168657765, "grad_norm": 0.2762196362018585, "learning_rate": 2.951979386273132e-05, "loss": 0.0265, "step": 8144 }, { "epoch": 5.723822909346451, "grad_norm": 0.2597278654575348, "learning_rate": 2.9519325368938863e-05, "loss": 0.0261, "step": 8145 }, { "epoch": 5.724525650035137, "grad_norm": 0.16200388967990875, "learning_rate": 2.9518856875146404e-05, "loss": 0.023, "step": 8146 }, { "epoch": 5.725228390723823, "grad_norm": 0.4326915740966797, "learning_rate": 2.9518388381353947e-05, "loss": 0.0203, "step": 8147 }, { "epoch": 5.7259311314125085, "grad_norm": 0.17181526124477386, "learning_rate": 2.951791988756149e-05, "loss": 0.0241, "step": 8148 }, { "epoch": 5.7266338721011945, "grad_norm": 0.16769343614578247, "learning_rate": 2.9517451393769035e-05, "loss": 0.0258, "step": 8149 }, { "epoch": 5.72733661278988, "grad_norm": 0.1855269819498062, "learning_rate": 2.9516982899976575e-05, "loss": 0.0425, "step": 8150 }, { "epoch": 5.728039353478566, "grad_norm": 0.20873185992240906, "learning_rate": 2.951651440618412e-05, "loss": 0.0424, "step": 8151 }, { "epoch": 5.728742094167252, "grad_norm": 0.1810465157032013, "learning_rate": 2.9516045912391663e-05, "loss": 0.0139, "step": 8152 }, { "epoch": 5.729444834855938, "grad_norm": 0.1978054791688919, "learning_rate": 2.9515577418599206e-05, "loss": 0.0479, "step": 8153 }, { "epoch": 5.730147575544624, "grad_norm": 0.18911117315292358, "learning_rate": 2.951510892480675e-05, "loss": 0.0275, "step": 8154 }, { "epoch": 5.73085031623331, "grad_norm": 0.44027936458587646, "learning_rate": 2.951464043101429e-05, "loss": 0.0328, "step": 8155 }, { "epoch": 5.731553056921996, "grad_norm": 0.4687618613243103, "learning_rate": 2.951417193722183e-05, "loss": 0.0514, "step": 8156 }, { "epoch": 5.732255797610682, "grad_norm": 0.3307898938655853, "learning_rate": 2.9513703443429374e-05, "loss": 0.024, "step": 8157 }, { "epoch": 5.732958538299368, "grad_norm": 0.4905763864517212, "learning_rate": 2.9513234949636918e-05, "loss": 0.0478, "step": 8158 }, { "epoch": 5.733661278988054, "grad_norm": 0.33018484711647034, "learning_rate": 2.951276645584446e-05, "loss": 0.0832, "step": 8159 }, { "epoch": 5.73436401967674, "grad_norm": 0.6151742339134216, "learning_rate": 2.9512297962052002e-05, "loss": 0.0655, "step": 8160 }, { "epoch": 5.7350667603654255, "grad_norm": 0.5213577151298523, "learning_rate": 2.9511829468259546e-05, "loss": 0.1167, "step": 8161 }, { "epoch": 5.735769501054111, "grad_norm": 0.5311123132705688, "learning_rate": 2.951136097446709e-05, "loss": 0.1366, "step": 8162 }, { "epoch": 5.736472241742797, "grad_norm": 0.794730544090271, "learning_rate": 2.951089248067463e-05, "loss": 0.1943, "step": 8163 }, { "epoch": 5.737174982431483, "grad_norm": 1.0429894924163818, "learning_rate": 2.9510423986882174e-05, "loss": 0.2292, "step": 8164 }, { "epoch": 5.737877723120168, "grad_norm": 0.93943852186203, "learning_rate": 2.9509955493089717e-05, "loss": 0.242, "step": 8165 }, { "epoch": 5.738580463808854, "grad_norm": 0.3830850124359131, "learning_rate": 2.950948699929726e-05, "loss": 0.0933, "step": 8166 }, { "epoch": 5.73928320449754, "grad_norm": 0.17178067564964294, "learning_rate": 2.9509018505504805e-05, "loss": 0.0307, "step": 8167 }, { "epoch": 5.739985945186226, "grad_norm": 0.1266220659017563, "learning_rate": 2.9508550011712345e-05, "loss": 0.0224, "step": 8168 }, { "epoch": 5.740688685874912, "grad_norm": 0.16530165076255798, "learning_rate": 2.950808151791989e-05, "loss": 0.0212, "step": 8169 }, { "epoch": 5.741391426563598, "grad_norm": 0.16182059049606323, "learning_rate": 2.9507613024127433e-05, "loss": 0.031, "step": 8170 }, { "epoch": 5.742094167252284, "grad_norm": 0.1662081927061081, "learning_rate": 2.9507144530334976e-05, "loss": 0.0187, "step": 8171 }, { "epoch": 5.74279690794097, "grad_norm": 0.2990858852863312, "learning_rate": 2.9506676036542517e-05, "loss": 0.0217, "step": 8172 }, { "epoch": 5.743499648629656, "grad_norm": 0.2653956413269043, "learning_rate": 2.9506207542750057e-05, "loss": 0.0327, "step": 8173 }, { "epoch": 5.744202389318342, "grad_norm": 0.129507914185524, "learning_rate": 2.95057390489576e-05, "loss": 0.0234, "step": 8174 }, { "epoch": 5.7449051300070275, "grad_norm": 0.23947173357009888, "learning_rate": 2.9505270555165145e-05, "loss": 0.021, "step": 8175 }, { "epoch": 5.745607870695713, "grad_norm": 0.20545506477355957, "learning_rate": 2.9504802061372685e-05, "loss": 0.0342, "step": 8176 }, { "epoch": 5.746310611384399, "grad_norm": 0.24609023332595825, "learning_rate": 2.950433356758023e-05, "loss": 0.0265, "step": 8177 }, { "epoch": 5.747013352073085, "grad_norm": 0.641899824142456, "learning_rate": 2.9503865073787772e-05, "loss": 0.0302, "step": 8178 }, { "epoch": 5.747716092761771, "grad_norm": 0.3004903495311737, "learning_rate": 2.9503396579995316e-05, "loss": 0.0284, "step": 8179 }, { "epoch": 5.748418833450457, "grad_norm": 0.217339888215065, "learning_rate": 2.950292808620286e-05, "loss": 0.0335, "step": 8180 }, { "epoch": 5.749121574139143, "grad_norm": 0.23240765929222107, "learning_rate": 2.95024595924104e-05, "loss": 0.03, "step": 8181 }, { "epoch": 5.749824314827828, "grad_norm": 0.1748507171869278, "learning_rate": 2.9501991098617944e-05, "loss": 0.0274, "step": 8182 }, { "epoch": 5.750527055516514, "grad_norm": 0.2660323977470398, "learning_rate": 2.9501522604825488e-05, "loss": 0.0291, "step": 8183 }, { "epoch": 5.7512297962052, "grad_norm": 0.27771058678627014, "learning_rate": 2.950105411103303e-05, "loss": 0.05, "step": 8184 }, { "epoch": 5.751932536893886, "grad_norm": 0.6034627556800842, "learning_rate": 2.950058561724057e-05, "loss": 0.0738, "step": 8185 }, { "epoch": 5.752635277582572, "grad_norm": 0.29542213678359985, "learning_rate": 2.9500117123448115e-05, "loss": 0.0812, "step": 8186 }, { "epoch": 5.753338018271258, "grad_norm": 0.7336392998695374, "learning_rate": 2.949964862965566e-05, "loss": 0.1278, "step": 8187 }, { "epoch": 5.754040758959944, "grad_norm": 1.4316681623458862, "learning_rate": 2.9499180135863203e-05, "loss": 0.2093, "step": 8188 }, { "epoch": 5.7547434996486295, "grad_norm": 0.8560253977775574, "learning_rate": 2.9498711642070743e-05, "loss": 0.2138, "step": 8189 }, { "epoch": 5.7554462403373154, "grad_norm": 1.2104336023330688, "learning_rate": 2.9498243148278287e-05, "loss": 0.2976, "step": 8190 }, { "epoch": 5.756148981026001, "grad_norm": 0.32119858264923096, "learning_rate": 2.9497774654485827e-05, "loss": 0.1004, "step": 8191 }, { "epoch": 5.756851721714687, "grad_norm": 0.14260661602020264, "learning_rate": 2.949730616069337e-05, "loss": 0.04, "step": 8192 }, { "epoch": 5.757554462403373, "grad_norm": 0.24841317534446716, "learning_rate": 2.9496837666900915e-05, "loss": 0.0316, "step": 8193 }, { "epoch": 5.758257203092059, "grad_norm": 0.1592891365289688, "learning_rate": 2.9496369173108455e-05, "loss": 0.0159, "step": 8194 }, { "epoch": 5.758959943780745, "grad_norm": 0.17221510410308838, "learning_rate": 2.9495900679316e-05, "loss": 0.0145, "step": 8195 }, { "epoch": 5.759662684469431, "grad_norm": 0.127010315656662, "learning_rate": 2.9495432185523542e-05, "loss": 0.0189, "step": 8196 }, { "epoch": 5.760365425158117, "grad_norm": 0.13629472255706787, "learning_rate": 2.9494963691731086e-05, "loss": 0.0197, "step": 8197 }, { "epoch": 5.761068165846803, "grad_norm": 0.1753169447183609, "learning_rate": 2.9494495197938627e-05, "loss": 0.0246, "step": 8198 }, { "epoch": 5.761770906535489, "grad_norm": 0.32403507828712463, "learning_rate": 2.949402670414617e-05, "loss": 0.0196, "step": 8199 }, { "epoch": 5.762473647224175, "grad_norm": 0.24784314632415771, "learning_rate": 2.9493558210353714e-05, "loss": 0.0223, "step": 8200 }, { "epoch": 5.763176387912861, "grad_norm": 0.17503085732460022, "learning_rate": 2.9493089716561258e-05, "loss": 0.0273, "step": 8201 }, { "epoch": 5.7638791286015465, "grad_norm": 0.15761125087738037, "learning_rate": 2.9492621222768798e-05, "loss": 0.0188, "step": 8202 }, { "epoch": 5.7645818692902315, "grad_norm": 0.24897879362106323, "learning_rate": 2.9492152728976342e-05, "loss": 0.0236, "step": 8203 }, { "epoch": 5.7652846099789175, "grad_norm": 0.4746558666229248, "learning_rate": 2.9491684235183885e-05, "loss": 0.0431, "step": 8204 }, { "epoch": 5.765987350667603, "grad_norm": 0.20626011490821838, "learning_rate": 2.949121574139143e-05, "loss": 0.0365, "step": 8205 }, { "epoch": 5.766690091356289, "grad_norm": 0.2884933352470398, "learning_rate": 2.9490747247598973e-05, "loss": 0.0331, "step": 8206 }, { "epoch": 5.767392832044975, "grad_norm": 0.38403037190437317, "learning_rate": 2.9490278753806513e-05, "loss": 0.0453, "step": 8207 }, { "epoch": 5.768095572733661, "grad_norm": 0.21519355475902557, "learning_rate": 2.9489810260014054e-05, "loss": 0.0544, "step": 8208 }, { "epoch": 5.768798313422347, "grad_norm": 0.27444642782211304, "learning_rate": 2.9489341766221597e-05, "loss": 0.0463, "step": 8209 }, { "epoch": 5.769501054111033, "grad_norm": 0.3173092007637024, "learning_rate": 2.948887327242914e-05, "loss": 0.0501, "step": 8210 }, { "epoch": 5.770203794799719, "grad_norm": 1.1043416261672974, "learning_rate": 2.948840477863668e-05, "loss": 0.0832, "step": 8211 }, { "epoch": 5.770906535488405, "grad_norm": 0.41407445073127747, "learning_rate": 2.9487936284844225e-05, "loss": 0.1416, "step": 8212 }, { "epoch": 5.771609276177091, "grad_norm": 0.6994579434394836, "learning_rate": 2.948746779105177e-05, "loss": 0.1434, "step": 8213 }, { "epoch": 5.772312016865777, "grad_norm": 1.5600908994674683, "learning_rate": 2.9486999297259313e-05, "loss": 0.2072, "step": 8214 }, { "epoch": 5.773014757554463, "grad_norm": 2.2134530544281006, "learning_rate": 2.9486530803466853e-05, "loss": 0.2885, "step": 8215 }, { "epoch": 5.7737174982431485, "grad_norm": 0.2891102135181427, "learning_rate": 2.9486062309674397e-05, "loss": 0.0686, "step": 8216 }, { "epoch": 5.774420238931834, "grad_norm": 0.19127093255519867, "learning_rate": 2.948559381588194e-05, "loss": 0.0282, "step": 8217 }, { "epoch": 5.77512297962052, "grad_norm": 0.16923308372497559, "learning_rate": 2.9485125322089484e-05, "loss": 0.0193, "step": 8218 }, { "epoch": 5.775825720309206, "grad_norm": 0.16317160427570343, "learning_rate": 2.9484656828297028e-05, "loss": 0.0247, "step": 8219 }, { "epoch": 5.776528460997891, "grad_norm": 0.1972634494304657, "learning_rate": 2.9484188334504568e-05, "loss": 0.0255, "step": 8220 }, { "epoch": 5.777231201686577, "grad_norm": 0.1808781623840332, "learning_rate": 2.9483719840712112e-05, "loss": 0.0153, "step": 8221 }, { "epoch": 5.777933942375263, "grad_norm": 0.14174652099609375, "learning_rate": 2.9483251346919656e-05, "loss": 0.0168, "step": 8222 }, { "epoch": 5.778636683063949, "grad_norm": 0.19245991110801697, "learning_rate": 2.94827828531272e-05, "loss": 0.0388, "step": 8223 }, { "epoch": 5.779339423752635, "grad_norm": 0.264769583940506, "learning_rate": 2.948231435933474e-05, "loss": 0.0301, "step": 8224 }, { "epoch": 5.780042164441321, "grad_norm": 0.11158327758312225, "learning_rate": 2.948184586554228e-05, "loss": 0.0126, "step": 8225 }, { "epoch": 5.780744905130007, "grad_norm": 0.2525594234466553, "learning_rate": 2.9481377371749824e-05, "loss": 0.0354, "step": 8226 }, { "epoch": 5.781447645818693, "grad_norm": 0.14408892393112183, "learning_rate": 2.9480908877957367e-05, "loss": 0.0229, "step": 8227 }, { "epoch": 5.782150386507379, "grad_norm": 1.152388334274292, "learning_rate": 2.948044038416491e-05, "loss": 0.0382, "step": 8228 }, { "epoch": 5.782853127196065, "grad_norm": 0.11382536590099335, "learning_rate": 2.947997189037245e-05, "loss": 0.0086, "step": 8229 }, { "epoch": 5.7835558678847505, "grad_norm": 0.20203684270381927, "learning_rate": 2.9479503396579995e-05, "loss": 0.0364, "step": 8230 }, { "epoch": 5.784258608573436, "grad_norm": 0.24937494099140167, "learning_rate": 2.947903490278754e-05, "loss": 0.0265, "step": 8231 }, { "epoch": 5.784961349262122, "grad_norm": 0.1480906903743744, "learning_rate": 2.9478566408995083e-05, "loss": 0.0253, "step": 8232 }, { "epoch": 5.785664089950808, "grad_norm": 0.22764241695404053, "learning_rate": 2.9478097915202623e-05, "loss": 0.0398, "step": 8233 }, { "epoch": 5.786366830639494, "grad_norm": 0.26890262961387634, "learning_rate": 2.9477629421410167e-05, "loss": 0.0647, "step": 8234 }, { "epoch": 5.78706957132818, "grad_norm": 0.34639114141464233, "learning_rate": 2.947716092761771e-05, "loss": 0.0659, "step": 8235 }, { "epoch": 5.787772312016866, "grad_norm": 0.2899589240550995, "learning_rate": 2.9476692433825254e-05, "loss": 0.0775, "step": 8236 }, { "epoch": 5.788475052705552, "grad_norm": 0.5088305473327637, "learning_rate": 2.9476223940032795e-05, "loss": 0.16, "step": 8237 }, { "epoch": 5.789177793394238, "grad_norm": 0.6843448877334595, "learning_rate": 2.9475755446240338e-05, "loss": 0.1851, "step": 8238 }, { "epoch": 5.789880534082924, "grad_norm": 0.6970930695533752, "learning_rate": 2.9475286952447882e-05, "loss": 0.2263, "step": 8239 }, { "epoch": 5.79058327477161, "grad_norm": 1.567733883857727, "learning_rate": 2.9474818458655426e-05, "loss": 0.2452, "step": 8240 }, { "epoch": 5.791286015460296, "grad_norm": 0.2765332758426666, "learning_rate": 2.9474349964862966e-05, "loss": 0.086, "step": 8241 }, { "epoch": 5.791988756148981, "grad_norm": 0.5056596398353577, "learning_rate": 2.947388147107051e-05, "loss": 0.0277, "step": 8242 }, { "epoch": 5.792691496837667, "grad_norm": 0.3351539969444275, "learning_rate": 2.947341297727805e-05, "loss": 0.0204, "step": 8243 }, { "epoch": 5.7933942375263525, "grad_norm": 0.14953657984733582, "learning_rate": 2.9472944483485594e-05, "loss": 0.0237, "step": 8244 }, { "epoch": 5.794096978215038, "grad_norm": 0.2692054808139801, "learning_rate": 2.9472475989693138e-05, "loss": 0.0257, "step": 8245 }, { "epoch": 5.794799718903724, "grad_norm": 0.19792459905147552, "learning_rate": 2.9472007495900678e-05, "loss": 0.0194, "step": 8246 }, { "epoch": 5.79550245959241, "grad_norm": 0.09299275279045105, "learning_rate": 2.947153900210822e-05, "loss": 0.0093, "step": 8247 }, { "epoch": 5.796205200281096, "grad_norm": 0.219338521361351, "learning_rate": 2.9471070508315765e-05, "loss": 0.0211, "step": 8248 }, { "epoch": 5.796907940969782, "grad_norm": 0.16696001589298248, "learning_rate": 2.947060201452331e-05, "loss": 0.0312, "step": 8249 }, { "epoch": 5.797610681658468, "grad_norm": 0.14342190325260162, "learning_rate": 2.947013352073085e-05, "loss": 0.0122, "step": 8250 }, { "epoch": 5.798313422347154, "grad_norm": 0.4249683618545532, "learning_rate": 2.9469665026938393e-05, "loss": 0.0327, "step": 8251 }, { "epoch": 5.79901616303584, "grad_norm": 0.15126661956310272, "learning_rate": 2.9469196533145937e-05, "loss": 0.0228, "step": 8252 }, { "epoch": 5.799718903724526, "grad_norm": 0.19943580031394958, "learning_rate": 2.946872803935348e-05, "loss": 0.0362, "step": 8253 }, { "epoch": 5.800421644413212, "grad_norm": 0.24757397174835205, "learning_rate": 2.9468259545561024e-05, "loss": 0.0265, "step": 8254 }, { "epoch": 5.801124385101898, "grad_norm": 0.23892687261104584, "learning_rate": 2.9467791051768565e-05, "loss": 0.0438, "step": 8255 }, { "epoch": 5.801827125790584, "grad_norm": 0.25927653908729553, "learning_rate": 2.946732255797611e-05, "loss": 0.0652, "step": 8256 }, { "epoch": 5.8025298664792695, "grad_norm": 0.2680898904800415, "learning_rate": 2.9466854064183652e-05, "loss": 0.0229, "step": 8257 }, { "epoch": 5.8032326071679545, "grad_norm": 0.1568201333284378, "learning_rate": 2.9466385570391196e-05, "loss": 0.0251, "step": 8258 }, { "epoch": 5.8039353478566404, "grad_norm": 0.2484547346830368, "learning_rate": 2.9465917076598736e-05, "loss": 0.0485, "step": 8259 }, { "epoch": 5.804638088545326, "grad_norm": 0.5761639475822449, "learning_rate": 2.9465448582806277e-05, "loss": 0.0554, "step": 8260 }, { "epoch": 5.805340829234012, "grad_norm": 0.9417275190353394, "learning_rate": 2.946498008901382e-05, "loss": 0.0874, "step": 8261 }, { "epoch": 5.806043569922698, "grad_norm": 0.7827412486076355, "learning_rate": 2.9464511595221364e-05, "loss": 0.1075, "step": 8262 }, { "epoch": 5.806746310611384, "grad_norm": 0.5844250321388245, "learning_rate": 2.9464043101428904e-05, "loss": 0.2406, "step": 8263 }, { "epoch": 5.80744905130007, "grad_norm": 0.8214805722236633, "learning_rate": 2.9463574607636448e-05, "loss": 0.2193, "step": 8264 }, { "epoch": 5.808151791988756, "grad_norm": 1.0052635669708252, "learning_rate": 2.9463106113843992e-05, "loss": 0.2451, "step": 8265 }, { "epoch": 5.808854532677442, "grad_norm": 0.3014949560165405, "learning_rate": 2.9462637620051535e-05, "loss": 0.0688, "step": 8266 }, { "epoch": 5.809557273366128, "grad_norm": 0.14099153876304626, "learning_rate": 2.946216912625908e-05, "loss": 0.0206, "step": 8267 }, { "epoch": 5.810260014054814, "grad_norm": 0.08831852674484253, "learning_rate": 2.946170063246662e-05, "loss": 0.0179, "step": 8268 }, { "epoch": 5.8109627547435, "grad_norm": 0.17552393674850464, "learning_rate": 2.9461232138674163e-05, "loss": 0.0216, "step": 8269 }, { "epoch": 5.811665495432186, "grad_norm": 0.2678379714488983, "learning_rate": 2.9460763644881707e-05, "loss": 0.0248, "step": 8270 }, { "epoch": 5.8123682361208715, "grad_norm": 0.1693163961172104, "learning_rate": 2.946029515108925e-05, "loss": 0.0214, "step": 8271 }, { "epoch": 5.813070976809557, "grad_norm": 0.1730494648218155, "learning_rate": 2.945982665729679e-05, "loss": 0.024, "step": 8272 }, { "epoch": 5.813773717498243, "grad_norm": 0.23770225048065186, "learning_rate": 2.9459358163504335e-05, "loss": 0.0288, "step": 8273 }, { "epoch": 5.814476458186929, "grad_norm": 0.4070156216621399, "learning_rate": 2.945888966971188e-05, "loss": 0.0459, "step": 8274 }, { "epoch": 5.815179198875615, "grad_norm": 0.21484258770942688, "learning_rate": 2.9458421175919422e-05, "loss": 0.0169, "step": 8275 }, { "epoch": 5.815881939564301, "grad_norm": 0.4405139088630676, "learning_rate": 2.9457952682126963e-05, "loss": 0.0435, "step": 8276 }, { "epoch": 5.816584680252987, "grad_norm": 0.12664629518985748, "learning_rate": 2.9457484188334506e-05, "loss": 0.0188, "step": 8277 }, { "epoch": 5.817287420941673, "grad_norm": 0.4292442500591278, "learning_rate": 2.9457015694542047e-05, "loss": 0.0302, "step": 8278 }, { "epoch": 5.817990161630359, "grad_norm": 0.25118133425712585, "learning_rate": 2.945654720074959e-05, "loss": 0.0281, "step": 8279 }, { "epoch": 5.818692902319044, "grad_norm": 0.21175959706306458, "learning_rate": 2.9456078706957134e-05, "loss": 0.0287, "step": 8280 }, { "epoch": 5.81939564300773, "grad_norm": 0.18322446942329407, "learning_rate": 2.9455610213164674e-05, "loss": 0.0333, "step": 8281 }, { "epoch": 5.820098383696416, "grad_norm": 0.3501269817352295, "learning_rate": 2.9455141719372218e-05, "loss": 0.0215, "step": 8282 }, { "epoch": 5.820801124385102, "grad_norm": 0.30065229535102844, "learning_rate": 2.9454673225579762e-05, "loss": 0.091, "step": 8283 }, { "epoch": 5.821503865073788, "grad_norm": 0.19983163475990295, "learning_rate": 2.9454204731787306e-05, "loss": 0.0394, "step": 8284 }, { "epoch": 5.8222066057624735, "grad_norm": 0.37098777294158936, "learning_rate": 2.9453736237994846e-05, "loss": 0.0541, "step": 8285 }, { "epoch": 5.822909346451159, "grad_norm": 0.36271369457244873, "learning_rate": 2.945326774420239e-05, "loss": 0.086, "step": 8286 }, { "epoch": 5.823612087139845, "grad_norm": 0.5809299349784851, "learning_rate": 2.9452799250409933e-05, "loss": 0.1317, "step": 8287 }, { "epoch": 5.824314827828531, "grad_norm": 0.44047069549560547, "learning_rate": 2.9452330756617477e-05, "loss": 0.1801, "step": 8288 }, { "epoch": 5.825017568517217, "grad_norm": 0.8981974124908447, "learning_rate": 2.9451862262825017e-05, "loss": 0.2282, "step": 8289 }, { "epoch": 5.825720309205903, "grad_norm": 2.3025643825531006, "learning_rate": 2.945139376903256e-05, "loss": 0.2722, "step": 8290 }, { "epoch": 5.826423049894589, "grad_norm": 0.25215429067611694, "learning_rate": 2.9450925275240105e-05, "loss": 0.0824, "step": 8291 }, { "epoch": 5.827125790583275, "grad_norm": 0.1657032072544098, "learning_rate": 2.945045678144765e-05, "loss": 0.0273, "step": 8292 }, { "epoch": 5.827828531271961, "grad_norm": 0.19542524218559265, "learning_rate": 2.9449988287655192e-05, "loss": 0.0254, "step": 8293 }, { "epoch": 5.828531271960647, "grad_norm": 0.18136759102344513, "learning_rate": 2.9449519793862733e-05, "loss": 0.0319, "step": 8294 }, { "epoch": 5.829234012649333, "grad_norm": 0.1780354082584381, "learning_rate": 2.9449051300070273e-05, "loss": 0.019, "step": 8295 }, { "epoch": 5.829936753338019, "grad_norm": 0.222296342253685, "learning_rate": 2.9448582806277817e-05, "loss": 0.0472, "step": 8296 }, { "epoch": 5.830639494026704, "grad_norm": 0.21081659197807312, "learning_rate": 2.944811431248536e-05, "loss": 0.0221, "step": 8297 }, { "epoch": 5.83134223471539, "grad_norm": 0.14468294382095337, "learning_rate": 2.94476458186929e-05, "loss": 0.018, "step": 8298 }, { "epoch": 5.8320449754040755, "grad_norm": 0.17037226259708405, "learning_rate": 2.9447177324900445e-05, "loss": 0.0259, "step": 8299 }, { "epoch": 5.832747716092761, "grad_norm": 0.24306431412696838, "learning_rate": 2.9446708831107988e-05, "loss": 0.0234, "step": 8300 }, { "epoch": 5.833450456781447, "grad_norm": 0.15599921345710754, "learning_rate": 2.9446240337315532e-05, "loss": 0.0282, "step": 8301 }, { "epoch": 5.834153197470133, "grad_norm": 0.15382109582424164, "learning_rate": 2.9445771843523072e-05, "loss": 0.0204, "step": 8302 }, { "epoch": 5.834855938158819, "grad_norm": 0.4284968376159668, "learning_rate": 2.9445303349730616e-05, "loss": 0.0373, "step": 8303 }, { "epoch": 5.835558678847505, "grad_norm": 0.2202412486076355, "learning_rate": 2.944483485593816e-05, "loss": 0.0256, "step": 8304 }, { "epoch": 5.836261419536191, "grad_norm": 0.318062424659729, "learning_rate": 2.9444366362145703e-05, "loss": 0.0316, "step": 8305 }, { "epoch": 5.836964160224877, "grad_norm": 0.31974899768829346, "learning_rate": 2.9443897868353247e-05, "loss": 0.0373, "step": 8306 }, { "epoch": 5.837666900913563, "grad_norm": 0.29258614778518677, "learning_rate": 2.9443429374560788e-05, "loss": 0.0269, "step": 8307 }, { "epoch": 5.838369641602249, "grad_norm": 0.23861229419708252, "learning_rate": 2.944296088076833e-05, "loss": 0.0614, "step": 8308 }, { "epoch": 5.839072382290935, "grad_norm": 0.22753751277923584, "learning_rate": 2.9442492386975875e-05, "loss": 0.0442, "step": 8309 }, { "epoch": 5.839775122979621, "grad_norm": 0.2900921702384949, "learning_rate": 2.944202389318342e-05, "loss": 0.084, "step": 8310 }, { "epoch": 5.840477863668307, "grad_norm": 0.554266095161438, "learning_rate": 2.944155539939096e-05, "loss": 0.0883, "step": 8311 }, { "epoch": 5.8411806043569925, "grad_norm": 0.46690553426742554, "learning_rate": 2.94410869055985e-05, "loss": 0.117, "step": 8312 }, { "epoch": 5.841883345045678, "grad_norm": 0.8286277055740356, "learning_rate": 2.9440618411806043e-05, "loss": 0.1854, "step": 8313 }, { "epoch": 5.842586085734364, "grad_norm": 0.9295296669006348, "learning_rate": 2.9440149918013587e-05, "loss": 0.1996, "step": 8314 }, { "epoch": 5.84328882642305, "grad_norm": 1.436177372932434, "learning_rate": 2.9439681424221127e-05, "loss": 0.2822, "step": 8315 }, { "epoch": 5.843991567111736, "grad_norm": 0.33873844146728516, "learning_rate": 2.943921293042867e-05, "loss": 0.0807, "step": 8316 }, { "epoch": 5.844694307800422, "grad_norm": 0.17212267220020294, "learning_rate": 2.9438744436636215e-05, "loss": 0.0297, "step": 8317 }, { "epoch": 5.845397048489108, "grad_norm": 0.1864657700061798, "learning_rate": 2.943827594284376e-05, "loss": 0.03, "step": 8318 }, { "epoch": 5.846099789177793, "grad_norm": 0.20213478803634644, "learning_rate": 2.9437807449051302e-05, "loss": 0.0361, "step": 8319 }, { "epoch": 5.846802529866479, "grad_norm": 0.19085703790187836, "learning_rate": 2.9437338955258842e-05, "loss": 0.0302, "step": 8320 }, { "epoch": 5.847505270555165, "grad_norm": 0.11113518476486206, "learning_rate": 2.9436870461466386e-05, "loss": 0.0089, "step": 8321 }, { "epoch": 5.848208011243851, "grad_norm": 0.20263168215751648, "learning_rate": 2.943640196767393e-05, "loss": 0.015, "step": 8322 }, { "epoch": 5.848910751932537, "grad_norm": 0.18979959189891815, "learning_rate": 2.9435933473881474e-05, "loss": 0.0236, "step": 8323 }, { "epoch": 5.849613492621223, "grad_norm": 0.15621574223041534, "learning_rate": 2.9435464980089014e-05, "loss": 0.0229, "step": 8324 }, { "epoch": 5.850316233309909, "grad_norm": 0.24746282398700714, "learning_rate": 2.9434996486296558e-05, "loss": 0.0465, "step": 8325 }, { "epoch": 5.8510189739985945, "grad_norm": 0.15242867171764374, "learning_rate": 2.94345279925041e-05, "loss": 0.0215, "step": 8326 }, { "epoch": 5.85172171468728, "grad_norm": 0.24054504930973053, "learning_rate": 2.9434059498711645e-05, "loss": 0.0207, "step": 8327 }, { "epoch": 5.852424455375966, "grad_norm": 0.29802605509757996, "learning_rate": 2.9433591004919185e-05, "loss": 0.0296, "step": 8328 }, { "epoch": 5.853127196064652, "grad_norm": 0.1762913465499878, "learning_rate": 2.943312251112673e-05, "loss": 0.0166, "step": 8329 }, { "epoch": 5.853829936753338, "grad_norm": 0.1954304277896881, "learning_rate": 2.943265401733427e-05, "loss": 0.0454, "step": 8330 }, { "epoch": 5.854532677442024, "grad_norm": 0.26877251267433167, "learning_rate": 2.9432185523541813e-05, "loss": 0.051, "step": 8331 }, { "epoch": 5.85523541813071, "grad_norm": 0.23065030574798584, "learning_rate": 2.9431717029749357e-05, "loss": 0.0321, "step": 8332 }, { "epoch": 5.855938158819396, "grad_norm": 0.4006798565387726, "learning_rate": 2.9431248535956897e-05, "loss": 0.0473, "step": 8333 }, { "epoch": 5.856640899508082, "grad_norm": 0.3639979064464569, "learning_rate": 2.943078004216444e-05, "loss": 0.0638, "step": 8334 }, { "epoch": 5.857343640196767, "grad_norm": 0.37036556005477905, "learning_rate": 2.9430311548371985e-05, "loss": 0.0674, "step": 8335 }, { "epoch": 5.858046380885453, "grad_norm": 0.6835847496986389, "learning_rate": 2.942984305457953e-05, "loss": 0.0829, "step": 8336 }, { "epoch": 5.858749121574139, "grad_norm": 0.9317834973335266, "learning_rate": 2.942937456078707e-05, "loss": 0.1753, "step": 8337 }, { "epoch": 5.859451862262825, "grad_norm": 0.7037867307662964, "learning_rate": 2.9428906066994613e-05, "loss": 0.1729, "step": 8338 }, { "epoch": 5.860154602951511, "grad_norm": 0.7358872890472412, "learning_rate": 2.9428437573202156e-05, "loss": 0.187, "step": 8339 }, { "epoch": 5.8608573436401965, "grad_norm": 1.5237032175064087, "learning_rate": 2.94279690794097e-05, "loss": 0.2895, "step": 8340 }, { "epoch": 5.861560084328882, "grad_norm": 0.2997601330280304, "learning_rate": 2.942750058561724e-05, "loss": 0.0738, "step": 8341 }, { "epoch": 5.862262825017568, "grad_norm": 1.5966498851776123, "learning_rate": 2.9427032091824784e-05, "loss": 0.0264, "step": 8342 }, { "epoch": 5.862965565706254, "grad_norm": 0.18000420928001404, "learning_rate": 2.9426563598032328e-05, "loss": 0.0397, "step": 8343 }, { "epoch": 5.86366830639494, "grad_norm": 0.17244896292686462, "learning_rate": 2.942609510423987e-05, "loss": 0.0373, "step": 8344 }, { "epoch": 5.864371047083626, "grad_norm": 0.15690724551677704, "learning_rate": 2.9425626610447415e-05, "loss": 0.0149, "step": 8345 }, { "epoch": 5.865073787772312, "grad_norm": 0.12480282783508301, "learning_rate": 2.9425158116654956e-05, "loss": 0.0269, "step": 8346 }, { "epoch": 5.865776528460998, "grad_norm": 0.15433306992053986, "learning_rate": 2.9424689622862496e-05, "loss": 0.0227, "step": 8347 }, { "epoch": 5.866479269149684, "grad_norm": 0.08915864676237106, "learning_rate": 2.942422112907004e-05, "loss": 0.0126, "step": 8348 }, { "epoch": 5.86718200983837, "grad_norm": 0.5591386556625366, "learning_rate": 2.9423752635277583e-05, "loss": 0.0433, "step": 8349 }, { "epoch": 5.867884750527056, "grad_norm": 0.23106923699378967, "learning_rate": 2.9423284141485124e-05, "loss": 0.0214, "step": 8350 }, { "epoch": 5.868587491215742, "grad_norm": 0.46509042382240295, "learning_rate": 2.9422815647692667e-05, "loss": 0.0264, "step": 8351 }, { "epoch": 5.869290231904428, "grad_norm": 0.22304004430770874, "learning_rate": 2.942234715390021e-05, "loss": 0.0141, "step": 8352 }, { "epoch": 5.8699929725931135, "grad_norm": 0.27980858087539673, "learning_rate": 2.9421878660107755e-05, "loss": 0.0344, "step": 8353 }, { "epoch": 5.870695713281799, "grad_norm": 0.14397020637989044, "learning_rate": 2.9421410166315295e-05, "loss": 0.0231, "step": 8354 }, { "epoch": 5.871398453970485, "grad_norm": 0.18814904987812042, "learning_rate": 2.942094167252284e-05, "loss": 0.0238, "step": 8355 }, { "epoch": 5.872101194659171, "grad_norm": 0.2574375867843628, "learning_rate": 2.9420473178730383e-05, "loss": 0.0467, "step": 8356 }, { "epoch": 5.872803935347856, "grad_norm": 0.18973411619663239, "learning_rate": 2.9420004684937926e-05, "loss": 0.0275, "step": 8357 }, { "epoch": 5.873506676036542, "grad_norm": 0.24622222781181335, "learning_rate": 2.941953619114547e-05, "loss": 0.0434, "step": 8358 }, { "epoch": 5.874209416725228, "grad_norm": 0.430070698261261, "learning_rate": 2.941906769735301e-05, "loss": 0.0378, "step": 8359 }, { "epoch": 5.874912157413914, "grad_norm": 0.35601097345352173, "learning_rate": 2.9418599203560554e-05, "loss": 0.0632, "step": 8360 }, { "epoch": 5.8756148981026, "grad_norm": 0.43596887588500977, "learning_rate": 2.9418130709768098e-05, "loss": 0.0758, "step": 8361 }, { "epoch": 5.876317638791286, "grad_norm": 0.4769582748413086, "learning_rate": 2.941766221597564e-05, "loss": 0.1218, "step": 8362 }, { "epoch": 5.877020379479972, "grad_norm": 0.8370792269706726, "learning_rate": 2.9417193722183182e-05, "loss": 0.1778, "step": 8363 }, { "epoch": 5.877723120168658, "grad_norm": 0.8940321803092957, "learning_rate": 2.9416725228390726e-05, "loss": 0.2292, "step": 8364 }, { "epoch": 5.878425860857344, "grad_norm": 2.250673770904541, "learning_rate": 2.9416256734598266e-05, "loss": 0.2233, "step": 8365 }, { "epoch": 5.87912860154603, "grad_norm": 0.18389618396759033, "learning_rate": 2.941578824080581e-05, "loss": 0.0547, "step": 8366 }, { "epoch": 5.8798313422347155, "grad_norm": 0.1721685230731964, "learning_rate": 2.941531974701335e-05, "loss": 0.0286, "step": 8367 }, { "epoch": 5.880534082923401, "grad_norm": 0.2027045041322708, "learning_rate": 2.9414851253220894e-05, "loss": 0.0407, "step": 8368 }, { "epoch": 5.881236823612087, "grad_norm": 0.3619556427001953, "learning_rate": 2.9414382759428438e-05, "loss": 0.0217, "step": 8369 }, { "epoch": 5.881939564300773, "grad_norm": 0.1468515396118164, "learning_rate": 2.941391426563598e-05, "loss": 0.0223, "step": 8370 }, { "epoch": 5.882642304989459, "grad_norm": 0.10420674085617065, "learning_rate": 2.9413445771843525e-05, "loss": 0.0143, "step": 8371 }, { "epoch": 5.883345045678145, "grad_norm": 0.23586806654930115, "learning_rate": 2.9412977278051065e-05, "loss": 0.0257, "step": 8372 }, { "epoch": 5.884047786366831, "grad_norm": 0.17038805782794952, "learning_rate": 2.941250878425861e-05, "loss": 0.0245, "step": 8373 }, { "epoch": 5.884750527055516, "grad_norm": 0.19038382172584534, "learning_rate": 2.9412040290466153e-05, "loss": 0.0274, "step": 8374 }, { "epoch": 5.885453267744202, "grad_norm": 0.11916481703519821, "learning_rate": 2.9411571796673696e-05, "loss": 0.0161, "step": 8375 }, { "epoch": 5.886156008432888, "grad_norm": 0.21241506934165955, "learning_rate": 2.9411103302881237e-05, "loss": 0.0569, "step": 8376 }, { "epoch": 5.886858749121574, "grad_norm": 0.32216373085975647, "learning_rate": 2.941063480908878e-05, "loss": 0.0252, "step": 8377 }, { "epoch": 5.88756148981026, "grad_norm": 0.23578676581382751, "learning_rate": 2.9410166315296324e-05, "loss": 0.0342, "step": 8378 }, { "epoch": 5.888264230498946, "grad_norm": 0.13172362744808197, "learning_rate": 2.9409697821503868e-05, "loss": 0.0257, "step": 8379 }, { "epoch": 5.888966971187632, "grad_norm": 0.2873116731643677, "learning_rate": 2.940922932771141e-05, "loss": 0.0424, "step": 8380 }, { "epoch": 5.8896697118763175, "grad_norm": 0.19227072596549988, "learning_rate": 2.9408760833918952e-05, "loss": 0.0374, "step": 8381 }, { "epoch": 5.890372452565003, "grad_norm": 0.16208332777023315, "learning_rate": 2.9408292340126492e-05, "loss": 0.0275, "step": 8382 }, { "epoch": 5.891075193253689, "grad_norm": 0.28155994415283203, "learning_rate": 2.9407823846334036e-05, "loss": 0.0514, "step": 8383 }, { "epoch": 5.891777933942375, "grad_norm": 0.386640727519989, "learning_rate": 2.940735535254158e-05, "loss": 0.058, "step": 8384 }, { "epoch": 5.892480674631061, "grad_norm": 0.3815712332725525, "learning_rate": 2.940688685874912e-05, "loss": 0.0722, "step": 8385 }, { "epoch": 5.893183415319747, "grad_norm": 0.5702530145645142, "learning_rate": 2.9406418364956664e-05, "loss": 0.0636, "step": 8386 }, { "epoch": 5.893886156008433, "grad_norm": 0.5120143890380859, "learning_rate": 2.9405949871164208e-05, "loss": 0.1357, "step": 8387 }, { "epoch": 5.894588896697119, "grad_norm": 0.5447353720664978, "learning_rate": 2.940548137737175e-05, "loss": 0.1993, "step": 8388 }, { "epoch": 5.895291637385805, "grad_norm": 0.6381738781929016, "learning_rate": 2.9405012883579292e-05, "loss": 0.1823, "step": 8389 }, { "epoch": 5.895994378074491, "grad_norm": 0.975430965423584, "learning_rate": 2.9404544389786835e-05, "loss": 0.3248, "step": 8390 }, { "epoch": 5.896697118763177, "grad_norm": 0.26884710788726807, "learning_rate": 2.940407589599438e-05, "loss": 0.0677, "step": 8391 }, { "epoch": 5.897399859451863, "grad_norm": 0.1595492660999298, "learning_rate": 2.9403607402201923e-05, "loss": 0.0238, "step": 8392 }, { "epoch": 5.8981026001405485, "grad_norm": 0.2437443882226944, "learning_rate": 2.9403138908409463e-05, "loss": 0.0352, "step": 8393 }, { "epoch": 5.8988053408292345, "grad_norm": 0.20445509254932404, "learning_rate": 2.9402670414617007e-05, "loss": 0.0194, "step": 8394 }, { "epoch": 5.8995080815179195, "grad_norm": 0.12687471508979797, "learning_rate": 2.940220192082455e-05, "loss": 0.0317, "step": 8395 }, { "epoch": 5.900210822206605, "grad_norm": 0.1481902003288269, "learning_rate": 2.9401733427032094e-05, "loss": 0.0211, "step": 8396 }, { "epoch": 5.900913562895291, "grad_norm": 0.1729431301355362, "learning_rate": 2.9401264933239638e-05, "loss": 0.0239, "step": 8397 }, { "epoch": 5.901616303583977, "grad_norm": 0.19241252541542053, "learning_rate": 2.940079643944718e-05, "loss": 0.0324, "step": 8398 }, { "epoch": 5.902319044272663, "grad_norm": 0.13497799634933472, "learning_rate": 2.9400327945654722e-05, "loss": 0.0307, "step": 8399 }, { "epoch": 5.903021784961349, "grad_norm": 0.1120583787560463, "learning_rate": 2.9399859451862263e-05, "loss": 0.0197, "step": 8400 }, { "epoch": 5.903724525650035, "grad_norm": 0.2272997349500656, "learning_rate": 2.9399390958069806e-05, "loss": 0.0238, "step": 8401 }, { "epoch": 5.904427266338721, "grad_norm": 0.14771126210689545, "learning_rate": 2.9398922464277347e-05, "loss": 0.0456, "step": 8402 }, { "epoch": 5.905130007027407, "grad_norm": 0.21223725378513336, "learning_rate": 2.939845397048489e-05, "loss": 0.0394, "step": 8403 }, { "epoch": 5.905832747716093, "grad_norm": 0.12910233438014984, "learning_rate": 2.9397985476692434e-05, "loss": 0.0202, "step": 8404 }, { "epoch": 5.906535488404779, "grad_norm": 0.1764489710330963, "learning_rate": 2.9397516982899978e-05, "loss": 0.0423, "step": 8405 }, { "epoch": 5.907238229093465, "grad_norm": 0.25541821122169495, "learning_rate": 2.9397048489107518e-05, "loss": 0.0444, "step": 8406 }, { "epoch": 5.9079409697821506, "grad_norm": 0.14189274609088898, "learning_rate": 2.9396579995315062e-05, "loss": 0.0192, "step": 8407 }, { "epoch": 5.9086437104708365, "grad_norm": 0.1792270988225937, "learning_rate": 2.9396111501522606e-05, "loss": 0.0445, "step": 8408 }, { "epoch": 5.909346451159522, "grad_norm": 0.21830914914608002, "learning_rate": 2.939564300773015e-05, "loss": 0.0365, "step": 8409 }, { "epoch": 5.910049191848208, "grad_norm": 0.3292257487773895, "learning_rate": 2.9395174513937693e-05, "loss": 0.0728, "step": 8410 }, { "epoch": 5.910751932536894, "grad_norm": 0.5770666003227234, "learning_rate": 2.9394706020145233e-05, "loss": 0.1419, "step": 8411 }, { "epoch": 5.911454673225579, "grad_norm": 0.6142613291740417, "learning_rate": 2.9394237526352777e-05, "loss": 0.1125, "step": 8412 }, { "epoch": 5.912157413914265, "grad_norm": 0.5461273789405823, "learning_rate": 2.939376903256032e-05, "loss": 0.1647, "step": 8413 }, { "epoch": 5.912860154602951, "grad_norm": 0.7400493621826172, "learning_rate": 2.9393300538767864e-05, "loss": 0.2244, "step": 8414 }, { "epoch": 5.913562895291637, "grad_norm": 1.8602402210235596, "learning_rate": 2.9392832044975405e-05, "loss": 0.2827, "step": 8415 }, { "epoch": 5.914265635980323, "grad_norm": 0.26445290446281433, "learning_rate": 2.939236355118295e-05, "loss": 0.0891, "step": 8416 }, { "epoch": 5.914968376669009, "grad_norm": 0.09221865981817245, "learning_rate": 2.939189505739049e-05, "loss": 0.0203, "step": 8417 }, { "epoch": 5.915671117357695, "grad_norm": 0.1376316398382187, "learning_rate": 2.9391426563598033e-05, "loss": 0.0233, "step": 8418 }, { "epoch": 5.916373858046381, "grad_norm": 0.15923160314559937, "learning_rate": 2.9390958069805573e-05, "loss": 0.0224, "step": 8419 }, { "epoch": 5.917076598735067, "grad_norm": 0.16923458874225616, "learning_rate": 2.9390489576013117e-05, "loss": 0.0289, "step": 8420 }, { "epoch": 5.917779339423753, "grad_norm": 0.24077707529067993, "learning_rate": 2.939002108222066e-05, "loss": 0.0145, "step": 8421 }, { "epoch": 5.9184820801124385, "grad_norm": 0.15188413858413696, "learning_rate": 2.9389552588428204e-05, "loss": 0.0216, "step": 8422 }, { "epoch": 5.919184820801124, "grad_norm": 0.14933520555496216, "learning_rate": 2.9389084094635748e-05, "loss": 0.0216, "step": 8423 }, { "epoch": 5.91988756148981, "grad_norm": 0.15555034577846527, "learning_rate": 2.9388615600843288e-05, "loss": 0.0264, "step": 8424 }, { "epoch": 5.920590302178496, "grad_norm": 0.18351124227046967, "learning_rate": 2.9388147107050832e-05, "loss": 0.0161, "step": 8425 }, { "epoch": 5.921293042867182, "grad_norm": 0.17033109068870544, "learning_rate": 2.9387678613258376e-05, "loss": 0.032, "step": 8426 }, { "epoch": 5.921995783555868, "grad_norm": 0.19322435557842255, "learning_rate": 2.938721011946592e-05, "loss": 0.0205, "step": 8427 }, { "epoch": 5.922698524244554, "grad_norm": 0.2522154450416565, "learning_rate": 2.938674162567346e-05, "loss": 0.0438, "step": 8428 }, { "epoch": 5.92340126493324, "grad_norm": 0.1408717930316925, "learning_rate": 2.9386273131881003e-05, "loss": 0.0182, "step": 8429 }, { "epoch": 5.924104005621926, "grad_norm": 0.19582882523536682, "learning_rate": 2.9385804638088547e-05, "loss": 0.0291, "step": 8430 }, { "epoch": 5.924806746310612, "grad_norm": 0.17487534880638123, "learning_rate": 2.938533614429609e-05, "loss": 0.0358, "step": 8431 }, { "epoch": 5.925509486999298, "grad_norm": 0.17906758189201355, "learning_rate": 2.938486765050363e-05, "loss": 0.0216, "step": 8432 }, { "epoch": 5.926212227687984, "grad_norm": 0.23714478313922882, "learning_rate": 2.9384399156711175e-05, "loss": 0.0431, "step": 8433 }, { "epoch": 5.926914968376669, "grad_norm": 0.361073762178421, "learning_rate": 2.9383930662918715e-05, "loss": 0.0545, "step": 8434 }, { "epoch": 5.927617709065355, "grad_norm": 0.6382583975791931, "learning_rate": 2.938346216912626e-05, "loss": 0.0739, "step": 8435 }, { "epoch": 5.9283204497540405, "grad_norm": 0.9763234853744507, "learning_rate": 2.9382993675333803e-05, "loss": 0.0957, "step": 8436 }, { "epoch": 5.929023190442726, "grad_norm": 1.8497620820999146, "learning_rate": 2.9382525181541343e-05, "loss": 0.129, "step": 8437 }, { "epoch": 5.929725931131412, "grad_norm": 0.6537178754806519, "learning_rate": 2.9382056687748887e-05, "loss": 0.1731, "step": 8438 }, { "epoch": 5.930428671820098, "grad_norm": 1.2665302753448486, "learning_rate": 2.938158819395643e-05, "loss": 0.229, "step": 8439 }, { "epoch": 5.931131412508784, "grad_norm": 1.8692309856414795, "learning_rate": 2.9381119700163974e-05, "loss": 0.3473, "step": 8440 }, { "epoch": 5.93183415319747, "grad_norm": 0.23401153087615967, "learning_rate": 2.9380651206371515e-05, "loss": 0.0717, "step": 8441 }, { "epoch": 5.932536893886156, "grad_norm": 0.15158236026763916, "learning_rate": 2.938018271257906e-05, "loss": 0.0299, "step": 8442 }, { "epoch": 5.933239634574842, "grad_norm": 0.18693974614143372, "learning_rate": 2.9379714218786602e-05, "loss": 0.0301, "step": 8443 }, { "epoch": 5.933942375263528, "grad_norm": 0.18015968799591064, "learning_rate": 2.9379245724994146e-05, "loss": 0.0241, "step": 8444 }, { "epoch": 5.934645115952214, "grad_norm": 0.14677387475967407, "learning_rate": 2.937877723120169e-05, "loss": 0.0422, "step": 8445 }, { "epoch": 5.9353478566409, "grad_norm": 0.10426366329193115, "learning_rate": 2.937830873740923e-05, "loss": 0.0123, "step": 8446 }, { "epoch": 5.936050597329586, "grad_norm": 0.20432454347610474, "learning_rate": 2.9377840243616774e-05, "loss": 0.0468, "step": 8447 }, { "epoch": 5.9367533380182715, "grad_norm": 0.13634289801120758, "learning_rate": 2.9377371749824317e-05, "loss": 0.0161, "step": 8448 }, { "epoch": 5.9374560787069575, "grad_norm": 0.13517339527606964, "learning_rate": 2.937690325603186e-05, "loss": 0.0284, "step": 8449 }, { "epoch": 5.938158819395643, "grad_norm": 0.14267376065254211, "learning_rate": 2.93764347622394e-05, "loss": 0.0221, "step": 8450 }, { "epoch": 5.938861560084328, "grad_norm": 0.26116493344306946, "learning_rate": 2.9375966268446945e-05, "loss": 0.0286, "step": 8451 }, { "epoch": 5.939564300773014, "grad_norm": 0.0800786018371582, "learning_rate": 2.9375497774654485e-05, "loss": 0.0053, "step": 8452 }, { "epoch": 5.9402670414617, "grad_norm": 0.1937081664800644, "learning_rate": 2.937502928086203e-05, "loss": 0.0261, "step": 8453 }, { "epoch": 5.940969782150386, "grad_norm": 0.2537393271923065, "learning_rate": 2.937456078706957e-05, "loss": 0.05, "step": 8454 }, { "epoch": 5.941672522839072, "grad_norm": 0.16655124723911285, "learning_rate": 2.9374092293277113e-05, "loss": 0.0198, "step": 8455 }, { "epoch": 5.942375263527758, "grad_norm": 0.25895971059799194, "learning_rate": 2.9373623799484657e-05, "loss": 0.0538, "step": 8456 }, { "epoch": 5.943078004216444, "grad_norm": 0.408674031496048, "learning_rate": 2.93731553056922e-05, "loss": 0.0629, "step": 8457 }, { "epoch": 5.94378074490513, "grad_norm": 0.21672241389751434, "learning_rate": 2.9372686811899744e-05, "loss": 0.0605, "step": 8458 }, { "epoch": 5.944483485593816, "grad_norm": 0.21910004317760468, "learning_rate": 2.9372218318107285e-05, "loss": 0.0461, "step": 8459 }, { "epoch": 5.945186226282502, "grad_norm": 1.1849733591079712, "learning_rate": 2.937174982431483e-05, "loss": 0.0707, "step": 8460 }, { "epoch": 5.945888966971188, "grad_norm": 0.41903188824653625, "learning_rate": 2.9371281330522372e-05, "loss": 0.091, "step": 8461 }, { "epoch": 5.9465917076598735, "grad_norm": 0.512885570526123, "learning_rate": 2.9370812836729916e-05, "loss": 0.1223, "step": 8462 }, { "epoch": 5.9472944483485595, "grad_norm": 0.5652691125869751, "learning_rate": 2.9370344342937456e-05, "loss": 0.1471, "step": 8463 }, { "epoch": 5.947997189037245, "grad_norm": 2.3628463745117188, "learning_rate": 2.9369875849145e-05, "loss": 0.2021, "step": 8464 }, { "epoch": 5.948699929725931, "grad_norm": 1.2093473672866821, "learning_rate": 2.9369407355352544e-05, "loss": 0.308, "step": 8465 }, { "epoch": 5.949402670414617, "grad_norm": 0.3711034655570984, "learning_rate": 2.9368938861560087e-05, "loss": 0.0911, "step": 8466 }, { "epoch": 5.950105411103303, "grad_norm": 0.25578975677490234, "learning_rate": 2.9368470367767628e-05, "loss": 0.0403, "step": 8467 }, { "epoch": 5.950808151791989, "grad_norm": 0.20450125634670258, "learning_rate": 2.936800187397517e-05, "loss": 0.0457, "step": 8468 }, { "epoch": 5.951510892480675, "grad_norm": 0.2402285635471344, "learning_rate": 2.9367533380182712e-05, "loss": 0.0174, "step": 8469 }, { "epoch": 5.952213633169361, "grad_norm": 0.30279088020324707, "learning_rate": 2.9367064886390256e-05, "loss": 0.0218, "step": 8470 }, { "epoch": 5.952916373858047, "grad_norm": 0.16320092976093292, "learning_rate": 2.93665963925978e-05, "loss": 0.0117, "step": 8471 }, { "epoch": 5.953619114546732, "grad_norm": 0.12547658383846283, "learning_rate": 2.936612789880534e-05, "loss": 0.0237, "step": 8472 }, { "epoch": 5.954321855235418, "grad_norm": 0.1540539413690567, "learning_rate": 2.9365659405012883e-05, "loss": 0.0505, "step": 8473 }, { "epoch": 5.955024595924104, "grad_norm": 0.22384406626224518, "learning_rate": 2.9365190911220427e-05, "loss": 0.0326, "step": 8474 }, { "epoch": 5.95572733661279, "grad_norm": 0.14462478458881378, "learning_rate": 2.936472241742797e-05, "loss": 0.0143, "step": 8475 }, { "epoch": 5.956430077301476, "grad_norm": 0.29698875546455383, "learning_rate": 2.936425392363551e-05, "loss": 0.0298, "step": 8476 }, { "epoch": 5.9571328179901615, "grad_norm": 0.10750545561313629, "learning_rate": 2.9363785429843055e-05, "loss": 0.011, "step": 8477 }, { "epoch": 5.957835558678847, "grad_norm": 0.24806496500968933, "learning_rate": 2.93633169360506e-05, "loss": 0.0442, "step": 8478 }, { "epoch": 5.958538299367533, "grad_norm": 0.1934739202260971, "learning_rate": 2.9362848442258142e-05, "loss": 0.0204, "step": 8479 }, { "epoch": 5.959241040056219, "grad_norm": 0.2837485373020172, "learning_rate": 2.9362379948465683e-05, "loss": 0.0345, "step": 8480 }, { "epoch": 5.959943780744905, "grad_norm": 0.25019609928131104, "learning_rate": 2.9361911454673226e-05, "loss": 0.0327, "step": 8481 }, { "epoch": 5.960646521433591, "grad_norm": 0.32521748542785645, "learning_rate": 2.936144296088077e-05, "loss": 0.0362, "step": 8482 }, { "epoch": 5.961349262122277, "grad_norm": 0.40909865498542786, "learning_rate": 2.9360974467088314e-05, "loss": 0.0521, "step": 8483 }, { "epoch": 5.962052002810963, "grad_norm": 0.2577090859413147, "learning_rate": 2.9360505973295857e-05, "loss": 0.041, "step": 8484 }, { "epoch": 5.962754743499649, "grad_norm": 0.32494986057281494, "learning_rate": 2.9360037479503398e-05, "loss": 0.0592, "step": 8485 }, { "epoch": 5.963457484188335, "grad_norm": 1.2149678468704224, "learning_rate": 2.935956898571094e-05, "loss": 0.1061, "step": 8486 }, { "epoch": 5.964160224877021, "grad_norm": 0.5098801851272583, "learning_rate": 2.9359100491918482e-05, "loss": 0.1151, "step": 8487 }, { "epoch": 5.964862965565707, "grad_norm": 0.9606742858886719, "learning_rate": 2.9358631998126026e-05, "loss": 0.1649, "step": 8488 }, { "epoch": 5.965565706254392, "grad_norm": 0.729182779788971, "learning_rate": 2.9358163504333566e-05, "loss": 0.2273, "step": 8489 }, { "epoch": 5.966268446943078, "grad_norm": 1.2698413133621216, "learning_rate": 2.935769501054111e-05, "loss": 0.2818, "step": 8490 }, { "epoch": 5.9669711876317635, "grad_norm": 0.2309785783290863, "learning_rate": 2.9357226516748653e-05, "loss": 0.0619, "step": 8491 }, { "epoch": 5.967673928320449, "grad_norm": 0.15790295600891113, "learning_rate": 2.9356758022956197e-05, "loss": 0.049, "step": 8492 }, { "epoch": 5.968376669009135, "grad_norm": 0.20886224508285522, "learning_rate": 2.9356289529163737e-05, "loss": 0.048, "step": 8493 }, { "epoch": 5.969079409697821, "grad_norm": 0.1562277227640152, "learning_rate": 2.935582103537128e-05, "loss": 0.0281, "step": 8494 }, { "epoch": 5.969782150386507, "grad_norm": 0.1698487102985382, "learning_rate": 2.9355352541578825e-05, "loss": 0.0245, "step": 8495 }, { "epoch": 5.970484891075193, "grad_norm": 0.20445850491523743, "learning_rate": 2.935488404778637e-05, "loss": 0.025, "step": 8496 }, { "epoch": 5.971187631763879, "grad_norm": 0.09979090839624405, "learning_rate": 2.9354415553993912e-05, "loss": 0.0176, "step": 8497 }, { "epoch": 5.971890372452565, "grad_norm": 0.14498738944530487, "learning_rate": 2.9353947060201453e-05, "loss": 0.0263, "step": 8498 }, { "epoch": 5.972593113141251, "grad_norm": 0.21358399093151093, "learning_rate": 2.9353478566408996e-05, "loss": 0.0218, "step": 8499 }, { "epoch": 5.973295853829937, "grad_norm": 0.12336172163486481, "learning_rate": 2.935301007261654e-05, "loss": 0.0226, "step": 8500 }, { "epoch": 5.973998594518623, "grad_norm": 0.10797759890556335, "learning_rate": 2.9352541578824084e-05, "loss": 0.0197, "step": 8501 }, { "epoch": 5.974701335207309, "grad_norm": 0.15097004175186157, "learning_rate": 2.9352073085031624e-05, "loss": 0.0158, "step": 8502 }, { "epoch": 5.9754040758959945, "grad_norm": 0.23141297698020935, "learning_rate": 2.9351604591239168e-05, "loss": 0.0569, "step": 8503 }, { "epoch": 5.9761068165846805, "grad_norm": 0.15415482223033905, "learning_rate": 2.9351136097446708e-05, "loss": 0.0169, "step": 8504 }, { "epoch": 5.976809557273366, "grad_norm": 0.17863504588603973, "learning_rate": 2.9350667603654252e-05, "loss": 0.0424, "step": 8505 }, { "epoch": 5.977512297962052, "grad_norm": 0.2548459470272064, "learning_rate": 2.9350199109861792e-05, "loss": 0.0403, "step": 8506 }, { "epoch": 5.978215038650738, "grad_norm": 0.18769031763076782, "learning_rate": 2.9349730616069336e-05, "loss": 0.0359, "step": 8507 }, { "epoch": 5.978917779339424, "grad_norm": 0.29811957478523254, "learning_rate": 2.934926212227688e-05, "loss": 0.0441, "step": 8508 }, { "epoch": 5.97962052002811, "grad_norm": 0.3421918749809265, "learning_rate": 2.9348793628484424e-05, "loss": 0.0601, "step": 8509 }, { "epoch": 5.980323260716796, "grad_norm": 0.7016676068305969, "learning_rate": 2.9348325134691967e-05, "loss": 0.0731, "step": 8510 }, { "epoch": 5.981026001405481, "grad_norm": 0.4125620126724243, "learning_rate": 2.9347856640899508e-05, "loss": 0.1121, "step": 8511 }, { "epoch": 5.981728742094167, "grad_norm": 0.47401052713394165, "learning_rate": 2.934738814710705e-05, "loss": 0.1358, "step": 8512 }, { "epoch": 5.982431482782853, "grad_norm": 0.6999518871307373, "learning_rate": 2.9346919653314595e-05, "loss": 0.186, "step": 8513 }, { "epoch": 5.983134223471539, "grad_norm": 0.8642088174819946, "learning_rate": 2.934645115952214e-05, "loss": 0.23, "step": 8514 }, { "epoch": 5.983836964160225, "grad_norm": 1.1686252355575562, "learning_rate": 2.934598266572968e-05, "loss": 0.287, "step": 8515 }, { "epoch": 5.984539704848911, "grad_norm": 0.26507145166397095, "learning_rate": 2.9345514171937223e-05, "loss": 0.0987, "step": 8516 }, { "epoch": 5.9852424455375965, "grad_norm": 0.28294646739959717, "learning_rate": 2.9345045678144767e-05, "loss": 0.0546, "step": 8517 }, { "epoch": 5.9859451862262825, "grad_norm": 0.15188397467136383, "learning_rate": 2.934457718435231e-05, "loss": 0.0279, "step": 8518 }, { "epoch": 5.986647926914968, "grad_norm": 0.42324772477149963, "learning_rate": 2.934410869055985e-05, "loss": 0.0147, "step": 8519 }, { "epoch": 5.987350667603654, "grad_norm": 0.14291729032993317, "learning_rate": 2.9343640196767394e-05, "loss": 0.0263, "step": 8520 }, { "epoch": 5.98805340829234, "grad_norm": 0.14036700129508972, "learning_rate": 2.9343171702974935e-05, "loss": 0.0192, "step": 8521 }, { "epoch": 5.988756148981026, "grad_norm": 0.2207208275794983, "learning_rate": 2.934270320918248e-05, "loss": 0.019, "step": 8522 }, { "epoch": 5.989458889669712, "grad_norm": 0.21532003581523895, "learning_rate": 2.9342234715390022e-05, "loss": 0.0315, "step": 8523 }, { "epoch": 5.990161630358398, "grad_norm": 0.19495797157287598, "learning_rate": 2.9341766221597562e-05, "loss": 0.019, "step": 8524 }, { "epoch": 5.990864371047084, "grad_norm": 0.15957222878932953, "learning_rate": 2.9341297727805106e-05, "loss": 0.0254, "step": 8525 }, { "epoch": 5.99156711173577, "grad_norm": 0.3940719664096832, "learning_rate": 2.934082923401265e-05, "loss": 0.0257, "step": 8526 }, { "epoch": 5.992269852424456, "grad_norm": 0.23593102395534515, "learning_rate": 2.9340360740220194e-05, "loss": 0.0262, "step": 8527 }, { "epoch": 5.992972593113141, "grad_norm": 0.18624362349510193, "learning_rate": 2.9339892246427734e-05, "loss": 0.029, "step": 8528 }, { "epoch": 5.993675333801827, "grad_norm": 0.21698759496212006, "learning_rate": 2.9339423752635278e-05, "loss": 0.0314, "step": 8529 }, { "epoch": 5.994378074490513, "grad_norm": 0.3490526080131531, "learning_rate": 2.933895525884282e-05, "loss": 0.0385, "step": 8530 }, { "epoch": 5.9950808151791986, "grad_norm": 0.21198342740535736, "learning_rate": 2.9338486765050365e-05, "loss": 0.0561, "step": 8531 }, { "epoch": 5.9957835558678845, "grad_norm": 0.3190322518348694, "learning_rate": 2.9338018271257905e-05, "loss": 0.0645, "step": 8532 }, { "epoch": 5.99648629655657, "grad_norm": 0.35092562437057495, "learning_rate": 2.933754977746545e-05, "loss": 0.0565, "step": 8533 }, { "epoch": 5.997189037245256, "grad_norm": 0.671768307685852, "learning_rate": 2.9337081283672993e-05, "loss": 0.1011, "step": 8534 }, { "epoch": 5.997891777933942, "grad_norm": 0.4513382613658905, "learning_rate": 2.9336612789880537e-05, "loss": 0.123, "step": 8535 }, { "epoch": 5.998594518622628, "grad_norm": 0.6761175394058228, "learning_rate": 2.933614429608808e-05, "loss": 0.2034, "step": 8536 }, { "epoch": 5.999297259311314, "grad_norm": 1.5300947427749634, "learning_rate": 2.933567580229562e-05, "loss": 0.2822, "step": 8537 }, { "epoch": 6.0, "grad_norm": 1.3084322214126587, "learning_rate": 2.9335207308503164e-05, "loss": 0.1704, "step": 8538 }, { "epoch": 6.000702740688686, "grad_norm": 0.2753283977508545, "learning_rate": 2.9334738814710705e-05, "loss": 0.0726, "step": 8539 }, { "epoch": 6.001405481377372, "grad_norm": 0.12004389613866806, "learning_rate": 2.933427032091825e-05, "loss": 0.0267, "step": 8540 }, { "epoch": 6.002108222066058, "grad_norm": 0.12339656054973602, "learning_rate": 2.933380182712579e-05, "loss": 0.0224, "step": 8541 }, { "epoch": 6.002810962754744, "grad_norm": 0.1962108612060547, "learning_rate": 2.9333333333333333e-05, "loss": 0.0298, "step": 8542 }, { "epoch": 6.00351370344343, "grad_norm": 0.22317840158939362, "learning_rate": 2.9332864839540876e-05, "loss": 0.0129, "step": 8543 }, { "epoch": 6.0042164441321155, "grad_norm": 0.20219257473945618, "learning_rate": 2.933239634574842e-05, "loss": 0.0394, "step": 8544 }, { "epoch": 6.0049191848208014, "grad_norm": 1.186840295791626, "learning_rate": 2.933192785195596e-05, "loss": 0.038, "step": 8545 }, { "epoch": 6.005621925509487, "grad_norm": 0.2581048309803009, "learning_rate": 2.9331459358163504e-05, "loss": 0.0123, "step": 8546 }, { "epoch": 6.006324666198173, "grad_norm": 0.2630707025527954, "learning_rate": 2.9330990864371048e-05, "loss": 0.0353, "step": 8547 }, { "epoch": 6.007027406886858, "grad_norm": 0.23641476035118103, "learning_rate": 2.933052237057859e-05, "loss": 0.06, "step": 8548 }, { "epoch": 6.007730147575544, "grad_norm": 0.23049449920654297, "learning_rate": 2.9330053876786135e-05, "loss": 0.0352, "step": 8549 }, { "epoch": 6.00843288826423, "grad_norm": 0.2630433738231659, "learning_rate": 2.9329585382993676e-05, "loss": 0.0192, "step": 8550 }, { "epoch": 6.009135628952916, "grad_norm": 0.2479848563671112, "learning_rate": 2.932911688920122e-05, "loss": 0.0396, "step": 8551 }, { "epoch": 6.009838369641602, "grad_norm": 0.1887999325990677, "learning_rate": 2.9328648395408763e-05, "loss": 0.0156, "step": 8552 }, { "epoch": 6.010541110330288, "grad_norm": 0.2793051302433014, "learning_rate": 2.9328179901616307e-05, "loss": 0.0612, "step": 8553 }, { "epoch": 6.011243851018974, "grad_norm": 0.17853178083896637, "learning_rate": 2.9327711407823847e-05, "loss": 0.0313, "step": 8554 }, { "epoch": 6.01194659170766, "grad_norm": 0.1633952111005783, "learning_rate": 2.932724291403139e-05, "loss": 0.0221, "step": 8555 }, { "epoch": 6.012649332396346, "grad_norm": 0.5639838576316833, "learning_rate": 2.932677442023893e-05, "loss": 0.0442, "step": 8556 }, { "epoch": 6.013352073085032, "grad_norm": 0.2744220197200775, "learning_rate": 2.9326305926446475e-05, "loss": 0.0438, "step": 8557 }, { "epoch": 6.0140548137737175, "grad_norm": 0.3505035638809204, "learning_rate": 2.9325837432654015e-05, "loss": 0.0606, "step": 8558 }, { "epoch": 6.0147575544624035, "grad_norm": 0.5425015687942505, "learning_rate": 2.932536893886156e-05, "loss": 0.0811, "step": 8559 }, { "epoch": 6.015460295151089, "grad_norm": 0.7479525804519653, "learning_rate": 2.9324900445069103e-05, "loss": 0.1331, "step": 8560 }, { "epoch": 6.016163035839775, "grad_norm": 2.654320478439331, "learning_rate": 2.9324431951276646e-05, "loss": 0.1658, "step": 8561 }, { "epoch": 6.016865776528461, "grad_norm": 1.3077099323272705, "learning_rate": 2.932396345748419e-05, "loss": 0.2146, "step": 8562 }, { "epoch": 6.017568517217147, "grad_norm": 1.2596824169158936, "learning_rate": 2.932349496369173e-05, "loss": 0.2392, "step": 8563 }, { "epoch": 6.018271257905833, "grad_norm": 0.34634679555892944, "learning_rate": 2.9323026469899274e-05, "loss": 0.081, "step": 8564 }, { "epoch": 6.018973998594519, "grad_norm": 0.17821893095970154, "learning_rate": 2.9322557976106818e-05, "loss": 0.0332, "step": 8565 }, { "epoch": 6.019676739283205, "grad_norm": 0.293690025806427, "learning_rate": 2.932208948231436e-05, "loss": 0.0317, "step": 8566 }, { "epoch": 6.02037947997189, "grad_norm": 0.17745526134967804, "learning_rate": 2.9321620988521902e-05, "loss": 0.026, "step": 8567 }, { "epoch": 6.021082220660576, "grad_norm": 0.10880807042121887, "learning_rate": 2.9321152494729446e-05, "loss": 0.0128, "step": 8568 }, { "epoch": 6.021784961349262, "grad_norm": 0.11025947332382202, "learning_rate": 2.932068400093699e-05, "loss": 0.0115, "step": 8569 }, { "epoch": 6.022487702037948, "grad_norm": 0.09953609853982925, "learning_rate": 2.9320215507144533e-05, "loss": 0.0127, "step": 8570 }, { "epoch": 6.023190442726634, "grad_norm": 0.10465884953737259, "learning_rate": 2.9319747013352073e-05, "loss": 0.0207, "step": 8571 }, { "epoch": 6.0238931834153195, "grad_norm": 0.16160985827445984, "learning_rate": 2.9319278519559617e-05, "loss": 0.0264, "step": 8572 }, { "epoch": 6.0245959241040055, "grad_norm": 0.11548969149589539, "learning_rate": 2.931881002576716e-05, "loss": 0.0113, "step": 8573 }, { "epoch": 6.025298664792691, "grad_norm": 0.13545988500118256, "learning_rate": 2.93183415319747e-05, "loss": 0.0246, "step": 8574 }, { "epoch": 6.026001405481377, "grad_norm": 0.2099190652370453, "learning_rate": 2.9317873038182245e-05, "loss": 0.0273, "step": 8575 }, { "epoch": 6.026704146170063, "grad_norm": 0.15764020383358002, "learning_rate": 2.9317404544389785e-05, "loss": 0.0433, "step": 8576 }, { "epoch": 6.027406886858749, "grad_norm": 0.16355903446674347, "learning_rate": 2.931693605059733e-05, "loss": 0.016, "step": 8577 }, { "epoch": 6.028109627547435, "grad_norm": 0.1959577202796936, "learning_rate": 2.9316467556804873e-05, "loss": 0.029, "step": 8578 }, { "epoch": 6.028812368236121, "grad_norm": 0.29137012362480164, "learning_rate": 2.9315999063012417e-05, "loss": 0.0302, "step": 8579 }, { "epoch": 6.029515108924807, "grad_norm": 0.19465254247188568, "learning_rate": 2.9315530569219957e-05, "loss": 0.0272, "step": 8580 }, { "epoch": 6.030217849613493, "grad_norm": 0.2751631736755371, "learning_rate": 2.93150620754275e-05, "loss": 0.0487, "step": 8581 }, { "epoch": 6.030920590302179, "grad_norm": 0.35231679677963257, "learning_rate": 2.9314593581635044e-05, "loss": 0.05, "step": 8582 }, { "epoch": 6.031623330990865, "grad_norm": 0.2873775362968445, "learning_rate": 2.9314125087842588e-05, "loss": 0.0425, "step": 8583 }, { "epoch": 6.032326071679551, "grad_norm": 0.3109924793243408, "learning_rate": 2.931365659405013e-05, "loss": 0.0701, "step": 8584 }, { "epoch": 6.0330288123682365, "grad_norm": 0.5996108055114746, "learning_rate": 2.9313188100257672e-05, "loss": 0.1128, "step": 8585 }, { "epoch": 6.033731553056922, "grad_norm": 0.6863152384757996, "learning_rate": 2.9312719606465216e-05, "loss": 0.1707, "step": 8586 }, { "epoch": 6.0344342937456075, "grad_norm": 3.682049036026001, "learning_rate": 2.931225111267276e-05, "loss": 0.2349, "step": 8587 }, { "epoch": 6.035137034434293, "grad_norm": 0.9765678644180298, "learning_rate": 2.9311782618880303e-05, "loss": 0.2592, "step": 8588 }, { "epoch": 6.035839775122979, "grad_norm": 0.28129008412361145, "learning_rate": 2.9311314125087844e-05, "loss": 0.0904, "step": 8589 }, { "epoch": 6.036542515811665, "grad_norm": 0.5495396852493286, "learning_rate": 2.9310845631295387e-05, "loss": 0.0258, "step": 8590 }, { "epoch": 6.037245256500351, "grad_norm": 0.12101740390062332, "learning_rate": 2.9310377137502928e-05, "loss": 0.0273, "step": 8591 }, { "epoch": 6.037947997189037, "grad_norm": 0.09000387787818909, "learning_rate": 2.930990864371047e-05, "loss": 0.0155, "step": 8592 }, { "epoch": 6.038650737877723, "grad_norm": 0.15167389810085297, "learning_rate": 2.9309440149918012e-05, "loss": 0.0132, "step": 8593 }, { "epoch": 6.039353478566409, "grad_norm": 0.2012782096862793, "learning_rate": 2.9308971656125555e-05, "loss": 0.0217, "step": 8594 }, { "epoch": 6.040056219255095, "grad_norm": 0.10548390448093414, "learning_rate": 2.93085031623331e-05, "loss": 0.0136, "step": 8595 }, { "epoch": 6.040758959943781, "grad_norm": 0.18931400775909424, "learning_rate": 2.9308034668540643e-05, "loss": 0.0191, "step": 8596 }, { "epoch": 6.041461700632467, "grad_norm": 0.1676899790763855, "learning_rate": 2.9307566174748183e-05, "loss": 0.0226, "step": 8597 }, { "epoch": 6.042164441321153, "grad_norm": 0.13611170649528503, "learning_rate": 2.9307097680955727e-05, "loss": 0.0196, "step": 8598 }, { "epoch": 6.0428671820098385, "grad_norm": 0.1335482895374298, "learning_rate": 2.930662918716327e-05, "loss": 0.0221, "step": 8599 }, { "epoch": 6.043569922698524, "grad_norm": 0.23471620678901672, "learning_rate": 2.9306160693370814e-05, "loss": 0.0208, "step": 8600 }, { "epoch": 6.04427266338721, "grad_norm": 0.3413837254047394, "learning_rate": 2.9305692199578358e-05, "loss": 0.039, "step": 8601 }, { "epoch": 6.044975404075896, "grad_norm": 0.1081121414899826, "learning_rate": 2.93052237057859e-05, "loss": 0.0142, "step": 8602 }, { "epoch": 6.045678144764582, "grad_norm": 0.1515457034111023, "learning_rate": 2.9304755211993442e-05, "loss": 0.0319, "step": 8603 }, { "epoch": 6.046380885453268, "grad_norm": 0.24431796371936798, "learning_rate": 2.9304286718200986e-05, "loss": 0.0253, "step": 8604 }, { "epoch": 6.047083626141954, "grad_norm": 0.3175743818283081, "learning_rate": 2.930381822440853e-05, "loss": 0.0372, "step": 8605 }, { "epoch": 6.047786366830639, "grad_norm": 0.18842622637748718, "learning_rate": 2.930334973061607e-05, "loss": 0.0243, "step": 8606 }, { "epoch": 6.048489107519325, "grad_norm": 0.3220403492450714, "learning_rate": 2.9302881236823614e-05, "loss": 0.0761, "step": 8607 }, { "epoch": 6.049191848208011, "grad_norm": 0.37170830368995667, "learning_rate": 2.9302412743031154e-05, "loss": 0.0444, "step": 8608 }, { "epoch": 6.049894588896697, "grad_norm": 0.400633305311203, "learning_rate": 2.9301944249238698e-05, "loss": 0.0715, "step": 8609 }, { "epoch": 6.050597329585383, "grad_norm": 0.6645165681838989, "learning_rate": 2.9301475755446238e-05, "loss": 0.1149, "step": 8610 }, { "epoch": 6.051300070274069, "grad_norm": 0.6505870819091797, "learning_rate": 2.9301007261653782e-05, "loss": 0.1738, "step": 8611 }, { "epoch": 6.052002810962755, "grad_norm": 0.9461361765861511, "learning_rate": 2.9300538767861326e-05, "loss": 0.2179, "step": 8612 }, { "epoch": 6.0527055516514405, "grad_norm": 1.0633375644683838, "learning_rate": 2.930007027406887e-05, "loss": 0.2557, "step": 8613 }, { "epoch": 6.0534082923401265, "grad_norm": 0.31917378306388855, "learning_rate": 2.9299601780276413e-05, "loss": 0.0881, "step": 8614 }, { "epoch": 6.054111033028812, "grad_norm": 0.16532044112682343, "learning_rate": 2.9299133286483953e-05, "loss": 0.0261, "step": 8615 }, { "epoch": 6.054813773717498, "grad_norm": 0.09261824190616608, "learning_rate": 2.9298664792691497e-05, "loss": 0.0163, "step": 8616 }, { "epoch": 6.055516514406184, "grad_norm": 0.12187018990516663, "learning_rate": 2.929819629889904e-05, "loss": 0.0213, "step": 8617 }, { "epoch": 6.05621925509487, "grad_norm": 0.11802177131175995, "learning_rate": 2.9297727805106585e-05, "loss": 0.0209, "step": 8618 }, { "epoch": 6.056921995783556, "grad_norm": 0.13925732672214508, "learning_rate": 2.9297259311314125e-05, "loss": 0.0153, "step": 8619 }, { "epoch": 6.057624736472242, "grad_norm": 0.13829033076763153, "learning_rate": 2.929679081752167e-05, "loss": 0.0221, "step": 8620 }, { "epoch": 6.058327477160928, "grad_norm": 0.2529347538948059, "learning_rate": 2.9296322323729212e-05, "loss": 0.0283, "step": 8621 }, { "epoch": 6.059030217849614, "grad_norm": 0.4372801184654236, "learning_rate": 2.9295853829936756e-05, "loss": 0.0205, "step": 8622 }, { "epoch": 6.0597329585383, "grad_norm": 0.23866280913352966, "learning_rate": 2.9295385336144296e-05, "loss": 0.0231, "step": 8623 }, { "epoch": 6.060435699226986, "grad_norm": 0.2769959568977356, "learning_rate": 2.929491684235184e-05, "loss": 0.0483, "step": 8624 }, { "epoch": 6.061138439915671, "grad_norm": 0.2087029665708542, "learning_rate": 2.9294448348559384e-05, "loss": 0.0244, "step": 8625 }, { "epoch": 6.061841180604357, "grad_norm": 0.19866439700126648, "learning_rate": 2.9293979854766924e-05, "loss": 0.0257, "step": 8626 }, { "epoch": 6.0625439212930425, "grad_norm": 0.14816924929618835, "learning_rate": 2.9293511360974468e-05, "loss": 0.0133, "step": 8627 }, { "epoch": 6.0632466619817285, "grad_norm": 0.1950886845588684, "learning_rate": 2.9293042867182008e-05, "loss": 0.0244, "step": 8628 }, { "epoch": 6.063949402670414, "grad_norm": 0.24736951291561127, "learning_rate": 2.9292574373389552e-05, "loss": 0.0299, "step": 8629 }, { "epoch": 6.0646521433591, "grad_norm": 0.38909077644348145, "learning_rate": 2.9292105879597096e-05, "loss": 0.0431, "step": 8630 }, { "epoch": 6.065354884047786, "grad_norm": 0.2135562151670456, "learning_rate": 2.929163738580464e-05, "loss": 0.0391, "step": 8631 }, { "epoch": 6.066057624736472, "grad_norm": 0.3326271176338196, "learning_rate": 2.929116889201218e-05, "loss": 0.0615, "step": 8632 }, { "epoch": 6.066760365425158, "grad_norm": 0.29215678572654724, "learning_rate": 2.9290700398219723e-05, "loss": 0.0517, "step": 8633 }, { "epoch": 6.067463106113844, "grad_norm": 0.49652889370918274, "learning_rate": 2.9290231904427267e-05, "loss": 0.0966, "step": 8634 }, { "epoch": 6.06816584680253, "grad_norm": 0.6004664897918701, "learning_rate": 2.928976341063481e-05, "loss": 0.1255, "step": 8635 }, { "epoch": 6.068868587491216, "grad_norm": 0.6467068791389465, "learning_rate": 2.928929491684235e-05, "loss": 0.1927, "step": 8636 }, { "epoch": 6.069571328179902, "grad_norm": 1.156914472579956, "learning_rate": 2.9288826423049895e-05, "loss": 0.2225, "step": 8637 }, { "epoch": 6.070274068868588, "grad_norm": 2.39579701423645, "learning_rate": 2.928835792925744e-05, "loss": 0.2257, "step": 8638 }, { "epoch": 6.070976809557274, "grad_norm": 0.3890579044818878, "learning_rate": 2.9287889435464982e-05, "loss": 0.084, "step": 8639 }, { "epoch": 6.0716795502459595, "grad_norm": 0.15577566623687744, "learning_rate": 2.9287420941672526e-05, "loss": 0.0373, "step": 8640 }, { "epoch": 6.072382290934645, "grad_norm": 0.13296882808208466, "learning_rate": 2.9286952447880066e-05, "loss": 0.0286, "step": 8641 }, { "epoch": 6.073085031623331, "grad_norm": 0.27493929862976074, "learning_rate": 2.928648395408761e-05, "loss": 0.0195, "step": 8642 }, { "epoch": 6.073787772312017, "grad_norm": 0.16038477420806885, "learning_rate": 2.928601546029515e-05, "loss": 0.0218, "step": 8643 }, { "epoch": 6.074490513000702, "grad_norm": 0.13256920874118805, "learning_rate": 2.9285546966502694e-05, "loss": 0.0106, "step": 8644 }, { "epoch": 6.075193253689388, "grad_norm": 0.2291109263896942, "learning_rate": 2.9285078472710235e-05, "loss": 0.0308, "step": 8645 }, { "epoch": 6.075895994378074, "grad_norm": 0.22289440035820007, "learning_rate": 2.928460997891778e-05, "loss": 0.026, "step": 8646 }, { "epoch": 6.07659873506676, "grad_norm": 0.4697003662586212, "learning_rate": 2.9284141485125322e-05, "loss": 0.026, "step": 8647 }, { "epoch": 6.077301475755446, "grad_norm": 0.21650861203670502, "learning_rate": 2.9283672991332866e-05, "loss": 0.0129, "step": 8648 }, { "epoch": 6.078004216444132, "grad_norm": 0.34003254771232605, "learning_rate": 2.928320449754041e-05, "loss": 0.0383, "step": 8649 }, { "epoch": 6.078706957132818, "grad_norm": 0.2063484787940979, "learning_rate": 2.928273600374795e-05, "loss": 0.0237, "step": 8650 }, { "epoch": 6.079409697821504, "grad_norm": 0.17370450496673584, "learning_rate": 2.9282267509955494e-05, "loss": 0.0251, "step": 8651 }, { "epoch": 6.08011243851019, "grad_norm": 0.14139650762081146, "learning_rate": 2.9281799016163037e-05, "loss": 0.0179, "step": 8652 }, { "epoch": 6.080815179198876, "grad_norm": 0.20655129849910736, "learning_rate": 2.928133052237058e-05, "loss": 0.0388, "step": 8653 }, { "epoch": 6.0815179198875615, "grad_norm": 0.2114216536283493, "learning_rate": 2.928086202857812e-05, "loss": 0.0486, "step": 8654 }, { "epoch": 6.082220660576247, "grad_norm": 0.34607821702957153, "learning_rate": 2.9280393534785665e-05, "loss": 0.0409, "step": 8655 }, { "epoch": 6.082923401264933, "grad_norm": 0.2187211960554123, "learning_rate": 2.927992504099321e-05, "loss": 0.033, "step": 8656 }, { "epoch": 6.083626141953619, "grad_norm": 0.2176651805639267, "learning_rate": 2.9279456547200753e-05, "loss": 0.063, "step": 8657 }, { "epoch": 6.084328882642305, "grad_norm": 0.2552397549152374, "learning_rate": 2.9278988053408293e-05, "loss": 0.0664, "step": 8658 }, { "epoch": 6.085031623330991, "grad_norm": 0.5585746169090271, "learning_rate": 2.9278519559615837e-05, "loss": 0.134, "step": 8659 }, { "epoch": 6.085734364019677, "grad_norm": 0.5469229221343994, "learning_rate": 2.927805106582338e-05, "loss": 0.1381, "step": 8660 }, { "epoch": 6.086437104708363, "grad_norm": 0.8201650381088257, "learning_rate": 2.927758257203092e-05, "loss": 0.2006, "step": 8661 }, { "epoch": 6.087139845397049, "grad_norm": 1.7918775081634521, "learning_rate": 2.9277114078238464e-05, "loss": 0.1993, "step": 8662 }, { "epoch": 6.087842586085735, "grad_norm": 1.0111839771270752, "learning_rate": 2.9276645584446005e-05, "loss": 0.2601, "step": 8663 }, { "epoch": 6.08854532677442, "grad_norm": 0.6554032564163208, "learning_rate": 2.927617709065355e-05, "loss": 0.0715, "step": 8664 }, { "epoch": 6.089248067463106, "grad_norm": 0.27749156951904297, "learning_rate": 2.9275708596861092e-05, "loss": 0.0359, "step": 8665 }, { "epoch": 6.089950808151792, "grad_norm": 0.17619061470031738, "learning_rate": 2.9275240103068636e-05, "loss": 0.0343, "step": 8666 }, { "epoch": 6.090653548840478, "grad_norm": 0.1321638971567154, "learning_rate": 2.9274771609276176e-05, "loss": 0.0182, "step": 8667 }, { "epoch": 6.0913562895291635, "grad_norm": 0.18743132054805756, "learning_rate": 2.927430311548372e-05, "loss": 0.0168, "step": 8668 }, { "epoch": 6.0920590302178494, "grad_norm": 0.1967185139656067, "learning_rate": 2.9273834621691264e-05, "loss": 0.0211, "step": 8669 }, { "epoch": 6.092761770906535, "grad_norm": 0.22849838435649872, "learning_rate": 2.9273366127898807e-05, "loss": 0.0274, "step": 8670 }, { "epoch": 6.093464511595221, "grad_norm": 0.29557332396507263, "learning_rate": 2.9272897634106348e-05, "loss": 0.0438, "step": 8671 }, { "epoch": 6.094167252283907, "grad_norm": 0.15454864501953125, "learning_rate": 2.927242914031389e-05, "loss": 0.0261, "step": 8672 }, { "epoch": 6.094869992972593, "grad_norm": 0.16158373653888702, "learning_rate": 2.9271960646521435e-05, "loss": 0.0173, "step": 8673 }, { "epoch": 6.095572733661279, "grad_norm": 0.1543400138616562, "learning_rate": 2.927149215272898e-05, "loss": 0.0237, "step": 8674 }, { "epoch": 6.096275474349965, "grad_norm": 0.1465783715248108, "learning_rate": 2.9271023658936523e-05, "loss": 0.0122, "step": 8675 }, { "epoch": 6.096978215038651, "grad_norm": 0.26423412561416626, "learning_rate": 2.9270555165144063e-05, "loss": 0.0308, "step": 8676 }, { "epoch": 6.097680955727337, "grad_norm": 0.14682333171367645, "learning_rate": 2.9270086671351607e-05, "loss": 0.0186, "step": 8677 }, { "epoch": 6.098383696416023, "grad_norm": 0.3026275634765625, "learning_rate": 2.9269618177559147e-05, "loss": 0.0374, "step": 8678 }, { "epoch": 6.099086437104709, "grad_norm": 0.5645096302032471, "learning_rate": 2.926914968376669e-05, "loss": 0.045, "step": 8679 }, { "epoch": 6.099789177793395, "grad_norm": 0.5789812207221985, "learning_rate": 2.926868118997423e-05, "loss": 0.0183, "step": 8680 }, { "epoch": 6.1004919184820805, "grad_norm": 0.29387930035591125, "learning_rate": 2.9268212696181775e-05, "loss": 0.0476, "step": 8681 }, { "epoch": 6.101194659170766, "grad_norm": 0.1860331892967224, "learning_rate": 2.926774420238932e-05, "loss": 0.0333, "step": 8682 }, { "epoch": 6.1018973998594515, "grad_norm": 0.40814122557640076, "learning_rate": 2.9267275708596862e-05, "loss": 0.0711, "step": 8683 }, { "epoch": 6.102600140548137, "grad_norm": 0.3242815136909485, "learning_rate": 2.9266807214804403e-05, "loss": 0.082, "step": 8684 }, { "epoch": 6.103302881236823, "grad_norm": 0.5293605327606201, "learning_rate": 2.9266338721011946e-05, "loss": 0.1079, "step": 8685 }, { "epoch": 6.104005621925509, "grad_norm": 0.6516544222831726, "learning_rate": 2.926587022721949e-05, "loss": 0.1824, "step": 8686 }, { "epoch": 6.104708362614195, "grad_norm": 0.8740196228027344, "learning_rate": 2.9265401733427034e-05, "loss": 0.2138, "step": 8687 }, { "epoch": 6.105411103302881, "grad_norm": 1.0515836477279663, "learning_rate": 2.9264933239634578e-05, "loss": 0.2367, "step": 8688 }, { "epoch": 6.106113843991567, "grad_norm": 0.2809036076068878, "learning_rate": 2.9264464745842118e-05, "loss": 0.0666, "step": 8689 }, { "epoch": 6.106816584680253, "grad_norm": 0.13313531875610352, "learning_rate": 2.926399625204966e-05, "loss": 0.0211, "step": 8690 }, { "epoch": 6.107519325368939, "grad_norm": 0.1442517340183258, "learning_rate": 2.9263527758257205e-05, "loss": 0.0185, "step": 8691 }, { "epoch": 6.108222066057625, "grad_norm": 0.21061426401138306, "learning_rate": 2.926305926446475e-05, "loss": 0.027, "step": 8692 }, { "epoch": 6.108924806746311, "grad_norm": 0.41989946365356445, "learning_rate": 2.926259077067229e-05, "loss": 0.0262, "step": 8693 }, { "epoch": 6.109627547434997, "grad_norm": 0.13959695398807526, "learning_rate": 2.9262122276879833e-05, "loss": 0.0118, "step": 8694 }, { "epoch": 6.1103302881236825, "grad_norm": 0.22544890642166138, "learning_rate": 2.9261653783087377e-05, "loss": 0.0133, "step": 8695 }, { "epoch": 6.111033028812368, "grad_norm": 0.3190053701400757, "learning_rate": 2.9261185289294917e-05, "loss": 0.0227, "step": 8696 }, { "epoch": 6.111735769501054, "grad_norm": 0.6826529502868652, "learning_rate": 2.9260716795502458e-05, "loss": 0.0242, "step": 8697 }, { "epoch": 6.11243851018974, "grad_norm": 0.19176152348518372, "learning_rate": 2.926024830171e-05, "loss": 0.0138, "step": 8698 }, { "epoch": 6.113141250878426, "grad_norm": 0.25330495834350586, "learning_rate": 2.9259779807917545e-05, "loss": 0.0235, "step": 8699 }, { "epoch": 6.113843991567112, "grad_norm": 0.2593213617801666, "learning_rate": 2.925931131412509e-05, "loss": 0.0252, "step": 8700 }, { "epoch": 6.114546732255798, "grad_norm": 0.2575039565563202, "learning_rate": 2.9258842820332632e-05, "loss": 0.0295, "step": 8701 }, { "epoch": 6.115249472944483, "grad_norm": 0.10837004333734512, "learning_rate": 2.9258374326540173e-05, "loss": 0.0095, "step": 8702 }, { "epoch": 6.115952213633169, "grad_norm": 0.20354770123958588, "learning_rate": 2.9257905832747716e-05, "loss": 0.0344, "step": 8703 }, { "epoch": 6.116654954321855, "grad_norm": 0.23829445242881775, "learning_rate": 2.925743733895526e-05, "loss": 0.0432, "step": 8704 }, { "epoch": 6.117357695010541, "grad_norm": 0.19473356008529663, "learning_rate": 2.9256968845162804e-05, "loss": 0.0261, "step": 8705 }, { "epoch": 6.118060435699227, "grad_norm": 0.2565990686416626, "learning_rate": 2.9256500351370344e-05, "loss": 0.0259, "step": 8706 }, { "epoch": 6.118763176387913, "grad_norm": 0.30934926867485046, "learning_rate": 2.9256031857577888e-05, "loss": 0.0548, "step": 8707 }, { "epoch": 6.119465917076599, "grad_norm": 0.5724221467971802, "learning_rate": 2.9255563363785432e-05, "loss": 0.0677, "step": 8708 }, { "epoch": 6.1201686577652845, "grad_norm": 0.3092375695705414, "learning_rate": 2.9255094869992975e-05, "loss": 0.1049, "step": 8709 }, { "epoch": 6.12087139845397, "grad_norm": 0.6206330060958862, "learning_rate": 2.9254626376200516e-05, "loss": 0.1288, "step": 8710 }, { "epoch": 6.121574139142656, "grad_norm": 0.8156816959381104, "learning_rate": 2.925415788240806e-05, "loss": 0.171, "step": 8711 }, { "epoch": 6.122276879831342, "grad_norm": 0.9441401362419128, "learning_rate": 2.9253689388615603e-05, "loss": 0.1947, "step": 8712 }, { "epoch": 6.122979620520028, "grad_norm": 2.736109495162964, "learning_rate": 2.9253220894823144e-05, "loss": 0.2405, "step": 8713 }, { "epoch": 6.123682361208714, "grad_norm": 0.35014238953590393, "learning_rate": 2.9252752401030687e-05, "loss": 0.0846, "step": 8714 }, { "epoch": 6.1243851018974, "grad_norm": 0.2132989764213562, "learning_rate": 2.9252283907238228e-05, "loss": 0.0347, "step": 8715 }, { "epoch": 6.125087842586086, "grad_norm": 0.16328172385692596, "learning_rate": 2.925181541344577e-05, "loss": 0.0277, "step": 8716 }, { "epoch": 6.125790583274772, "grad_norm": 0.3779911696910858, "learning_rate": 2.9251346919653315e-05, "loss": 0.0186, "step": 8717 }, { "epoch": 6.126493323963458, "grad_norm": 0.15923471748828888, "learning_rate": 2.925087842586086e-05, "loss": 0.0193, "step": 8718 }, { "epoch": 6.127196064652144, "grad_norm": 0.1525190770626068, "learning_rate": 2.92504099320684e-05, "loss": 0.0181, "step": 8719 }, { "epoch": 6.12789880534083, "grad_norm": 0.18183699250221252, "learning_rate": 2.9249941438275943e-05, "loss": 0.0192, "step": 8720 }, { "epoch": 6.128601546029515, "grad_norm": 0.1318463236093521, "learning_rate": 2.9249472944483487e-05, "loss": 0.0165, "step": 8721 }, { "epoch": 6.129304286718201, "grad_norm": 0.2519892752170563, "learning_rate": 2.924900445069103e-05, "loss": 0.0243, "step": 8722 }, { "epoch": 6.1300070274068865, "grad_norm": 0.5096836686134338, "learning_rate": 2.924853595689857e-05, "loss": 0.0216, "step": 8723 }, { "epoch": 6.130709768095572, "grad_norm": 0.34696924686431885, "learning_rate": 2.9248067463106114e-05, "loss": 0.0441, "step": 8724 }, { "epoch": 6.131412508784258, "grad_norm": 0.30051541328430176, "learning_rate": 2.9247598969313658e-05, "loss": 0.0212, "step": 8725 }, { "epoch": 6.132115249472944, "grad_norm": 0.39733782410621643, "learning_rate": 2.9247130475521202e-05, "loss": 0.0316, "step": 8726 }, { "epoch": 6.13281799016163, "grad_norm": 0.2972732484340668, "learning_rate": 2.9246661981728746e-05, "loss": 0.0221, "step": 8727 }, { "epoch": 6.133520730850316, "grad_norm": 0.1872839778661728, "learning_rate": 2.9246193487936286e-05, "loss": 0.032, "step": 8728 }, { "epoch": 6.134223471539002, "grad_norm": 0.44292494654655457, "learning_rate": 2.924572499414383e-05, "loss": 0.0643, "step": 8729 }, { "epoch": 6.134926212227688, "grad_norm": 0.17402130365371704, "learning_rate": 2.924525650035137e-05, "loss": 0.0202, "step": 8730 }, { "epoch": 6.135628952916374, "grad_norm": 0.359533429145813, "learning_rate": 2.9244788006558914e-05, "loss": 0.0375, "step": 8731 }, { "epoch": 6.13633169360506, "grad_norm": 0.47660014033317566, "learning_rate": 2.9244319512766454e-05, "loss": 0.0698, "step": 8732 }, { "epoch": 6.137034434293746, "grad_norm": 0.95984947681427, "learning_rate": 2.9243851018973998e-05, "loss": 0.0483, "step": 8733 }, { "epoch": 6.137737174982432, "grad_norm": 0.48818331956863403, "learning_rate": 2.924338252518154e-05, "loss": 0.0901, "step": 8734 }, { "epoch": 6.138439915671118, "grad_norm": 2.5441408157348633, "learning_rate": 2.9242914031389085e-05, "loss": 0.1077, "step": 8735 }, { "epoch": 6.1391426563598035, "grad_norm": 1.2302992343902588, "learning_rate": 2.9242445537596626e-05, "loss": 0.1847, "step": 8736 }, { "epoch": 6.139845397048489, "grad_norm": 1.4244977235794067, "learning_rate": 2.924197704380417e-05, "loss": 0.1803, "step": 8737 }, { "epoch": 6.140548137737175, "grad_norm": 1.7084025144577026, "learning_rate": 2.9241508550011713e-05, "loss": 0.2665, "step": 8738 }, { "epoch": 6.141250878425861, "grad_norm": 0.2907974421977997, "learning_rate": 2.9241040056219257e-05, "loss": 0.0553, "step": 8739 }, { "epoch": 6.141953619114547, "grad_norm": 0.1438477337360382, "learning_rate": 2.92405715624268e-05, "loss": 0.0203, "step": 8740 }, { "epoch": 6.142656359803232, "grad_norm": 0.17650660872459412, "learning_rate": 2.924010306863434e-05, "loss": 0.0282, "step": 8741 }, { "epoch": 6.143359100491918, "grad_norm": 0.17296403646469116, "learning_rate": 2.9239634574841884e-05, "loss": 0.0249, "step": 8742 }, { "epoch": 6.144061841180604, "grad_norm": 0.8065077066421509, "learning_rate": 2.9239166081049428e-05, "loss": 0.0186, "step": 8743 }, { "epoch": 6.14476458186929, "grad_norm": 0.11403334885835648, "learning_rate": 2.9238697587256972e-05, "loss": 0.0175, "step": 8744 }, { "epoch": 6.145467322557976, "grad_norm": 0.16580966114997864, "learning_rate": 2.9238229093464512e-05, "loss": 0.0184, "step": 8745 }, { "epoch": 6.146170063246662, "grad_norm": 0.39982545375823975, "learning_rate": 2.9237760599672056e-05, "loss": 0.0296, "step": 8746 }, { "epoch": 6.146872803935348, "grad_norm": 0.17685522139072418, "learning_rate": 2.92372921058796e-05, "loss": 0.0248, "step": 8747 }, { "epoch": 6.147575544624034, "grad_norm": 0.19506898522377014, "learning_rate": 2.923682361208714e-05, "loss": 0.0186, "step": 8748 }, { "epoch": 6.14827828531272, "grad_norm": 0.2115054428577423, "learning_rate": 2.923635511829468e-05, "loss": 0.029, "step": 8749 }, { "epoch": 6.1489810260014055, "grad_norm": 0.2060287445783615, "learning_rate": 2.9235886624502224e-05, "loss": 0.0222, "step": 8750 }, { "epoch": 6.149683766690091, "grad_norm": 0.26139193773269653, "learning_rate": 2.9235418130709768e-05, "loss": 0.0369, "step": 8751 }, { "epoch": 6.150386507378777, "grad_norm": 0.2251446694135666, "learning_rate": 2.923494963691731e-05, "loss": 0.0295, "step": 8752 }, { "epoch": 6.151089248067463, "grad_norm": 0.1838938444852829, "learning_rate": 2.9234481143124855e-05, "loss": 0.0274, "step": 8753 }, { "epoch": 6.151791988756149, "grad_norm": 0.27131447196006775, "learning_rate": 2.9234012649332396e-05, "loss": 0.0482, "step": 8754 }, { "epoch": 6.152494729444835, "grad_norm": 0.36404597759246826, "learning_rate": 2.923354415553994e-05, "loss": 0.0204, "step": 8755 }, { "epoch": 6.153197470133521, "grad_norm": 0.24541188776493073, "learning_rate": 2.9233075661747483e-05, "loss": 0.0336, "step": 8756 }, { "epoch": 6.153900210822207, "grad_norm": 0.30298179388046265, "learning_rate": 2.9232607167955027e-05, "loss": 0.0613, "step": 8757 }, { "epoch": 6.154602951510893, "grad_norm": 0.6948968172073364, "learning_rate": 2.9232138674162567e-05, "loss": 0.0474, "step": 8758 }, { "epoch": 6.155305692199578, "grad_norm": 0.5424656271934509, "learning_rate": 2.923167018037011e-05, "loss": 0.0952, "step": 8759 }, { "epoch": 6.156008432888264, "grad_norm": 0.5165116190910339, "learning_rate": 2.9231201686577655e-05, "loss": 0.1422, "step": 8760 }, { "epoch": 6.15671117357695, "grad_norm": 1.2932274341583252, "learning_rate": 2.92307331927852e-05, "loss": 0.1578, "step": 8761 }, { "epoch": 6.157413914265636, "grad_norm": 0.7610073089599609, "learning_rate": 2.923026469899274e-05, "loss": 0.173, "step": 8762 }, { "epoch": 6.158116654954322, "grad_norm": 1.4310868978500366, "learning_rate": 2.9229796205200282e-05, "loss": 0.2298, "step": 8763 }, { "epoch": 6.1588193956430075, "grad_norm": 0.23352526128292084, "learning_rate": 2.9229327711407826e-05, "loss": 0.0778, "step": 8764 }, { "epoch": 6.159522136331693, "grad_norm": 0.14228969812393188, "learning_rate": 2.9228859217615366e-05, "loss": 0.023, "step": 8765 }, { "epoch": 6.160224877020379, "grad_norm": 0.14156204462051392, "learning_rate": 2.922839072382291e-05, "loss": 0.0178, "step": 8766 }, { "epoch": 6.160927617709065, "grad_norm": 0.16830335557460785, "learning_rate": 2.922792223003045e-05, "loss": 0.0146, "step": 8767 }, { "epoch": 6.161630358397751, "grad_norm": 0.19112150371074677, "learning_rate": 2.9227453736237994e-05, "loss": 0.0141, "step": 8768 }, { "epoch": 6.162333099086437, "grad_norm": 0.1961035579442978, "learning_rate": 2.9226985242445538e-05, "loss": 0.0146, "step": 8769 }, { "epoch": 6.163035839775123, "grad_norm": 0.18329495191574097, "learning_rate": 2.922651674865308e-05, "loss": 0.0247, "step": 8770 }, { "epoch": 6.163738580463809, "grad_norm": 0.22027616202831268, "learning_rate": 2.9226048254860622e-05, "loss": 0.0318, "step": 8771 }, { "epoch": 6.164441321152495, "grad_norm": 0.17783164978027344, "learning_rate": 2.9225579761068166e-05, "loss": 0.0319, "step": 8772 }, { "epoch": 6.165144061841181, "grad_norm": 0.1200866624712944, "learning_rate": 2.922511126727571e-05, "loss": 0.0136, "step": 8773 }, { "epoch": 6.165846802529867, "grad_norm": 0.1331419199705124, "learning_rate": 2.9224642773483253e-05, "loss": 0.0239, "step": 8774 }, { "epoch": 6.166549543218553, "grad_norm": 0.16448989510536194, "learning_rate": 2.9224174279690794e-05, "loss": 0.0181, "step": 8775 }, { "epoch": 6.167252283907239, "grad_norm": 0.4830683171749115, "learning_rate": 2.9223705785898337e-05, "loss": 0.0467, "step": 8776 }, { "epoch": 6.1679550245959245, "grad_norm": 0.23734457790851593, "learning_rate": 2.922323729210588e-05, "loss": 0.0213, "step": 8777 }, { "epoch": 6.16865776528461, "grad_norm": 0.3366363048553467, "learning_rate": 2.9222768798313425e-05, "loss": 0.0273, "step": 8778 }, { "epoch": 6.169360505973295, "grad_norm": 0.2665209174156189, "learning_rate": 2.922230030452097e-05, "loss": 0.0444, "step": 8779 }, { "epoch": 6.170063246661981, "grad_norm": 0.5363764762878418, "learning_rate": 2.922183181072851e-05, "loss": 0.029, "step": 8780 }, { "epoch": 6.170765987350667, "grad_norm": 0.2332274466753006, "learning_rate": 2.9221363316936052e-05, "loss": 0.0358, "step": 8781 }, { "epoch": 6.171468728039353, "grad_norm": 1.326353907585144, "learning_rate": 2.9220894823143596e-05, "loss": 0.055, "step": 8782 }, { "epoch": 6.172171468728039, "grad_norm": 0.3108578324317932, "learning_rate": 2.9220426329351137e-05, "loss": 0.0634, "step": 8783 }, { "epoch": 6.172874209416725, "grad_norm": 0.5306452512741089, "learning_rate": 2.9219957835558677e-05, "loss": 0.0649, "step": 8784 }, { "epoch": 6.173576950105411, "grad_norm": 1.2351622581481934, "learning_rate": 2.921948934176622e-05, "loss": 0.1204, "step": 8785 }, { "epoch": 6.174279690794097, "grad_norm": 0.6836054921150208, "learning_rate": 2.9219020847973764e-05, "loss": 0.1825, "step": 8786 }, { "epoch": 6.174982431482783, "grad_norm": 0.8135836124420166, "learning_rate": 2.9218552354181308e-05, "loss": 0.2143, "step": 8787 }, { "epoch": 6.175685172171469, "grad_norm": 1.3690310716629028, "learning_rate": 2.921808386038885e-05, "loss": 0.2742, "step": 8788 }, { "epoch": 6.176387912860155, "grad_norm": 0.38693761825561523, "learning_rate": 2.9217615366596392e-05, "loss": 0.071, "step": 8789 }, { "epoch": 6.177090653548841, "grad_norm": 0.23308831453323364, "learning_rate": 2.9217146872803936e-05, "loss": 0.0334, "step": 8790 }, { "epoch": 6.1777933942375265, "grad_norm": 0.1549321711063385, "learning_rate": 2.921667837901148e-05, "loss": 0.0323, "step": 8791 }, { "epoch": 6.178496134926212, "grad_norm": 0.11939599364995956, "learning_rate": 2.9216209885219023e-05, "loss": 0.0171, "step": 8792 }, { "epoch": 6.179198875614898, "grad_norm": 0.14715375006198883, "learning_rate": 2.9215741391426564e-05, "loss": 0.0168, "step": 8793 }, { "epoch": 6.179901616303584, "grad_norm": 0.11388027667999268, "learning_rate": 2.9215272897634107e-05, "loss": 0.0136, "step": 8794 }, { "epoch": 6.18060435699227, "grad_norm": 0.08985606580972672, "learning_rate": 2.921480440384165e-05, "loss": 0.0144, "step": 8795 }, { "epoch": 6.181307097680956, "grad_norm": 0.1933836191892624, "learning_rate": 2.9214335910049195e-05, "loss": 0.0423, "step": 8796 }, { "epoch": 6.182009838369642, "grad_norm": 0.18270236253738403, "learning_rate": 2.9213867416256735e-05, "loss": 0.0207, "step": 8797 }, { "epoch": 6.182712579058327, "grad_norm": 0.2376062422990799, "learning_rate": 2.921339892246428e-05, "loss": 0.0193, "step": 8798 }, { "epoch": 6.183415319747013, "grad_norm": 0.2191123217344284, "learning_rate": 2.9212930428671823e-05, "loss": 0.0273, "step": 8799 }, { "epoch": 6.184118060435699, "grad_norm": 0.11903293430805206, "learning_rate": 2.9212461934879363e-05, "loss": 0.0159, "step": 8800 }, { "epoch": 6.184820801124385, "grad_norm": 0.17681282758712769, "learning_rate": 2.9211993441086903e-05, "loss": 0.0254, "step": 8801 }, { "epoch": 6.185523541813071, "grad_norm": 0.2367803156375885, "learning_rate": 2.9211524947294447e-05, "loss": 0.0308, "step": 8802 }, { "epoch": 6.186226282501757, "grad_norm": 0.20379000902175903, "learning_rate": 2.921105645350199e-05, "loss": 0.0491, "step": 8803 }, { "epoch": 6.186929023190443, "grad_norm": 0.15158763527870178, "learning_rate": 2.9210587959709534e-05, "loss": 0.0336, "step": 8804 }, { "epoch": 6.1876317638791285, "grad_norm": 0.19921378791332245, "learning_rate": 2.9210119465917078e-05, "loss": 0.0277, "step": 8805 }, { "epoch": 6.188334504567814, "grad_norm": 0.6096354126930237, "learning_rate": 2.920965097212462e-05, "loss": 0.039, "step": 8806 }, { "epoch": 6.1890372452565, "grad_norm": 0.26698729395866394, "learning_rate": 2.9209182478332162e-05, "loss": 0.044, "step": 8807 }, { "epoch": 6.189739985945186, "grad_norm": 0.26593858003616333, "learning_rate": 2.9208713984539706e-05, "loss": 0.0485, "step": 8808 }, { "epoch": 6.190442726633872, "grad_norm": 0.3491763770580292, "learning_rate": 2.920824549074725e-05, "loss": 0.0861, "step": 8809 }, { "epoch": 6.191145467322558, "grad_norm": 0.5122763514518738, "learning_rate": 2.920777699695479e-05, "loss": 0.123, "step": 8810 }, { "epoch": 6.191848208011244, "grad_norm": 0.6996469497680664, "learning_rate": 2.9207308503162334e-05, "loss": 0.1921, "step": 8811 }, { "epoch": 6.19255094869993, "grad_norm": 0.8694784045219421, "learning_rate": 2.9206840009369877e-05, "loss": 0.2127, "step": 8812 }, { "epoch": 6.193253689388616, "grad_norm": 1.0079591274261475, "learning_rate": 2.920637151557742e-05, "loss": 0.2768, "step": 8813 }, { "epoch": 6.193956430077302, "grad_norm": 0.17937912046909332, "learning_rate": 2.920590302178496e-05, "loss": 0.0625, "step": 8814 }, { "epoch": 6.194659170765988, "grad_norm": 0.1221587210893631, "learning_rate": 2.9205434527992505e-05, "loss": 0.0237, "step": 8815 }, { "epoch": 6.195361911454674, "grad_norm": 0.21780361235141754, "learning_rate": 2.920496603420005e-05, "loss": 0.0273, "step": 8816 }, { "epoch": 6.1960646521433596, "grad_norm": 0.13366177678108215, "learning_rate": 2.920449754040759e-05, "loss": 0.0254, "step": 8817 }, { "epoch": 6.196767392832045, "grad_norm": 0.1639167219400406, "learning_rate": 2.9204029046615133e-05, "loss": 0.022, "step": 8818 }, { "epoch": 6.1974701335207305, "grad_norm": 0.11951695382595062, "learning_rate": 2.9203560552822673e-05, "loss": 0.0115, "step": 8819 }, { "epoch": 6.198172874209416, "grad_norm": 0.2149103432893753, "learning_rate": 2.9203092059030217e-05, "loss": 0.0236, "step": 8820 }, { "epoch": 6.198875614898102, "grad_norm": 0.15055984258651733, "learning_rate": 2.920262356523776e-05, "loss": 0.0218, "step": 8821 }, { "epoch": 6.199578355586788, "grad_norm": 1.1254210472106934, "learning_rate": 2.9202155071445305e-05, "loss": 0.0287, "step": 8822 }, { "epoch": 6.200281096275474, "grad_norm": 0.1420840173959732, "learning_rate": 2.9201686577652845e-05, "loss": 0.0175, "step": 8823 }, { "epoch": 6.20098383696416, "grad_norm": 0.2031646966934204, "learning_rate": 2.920121808386039e-05, "loss": 0.026, "step": 8824 }, { "epoch": 6.201686577652846, "grad_norm": 0.16927020251750946, "learning_rate": 2.9200749590067932e-05, "loss": 0.0181, "step": 8825 }, { "epoch": 6.202389318341532, "grad_norm": 0.2616788446903229, "learning_rate": 2.9200281096275476e-05, "loss": 0.0243, "step": 8826 }, { "epoch": 6.203092059030218, "grad_norm": 0.28296077251434326, "learning_rate": 2.9199812602483016e-05, "loss": 0.0309, "step": 8827 }, { "epoch": 6.203794799718904, "grad_norm": 0.3733835220336914, "learning_rate": 2.919934410869056e-05, "loss": 0.0383, "step": 8828 }, { "epoch": 6.20449754040759, "grad_norm": 0.24226057529449463, "learning_rate": 2.9198875614898104e-05, "loss": 0.0271, "step": 8829 }, { "epoch": 6.205200281096276, "grad_norm": 0.1273815780878067, "learning_rate": 2.9198407121105648e-05, "loss": 0.0195, "step": 8830 }, { "epoch": 6.205903021784962, "grad_norm": 0.2681165933609009, "learning_rate": 2.919793862731319e-05, "loss": 0.0387, "step": 8831 }, { "epoch": 6.2066057624736475, "grad_norm": 0.3860936462879181, "learning_rate": 2.919747013352073e-05, "loss": 0.0461, "step": 8832 }, { "epoch": 6.207308503162333, "grad_norm": 0.2599535286426544, "learning_rate": 2.9197001639728275e-05, "loss": 0.0485, "step": 8833 }, { "epoch": 6.208011243851019, "grad_norm": 0.32053929567337036, "learning_rate": 2.919653314593582e-05, "loss": 0.0757, "step": 8834 }, { "epoch": 6.208713984539705, "grad_norm": 0.42770397663116455, "learning_rate": 2.919606465214336e-05, "loss": 0.1222, "step": 8835 }, { "epoch": 6.20941672522839, "grad_norm": 0.6798434853553772, "learning_rate": 2.91955961583509e-05, "loss": 0.199, "step": 8836 }, { "epoch": 6.210119465917076, "grad_norm": 0.8374067544937134, "learning_rate": 2.9195127664558444e-05, "loss": 0.2144, "step": 8837 }, { "epoch": 6.210822206605762, "grad_norm": 0.9560675024986267, "learning_rate": 2.9194659170765987e-05, "loss": 0.2343, "step": 8838 }, { "epoch": 6.211524947294448, "grad_norm": 0.21575863659381866, "learning_rate": 2.919419067697353e-05, "loss": 0.0683, "step": 8839 }, { "epoch": 6.212227687983134, "grad_norm": 0.15800103545188904, "learning_rate": 2.9193722183181075e-05, "loss": 0.0278, "step": 8840 }, { "epoch": 6.21293042867182, "grad_norm": 0.18928922712802887, "learning_rate": 2.9193253689388615e-05, "loss": 0.0389, "step": 8841 }, { "epoch": 6.213633169360506, "grad_norm": 0.19928891956806183, "learning_rate": 2.919278519559616e-05, "loss": 0.0157, "step": 8842 }, { "epoch": 6.214335910049192, "grad_norm": 0.15642495453357697, "learning_rate": 2.9192316701803702e-05, "loss": 0.017, "step": 8843 }, { "epoch": 6.215038650737878, "grad_norm": 0.13351476192474365, "learning_rate": 2.9191848208011246e-05, "loss": 0.014, "step": 8844 }, { "epoch": 6.215741391426564, "grad_norm": 0.205906942486763, "learning_rate": 2.9191379714218787e-05, "loss": 0.0285, "step": 8845 }, { "epoch": 6.2164441321152495, "grad_norm": 0.1618771255016327, "learning_rate": 2.919091122042633e-05, "loss": 0.0239, "step": 8846 }, { "epoch": 6.217146872803935, "grad_norm": 0.2422921508550644, "learning_rate": 2.9190442726633874e-05, "loss": 0.0279, "step": 8847 }, { "epoch": 6.217849613492621, "grad_norm": 0.08922659605741501, "learning_rate": 2.9189974232841418e-05, "loss": 0.0111, "step": 8848 }, { "epoch": 6.218552354181307, "grad_norm": 0.1412867158651352, "learning_rate": 2.9189505739048958e-05, "loss": 0.0282, "step": 8849 }, { "epoch": 6.219255094869993, "grad_norm": 0.47886648774147034, "learning_rate": 2.9189037245256502e-05, "loss": 0.0232, "step": 8850 }, { "epoch": 6.219957835558679, "grad_norm": 0.24583323299884796, "learning_rate": 2.9188568751464045e-05, "loss": 0.0518, "step": 8851 }, { "epoch": 6.220660576247365, "grad_norm": 0.20054326951503754, "learning_rate": 2.9188100257671586e-05, "loss": 0.0177, "step": 8852 }, { "epoch": 6.221363316936051, "grad_norm": 0.14673960208892822, "learning_rate": 2.918763176387913e-05, "loss": 0.0277, "step": 8853 }, { "epoch": 6.222066057624737, "grad_norm": 0.18379883468151093, "learning_rate": 2.918716327008667e-05, "loss": 0.0468, "step": 8854 }, { "epoch": 6.222768798313423, "grad_norm": 0.25179657340049744, "learning_rate": 2.9186694776294214e-05, "loss": 0.0297, "step": 8855 }, { "epoch": 6.223471539002108, "grad_norm": 0.4660525918006897, "learning_rate": 2.9186226282501757e-05, "loss": 0.0505, "step": 8856 }, { "epoch": 6.224174279690794, "grad_norm": 0.39529016613960266, "learning_rate": 2.91857577887093e-05, "loss": 0.0459, "step": 8857 }, { "epoch": 6.22487702037948, "grad_norm": 0.9232004284858704, "learning_rate": 2.918528929491684e-05, "loss": 0.0609, "step": 8858 }, { "epoch": 6.225579761068166, "grad_norm": 0.6438657641410828, "learning_rate": 2.9184820801124385e-05, "loss": 0.0964, "step": 8859 }, { "epoch": 6.2262825017568515, "grad_norm": 0.43680253624916077, "learning_rate": 2.918435230733193e-05, "loss": 0.1063, "step": 8860 }, { "epoch": 6.226985242445537, "grad_norm": 0.674060046672821, "learning_rate": 2.9183883813539473e-05, "loss": 0.1958, "step": 8861 }, { "epoch": 6.227687983134223, "grad_norm": 1.1841775178909302, "learning_rate": 2.9183415319747013e-05, "loss": 0.219, "step": 8862 }, { "epoch": 6.228390723822909, "grad_norm": 0.9745957851409912, "learning_rate": 2.9182946825954557e-05, "loss": 0.2348, "step": 8863 }, { "epoch": 6.229093464511595, "grad_norm": 0.26704180240631104, "learning_rate": 2.91824783321621e-05, "loss": 0.0799, "step": 8864 }, { "epoch": 6.229796205200281, "grad_norm": 0.3575249910354614, "learning_rate": 2.9182009838369644e-05, "loss": 0.0334, "step": 8865 }, { "epoch": 6.230498945888967, "grad_norm": 0.14126437902450562, "learning_rate": 2.9181541344577188e-05, "loss": 0.0157, "step": 8866 }, { "epoch": 6.231201686577653, "grad_norm": 0.22709673643112183, "learning_rate": 2.9181072850784728e-05, "loss": 0.0251, "step": 8867 }, { "epoch": 6.231904427266339, "grad_norm": 0.2248002141714096, "learning_rate": 2.9180604356992272e-05, "loss": 0.0214, "step": 8868 }, { "epoch": 6.232607167955025, "grad_norm": 0.15598425269126892, "learning_rate": 2.9180135863199816e-05, "loss": 0.0149, "step": 8869 }, { "epoch": 6.233309908643711, "grad_norm": 0.22090399265289307, "learning_rate": 2.9179667369407356e-05, "loss": 0.024, "step": 8870 }, { "epoch": 6.234012649332397, "grad_norm": 0.14084193110466003, "learning_rate": 2.9179198875614896e-05, "loss": 0.0182, "step": 8871 }, { "epoch": 6.2347153900210825, "grad_norm": 0.38690710067749023, "learning_rate": 2.917873038182244e-05, "loss": 0.0238, "step": 8872 }, { "epoch": 6.2354181307097685, "grad_norm": 0.1330224573612213, "learning_rate": 2.9178261888029984e-05, "loss": 0.0123, "step": 8873 }, { "epoch": 6.236120871398454, "grad_norm": 0.17109958827495575, "learning_rate": 2.9177793394237527e-05, "loss": 0.0255, "step": 8874 }, { "epoch": 6.236823612087139, "grad_norm": 0.5246396064758301, "learning_rate": 2.9177324900445068e-05, "loss": 0.0295, "step": 8875 }, { "epoch": 6.237526352775825, "grad_norm": 0.2270120531320572, "learning_rate": 2.917685640665261e-05, "loss": 0.0308, "step": 8876 }, { "epoch": 6.238229093464511, "grad_norm": 0.13733941316604614, "learning_rate": 2.9176387912860155e-05, "loss": 0.0193, "step": 8877 }, { "epoch": 6.238931834153197, "grad_norm": 0.30828580260276794, "learning_rate": 2.91759194190677e-05, "loss": 0.0591, "step": 8878 }, { "epoch": 6.239634574841883, "grad_norm": 0.2246301919221878, "learning_rate": 2.9175450925275243e-05, "loss": 0.0359, "step": 8879 }, { "epoch": 6.240337315530569, "grad_norm": 0.2424510270357132, "learning_rate": 2.9174982431482783e-05, "loss": 0.0312, "step": 8880 }, { "epoch": 6.241040056219255, "grad_norm": 0.24327604472637177, "learning_rate": 2.9174513937690327e-05, "loss": 0.0374, "step": 8881 }, { "epoch": 6.241742796907941, "grad_norm": 0.478083997964859, "learning_rate": 2.917404544389787e-05, "loss": 0.0518, "step": 8882 }, { "epoch": 6.242445537596627, "grad_norm": 0.28147420287132263, "learning_rate": 2.9173576950105414e-05, "loss": 0.034, "step": 8883 }, { "epoch": 6.243148278285313, "grad_norm": 0.4989408552646637, "learning_rate": 2.9173108456312955e-05, "loss": 0.0941, "step": 8884 }, { "epoch": 6.243851018973999, "grad_norm": 0.4987172782421112, "learning_rate": 2.9172639962520498e-05, "loss": 0.1055, "step": 8885 }, { "epoch": 6.2445537596626846, "grad_norm": 0.6970508694648743, "learning_rate": 2.9172171468728042e-05, "loss": 0.1922, "step": 8886 }, { "epoch": 6.2452565003513705, "grad_norm": 2.0215063095092773, "learning_rate": 2.9171702974935582e-05, "loss": 0.1995, "step": 8887 }, { "epoch": 6.245959241040056, "grad_norm": 2.5921692848205566, "learning_rate": 2.9171234481143123e-05, "loss": 0.2505, "step": 8888 }, { "epoch": 6.246661981728742, "grad_norm": 0.33402568101882935, "learning_rate": 2.9170765987350666e-05, "loss": 0.0708, "step": 8889 }, { "epoch": 6.247364722417428, "grad_norm": 0.2513526380062103, "learning_rate": 2.917029749355821e-05, "loss": 0.024, "step": 8890 }, { "epoch": 6.248067463106114, "grad_norm": 0.1274450272321701, "learning_rate": 2.9169828999765754e-05, "loss": 0.0246, "step": 8891 }, { "epoch": 6.2487702037948, "grad_norm": 0.1082114726305008, "learning_rate": 2.9169360505973298e-05, "loss": 0.0232, "step": 8892 }, { "epoch": 6.249472944483486, "grad_norm": 0.16112422943115234, "learning_rate": 2.9168892012180838e-05, "loss": 0.0186, "step": 8893 }, { "epoch": 6.250175685172172, "grad_norm": 0.13655529916286469, "learning_rate": 2.916842351838838e-05, "loss": 0.0166, "step": 8894 }, { "epoch": 6.250878425860857, "grad_norm": 0.1165986880660057, "learning_rate": 2.9167955024595925e-05, "loss": 0.0139, "step": 8895 }, { "epoch": 6.251581166549543, "grad_norm": 0.19579939544200897, "learning_rate": 2.916748653080347e-05, "loss": 0.0292, "step": 8896 }, { "epoch": 6.252283907238229, "grad_norm": 0.2913512885570526, "learning_rate": 2.916701803701101e-05, "loss": 0.0389, "step": 8897 }, { "epoch": 6.252986647926915, "grad_norm": 0.1942673921585083, "learning_rate": 2.9166549543218553e-05, "loss": 0.0132, "step": 8898 }, { "epoch": 6.253689388615601, "grad_norm": 0.5441602468490601, "learning_rate": 2.9166081049426097e-05, "loss": 0.0212, "step": 8899 }, { "epoch": 6.254392129304287, "grad_norm": 0.1006690114736557, "learning_rate": 2.916561255563364e-05, "loss": 0.0144, "step": 8900 }, { "epoch": 6.2550948699929725, "grad_norm": 0.28932663798332214, "learning_rate": 2.916514406184118e-05, "loss": 0.0258, "step": 8901 }, { "epoch": 6.255797610681658, "grad_norm": 0.1459120661020279, "learning_rate": 2.9164675568048725e-05, "loss": 0.0129, "step": 8902 }, { "epoch": 6.256500351370344, "grad_norm": 0.3177570700645447, "learning_rate": 2.916420707425627e-05, "loss": 0.032, "step": 8903 }, { "epoch": 6.25720309205903, "grad_norm": 0.3870832026004791, "learning_rate": 2.9163738580463812e-05, "loss": 0.0393, "step": 8904 }, { "epoch": 6.257905832747716, "grad_norm": 0.3475031852722168, "learning_rate": 2.9163270086671352e-05, "loss": 0.0231, "step": 8905 }, { "epoch": 6.258608573436402, "grad_norm": 0.23457133769989014, "learning_rate": 2.9162801592878893e-05, "loss": 0.0643, "step": 8906 }, { "epoch": 6.259311314125088, "grad_norm": 0.30385512113571167, "learning_rate": 2.9162333099086437e-05, "loss": 0.054, "step": 8907 }, { "epoch": 6.260014054813774, "grad_norm": 0.29206886887550354, "learning_rate": 2.916186460529398e-05, "loss": 0.0728, "step": 8908 }, { "epoch": 6.26071679550246, "grad_norm": 0.39973175525665283, "learning_rate": 2.9161396111501524e-05, "loss": 0.089, "step": 8909 }, { "epoch": 6.261419536191146, "grad_norm": 0.7540439367294312, "learning_rate": 2.9160927617709064e-05, "loss": 0.0884, "step": 8910 }, { "epoch": 6.262122276879832, "grad_norm": 0.5949935913085938, "learning_rate": 2.9160459123916608e-05, "loss": 0.1523, "step": 8911 }, { "epoch": 6.262825017568518, "grad_norm": 1.229131817817688, "learning_rate": 2.9159990630124152e-05, "loss": 0.251, "step": 8912 }, { "epoch": 6.263527758257203, "grad_norm": 1.1951364278793335, "learning_rate": 2.9159522136331695e-05, "loss": 0.2249, "step": 8913 }, { "epoch": 6.264230498945889, "grad_norm": 0.19617488980293274, "learning_rate": 2.9159053642539236e-05, "loss": 0.0661, "step": 8914 }, { "epoch": 6.2649332396345745, "grad_norm": 0.1562088131904602, "learning_rate": 2.915858514874678e-05, "loss": 0.0263, "step": 8915 }, { "epoch": 6.26563598032326, "grad_norm": 0.17618714272975922, "learning_rate": 2.9158116654954323e-05, "loss": 0.0245, "step": 8916 }, { "epoch": 6.266338721011946, "grad_norm": 0.09572435915470123, "learning_rate": 2.9157648161161867e-05, "loss": 0.0148, "step": 8917 }, { "epoch": 6.267041461700632, "grad_norm": 0.22318384051322937, "learning_rate": 2.915717966736941e-05, "loss": 0.0243, "step": 8918 }, { "epoch": 6.267744202389318, "grad_norm": 0.16543318331241608, "learning_rate": 2.915671117357695e-05, "loss": 0.0161, "step": 8919 }, { "epoch": 6.268446943078004, "grad_norm": 0.18176336586475372, "learning_rate": 2.9156242679784495e-05, "loss": 0.0241, "step": 8920 }, { "epoch": 6.26914968376669, "grad_norm": 0.10963647067546844, "learning_rate": 2.915577418599204e-05, "loss": 0.0203, "step": 8921 }, { "epoch": 6.269852424455376, "grad_norm": 0.15218959748744965, "learning_rate": 2.915530569219958e-05, "loss": 0.028, "step": 8922 }, { "epoch": 6.270555165144062, "grad_norm": 0.15213032066822052, "learning_rate": 2.915483719840712e-05, "loss": 0.015, "step": 8923 }, { "epoch": 6.271257905832748, "grad_norm": 0.23647800087928772, "learning_rate": 2.9154368704614663e-05, "loss": 0.0358, "step": 8924 }, { "epoch": 6.271960646521434, "grad_norm": 0.11922690272331238, "learning_rate": 2.9153900210822207e-05, "loss": 0.0113, "step": 8925 }, { "epoch": 6.27266338721012, "grad_norm": 0.16640757024288177, "learning_rate": 2.915343171702975e-05, "loss": 0.0213, "step": 8926 }, { "epoch": 6.2733661278988055, "grad_norm": 0.15533921122550964, "learning_rate": 2.915296322323729e-05, "loss": 0.0165, "step": 8927 }, { "epoch": 6.2740688685874915, "grad_norm": 0.40444323420524597, "learning_rate": 2.9152494729444834e-05, "loss": 0.039, "step": 8928 }, { "epoch": 6.274771609276177, "grad_norm": 0.3346625864505768, "learning_rate": 2.9152026235652378e-05, "loss": 0.0358, "step": 8929 }, { "epoch": 6.275474349964863, "grad_norm": 0.28542402386665344, "learning_rate": 2.9151557741859922e-05, "loss": 0.0202, "step": 8930 }, { "epoch": 6.276177090653549, "grad_norm": 0.27158740162849426, "learning_rate": 2.9151089248067466e-05, "loss": 0.0416, "step": 8931 }, { "epoch": 6.276879831342235, "grad_norm": 0.3268435597419739, "learning_rate": 2.9150620754275006e-05, "loss": 0.0449, "step": 8932 }, { "epoch": 6.27758257203092, "grad_norm": 0.40794745087623596, "learning_rate": 2.915015226048255e-05, "loss": 0.0721, "step": 8933 }, { "epoch": 6.278285312719606, "grad_norm": 0.3302583694458008, "learning_rate": 2.9149683766690093e-05, "loss": 0.0908, "step": 8934 }, { "epoch": 6.278988053408292, "grad_norm": 0.4258479177951813, "learning_rate": 2.9149215272897637e-05, "loss": 0.1146, "step": 8935 }, { "epoch": 6.279690794096978, "grad_norm": 0.5760331153869629, "learning_rate": 2.9148746779105177e-05, "loss": 0.1556, "step": 8936 }, { "epoch": 6.280393534785664, "grad_norm": 1.1255172491073608, "learning_rate": 2.914827828531272e-05, "loss": 0.1944, "step": 8937 }, { "epoch": 6.28109627547435, "grad_norm": 1.1435606479644775, "learning_rate": 2.9147809791520265e-05, "loss": 0.2978, "step": 8938 }, { "epoch": 6.281799016163036, "grad_norm": 0.3421432673931122, "learning_rate": 2.9147341297727805e-05, "loss": 0.08, "step": 8939 }, { "epoch": 6.282501756851722, "grad_norm": 0.17219214141368866, "learning_rate": 2.9146872803935346e-05, "loss": 0.0251, "step": 8940 }, { "epoch": 6.2832044975404076, "grad_norm": 0.17283621430397034, "learning_rate": 2.914640431014289e-05, "loss": 0.0394, "step": 8941 }, { "epoch": 6.2839072382290935, "grad_norm": 0.13826854526996613, "learning_rate": 2.9145935816350433e-05, "loss": 0.0278, "step": 8942 }, { "epoch": 6.284609978917779, "grad_norm": 0.16418547928333282, "learning_rate": 2.9145467322557977e-05, "loss": 0.0146, "step": 8943 }, { "epoch": 6.285312719606465, "grad_norm": 0.15200072526931763, "learning_rate": 2.914499882876552e-05, "loss": 0.0196, "step": 8944 }, { "epoch": 6.286015460295151, "grad_norm": 0.10765952616930008, "learning_rate": 2.914453033497306e-05, "loss": 0.0114, "step": 8945 }, { "epoch": 6.286718200983837, "grad_norm": 0.12548047304153442, "learning_rate": 2.9144061841180605e-05, "loss": 0.0184, "step": 8946 }, { "epoch": 6.287420941672523, "grad_norm": 0.2667371332645416, "learning_rate": 2.9143593347388148e-05, "loss": 0.0261, "step": 8947 }, { "epoch": 6.288123682361209, "grad_norm": 0.1277703046798706, "learning_rate": 2.9143124853595692e-05, "loss": 0.0172, "step": 8948 }, { "epoch": 6.288826423049895, "grad_norm": 0.26225516200065613, "learning_rate": 2.9142656359803232e-05, "loss": 0.0256, "step": 8949 }, { "epoch": 6.289529163738581, "grad_norm": 0.14307983219623566, "learning_rate": 2.9142187866010776e-05, "loss": 0.0092, "step": 8950 }, { "epoch": 6.290231904427266, "grad_norm": 0.25315582752227783, "learning_rate": 2.914171937221832e-05, "loss": 0.0318, "step": 8951 }, { "epoch": 6.290934645115952, "grad_norm": 0.20005065202713013, "learning_rate": 2.9141250878425863e-05, "loss": 0.0347, "step": 8952 }, { "epoch": 6.291637385804638, "grad_norm": 0.18811935186386108, "learning_rate": 2.9140782384633404e-05, "loss": 0.0245, "step": 8953 }, { "epoch": 6.292340126493324, "grad_norm": 0.34542855620384216, "learning_rate": 2.9140313890840948e-05, "loss": 0.0432, "step": 8954 }, { "epoch": 6.29304286718201, "grad_norm": 0.5204459428787231, "learning_rate": 2.913984539704849e-05, "loss": 0.066, "step": 8955 }, { "epoch": 6.2937456078706955, "grad_norm": 0.44457921385765076, "learning_rate": 2.9139376903256035e-05, "loss": 0.0526, "step": 8956 }, { "epoch": 6.294448348559381, "grad_norm": 0.34039637446403503, "learning_rate": 2.9138908409463575e-05, "loss": 0.0574, "step": 8957 }, { "epoch": 6.295151089248067, "grad_norm": 0.3132966160774231, "learning_rate": 2.9138439915671116e-05, "loss": 0.0463, "step": 8958 }, { "epoch": 6.295853829936753, "grad_norm": 0.657998263835907, "learning_rate": 2.913797142187866e-05, "loss": 0.085, "step": 8959 }, { "epoch": 6.296556570625439, "grad_norm": 0.4668331444263458, "learning_rate": 2.9137502928086203e-05, "loss": 0.1234, "step": 8960 }, { "epoch": 6.297259311314125, "grad_norm": 0.8943856954574585, "learning_rate": 2.9137034434293747e-05, "loss": 0.1601, "step": 8961 }, { "epoch": 6.297962052002811, "grad_norm": 0.684947669506073, "learning_rate": 2.9136565940501287e-05, "loss": 0.2075, "step": 8962 }, { "epoch": 6.298664792691497, "grad_norm": 1.3216817378997803, "learning_rate": 2.913609744670883e-05, "loss": 0.3135, "step": 8963 }, { "epoch": 6.299367533380183, "grad_norm": 0.14010588824748993, "learning_rate": 2.9135628952916375e-05, "loss": 0.0665, "step": 8964 }, { "epoch": 6.300070274068869, "grad_norm": 0.24410751461982727, "learning_rate": 2.913516045912392e-05, "loss": 0.0368, "step": 8965 }, { "epoch": 6.300773014757555, "grad_norm": 0.40744608640670776, "learning_rate": 2.913469196533146e-05, "loss": 0.0663, "step": 8966 }, { "epoch": 6.301475755446241, "grad_norm": 0.09232664108276367, "learning_rate": 2.9134223471539002e-05, "loss": 0.0108, "step": 8967 }, { "epoch": 6.3021784961349265, "grad_norm": 0.15066948533058167, "learning_rate": 2.9133754977746546e-05, "loss": 0.0235, "step": 8968 }, { "epoch": 6.3028812368236125, "grad_norm": 0.1970035582780838, "learning_rate": 2.913328648395409e-05, "loss": 0.0217, "step": 8969 }, { "epoch": 6.303583977512298, "grad_norm": 0.18033552169799805, "learning_rate": 2.9132817990161634e-05, "loss": 0.0286, "step": 8970 }, { "epoch": 6.304286718200984, "grad_norm": 0.2797258198261261, "learning_rate": 2.9132349496369174e-05, "loss": 0.0356, "step": 8971 }, { "epoch": 6.304989458889669, "grad_norm": 0.1126505434513092, "learning_rate": 2.9131881002576718e-05, "loss": 0.014, "step": 8972 }, { "epoch": 6.305692199578355, "grad_norm": 0.1771802008152008, "learning_rate": 2.913141250878426e-05, "loss": 0.0244, "step": 8973 }, { "epoch": 6.306394940267041, "grad_norm": 0.16800205409526825, "learning_rate": 2.9130944014991802e-05, "loss": 0.0305, "step": 8974 }, { "epoch": 6.307097680955727, "grad_norm": 0.17370137572288513, "learning_rate": 2.9130475521199342e-05, "loss": 0.0238, "step": 8975 }, { "epoch": 6.307800421644413, "grad_norm": 0.5978227853775024, "learning_rate": 2.9130007027406886e-05, "loss": 0.0481, "step": 8976 }, { "epoch": 6.308503162333099, "grad_norm": 0.17988665401935577, "learning_rate": 2.912953853361443e-05, "loss": 0.0206, "step": 8977 }, { "epoch": 6.309205903021785, "grad_norm": 0.2763790786266327, "learning_rate": 2.9129070039821973e-05, "loss": 0.0153, "step": 8978 }, { "epoch": 6.309908643710471, "grad_norm": 0.20909196138381958, "learning_rate": 2.9128601546029514e-05, "loss": 0.0362, "step": 8979 }, { "epoch": 6.310611384399157, "grad_norm": 0.17441652715206146, "learning_rate": 2.9128133052237057e-05, "loss": 0.0168, "step": 8980 }, { "epoch": 6.311314125087843, "grad_norm": 0.3636568784713745, "learning_rate": 2.91276645584446e-05, "loss": 0.0651, "step": 8981 }, { "epoch": 6.3120168657765285, "grad_norm": 0.25037893652915955, "learning_rate": 2.9127196064652145e-05, "loss": 0.0365, "step": 8982 }, { "epoch": 6.3127196064652145, "grad_norm": 0.38159453868865967, "learning_rate": 2.912672757085969e-05, "loss": 0.0675, "step": 8983 }, { "epoch": 6.3134223471539, "grad_norm": 0.3461112082004547, "learning_rate": 2.912625907706723e-05, "loss": 0.0804, "step": 8984 }, { "epoch": 6.314125087842586, "grad_norm": 1.8649990558624268, "learning_rate": 2.9125790583274773e-05, "loss": 0.1149, "step": 8985 }, { "epoch": 6.314827828531272, "grad_norm": 0.7070709466934204, "learning_rate": 2.9125322089482316e-05, "loss": 0.1469, "step": 8986 }, { "epoch": 6.315530569219958, "grad_norm": 1.0318795442581177, "learning_rate": 2.912485359568986e-05, "loss": 0.2303, "step": 8987 }, { "epoch": 6.316233309908644, "grad_norm": 2.419480085372925, "learning_rate": 2.91243851018974e-05, "loss": 0.233, "step": 8988 }, { "epoch": 6.31693605059733, "grad_norm": 0.3074270188808441, "learning_rate": 2.9123916608104944e-05, "loss": 0.0656, "step": 8989 }, { "epoch": 6.317638791286015, "grad_norm": 0.14675916731357574, "learning_rate": 2.9123448114312488e-05, "loss": 0.0239, "step": 8990 }, { "epoch": 6.318341531974701, "grad_norm": 0.2490706592798233, "learning_rate": 2.912297962052003e-05, "loss": 0.0302, "step": 8991 }, { "epoch": 6.319044272663387, "grad_norm": 0.14467023313045502, "learning_rate": 2.912251112672757e-05, "loss": 0.0221, "step": 8992 }, { "epoch": 6.319747013352073, "grad_norm": 0.3135318160057068, "learning_rate": 2.9122042632935112e-05, "loss": 0.0202, "step": 8993 }, { "epoch": 6.320449754040759, "grad_norm": 0.1472903937101364, "learning_rate": 2.9121574139142656e-05, "loss": 0.0151, "step": 8994 }, { "epoch": 6.321152494729445, "grad_norm": 0.13500162959098816, "learning_rate": 2.91211056453502e-05, "loss": 0.0163, "step": 8995 }, { "epoch": 6.3218552354181305, "grad_norm": 0.5469825267791748, "learning_rate": 2.9120637151557743e-05, "loss": 0.0367, "step": 8996 }, { "epoch": 6.3225579761068165, "grad_norm": 0.1462426483631134, "learning_rate": 2.9120168657765284e-05, "loss": 0.0214, "step": 8997 }, { "epoch": 6.323260716795502, "grad_norm": 0.15736240148544312, "learning_rate": 2.9119700163972827e-05, "loss": 0.0097, "step": 8998 }, { "epoch": 6.323963457484188, "grad_norm": 0.16929404437541962, "learning_rate": 2.911923167018037e-05, "loss": 0.0357, "step": 8999 }, { "epoch": 6.324666198172874, "grad_norm": 0.12502282857894897, "learning_rate": 2.9118763176387915e-05, "loss": 0.014, "step": 9000 }, { "epoch": 6.324666198172874, "eval_cer": 0.19636563996814071, "eval_loss": 0.2826572358608246, "eval_runtime": 18.4147, "eval_samples_per_second": 246.434, "eval_steps_per_second": 0.815, "eval_wer": 0.35613530348314293, "step": 9000 }, { "epoch": 6.32536893886156, "grad_norm": 0.1911451518535614, "learning_rate": 2.9118294682595455e-05, "loss": 0.044, "step": 9001 }, { "epoch": 6.326071679550246, "grad_norm": 0.36675095558166504, "learning_rate": 2.9117826188803e-05, "loss": 0.0322, "step": 9002 }, { "epoch": 6.326774420238932, "grad_norm": 0.2356211245059967, "learning_rate": 2.9117357695010543e-05, "loss": 0.0394, "step": 9003 }, { "epoch": 6.327477160927618, "grad_norm": 0.2350027859210968, "learning_rate": 2.9116889201218086e-05, "loss": 0.029, "step": 9004 }, { "epoch": 6.328179901616304, "grad_norm": 0.19383357465267181, "learning_rate": 2.9116420707425627e-05, "loss": 0.0253, "step": 9005 }, { "epoch": 6.32888264230499, "grad_norm": 0.4376268982887268, "learning_rate": 2.911595221363317e-05, "loss": 0.0473, "step": 9006 }, { "epoch": 6.329585382993676, "grad_norm": 0.20750737190246582, "learning_rate": 2.9115483719840714e-05, "loss": 0.0322, "step": 9007 }, { "epoch": 6.330288123682362, "grad_norm": 0.4013862609863281, "learning_rate": 2.9115015226048258e-05, "loss": 0.0666, "step": 9008 }, { "epoch": 6.3309908643710475, "grad_norm": 0.3515356183052063, "learning_rate": 2.9114546732255798e-05, "loss": 0.0988, "step": 9009 }, { "epoch": 6.3316936050597326, "grad_norm": 0.44011732935905457, "learning_rate": 2.911407823846334e-05, "loss": 0.1281, "step": 9010 }, { "epoch": 6.3323963457484185, "grad_norm": 0.6408091187477112, "learning_rate": 2.9113609744670882e-05, "loss": 0.1504, "step": 9011 }, { "epoch": 6.333099086437104, "grad_norm": 0.7769792079925537, "learning_rate": 2.9113141250878426e-05, "loss": 0.2506, "step": 9012 }, { "epoch": 6.33380182712579, "grad_norm": 1.5920014381408691, "learning_rate": 2.911267275708597e-05, "loss": 0.2301, "step": 9013 }, { "epoch": 6.334504567814476, "grad_norm": 0.4935693144798279, "learning_rate": 2.911220426329351e-05, "loss": 0.0673, "step": 9014 }, { "epoch": 6.335207308503162, "grad_norm": 0.1274023950099945, "learning_rate": 2.9111735769501054e-05, "loss": 0.0195, "step": 9015 }, { "epoch": 6.335910049191848, "grad_norm": 0.17942757904529572, "learning_rate": 2.9111267275708598e-05, "loss": 0.0384, "step": 9016 }, { "epoch": 6.336612789880534, "grad_norm": 0.1966010481119156, "learning_rate": 2.911079878191614e-05, "loss": 0.0203, "step": 9017 }, { "epoch": 6.33731553056922, "grad_norm": 0.41549381613731384, "learning_rate": 2.911033028812368e-05, "loss": 0.0201, "step": 9018 }, { "epoch": 6.338018271257906, "grad_norm": 0.15583574771881104, "learning_rate": 2.9109861794331225e-05, "loss": 0.0183, "step": 9019 }, { "epoch": 6.338721011946592, "grad_norm": 0.2825527787208557, "learning_rate": 2.910939330053877e-05, "loss": 0.03, "step": 9020 }, { "epoch": 6.339423752635278, "grad_norm": 0.41885828971862793, "learning_rate": 2.9108924806746313e-05, "loss": 0.0355, "step": 9021 }, { "epoch": 6.340126493323964, "grad_norm": 0.29531821608543396, "learning_rate": 2.9108456312953856e-05, "loss": 0.035, "step": 9022 }, { "epoch": 6.3408292340126495, "grad_norm": 0.112885020673275, "learning_rate": 2.9107987819161397e-05, "loss": 0.0177, "step": 9023 }, { "epoch": 6.3415319747013355, "grad_norm": 0.21348391473293304, "learning_rate": 2.910751932536894e-05, "loss": 0.0291, "step": 9024 }, { "epoch": 6.342234715390021, "grad_norm": 0.24384140968322754, "learning_rate": 2.9107050831576484e-05, "loss": 0.0157, "step": 9025 }, { "epoch": 6.342937456078707, "grad_norm": 0.17701224982738495, "learning_rate": 2.9106582337784025e-05, "loss": 0.0376, "step": 9026 }, { "epoch": 6.343640196767393, "grad_norm": 0.20857380330562592, "learning_rate": 2.9106113843991565e-05, "loss": 0.031, "step": 9027 }, { "epoch": 6.344342937456078, "grad_norm": 0.5601036548614502, "learning_rate": 2.910564535019911e-05, "loss": 0.0282, "step": 9028 }, { "epoch": 6.345045678144764, "grad_norm": 0.16579373180866241, "learning_rate": 2.9105176856406652e-05, "loss": 0.0324, "step": 9029 }, { "epoch": 6.34574841883345, "grad_norm": 0.20928028225898743, "learning_rate": 2.9104708362614196e-05, "loss": 0.0365, "step": 9030 }, { "epoch": 6.346451159522136, "grad_norm": 0.3737085163593292, "learning_rate": 2.9104239868821736e-05, "loss": 0.0435, "step": 9031 }, { "epoch": 6.347153900210822, "grad_norm": 0.3043145537376404, "learning_rate": 2.910377137502928e-05, "loss": 0.0678, "step": 9032 }, { "epoch": 6.347856640899508, "grad_norm": 0.23413850367069244, "learning_rate": 2.9103302881236824e-05, "loss": 0.0521, "step": 9033 }, { "epoch": 6.348559381588194, "grad_norm": 0.47868722677230835, "learning_rate": 2.9102834387444368e-05, "loss": 0.0779, "step": 9034 }, { "epoch": 6.34926212227688, "grad_norm": 0.4246315360069275, "learning_rate": 2.910236589365191e-05, "loss": 0.1302, "step": 9035 }, { "epoch": 6.349964862965566, "grad_norm": 0.8561802506446838, "learning_rate": 2.9101897399859452e-05, "loss": 0.1935, "step": 9036 }, { "epoch": 6.3506676036542515, "grad_norm": 0.6578044295310974, "learning_rate": 2.9101428906066995e-05, "loss": 0.2115, "step": 9037 }, { "epoch": 6.3513703443429375, "grad_norm": 1.465445876121521, "learning_rate": 2.910096041227454e-05, "loss": 0.2615, "step": 9038 }, { "epoch": 6.352073085031623, "grad_norm": 0.18640337884426117, "learning_rate": 2.9100491918482083e-05, "loss": 0.0722, "step": 9039 }, { "epoch": 6.352775825720309, "grad_norm": 0.12284497916698456, "learning_rate": 2.9100023424689623e-05, "loss": 0.0231, "step": 9040 }, { "epoch": 6.353478566408995, "grad_norm": 0.17461714148521423, "learning_rate": 2.9099554930897167e-05, "loss": 0.0648, "step": 9041 }, { "epoch": 6.354181307097681, "grad_norm": 0.1395096480846405, "learning_rate": 2.909908643710471e-05, "loss": 0.0202, "step": 9042 }, { "epoch": 6.354884047786367, "grad_norm": 0.17321506142616272, "learning_rate": 2.9098617943312254e-05, "loss": 0.037, "step": 9043 }, { "epoch": 6.355586788475053, "grad_norm": 0.16031944751739502, "learning_rate": 2.9098149449519795e-05, "loss": 0.0146, "step": 9044 }, { "epoch": 6.356289529163739, "grad_norm": 0.1530914157629013, "learning_rate": 2.9097680955727335e-05, "loss": 0.0253, "step": 9045 }, { "epoch": 6.356992269852425, "grad_norm": 0.17505134642124176, "learning_rate": 2.909721246193488e-05, "loss": 0.0299, "step": 9046 }, { "epoch": 6.357695010541111, "grad_norm": 0.15485267341136932, "learning_rate": 2.9096743968142423e-05, "loss": 0.0348, "step": 9047 }, { "epoch": 6.358397751229797, "grad_norm": 0.18122071027755737, "learning_rate": 2.9096275474349966e-05, "loss": 0.0132, "step": 9048 }, { "epoch": 6.359100491918482, "grad_norm": 0.3368399143218994, "learning_rate": 2.9095806980557507e-05, "loss": 0.0222, "step": 9049 }, { "epoch": 6.359803232607168, "grad_norm": 0.09858211874961853, "learning_rate": 2.909533848676505e-05, "loss": 0.0211, "step": 9050 }, { "epoch": 6.3605059732958535, "grad_norm": 0.2773455083370209, "learning_rate": 2.9094869992972594e-05, "loss": 0.0395, "step": 9051 }, { "epoch": 6.3612087139845395, "grad_norm": 0.21217460930347443, "learning_rate": 2.9094401499180138e-05, "loss": 0.02, "step": 9052 }, { "epoch": 6.361911454673225, "grad_norm": 0.20320090651512146, "learning_rate": 2.9093933005387678e-05, "loss": 0.0485, "step": 9053 }, { "epoch": 6.362614195361911, "grad_norm": 0.3434203267097473, "learning_rate": 2.9093464511595222e-05, "loss": 0.0452, "step": 9054 }, { "epoch": 6.363316936050597, "grad_norm": 0.23863373696804047, "learning_rate": 2.9092996017802766e-05, "loss": 0.0208, "step": 9055 }, { "epoch": 6.364019676739283, "grad_norm": 0.28147485852241516, "learning_rate": 2.909252752401031e-05, "loss": 0.0446, "step": 9056 }, { "epoch": 6.364722417427969, "grad_norm": 0.26944077014923096, "learning_rate": 2.9092059030217853e-05, "loss": 0.0505, "step": 9057 }, { "epoch": 6.365425158116655, "grad_norm": 0.31879863142967224, "learning_rate": 2.9091590536425393e-05, "loss": 0.0647, "step": 9058 }, { "epoch": 6.366127898805341, "grad_norm": 0.43209561705589294, "learning_rate": 2.9091122042632937e-05, "loss": 0.1008, "step": 9059 }, { "epoch": 6.366830639494027, "grad_norm": 0.8076346516609192, "learning_rate": 2.909065354884048e-05, "loss": 0.1295, "step": 9060 }, { "epoch": 6.367533380182713, "grad_norm": 0.9980059862136841, "learning_rate": 2.909018505504802e-05, "loss": 0.1818, "step": 9061 }, { "epoch": 6.368236120871399, "grad_norm": 1.2807083129882812, "learning_rate": 2.908971656125556e-05, "loss": 0.2191, "step": 9062 }, { "epoch": 6.368938861560085, "grad_norm": 1.391114592552185, "learning_rate": 2.9089248067463105e-05, "loss": 0.2407, "step": 9063 }, { "epoch": 6.3696416022487705, "grad_norm": 0.22891701757907867, "learning_rate": 2.908877957367065e-05, "loss": 0.0891, "step": 9064 }, { "epoch": 6.370344342937456, "grad_norm": 0.1499871015548706, "learning_rate": 2.9088311079878193e-05, "loss": 0.0361, "step": 9065 }, { "epoch": 6.371047083626142, "grad_norm": 0.3653343915939331, "learning_rate": 2.9087842586085733e-05, "loss": 0.0289, "step": 9066 }, { "epoch": 6.371749824314827, "grad_norm": 0.15472272038459778, "learning_rate": 2.9087374092293277e-05, "loss": 0.0202, "step": 9067 }, { "epoch": 6.372452565003513, "grad_norm": 0.17005513608455658, "learning_rate": 2.908690559850082e-05, "loss": 0.0268, "step": 9068 }, { "epoch": 6.373155305692199, "grad_norm": 0.114859938621521, "learning_rate": 2.9086437104708364e-05, "loss": 0.0167, "step": 9069 }, { "epoch": 6.373858046380885, "grad_norm": 0.1374402493238449, "learning_rate": 2.9085968610915908e-05, "loss": 0.0133, "step": 9070 }, { "epoch": 6.374560787069571, "grad_norm": 0.21979254484176636, "learning_rate": 2.9085500117123448e-05, "loss": 0.0394, "step": 9071 }, { "epoch": 6.375263527758257, "grad_norm": 0.20088370144367218, "learning_rate": 2.9085031623330992e-05, "loss": 0.028, "step": 9072 }, { "epoch": 6.375966268446943, "grad_norm": 0.13740289211273193, "learning_rate": 2.9084563129538536e-05, "loss": 0.018, "step": 9073 }, { "epoch": 6.376669009135629, "grad_norm": 0.14622734487056732, "learning_rate": 2.908409463574608e-05, "loss": 0.0222, "step": 9074 }, { "epoch": 6.377371749824315, "grad_norm": 0.46199995279312134, "learning_rate": 2.908362614195362e-05, "loss": 0.025, "step": 9075 }, { "epoch": 6.378074490513001, "grad_norm": 0.20208632946014404, "learning_rate": 2.9083157648161163e-05, "loss": 0.0279, "step": 9076 }, { "epoch": 6.378777231201687, "grad_norm": 0.3587101697921753, "learning_rate": 2.9082689154368707e-05, "loss": 0.0319, "step": 9077 }, { "epoch": 6.3794799718903725, "grad_norm": 0.5743634700775146, "learning_rate": 2.908222066057625e-05, "loss": 0.0304, "step": 9078 }, { "epoch": 6.3801827125790584, "grad_norm": 0.34832265973091125, "learning_rate": 2.9081752166783788e-05, "loss": 0.0509, "step": 9079 }, { "epoch": 6.380885453267744, "grad_norm": 0.2342223823070526, "learning_rate": 2.908128367299133e-05, "loss": 0.0299, "step": 9080 }, { "epoch": 6.38158819395643, "grad_norm": 0.2029540091753006, "learning_rate": 2.9080815179198875e-05, "loss": 0.0474, "step": 9081 }, { "epoch": 6.382290934645116, "grad_norm": 0.20007115602493286, "learning_rate": 2.908034668540642e-05, "loss": 0.0328, "step": 9082 }, { "epoch": 6.382993675333802, "grad_norm": 0.27269551157951355, "learning_rate": 2.9079878191613963e-05, "loss": 0.0537, "step": 9083 }, { "epoch": 6.383696416022488, "grad_norm": 0.5659507513046265, "learning_rate": 2.9079409697821503e-05, "loss": 0.0924, "step": 9084 }, { "epoch": 6.384399156711174, "grad_norm": 0.6107679605484009, "learning_rate": 2.9078941204029047e-05, "loss": 0.124, "step": 9085 }, { "epoch": 6.38510189739986, "grad_norm": 0.7030133605003357, "learning_rate": 2.907847271023659e-05, "loss": 0.1712, "step": 9086 }, { "epoch": 6.385804638088545, "grad_norm": 1.2587265968322754, "learning_rate": 2.9078004216444134e-05, "loss": 0.2256, "step": 9087 }, { "epoch": 6.386507378777231, "grad_norm": 1.9234287738800049, "learning_rate": 2.9077535722651675e-05, "loss": 0.2879, "step": 9088 }, { "epoch": 6.387210119465917, "grad_norm": 0.20594510436058044, "learning_rate": 2.907706722885922e-05, "loss": 0.0612, "step": 9089 }, { "epoch": 6.387912860154603, "grad_norm": 0.218509241938591, "learning_rate": 2.9076598735066762e-05, "loss": 0.0372, "step": 9090 }, { "epoch": 6.388615600843289, "grad_norm": 0.18219496309757233, "learning_rate": 2.9076130241274306e-05, "loss": 0.0224, "step": 9091 }, { "epoch": 6.3893183415319745, "grad_norm": 0.16916409134864807, "learning_rate": 2.9075661747481846e-05, "loss": 0.0243, "step": 9092 }, { "epoch": 6.3900210822206605, "grad_norm": 0.1065065786242485, "learning_rate": 2.907519325368939e-05, "loss": 0.0172, "step": 9093 }, { "epoch": 6.390723822909346, "grad_norm": 0.31461524963378906, "learning_rate": 2.9074724759896934e-05, "loss": 0.0199, "step": 9094 }, { "epoch": 6.391426563598032, "grad_norm": 0.10912065953016281, "learning_rate": 2.9074256266104477e-05, "loss": 0.0139, "step": 9095 }, { "epoch": 6.392129304286718, "grad_norm": 0.1754787415266037, "learning_rate": 2.9073787772312018e-05, "loss": 0.0274, "step": 9096 }, { "epoch": 6.392832044975404, "grad_norm": 0.17684096097946167, "learning_rate": 2.9073319278519558e-05, "loss": 0.0153, "step": 9097 }, { "epoch": 6.39353478566409, "grad_norm": 0.1715114265680313, "learning_rate": 2.90728507847271e-05, "loss": 0.0387, "step": 9098 }, { "epoch": 6.394237526352776, "grad_norm": 0.26685285568237305, "learning_rate": 2.9072382290934645e-05, "loss": 0.0316, "step": 9099 }, { "epoch": 6.394940267041462, "grad_norm": 0.16600929200649261, "learning_rate": 2.907191379714219e-05, "loss": 0.0131, "step": 9100 }, { "epoch": 6.395643007730148, "grad_norm": 0.2214660942554474, "learning_rate": 2.907144530334973e-05, "loss": 0.0319, "step": 9101 }, { "epoch": 6.396345748418834, "grad_norm": 0.18794073164463043, "learning_rate": 2.9070976809557273e-05, "loss": 0.0179, "step": 9102 }, { "epoch": 6.39704848910752, "grad_norm": 0.2606423795223236, "learning_rate": 2.9070508315764817e-05, "loss": 0.0469, "step": 9103 }, { "epoch": 6.397751229796206, "grad_norm": 0.21296574175357819, "learning_rate": 2.907003982197236e-05, "loss": 0.0382, "step": 9104 }, { "epoch": 6.398453970484891, "grad_norm": 0.174438014626503, "learning_rate": 2.90695713281799e-05, "loss": 0.0155, "step": 9105 }, { "epoch": 6.3991567111735765, "grad_norm": 0.21804705262184143, "learning_rate": 2.9069102834387445e-05, "loss": 0.0538, "step": 9106 }, { "epoch": 6.3998594518622625, "grad_norm": 0.3155151307582855, "learning_rate": 2.906863434059499e-05, "loss": 0.0613, "step": 9107 }, { "epoch": 6.400562192550948, "grad_norm": 0.2648448050022125, "learning_rate": 2.9068165846802532e-05, "loss": 0.0493, "step": 9108 }, { "epoch": 6.401264933239634, "grad_norm": 0.310188889503479, "learning_rate": 2.9067697353010076e-05, "loss": 0.0722, "step": 9109 }, { "epoch": 6.40196767392832, "grad_norm": 0.6433690190315247, "learning_rate": 2.9067228859217616e-05, "loss": 0.1102, "step": 9110 }, { "epoch": 6.402670414617006, "grad_norm": 0.9354378581047058, "learning_rate": 2.906676036542516e-05, "loss": 0.1568, "step": 9111 }, { "epoch": 6.403373155305692, "grad_norm": 1.0037070512771606, "learning_rate": 2.9066291871632704e-05, "loss": 0.1805, "step": 9112 }, { "epoch": 6.404075895994378, "grad_norm": 2.3602635860443115, "learning_rate": 2.9065823377840247e-05, "loss": 0.2387, "step": 9113 }, { "epoch": 6.404778636683064, "grad_norm": 0.24593836069107056, "learning_rate": 2.9065354884047784e-05, "loss": 0.0826, "step": 9114 }, { "epoch": 6.40548137737175, "grad_norm": 0.3171508312225342, "learning_rate": 2.9064886390255328e-05, "loss": 0.0464, "step": 9115 }, { "epoch": 6.406184118060436, "grad_norm": 0.14711061120033264, "learning_rate": 2.9064417896462872e-05, "loss": 0.0336, "step": 9116 }, { "epoch": 6.406886858749122, "grad_norm": 0.1749747395515442, "learning_rate": 2.9063949402670416e-05, "loss": 0.0203, "step": 9117 }, { "epoch": 6.407589599437808, "grad_norm": 0.46804025769233704, "learning_rate": 2.9063480908877956e-05, "loss": 0.0211, "step": 9118 }, { "epoch": 6.4082923401264935, "grad_norm": 0.2785024642944336, "learning_rate": 2.90630124150855e-05, "loss": 0.0128, "step": 9119 }, { "epoch": 6.408995080815179, "grad_norm": 0.13573478162288666, "learning_rate": 2.9062543921293043e-05, "loss": 0.0175, "step": 9120 }, { "epoch": 6.409697821503865, "grad_norm": 0.14533668756484985, "learning_rate": 2.9062075427500587e-05, "loss": 0.0171, "step": 9121 }, { "epoch": 6.410400562192551, "grad_norm": 0.3458613455295563, "learning_rate": 2.906160693370813e-05, "loss": 0.0256, "step": 9122 }, { "epoch": 6.411103302881237, "grad_norm": 0.16799014806747437, "learning_rate": 2.906113843991567e-05, "loss": 0.0307, "step": 9123 }, { "epoch": 6.411806043569923, "grad_norm": 0.5359904766082764, "learning_rate": 2.9060669946123215e-05, "loss": 0.027, "step": 9124 }, { "epoch": 6.412508784258609, "grad_norm": 1.0195025205612183, "learning_rate": 2.906020145233076e-05, "loss": 0.0267, "step": 9125 }, { "epoch": 6.413211524947294, "grad_norm": 0.1454629898071289, "learning_rate": 2.9059732958538302e-05, "loss": 0.0219, "step": 9126 }, { "epoch": 6.41391426563598, "grad_norm": 0.2108156830072403, "learning_rate": 2.9059264464745843e-05, "loss": 0.0267, "step": 9127 }, { "epoch": 6.414617006324666, "grad_norm": 0.15048369765281677, "learning_rate": 2.9058795970953386e-05, "loss": 0.0324, "step": 9128 }, { "epoch": 6.415319747013352, "grad_norm": 0.49132075905799866, "learning_rate": 2.905832747716093e-05, "loss": 0.0325, "step": 9129 }, { "epoch": 6.416022487702038, "grad_norm": 0.19873984158039093, "learning_rate": 2.9057858983368474e-05, "loss": 0.0264, "step": 9130 }, { "epoch": 6.416725228390724, "grad_norm": 0.25929298996925354, "learning_rate": 2.905739048957601e-05, "loss": 0.0426, "step": 9131 }, { "epoch": 6.41742796907941, "grad_norm": 0.48118364810943604, "learning_rate": 2.9056921995783554e-05, "loss": 0.0509, "step": 9132 }, { "epoch": 6.4181307097680955, "grad_norm": 0.33910471200942993, "learning_rate": 2.9056453501991098e-05, "loss": 0.0483, "step": 9133 }, { "epoch": 6.418833450456781, "grad_norm": 0.826989471912384, "learning_rate": 2.9055985008198642e-05, "loss": 0.0673, "step": 9134 }, { "epoch": 6.419536191145467, "grad_norm": 0.6648470759391785, "learning_rate": 2.9055516514406186e-05, "loss": 0.1305, "step": 9135 }, { "epoch": 6.420238931834153, "grad_norm": 1.0011813640594482, "learning_rate": 2.9055048020613726e-05, "loss": 0.1904, "step": 9136 }, { "epoch": 6.420941672522839, "grad_norm": 1.1164066791534424, "learning_rate": 2.905457952682127e-05, "loss": 0.181, "step": 9137 }, { "epoch": 6.421644413211525, "grad_norm": 1.0968613624572754, "learning_rate": 2.9054111033028813e-05, "loss": 0.2434, "step": 9138 }, { "epoch": 6.422347153900211, "grad_norm": 0.230974018573761, "learning_rate": 2.9053642539236357e-05, "loss": 0.0674, "step": 9139 }, { "epoch": 6.423049894588897, "grad_norm": 0.17583250999450684, "learning_rate": 2.9053174045443897e-05, "loss": 0.0254, "step": 9140 }, { "epoch": 6.423752635277583, "grad_norm": 0.23081859946250916, "learning_rate": 2.905270555165144e-05, "loss": 0.0244, "step": 9141 }, { "epoch": 6.424455375966269, "grad_norm": 0.1417665183544159, "learning_rate": 2.9052237057858985e-05, "loss": 0.0275, "step": 9142 }, { "epoch": 6.425158116654955, "grad_norm": 0.16428841650485992, "learning_rate": 2.905176856406653e-05, "loss": 0.034, "step": 9143 }, { "epoch": 6.42586085734364, "grad_norm": 0.44967734813690186, "learning_rate": 2.905130007027407e-05, "loss": 0.0079, "step": 9144 }, { "epoch": 6.426563598032326, "grad_norm": 0.15275326371192932, "learning_rate": 2.9050831576481613e-05, "loss": 0.0202, "step": 9145 }, { "epoch": 6.427266338721012, "grad_norm": 0.18948745727539062, "learning_rate": 2.9050363082689156e-05, "loss": 0.0251, "step": 9146 }, { "epoch": 6.4279690794096975, "grad_norm": 0.27949103713035583, "learning_rate": 2.90498945888967e-05, "loss": 0.0311, "step": 9147 }, { "epoch": 6.4286718200983834, "grad_norm": 0.13859973847866058, "learning_rate": 2.904942609510424e-05, "loss": 0.0199, "step": 9148 }, { "epoch": 6.429374560787069, "grad_norm": 0.2816750109195709, "learning_rate": 2.904895760131178e-05, "loss": 0.0293, "step": 9149 }, { "epoch": 6.430077301475755, "grad_norm": 0.22423125803470612, "learning_rate": 2.9048489107519325e-05, "loss": 0.0191, "step": 9150 }, { "epoch": 6.430780042164441, "grad_norm": 0.18432337045669556, "learning_rate": 2.9048020613726868e-05, "loss": 0.0329, "step": 9151 }, { "epoch": 6.431482782853127, "grad_norm": 0.29221388697624207, "learning_rate": 2.9047552119934412e-05, "loss": 0.0334, "step": 9152 }, { "epoch": 6.432185523541813, "grad_norm": 0.20340053737163544, "learning_rate": 2.9047083626141952e-05, "loss": 0.0411, "step": 9153 }, { "epoch": 6.432888264230499, "grad_norm": 0.3103979229927063, "learning_rate": 2.9046615132349496e-05, "loss": 0.031, "step": 9154 }, { "epoch": 6.433591004919185, "grad_norm": 0.4626266658306122, "learning_rate": 2.904614663855704e-05, "loss": 0.0361, "step": 9155 }, { "epoch": 6.434293745607871, "grad_norm": 1.0669935941696167, "learning_rate": 2.9045678144764584e-05, "loss": 0.055, "step": 9156 }, { "epoch": 6.434996486296557, "grad_norm": 0.6221303343772888, "learning_rate": 2.9045209650972124e-05, "loss": 0.0484, "step": 9157 }, { "epoch": 6.435699226985243, "grad_norm": 0.2677699327468872, "learning_rate": 2.9044741157179668e-05, "loss": 0.0658, "step": 9158 }, { "epoch": 6.436401967673929, "grad_norm": 0.47941282391548157, "learning_rate": 2.904427266338721e-05, "loss": 0.0796, "step": 9159 }, { "epoch": 6.4371047083626145, "grad_norm": 0.48338645696640015, "learning_rate": 2.9043804169594755e-05, "loss": 0.1456, "step": 9160 }, { "epoch": 6.4378074490513, "grad_norm": 1.2814112901687622, "learning_rate": 2.90433356758023e-05, "loss": 0.1912, "step": 9161 }, { "epoch": 6.438510189739986, "grad_norm": 0.9923827052116394, "learning_rate": 2.904286718200984e-05, "loss": 0.226, "step": 9162 }, { "epoch": 6.439212930428672, "grad_norm": 2.2370989322662354, "learning_rate": 2.9042398688217383e-05, "loss": 0.2534, "step": 9163 }, { "epoch": 6.439915671117357, "grad_norm": 0.21439945697784424, "learning_rate": 2.9041930194424927e-05, "loss": 0.0674, "step": 9164 }, { "epoch": 6.440618411806043, "grad_norm": 0.18586242198944092, "learning_rate": 2.904146170063247e-05, "loss": 0.0354, "step": 9165 }, { "epoch": 6.441321152494729, "grad_norm": 0.1770576536655426, "learning_rate": 2.9040993206840007e-05, "loss": 0.0188, "step": 9166 }, { "epoch": 6.442023893183415, "grad_norm": 0.14179812371730804, "learning_rate": 2.904052471304755e-05, "loss": 0.0165, "step": 9167 }, { "epoch": 6.442726633872101, "grad_norm": 0.3505247235298157, "learning_rate": 2.9040056219255095e-05, "loss": 0.0289, "step": 9168 }, { "epoch": 6.443429374560787, "grad_norm": 0.11819338798522949, "learning_rate": 2.903958772546264e-05, "loss": 0.0151, "step": 9169 }, { "epoch": 6.444132115249473, "grad_norm": 0.17754505574703217, "learning_rate": 2.903911923167018e-05, "loss": 0.0219, "step": 9170 }, { "epoch": 6.444834855938159, "grad_norm": 0.12557782232761383, "learning_rate": 2.9038650737877722e-05, "loss": 0.0118, "step": 9171 }, { "epoch": 6.445537596626845, "grad_norm": 0.15393979847431183, "learning_rate": 2.9038182244085266e-05, "loss": 0.0255, "step": 9172 }, { "epoch": 6.446240337315531, "grad_norm": 0.16648118197917938, "learning_rate": 2.903771375029281e-05, "loss": 0.0133, "step": 9173 }, { "epoch": 6.4469430780042165, "grad_norm": 0.20408138632774353, "learning_rate": 2.9037245256500354e-05, "loss": 0.0353, "step": 9174 }, { "epoch": 6.447645818692902, "grad_norm": 0.2561580240726471, "learning_rate": 2.9036776762707894e-05, "loss": 0.0394, "step": 9175 }, { "epoch": 6.448348559381588, "grad_norm": 0.28300538659095764, "learning_rate": 2.9036308268915438e-05, "loss": 0.0335, "step": 9176 }, { "epoch": 6.449051300070274, "grad_norm": 0.2725127339363098, "learning_rate": 2.903583977512298e-05, "loss": 0.0442, "step": 9177 }, { "epoch": 6.44975404075896, "grad_norm": 0.26839953660964966, "learning_rate": 2.9035371281330525e-05, "loss": 0.0355, "step": 9178 }, { "epoch": 6.450456781447646, "grad_norm": 0.21155165135860443, "learning_rate": 2.9034902787538065e-05, "loss": 0.0213, "step": 9179 }, { "epoch": 6.451159522136332, "grad_norm": 0.28514745831489563, "learning_rate": 2.903443429374561e-05, "loss": 0.0355, "step": 9180 }, { "epoch": 6.451862262825018, "grad_norm": 0.3047025799751282, "learning_rate": 2.9033965799953153e-05, "loss": 0.0382, "step": 9181 }, { "epoch": 6.452565003513703, "grad_norm": 0.28978535532951355, "learning_rate": 2.9033497306160697e-05, "loss": 0.0447, "step": 9182 }, { "epoch": 6.453267744202389, "grad_norm": 0.3675890266895294, "learning_rate": 2.9033028812368234e-05, "loss": 0.0557, "step": 9183 }, { "epoch": 6.453970484891075, "grad_norm": 0.3558635115623474, "learning_rate": 2.9032560318575777e-05, "loss": 0.0614, "step": 9184 }, { "epoch": 6.454673225579761, "grad_norm": 1.1607383489608765, "learning_rate": 2.903209182478332e-05, "loss": 0.1202, "step": 9185 }, { "epoch": 6.455375966268447, "grad_norm": 0.8308154940605164, "learning_rate": 2.9031623330990865e-05, "loss": 0.1653, "step": 9186 }, { "epoch": 6.456078706957133, "grad_norm": 0.896079957485199, "learning_rate": 2.903115483719841e-05, "loss": 0.2069, "step": 9187 }, { "epoch": 6.4567814476458185, "grad_norm": 1.4870051145553589, "learning_rate": 2.903068634340595e-05, "loss": 0.2437, "step": 9188 }, { "epoch": 6.457484188334504, "grad_norm": 0.21309973299503326, "learning_rate": 2.9030217849613493e-05, "loss": 0.0671, "step": 9189 }, { "epoch": 6.45818692902319, "grad_norm": 0.2094983160495758, "learning_rate": 2.9029749355821036e-05, "loss": 0.0475, "step": 9190 }, { "epoch": 6.458889669711876, "grad_norm": 0.11856111884117126, "learning_rate": 2.902928086202858e-05, "loss": 0.0179, "step": 9191 }, { "epoch": 6.459592410400562, "grad_norm": 0.13647720217704773, "learning_rate": 2.902881236823612e-05, "loss": 0.0262, "step": 9192 }, { "epoch": 6.460295151089248, "grad_norm": 0.13963831961154938, "learning_rate": 2.9028343874443664e-05, "loss": 0.0192, "step": 9193 }, { "epoch": 6.460997891777934, "grad_norm": 0.11122387647628784, "learning_rate": 2.9027875380651208e-05, "loss": 0.0139, "step": 9194 }, { "epoch": 6.46170063246662, "grad_norm": 0.23248539865016937, "learning_rate": 2.902740688685875e-05, "loss": 0.0222, "step": 9195 }, { "epoch": 6.462403373155306, "grad_norm": 0.12694129347801208, "learning_rate": 2.9026938393066292e-05, "loss": 0.0185, "step": 9196 }, { "epoch": 6.463106113843992, "grad_norm": 0.1559493988752365, "learning_rate": 2.9026469899273836e-05, "loss": 0.0167, "step": 9197 }, { "epoch": 6.463808854532678, "grad_norm": 0.23240676522254944, "learning_rate": 2.902600140548138e-05, "loss": 0.0236, "step": 9198 }, { "epoch": 6.464511595221364, "grad_norm": 0.19726860523223877, "learning_rate": 2.9025532911688923e-05, "loss": 0.0331, "step": 9199 }, { "epoch": 6.46521433591005, "grad_norm": 0.30147916078567505, "learning_rate": 2.9025064417896467e-05, "loss": 0.0196, "step": 9200 }, { "epoch": 6.4659170765987355, "grad_norm": 0.18948496878147125, "learning_rate": 2.9024595924104004e-05, "loss": 0.0324, "step": 9201 }, { "epoch": 6.466619817287421, "grad_norm": 0.26502934098243713, "learning_rate": 2.9024127430311547e-05, "loss": 0.0219, "step": 9202 }, { "epoch": 6.4673225579761064, "grad_norm": 0.2384045273065567, "learning_rate": 2.902365893651909e-05, "loss": 0.0427, "step": 9203 }, { "epoch": 6.468025298664792, "grad_norm": 0.49030154943466187, "learning_rate": 2.9023190442726635e-05, "loss": 0.0344, "step": 9204 }, { "epoch": 6.468728039353478, "grad_norm": 0.15811137855052948, "learning_rate": 2.9022721948934175e-05, "loss": 0.0369, "step": 9205 }, { "epoch": 6.469430780042164, "grad_norm": 0.19327767193317413, "learning_rate": 2.902225345514172e-05, "loss": 0.0154, "step": 9206 }, { "epoch": 6.47013352073085, "grad_norm": 0.23996832966804504, "learning_rate": 2.9021784961349263e-05, "loss": 0.0569, "step": 9207 }, { "epoch": 6.470836261419536, "grad_norm": 0.46828728914260864, "learning_rate": 2.9021316467556806e-05, "loss": 0.0715, "step": 9208 }, { "epoch": 6.471539002108222, "grad_norm": 0.3139974772930145, "learning_rate": 2.9020847973764347e-05, "loss": 0.0803, "step": 9209 }, { "epoch": 6.472241742796908, "grad_norm": 0.44075465202331543, "learning_rate": 2.902037947997189e-05, "loss": 0.1241, "step": 9210 }, { "epoch": 6.472944483485594, "grad_norm": 0.7142650485038757, "learning_rate": 2.9019910986179434e-05, "loss": 0.1723, "step": 9211 }, { "epoch": 6.47364722417428, "grad_norm": 0.9862667322158813, "learning_rate": 2.9019442492386978e-05, "loss": 0.2315, "step": 9212 }, { "epoch": 6.474349964862966, "grad_norm": 1.4713276624679565, "learning_rate": 2.901897399859452e-05, "loss": 0.2503, "step": 9213 }, { "epoch": 6.475052705551652, "grad_norm": 0.4997117817401886, "learning_rate": 2.9018505504802062e-05, "loss": 0.07, "step": 9214 }, { "epoch": 6.4757554462403375, "grad_norm": 0.40608757734298706, "learning_rate": 2.9018037011009606e-05, "loss": 0.0286, "step": 9215 }, { "epoch": 6.476458186929023, "grad_norm": 0.13689649105072021, "learning_rate": 2.901756851721715e-05, "loss": 0.0196, "step": 9216 }, { "epoch": 6.477160927617709, "grad_norm": 0.13862521946430206, "learning_rate": 2.9017100023424693e-05, "loss": 0.0255, "step": 9217 }, { "epoch": 6.477863668306395, "grad_norm": 0.15889841318130493, "learning_rate": 2.901663152963223e-05, "loss": 0.0174, "step": 9218 }, { "epoch": 6.478566408995081, "grad_norm": 0.14744360744953156, "learning_rate": 2.9016163035839774e-05, "loss": 0.0163, "step": 9219 }, { "epoch": 6.479269149683767, "grad_norm": 0.1679651290178299, "learning_rate": 2.9015694542047318e-05, "loss": 0.0263, "step": 9220 }, { "epoch": 6.479971890372452, "grad_norm": 0.09808694571256638, "learning_rate": 2.901522604825486e-05, "loss": 0.0154, "step": 9221 }, { "epoch": 6.480674631061138, "grad_norm": 0.26415738463401794, "learning_rate": 2.90147575544624e-05, "loss": 0.0191, "step": 9222 }, { "epoch": 6.481377371749824, "grad_norm": 0.15081652998924255, "learning_rate": 2.9014289060669945e-05, "loss": 0.0208, "step": 9223 }, { "epoch": 6.48208011243851, "grad_norm": 0.39924126863479614, "learning_rate": 2.901382056687749e-05, "loss": 0.0307, "step": 9224 }, { "epoch": 6.482782853127196, "grad_norm": 0.16205069422721863, "learning_rate": 2.9013352073085033e-05, "loss": 0.0202, "step": 9225 }, { "epoch": 6.483485593815882, "grad_norm": 0.12956635653972626, "learning_rate": 2.9012883579292577e-05, "loss": 0.0232, "step": 9226 }, { "epoch": 6.484188334504568, "grad_norm": 0.16500817239284515, "learning_rate": 2.9012415085500117e-05, "loss": 0.0178, "step": 9227 }, { "epoch": 6.484891075193254, "grad_norm": 0.17592588067054749, "learning_rate": 2.901194659170766e-05, "loss": 0.0241, "step": 9228 }, { "epoch": 6.4855938158819395, "grad_norm": 0.15121084451675415, "learning_rate": 2.9011478097915204e-05, "loss": 0.0262, "step": 9229 }, { "epoch": 6.486296556570625, "grad_norm": 0.3086799085140228, "learning_rate": 2.9011009604122748e-05, "loss": 0.0359, "step": 9230 }, { "epoch": 6.486999297259311, "grad_norm": 1.508976936340332, "learning_rate": 2.901054111033029e-05, "loss": 0.0422, "step": 9231 }, { "epoch": 6.487702037947997, "grad_norm": 3.5007119178771973, "learning_rate": 2.9010072616537832e-05, "loss": 0.0521, "step": 9232 }, { "epoch": 6.488404778636683, "grad_norm": 0.3108040988445282, "learning_rate": 2.9009604122745376e-05, "loss": 0.0623, "step": 9233 }, { "epoch": 6.489107519325369, "grad_norm": 0.3549756109714508, "learning_rate": 2.900913562895292e-05, "loss": 0.0858, "step": 9234 }, { "epoch": 6.489810260014055, "grad_norm": 0.7889984846115112, "learning_rate": 2.900866713516046e-05, "loss": 0.1246, "step": 9235 }, { "epoch": 6.490513000702741, "grad_norm": 0.8159617185592651, "learning_rate": 2.9008198641368e-05, "loss": 0.1699, "step": 9236 }, { "epoch": 6.491215741391427, "grad_norm": 0.7018150687217712, "learning_rate": 2.9007730147575544e-05, "loss": 0.1944, "step": 9237 }, { "epoch": 6.491918482080113, "grad_norm": 2.4150729179382324, "learning_rate": 2.9007261653783088e-05, "loss": 0.242, "step": 9238 }, { "epoch": 6.492621222768799, "grad_norm": 1.1215802431106567, "learning_rate": 2.900679315999063e-05, "loss": 0.0745, "step": 9239 }, { "epoch": 6.493323963457485, "grad_norm": 0.19484944641590118, "learning_rate": 2.9006324666198172e-05, "loss": 0.0335, "step": 9240 }, { "epoch": 6.49402670414617, "grad_norm": 0.3565010726451874, "learning_rate": 2.9005856172405715e-05, "loss": 0.0496, "step": 9241 }, { "epoch": 6.494729444834856, "grad_norm": 0.16381685435771942, "learning_rate": 2.900538767861326e-05, "loss": 0.0257, "step": 9242 }, { "epoch": 6.4954321855235415, "grad_norm": 0.4173565208911896, "learning_rate": 2.9004919184820803e-05, "loss": 0.0391, "step": 9243 }, { "epoch": 6.496134926212227, "grad_norm": 0.1649831086397171, "learning_rate": 2.9004450691028343e-05, "loss": 0.0123, "step": 9244 }, { "epoch": 6.496837666900913, "grad_norm": 0.15172405540943146, "learning_rate": 2.9003982197235887e-05, "loss": 0.0146, "step": 9245 }, { "epoch": 6.497540407589599, "grad_norm": 0.19217167794704437, "learning_rate": 2.900351370344343e-05, "loss": 0.0546, "step": 9246 }, { "epoch": 6.498243148278285, "grad_norm": 0.15271194279193878, "learning_rate": 2.9003045209650974e-05, "loss": 0.0194, "step": 9247 }, { "epoch": 6.498945888966971, "grad_norm": 0.17265602946281433, "learning_rate": 2.9002576715858515e-05, "loss": 0.0139, "step": 9248 }, { "epoch": 6.499648629655657, "grad_norm": 0.27949339151382446, "learning_rate": 2.900210822206606e-05, "loss": 0.0277, "step": 9249 }, { "epoch": 6.500351370344343, "grad_norm": 0.13896436989307404, "learning_rate": 2.9001639728273602e-05, "loss": 0.0204, "step": 9250 }, { "epoch": 6.501054111033029, "grad_norm": 0.1596459299325943, "learning_rate": 2.9001171234481146e-05, "loss": 0.0202, "step": 9251 }, { "epoch": 6.501756851721715, "grad_norm": 0.10634652525186539, "learning_rate": 2.900070274068869e-05, "loss": 0.0122, "step": 9252 }, { "epoch": 6.502459592410401, "grad_norm": 0.17498140037059784, "learning_rate": 2.9000234246896227e-05, "loss": 0.0387, "step": 9253 }, { "epoch": 6.503162333099087, "grad_norm": 0.45433276891708374, "learning_rate": 2.899976575310377e-05, "loss": 0.039, "step": 9254 }, { "epoch": 6.503865073787773, "grad_norm": 0.3536267578601837, "learning_rate": 2.8999297259311314e-05, "loss": 0.0175, "step": 9255 }, { "epoch": 6.5045678144764585, "grad_norm": 0.27374839782714844, "learning_rate": 2.8998828765518858e-05, "loss": 0.0345, "step": 9256 }, { "epoch": 6.505270555165144, "grad_norm": 0.24310505390167236, "learning_rate": 2.8998360271726398e-05, "loss": 0.0459, "step": 9257 }, { "epoch": 6.505973295853829, "grad_norm": 0.7147060036659241, "learning_rate": 2.8997891777933942e-05, "loss": 0.1112, "step": 9258 }, { "epoch": 6.506676036542515, "grad_norm": 0.2859046459197998, "learning_rate": 2.8997423284141486e-05, "loss": 0.0654, "step": 9259 }, { "epoch": 6.507378777231201, "grad_norm": 1.2537121772766113, "learning_rate": 2.899695479034903e-05, "loss": 0.1249, "step": 9260 }, { "epoch": 6.508081517919887, "grad_norm": 1.0466748476028442, "learning_rate": 2.8996486296556573e-05, "loss": 0.1629, "step": 9261 }, { "epoch": 6.508784258608573, "grad_norm": 1.3426239490509033, "learning_rate": 2.8996017802764113e-05, "loss": 0.2388, "step": 9262 }, { "epoch": 6.509486999297259, "grad_norm": 1.5023908615112305, "learning_rate": 2.8995549308971657e-05, "loss": 0.2741, "step": 9263 }, { "epoch": 6.510189739985945, "grad_norm": 0.2312338799238205, "learning_rate": 2.89950808151792e-05, "loss": 0.0676, "step": 9264 }, { "epoch": 6.510892480674631, "grad_norm": 0.7531131505966187, "learning_rate": 2.8994612321386745e-05, "loss": 0.0335, "step": 9265 }, { "epoch": 6.511595221363317, "grad_norm": 0.13287416100502014, "learning_rate": 2.8994143827594285e-05, "loss": 0.0238, "step": 9266 }, { "epoch": 6.512297962052003, "grad_norm": 0.19261431694030762, "learning_rate": 2.899367533380183e-05, "loss": 0.033, "step": 9267 }, { "epoch": 6.513000702740689, "grad_norm": 0.15467387437820435, "learning_rate": 2.8993206840009372e-05, "loss": 0.0192, "step": 9268 }, { "epoch": 6.513703443429375, "grad_norm": 0.16636137664318085, "learning_rate": 2.8992738346216916e-05, "loss": 0.0198, "step": 9269 }, { "epoch": 6.5144061841180605, "grad_norm": 0.18901948630809784, "learning_rate": 2.8992269852424453e-05, "loss": 0.0219, "step": 9270 }, { "epoch": 6.515108924806746, "grad_norm": 0.09567204117774963, "learning_rate": 2.8991801358631997e-05, "loss": 0.0146, "step": 9271 }, { "epoch": 6.515811665495432, "grad_norm": 0.2370425909757614, "learning_rate": 2.899133286483954e-05, "loss": 0.0303, "step": 9272 }, { "epoch": 6.516514406184118, "grad_norm": 0.31381461024284363, "learning_rate": 2.8990864371047084e-05, "loss": 0.014, "step": 9273 }, { "epoch": 6.517217146872804, "grad_norm": 0.36488276720046997, "learning_rate": 2.8990395877254628e-05, "loss": 0.0305, "step": 9274 }, { "epoch": 6.51791988756149, "grad_norm": 0.18841016292572021, "learning_rate": 2.8989927383462168e-05, "loss": 0.0186, "step": 9275 }, { "epoch": 6.518622628250176, "grad_norm": 0.1909862756729126, "learning_rate": 2.8989458889669712e-05, "loss": 0.0288, "step": 9276 }, { "epoch": 6.519325368938862, "grad_norm": 0.7734382748603821, "learning_rate": 2.8988990395877256e-05, "loss": 0.0395, "step": 9277 }, { "epoch": 6.520028109627548, "grad_norm": 0.21404805779457092, "learning_rate": 2.89885219020848e-05, "loss": 0.0308, "step": 9278 }, { "epoch": 6.520730850316234, "grad_norm": 0.5548146367073059, "learning_rate": 2.898805340829234e-05, "loss": 0.0522, "step": 9279 }, { "epoch": 6.521433591004919, "grad_norm": 0.3006534278392792, "learning_rate": 2.8987584914499883e-05, "loss": 0.0292, "step": 9280 }, { "epoch": 6.522136331693605, "grad_norm": 0.1942722499370575, "learning_rate": 2.8987116420707427e-05, "loss": 0.0418, "step": 9281 }, { "epoch": 6.522839072382291, "grad_norm": 0.2527012526988983, "learning_rate": 2.898664792691497e-05, "loss": 0.0346, "step": 9282 }, { "epoch": 6.523541813070977, "grad_norm": 0.23224517703056335, "learning_rate": 2.898617943312251e-05, "loss": 0.05, "step": 9283 }, { "epoch": 6.5242445537596625, "grad_norm": 0.35266363620758057, "learning_rate": 2.8985710939330055e-05, "loss": 0.0829, "step": 9284 }, { "epoch": 6.524947294448348, "grad_norm": 0.4884594678878784, "learning_rate": 2.89852424455376e-05, "loss": 0.1379, "step": 9285 }, { "epoch": 6.525650035137034, "grad_norm": 0.5606582760810852, "learning_rate": 2.8984773951745142e-05, "loss": 0.1302, "step": 9286 }, { "epoch": 6.52635277582572, "grad_norm": 1.558461308479309, "learning_rate": 2.8984305457952686e-05, "loss": 0.1979, "step": 9287 }, { "epoch": 6.527055516514406, "grad_norm": 1.1699390411376953, "learning_rate": 2.8983836964160223e-05, "loss": 0.2324, "step": 9288 }, { "epoch": 6.527758257203092, "grad_norm": 0.6375939249992371, "learning_rate": 2.8983368470367767e-05, "loss": 0.0768, "step": 9289 }, { "epoch": 6.528460997891778, "grad_norm": 0.24959856271743774, "learning_rate": 2.898289997657531e-05, "loss": 0.0413, "step": 9290 }, { "epoch": 6.529163738580464, "grad_norm": 0.1128428503870964, "learning_rate": 2.8982431482782854e-05, "loss": 0.021, "step": 9291 }, { "epoch": 6.52986647926915, "grad_norm": 0.19877158105373383, "learning_rate": 2.8981962988990395e-05, "loss": 0.0182, "step": 9292 }, { "epoch": 6.530569219957836, "grad_norm": 0.14244544506072998, "learning_rate": 2.898149449519794e-05, "loss": 0.0244, "step": 9293 }, { "epoch": 6.531271960646522, "grad_norm": 0.3835909962654114, "learning_rate": 2.8981026001405482e-05, "loss": 0.0123, "step": 9294 }, { "epoch": 6.531974701335208, "grad_norm": 0.1761181503534317, "learning_rate": 2.8980557507613026e-05, "loss": 0.0215, "step": 9295 }, { "epoch": 6.5326774420238936, "grad_norm": 0.1661163866519928, "learning_rate": 2.8980089013820566e-05, "loss": 0.0259, "step": 9296 }, { "epoch": 6.533380182712579, "grad_norm": 0.1873033344745636, "learning_rate": 2.897962052002811e-05, "loss": 0.0289, "step": 9297 }, { "epoch": 6.5340829234012645, "grad_norm": 0.12460818886756897, "learning_rate": 2.8979152026235654e-05, "loss": 0.01, "step": 9298 }, { "epoch": 6.53478566408995, "grad_norm": 0.1568542718887329, "learning_rate": 2.8978683532443197e-05, "loss": 0.024, "step": 9299 }, { "epoch": 6.535488404778636, "grad_norm": 0.1792498230934143, "learning_rate": 2.897821503865074e-05, "loss": 0.031, "step": 9300 }, { "epoch": 6.536191145467322, "grad_norm": 0.2644905149936676, "learning_rate": 2.897774654485828e-05, "loss": 0.0266, "step": 9301 }, { "epoch": 6.536893886156008, "grad_norm": 0.1513611376285553, "learning_rate": 2.8977278051065825e-05, "loss": 0.0164, "step": 9302 }, { "epoch": 6.537596626844694, "grad_norm": 0.3757247030735016, "learning_rate": 2.897680955727337e-05, "loss": 0.0363, "step": 9303 }, { "epoch": 6.53829936753338, "grad_norm": 0.413494348526001, "learning_rate": 2.8976341063480913e-05, "loss": 0.053, "step": 9304 }, { "epoch": 6.539002108222066, "grad_norm": 0.20449595153331757, "learning_rate": 2.897587256968845e-05, "loss": 0.02, "step": 9305 }, { "epoch": 6.539704848910752, "grad_norm": 0.2957925498485565, "learning_rate": 2.8975404075895993e-05, "loss": 0.0477, "step": 9306 }, { "epoch": 6.540407589599438, "grad_norm": 0.20193572342395782, "learning_rate": 2.8974935582103537e-05, "loss": 0.0509, "step": 9307 }, { "epoch": 6.541110330288124, "grad_norm": 0.42114749550819397, "learning_rate": 2.897446708831108e-05, "loss": 0.0881, "step": 9308 }, { "epoch": 6.54181307097681, "grad_norm": 0.4857771098613739, "learning_rate": 2.897399859451862e-05, "loss": 0.0878, "step": 9309 }, { "epoch": 6.542515811665496, "grad_norm": 0.4665406346321106, "learning_rate": 2.8973530100726165e-05, "loss": 0.1314, "step": 9310 }, { "epoch": 6.5432185523541815, "grad_norm": 1.4342327117919922, "learning_rate": 2.897306160693371e-05, "loss": 0.1626, "step": 9311 }, { "epoch": 6.543921293042867, "grad_norm": 1.1817471981048584, "learning_rate": 2.8972593113141252e-05, "loss": 0.1948, "step": 9312 }, { "epoch": 6.544624033731553, "grad_norm": 1.6178374290466309, "learning_rate": 2.8972124619348796e-05, "loss": 0.282, "step": 9313 }, { "epoch": 6.545326774420239, "grad_norm": 0.2406112253665924, "learning_rate": 2.8971656125556336e-05, "loss": 0.0621, "step": 9314 }, { "epoch": 6.546029515108925, "grad_norm": 0.16585488617420197, "learning_rate": 2.897118763176388e-05, "loss": 0.0332, "step": 9315 }, { "epoch": 6.546732255797611, "grad_norm": 0.2138148546218872, "learning_rate": 2.8970719137971424e-05, "loss": 0.0397, "step": 9316 }, { "epoch": 6.547434996486297, "grad_norm": 0.2435459941625595, "learning_rate": 2.8970250644178967e-05, "loss": 0.0302, "step": 9317 }, { "epoch": 6.548137737174983, "grad_norm": 0.12497930973768234, "learning_rate": 2.8969782150386508e-05, "loss": 0.0235, "step": 9318 }, { "epoch": 6.548840477863668, "grad_norm": 0.16873975098133087, "learning_rate": 2.896931365659405e-05, "loss": 0.0276, "step": 9319 }, { "epoch": 6.549543218552354, "grad_norm": 0.11891785264015198, "learning_rate": 2.8968845162801595e-05, "loss": 0.0165, "step": 9320 }, { "epoch": 6.55024595924104, "grad_norm": 0.19951380789279938, "learning_rate": 2.896837666900914e-05, "loss": 0.023, "step": 9321 }, { "epoch": 6.550948699929726, "grad_norm": 0.14106406271457672, "learning_rate": 2.8967908175216676e-05, "loss": 0.0216, "step": 9322 }, { "epoch": 6.551651440618412, "grad_norm": 0.1387389451265335, "learning_rate": 2.896743968142422e-05, "loss": 0.018, "step": 9323 }, { "epoch": 6.552354181307098, "grad_norm": 0.1235949769616127, "learning_rate": 2.8966971187631763e-05, "loss": 0.0215, "step": 9324 }, { "epoch": 6.5530569219957835, "grad_norm": 0.1986522674560547, "learning_rate": 2.8966502693839307e-05, "loss": 0.022, "step": 9325 }, { "epoch": 6.553759662684469, "grad_norm": 0.17136172950267792, "learning_rate": 2.896603420004685e-05, "loss": 0.0315, "step": 9326 }, { "epoch": 6.554462403373155, "grad_norm": 0.4811991751194, "learning_rate": 2.896556570625439e-05, "loss": 0.0225, "step": 9327 }, { "epoch": 6.555165144061841, "grad_norm": 0.2385147362947464, "learning_rate": 2.8965097212461935e-05, "loss": 0.0385, "step": 9328 }, { "epoch": 6.555867884750527, "grad_norm": 0.26407819986343384, "learning_rate": 2.896462871866948e-05, "loss": 0.0352, "step": 9329 }, { "epoch": 6.556570625439213, "grad_norm": 0.14366595447063446, "learning_rate": 2.8964160224877022e-05, "loss": 0.0217, "step": 9330 }, { "epoch": 6.557273366127899, "grad_norm": 0.29753369092941284, "learning_rate": 2.8963691731084563e-05, "loss": 0.0377, "step": 9331 }, { "epoch": 6.557976106816585, "grad_norm": 0.32510262727737427, "learning_rate": 2.8963223237292106e-05, "loss": 0.0572, "step": 9332 }, { "epoch": 6.558678847505271, "grad_norm": 0.2732398211956024, "learning_rate": 2.896275474349965e-05, "loss": 0.039, "step": 9333 }, { "epoch": 6.559381588193957, "grad_norm": 1.022601842880249, "learning_rate": 2.8962286249707194e-05, "loss": 0.068, "step": 9334 }, { "epoch": 6.560084328882642, "grad_norm": 0.4556194543838501, "learning_rate": 2.8961817755914734e-05, "loss": 0.106, "step": 9335 }, { "epoch": 6.560787069571328, "grad_norm": 2.641674280166626, "learning_rate": 2.8961349262122278e-05, "loss": 0.1538, "step": 9336 }, { "epoch": 6.561489810260014, "grad_norm": 1.3098912239074707, "learning_rate": 2.896088076832982e-05, "loss": 0.2342, "step": 9337 }, { "epoch": 6.5621925509487, "grad_norm": 1.611664056777954, "learning_rate": 2.8960412274537365e-05, "loss": 0.2435, "step": 9338 }, { "epoch": 6.5628952916373855, "grad_norm": 0.24672801792621613, "learning_rate": 2.895994378074491e-05, "loss": 0.072, "step": 9339 }, { "epoch": 6.563598032326071, "grad_norm": 0.1752435266971588, "learning_rate": 2.8959475286952446e-05, "loss": 0.0318, "step": 9340 }, { "epoch": 6.564300773014757, "grad_norm": 0.1344715803861618, "learning_rate": 2.895900679315999e-05, "loss": 0.0227, "step": 9341 }, { "epoch": 6.565003513703443, "grad_norm": 0.20926348865032196, "learning_rate": 2.8958538299367533e-05, "loss": 0.0169, "step": 9342 }, { "epoch": 6.565706254392129, "grad_norm": 0.0992153063416481, "learning_rate": 2.8958069805575077e-05, "loss": 0.0154, "step": 9343 }, { "epoch": 6.566408995080815, "grad_norm": 0.14331558346748352, "learning_rate": 2.8957601311782618e-05, "loss": 0.0128, "step": 9344 }, { "epoch": 6.567111735769501, "grad_norm": 0.14056231081485748, "learning_rate": 2.895713281799016e-05, "loss": 0.0227, "step": 9345 }, { "epoch": 6.567814476458187, "grad_norm": 0.16799531877040863, "learning_rate": 2.8956664324197705e-05, "loss": 0.022, "step": 9346 }, { "epoch": 6.568517217146873, "grad_norm": 0.16035661101341248, "learning_rate": 2.895619583040525e-05, "loss": 0.016, "step": 9347 }, { "epoch": 6.569219957835559, "grad_norm": 0.11077723652124405, "learning_rate": 2.895572733661279e-05, "loss": 0.0168, "step": 9348 }, { "epoch": 6.569922698524245, "grad_norm": 0.5643154382705688, "learning_rate": 2.8955258842820333e-05, "loss": 0.031, "step": 9349 }, { "epoch": 6.570625439212931, "grad_norm": 0.2680622935295105, "learning_rate": 2.8954790349027876e-05, "loss": 0.0225, "step": 9350 }, { "epoch": 6.5713281799016166, "grad_norm": 0.24460144340991974, "learning_rate": 2.895432185523542e-05, "loss": 0.0273, "step": 9351 }, { "epoch": 6.5720309205903025, "grad_norm": 0.2558692693710327, "learning_rate": 2.8953853361442964e-05, "loss": 0.0352, "step": 9352 }, { "epoch": 6.572733661278988, "grad_norm": 0.337309330701828, "learning_rate": 2.8953384867650504e-05, "loss": 0.0402, "step": 9353 }, { "epoch": 6.573436401967674, "grad_norm": 0.3243701756000519, "learning_rate": 2.8952916373858048e-05, "loss": 0.0727, "step": 9354 }, { "epoch": 6.57413914265636, "grad_norm": 0.15070740878582, "learning_rate": 2.8952447880065592e-05, "loss": 0.0205, "step": 9355 }, { "epoch": 6.574841883345046, "grad_norm": 0.30733978748321533, "learning_rate": 2.8951979386273135e-05, "loss": 0.0329, "step": 9356 }, { "epoch": 6.575544624033731, "grad_norm": 0.308301717042923, "learning_rate": 2.8951510892480672e-05, "loss": 0.0347, "step": 9357 }, { "epoch": 6.576247364722417, "grad_norm": 0.21055860817432404, "learning_rate": 2.8951042398688216e-05, "loss": 0.0565, "step": 9358 }, { "epoch": 6.576950105411103, "grad_norm": 0.3047139346599579, "learning_rate": 2.895057390489576e-05, "loss": 0.0687, "step": 9359 }, { "epoch": 6.577652846099789, "grad_norm": 0.44916996359825134, "learning_rate": 2.8950105411103304e-05, "loss": 0.1008, "step": 9360 }, { "epoch": 6.578355586788475, "grad_norm": 0.9584413170814514, "learning_rate": 2.8949636917310844e-05, "loss": 0.1418, "step": 9361 }, { "epoch": 6.579058327477161, "grad_norm": 2.6511592864990234, "learning_rate": 2.8949168423518388e-05, "loss": 0.2423, "step": 9362 }, { "epoch": 6.579761068165847, "grad_norm": 1.2630834579467773, "learning_rate": 2.894869992972593e-05, "loss": 0.1936, "step": 9363 }, { "epoch": 6.580463808854533, "grad_norm": 0.5559926629066467, "learning_rate": 2.8948231435933475e-05, "loss": 0.0702, "step": 9364 }, { "epoch": 6.581166549543219, "grad_norm": 0.2324623465538025, "learning_rate": 2.894776294214102e-05, "loss": 0.0229, "step": 9365 }, { "epoch": 6.5818692902319045, "grad_norm": 0.21662846207618713, "learning_rate": 2.894729444834856e-05, "loss": 0.0296, "step": 9366 }, { "epoch": 6.58257203092059, "grad_norm": 0.201810821890831, "learning_rate": 2.8946825954556103e-05, "loss": 0.0404, "step": 9367 }, { "epoch": 6.583274771609276, "grad_norm": 0.21104682981967926, "learning_rate": 2.8946357460763647e-05, "loss": 0.0295, "step": 9368 }, { "epoch": 6.583977512297962, "grad_norm": 0.23297575116157532, "learning_rate": 2.894588896697119e-05, "loss": 0.0111, "step": 9369 }, { "epoch": 6.584680252986648, "grad_norm": 0.11651632189750671, "learning_rate": 2.894542047317873e-05, "loss": 0.0171, "step": 9370 }, { "epoch": 6.585382993675334, "grad_norm": 0.28789764642715454, "learning_rate": 2.8944951979386274e-05, "loss": 0.019, "step": 9371 }, { "epoch": 6.58608573436402, "grad_norm": 0.6044554114341736, "learning_rate": 2.8944483485593818e-05, "loss": 0.0266, "step": 9372 }, { "epoch": 6.586788475052706, "grad_norm": 0.2775452733039856, "learning_rate": 2.8944014991801362e-05, "loss": 0.0374, "step": 9373 }, { "epoch": 6.587491215741391, "grad_norm": 0.154414102435112, "learning_rate": 2.8943546498008902e-05, "loss": 0.0186, "step": 9374 }, { "epoch": 6.588193956430077, "grad_norm": 0.14877204596996307, "learning_rate": 2.8943078004216443e-05, "loss": 0.014, "step": 9375 }, { "epoch": 6.588896697118763, "grad_norm": 0.17558477818965912, "learning_rate": 2.8942609510423986e-05, "loss": 0.033, "step": 9376 }, { "epoch": 6.589599437807449, "grad_norm": 0.10335543006658554, "learning_rate": 2.894214101663153e-05, "loss": 0.0141, "step": 9377 }, { "epoch": 6.590302178496135, "grad_norm": 0.28736311197280884, "learning_rate": 2.8941672522839074e-05, "loss": 0.0312, "step": 9378 }, { "epoch": 6.591004919184821, "grad_norm": 0.19424866139888763, "learning_rate": 2.8941204029046614e-05, "loss": 0.0303, "step": 9379 }, { "epoch": 6.5917076598735065, "grad_norm": 0.2706984877586365, "learning_rate": 2.8940735535254158e-05, "loss": 0.0288, "step": 9380 }, { "epoch": 6.592410400562192, "grad_norm": 0.3364379107952118, "learning_rate": 2.89402670414617e-05, "loss": 0.041, "step": 9381 }, { "epoch": 6.593113141250878, "grad_norm": 0.38501524925231934, "learning_rate": 2.8939798547669245e-05, "loss": 0.0542, "step": 9382 }, { "epoch": 6.593815881939564, "grad_norm": 0.6437138915061951, "learning_rate": 2.8939330053876786e-05, "loss": 0.0903, "step": 9383 }, { "epoch": 6.59451862262825, "grad_norm": 0.35412245988845825, "learning_rate": 2.893886156008433e-05, "loss": 0.0673, "step": 9384 }, { "epoch": 6.595221363316936, "grad_norm": 2.052842855453491, "learning_rate": 2.8938393066291873e-05, "loss": 0.1262, "step": 9385 }, { "epoch": 6.595924104005622, "grad_norm": 0.7484430074691772, "learning_rate": 2.8937924572499417e-05, "loss": 0.1636, "step": 9386 }, { "epoch": 6.596626844694308, "grad_norm": 1.201758623123169, "learning_rate": 2.8937456078706957e-05, "loss": 0.2296, "step": 9387 }, { "epoch": 6.597329585382994, "grad_norm": 1.8262627124786377, "learning_rate": 2.89369875849145e-05, "loss": 0.2663, "step": 9388 }, { "epoch": 6.59803232607168, "grad_norm": 0.30936071276664734, "learning_rate": 2.8936519091122044e-05, "loss": 0.078, "step": 9389 }, { "epoch": 6.598735066760366, "grad_norm": 0.4020874798297882, "learning_rate": 2.8936050597329588e-05, "loss": 0.0439, "step": 9390 }, { "epoch": 6.599437807449052, "grad_norm": 0.1271154135465622, "learning_rate": 2.8935582103537132e-05, "loss": 0.0237, "step": 9391 }, { "epoch": 6.6001405481377375, "grad_norm": 0.16116169095039368, "learning_rate": 2.893511360974467e-05, "loss": 0.0255, "step": 9392 }, { "epoch": 6.6008432888264235, "grad_norm": 0.20970091223716736, "learning_rate": 2.8934645115952213e-05, "loss": 0.0245, "step": 9393 }, { "epoch": 6.601546029515109, "grad_norm": 0.1765429675579071, "learning_rate": 2.8934176622159756e-05, "loss": 0.0157, "step": 9394 }, { "epoch": 6.602248770203794, "grad_norm": 0.19526275992393494, "learning_rate": 2.89337081283673e-05, "loss": 0.0172, "step": 9395 }, { "epoch": 6.60295151089248, "grad_norm": 0.14269782602787018, "learning_rate": 2.893323963457484e-05, "loss": 0.0198, "step": 9396 }, { "epoch": 6.603654251581166, "grad_norm": 0.09836975485086441, "learning_rate": 2.8932771140782384e-05, "loss": 0.0191, "step": 9397 }, { "epoch": 6.604356992269852, "grad_norm": 0.33244243264198303, "learning_rate": 2.8932302646989928e-05, "loss": 0.0144, "step": 9398 }, { "epoch": 6.605059732958538, "grad_norm": 0.233214870095253, "learning_rate": 2.893183415319747e-05, "loss": 0.0447, "step": 9399 }, { "epoch": 6.605762473647224, "grad_norm": 0.1666741818189621, "learning_rate": 2.8931365659405012e-05, "loss": 0.025, "step": 9400 }, { "epoch": 6.60646521433591, "grad_norm": 0.2368612140417099, "learning_rate": 2.8930897165612556e-05, "loss": 0.0227, "step": 9401 }, { "epoch": 6.607167955024596, "grad_norm": 0.31978654861450195, "learning_rate": 2.89304286718201e-05, "loss": 0.0136, "step": 9402 }, { "epoch": 6.607870695713282, "grad_norm": 1.4849985837936401, "learning_rate": 2.8929960178027643e-05, "loss": 0.0244, "step": 9403 }, { "epoch": 6.608573436401968, "grad_norm": 0.24072308838367462, "learning_rate": 2.8929491684235187e-05, "loss": 0.0365, "step": 9404 }, { "epoch": 6.609276177090654, "grad_norm": 0.1661127507686615, "learning_rate": 2.8929023190442727e-05, "loss": 0.0219, "step": 9405 }, { "epoch": 6.6099789177793395, "grad_norm": 0.3399726152420044, "learning_rate": 2.892855469665027e-05, "loss": 0.0769, "step": 9406 }, { "epoch": 6.6106816584680255, "grad_norm": 0.3331923186779022, "learning_rate": 2.8928086202857815e-05, "loss": 0.0511, "step": 9407 }, { "epoch": 6.611384399156711, "grad_norm": 0.30091241002082825, "learning_rate": 2.892761770906536e-05, "loss": 0.0567, "step": 9408 }, { "epoch": 6.612087139845397, "grad_norm": 0.7382115125656128, "learning_rate": 2.89271492152729e-05, "loss": 0.0878, "step": 9409 }, { "epoch": 6.612789880534083, "grad_norm": 0.5722347497940063, "learning_rate": 2.892668072148044e-05, "loss": 0.1206, "step": 9410 }, { "epoch": 6.613492621222769, "grad_norm": 0.9472599625587463, "learning_rate": 2.8926212227687983e-05, "loss": 0.1786, "step": 9411 }, { "epoch": 6.614195361911454, "grad_norm": 1.3372032642364502, "learning_rate": 2.8925743733895526e-05, "loss": 0.208, "step": 9412 }, { "epoch": 6.61489810260014, "grad_norm": 3.077453851699829, "learning_rate": 2.8925275240103067e-05, "loss": 0.2735, "step": 9413 }, { "epoch": 6.615600843288826, "grad_norm": 0.3760818541049957, "learning_rate": 2.892480674631061e-05, "loss": 0.0754, "step": 9414 }, { "epoch": 6.616303583977512, "grad_norm": 0.2958933711051941, "learning_rate": 2.8924338252518154e-05, "loss": 0.027, "step": 9415 }, { "epoch": 6.617006324666198, "grad_norm": 0.28095799684524536, "learning_rate": 2.8923869758725698e-05, "loss": 0.0398, "step": 9416 }, { "epoch": 6.617709065354884, "grad_norm": 0.13403597474098206, "learning_rate": 2.892340126493324e-05, "loss": 0.0177, "step": 9417 }, { "epoch": 6.61841180604357, "grad_norm": 0.1113133504986763, "learning_rate": 2.8922932771140782e-05, "loss": 0.0201, "step": 9418 }, { "epoch": 6.619114546732256, "grad_norm": 0.37479159235954285, "learning_rate": 2.8922464277348326e-05, "loss": 0.0497, "step": 9419 }, { "epoch": 6.6198172874209416, "grad_norm": 0.12055991590023041, "learning_rate": 2.892199578355587e-05, "loss": 0.0202, "step": 9420 }, { "epoch": 6.6205200281096275, "grad_norm": 0.15302209556102753, "learning_rate": 2.8921527289763413e-05, "loss": 0.0167, "step": 9421 }, { "epoch": 6.621222768798313, "grad_norm": 0.138951376080513, "learning_rate": 2.8921058795970954e-05, "loss": 0.0263, "step": 9422 }, { "epoch": 6.621925509486999, "grad_norm": 0.14270047843456268, "learning_rate": 2.8920590302178497e-05, "loss": 0.0192, "step": 9423 }, { "epoch": 6.622628250175685, "grad_norm": 0.12238546460866928, "learning_rate": 2.892012180838604e-05, "loss": 0.022, "step": 9424 }, { "epoch": 6.623330990864371, "grad_norm": 0.20306387543678284, "learning_rate": 2.8919653314593585e-05, "loss": 0.0149, "step": 9425 }, { "epoch": 6.624033731553057, "grad_norm": 0.2672509551048279, "learning_rate": 2.8919184820801125e-05, "loss": 0.0494, "step": 9426 }, { "epoch": 6.624736472241743, "grad_norm": 0.16732102632522583, "learning_rate": 2.8918716327008665e-05, "loss": 0.0203, "step": 9427 }, { "epoch": 6.625439212930429, "grad_norm": 0.3473697006702423, "learning_rate": 2.891824783321621e-05, "loss": 0.045, "step": 9428 }, { "epoch": 6.626141953619115, "grad_norm": 0.7280631065368652, "learning_rate": 2.8917779339423753e-05, "loss": 0.0403, "step": 9429 }, { "epoch": 6.626844694307801, "grad_norm": 0.29358887672424316, "learning_rate": 2.8917310845631297e-05, "loss": 0.0425, "step": 9430 }, { "epoch": 6.627547434996487, "grad_norm": 0.5007850527763367, "learning_rate": 2.8916842351838837e-05, "loss": 0.0528, "step": 9431 }, { "epoch": 6.628250175685173, "grad_norm": 0.29313647747039795, "learning_rate": 2.891637385804638e-05, "loss": 0.0363, "step": 9432 }, { "epoch": 6.6289529163738585, "grad_norm": 0.27782613039016724, "learning_rate": 2.8915905364253924e-05, "loss": 0.0476, "step": 9433 }, { "epoch": 6.629655657062544, "grad_norm": 0.43020355701446533, "learning_rate": 2.8915436870461468e-05, "loss": 0.0827, "step": 9434 }, { "epoch": 6.6303583977512295, "grad_norm": 0.6847310662269592, "learning_rate": 2.891496837666901e-05, "loss": 0.1043, "step": 9435 }, { "epoch": 6.631061138439915, "grad_norm": 0.852137565612793, "learning_rate": 2.8914499882876552e-05, "loss": 0.1767, "step": 9436 }, { "epoch": 6.631763879128601, "grad_norm": 0.825046181678772, "learning_rate": 2.8914031389084096e-05, "loss": 0.1996, "step": 9437 }, { "epoch": 6.632466619817287, "grad_norm": 1.1532262563705444, "learning_rate": 2.891356289529164e-05, "loss": 0.2147, "step": 9438 }, { "epoch": 6.633169360505973, "grad_norm": 0.3465594947338104, "learning_rate": 2.891309440149918e-05, "loss": 0.0825, "step": 9439 }, { "epoch": 6.633872101194659, "grad_norm": 0.24663756787776947, "learning_rate": 2.8912625907706724e-05, "loss": 0.0324, "step": 9440 }, { "epoch": 6.634574841883345, "grad_norm": 0.15179231762886047, "learning_rate": 2.8912157413914267e-05, "loss": 0.021, "step": 9441 }, { "epoch": 6.635277582572031, "grad_norm": 1.1982345581054688, "learning_rate": 2.891168892012181e-05, "loss": 0.0221, "step": 9442 }, { "epoch": 6.635980323260717, "grad_norm": 0.13030578196048737, "learning_rate": 2.8911220426329355e-05, "loss": 0.0223, "step": 9443 }, { "epoch": 6.636683063949403, "grad_norm": 0.2071724385023117, "learning_rate": 2.8910751932536892e-05, "loss": 0.0157, "step": 9444 }, { "epoch": 6.637385804638089, "grad_norm": 0.10218296200037003, "learning_rate": 2.8910283438744436e-05, "loss": 0.0139, "step": 9445 }, { "epoch": 6.638088545326775, "grad_norm": 0.18727248907089233, "learning_rate": 2.890981494495198e-05, "loss": 0.0252, "step": 9446 }, { "epoch": 6.6387912860154605, "grad_norm": 0.1723523736000061, "learning_rate": 2.8909346451159523e-05, "loss": 0.0162, "step": 9447 }, { "epoch": 6.6394940267041465, "grad_norm": 0.23811616003513336, "learning_rate": 2.8908877957367063e-05, "loss": 0.0166, "step": 9448 }, { "epoch": 6.640196767392832, "grad_norm": 0.20985350012779236, "learning_rate": 2.8908409463574607e-05, "loss": 0.0173, "step": 9449 }, { "epoch": 6.640899508081518, "grad_norm": 0.286598801612854, "learning_rate": 2.890794096978215e-05, "loss": 0.0208, "step": 9450 }, { "epoch": 6.641602248770203, "grad_norm": 0.4935636520385742, "learning_rate": 2.8907472475989694e-05, "loss": 0.0567, "step": 9451 }, { "epoch": 6.642304989458889, "grad_norm": 0.21496567130088806, "learning_rate": 2.8907003982197238e-05, "loss": 0.0243, "step": 9452 }, { "epoch": 6.643007730147575, "grad_norm": 0.32370805740356445, "learning_rate": 2.890653548840478e-05, "loss": 0.0404, "step": 9453 }, { "epoch": 6.643710470836261, "grad_norm": 0.2797255218029022, "learning_rate": 2.8906066994612322e-05, "loss": 0.0494, "step": 9454 }, { "epoch": 6.644413211524947, "grad_norm": 0.13098619878292084, "learning_rate": 2.8905598500819866e-05, "loss": 0.0221, "step": 9455 }, { "epoch": 6.645115952213633, "grad_norm": 0.2232225239276886, "learning_rate": 2.890513000702741e-05, "loss": 0.0497, "step": 9456 }, { "epoch": 6.645818692902319, "grad_norm": 0.3142901659011841, "learning_rate": 2.890466151323495e-05, "loss": 0.0581, "step": 9457 }, { "epoch": 6.646521433591005, "grad_norm": 0.48835405707359314, "learning_rate": 2.8904193019442494e-05, "loss": 0.0557, "step": 9458 }, { "epoch": 6.647224174279691, "grad_norm": 0.48993727564811707, "learning_rate": 2.8903724525650037e-05, "loss": 0.0779, "step": 9459 }, { "epoch": 6.647926914968377, "grad_norm": 0.9807680249214172, "learning_rate": 2.890325603185758e-05, "loss": 0.1202, "step": 9460 }, { "epoch": 6.6486296556570625, "grad_norm": 0.8030455112457275, "learning_rate": 2.890278753806512e-05, "loss": 0.1959, "step": 9461 }, { "epoch": 6.6493323963457485, "grad_norm": 1.1342204809188843, "learning_rate": 2.8902319044272662e-05, "loss": 0.1807, "step": 9462 }, { "epoch": 6.650035137034434, "grad_norm": 1.8821271657943726, "learning_rate": 2.8901850550480206e-05, "loss": 0.2827, "step": 9463 }, { "epoch": 6.65073787772312, "grad_norm": 0.3197602331638336, "learning_rate": 2.890138205668775e-05, "loss": 0.0701, "step": 9464 }, { "epoch": 6.651440618411806, "grad_norm": 0.3940140902996063, "learning_rate": 2.8900913562895293e-05, "loss": 0.0378, "step": 9465 }, { "epoch": 6.652143359100492, "grad_norm": 0.38671305775642395, "learning_rate": 2.8900445069102833e-05, "loss": 0.0229, "step": 9466 }, { "epoch": 6.652846099789178, "grad_norm": 0.13451850414276123, "learning_rate": 2.8899976575310377e-05, "loss": 0.0228, "step": 9467 }, { "epoch": 6.653548840477864, "grad_norm": 0.24781417846679688, "learning_rate": 2.889950808151792e-05, "loss": 0.0202, "step": 9468 }, { "epoch": 6.65425158116655, "grad_norm": 0.22352822124958038, "learning_rate": 2.8899039587725465e-05, "loss": 0.0104, "step": 9469 }, { "epoch": 6.654954321855236, "grad_norm": 0.5257812738418579, "learning_rate": 2.8898571093933005e-05, "loss": 0.0292, "step": 9470 }, { "epoch": 6.655657062543922, "grad_norm": 0.15231558680534363, "learning_rate": 2.889810260014055e-05, "loss": 0.0195, "step": 9471 }, { "epoch": 6.656359803232607, "grad_norm": 0.27526170015335083, "learning_rate": 2.8897634106348092e-05, "loss": 0.0232, "step": 9472 }, { "epoch": 6.657062543921293, "grad_norm": 0.17565982043743134, "learning_rate": 2.8897165612555636e-05, "loss": 0.0159, "step": 9473 }, { "epoch": 6.657765284609979, "grad_norm": 0.3368391692638397, "learning_rate": 2.8896697118763176e-05, "loss": 0.0393, "step": 9474 }, { "epoch": 6.6584680252986645, "grad_norm": 0.3674468994140625, "learning_rate": 2.889622862497072e-05, "loss": 0.0206, "step": 9475 }, { "epoch": 6.6591707659873505, "grad_norm": 0.25875037908554077, "learning_rate": 2.8895760131178264e-05, "loss": 0.0226, "step": 9476 }, { "epoch": 6.659873506676036, "grad_norm": 0.6613383293151855, "learning_rate": 2.8895291637385808e-05, "loss": 0.036, "step": 9477 }, { "epoch": 6.660576247364722, "grad_norm": 0.22254320979118347, "learning_rate": 2.889482314359335e-05, "loss": 0.0267, "step": 9478 }, { "epoch": 6.661278988053408, "grad_norm": 0.223870649933815, "learning_rate": 2.8894354649800888e-05, "loss": 0.0283, "step": 9479 }, { "epoch": 6.661981728742094, "grad_norm": 0.4497075080871582, "learning_rate": 2.8893886156008432e-05, "loss": 0.0409, "step": 9480 }, { "epoch": 6.66268446943078, "grad_norm": 0.4223490059375763, "learning_rate": 2.8893417662215976e-05, "loss": 0.0363, "step": 9481 }, { "epoch": 6.663387210119466, "grad_norm": 0.40573257207870483, "learning_rate": 2.889294916842352e-05, "loss": 0.0469, "step": 9482 }, { "epoch": 6.664089950808152, "grad_norm": 0.3707343637943268, "learning_rate": 2.889248067463106e-05, "loss": 0.0474, "step": 9483 }, { "epoch": 6.664792691496838, "grad_norm": 0.6418700814247131, "learning_rate": 2.8892012180838604e-05, "loss": 0.1035, "step": 9484 }, { "epoch": 6.665495432185524, "grad_norm": 0.556657075881958, "learning_rate": 2.8891543687046147e-05, "loss": 0.1539, "step": 9485 }, { "epoch": 6.66619817287421, "grad_norm": 1.0187140703201294, "learning_rate": 2.889107519325369e-05, "loss": 0.1613, "step": 9486 }, { "epoch": 6.666900913562896, "grad_norm": 0.9457624554634094, "learning_rate": 2.889060669946123e-05, "loss": 0.2129, "step": 9487 }, { "epoch": 6.6676036542515815, "grad_norm": 1.9899673461914062, "learning_rate": 2.8890138205668775e-05, "loss": 0.2813, "step": 9488 }, { "epoch": 6.668306394940267, "grad_norm": 0.19926786422729492, "learning_rate": 2.888966971187632e-05, "loss": 0.0679, "step": 9489 }, { "epoch": 6.6690091356289525, "grad_norm": 0.1486273854970932, "learning_rate": 2.8889201218083862e-05, "loss": 0.0298, "step": 9490 }, { "epoch": 6.669711876317638, "grad_norm": 0.16991694271564484, "learning_rate": 2.8888732724291406e-05, "loss": 0.0264, "step": 9491 }, { "epoch": 6.670414617006324, "grad_norm": 0.2813206613063812, "learning_rate": 2.8888264230498947e-05, "loss": 0.0265, "step": 9492 }, { "epoch": 6.67111735769501, "grad_norm": 0.1496962010860443, "learning_rate": 2.888779573670649e-05, "loss": 0.0172, "step": 9493 }, { "epoch": 6.671820098383696, "grad_norm": 0.1891138106584549, "learning_rate": 2.8887327242914034e-05, "loss": 0.0235, "step": 9494 }, { "epoch": 6.672522839072382, "grad_norm": 0.1494869738817215, "learning_rate": 2.8886858749121578e-05, "loss": 0.0284, "step": 9495 }, { "epoch": 6.673225579761068, "grad_norm": 0.1970357447862625, "learning_rate": 2.8886390255329118e-05, "loss": 0.0179, "step": 9496 }, { "epoch": 6.673928320449754, "grad_norm": 0.18883249163627625, "learning_rate": 2.888592176153666e-05, "loss": 0.042, "step": 9497 }, { "epoch": 6.67463106113844, "grad_norm": 0.1275314837694168, "learning_rate": 2.8885453267744202e-05, "loss": 0.0162, "step": 9498 }, { "epoch": 6.675333801827126, "grad_norm": 0.34983596205711365, "learning_rate": 2.8884984773951746e-05, "loss": 0.0282, "step": 9499 }, { "epoch": 6.676036542515812, "grad_norm": 0.15520071983337402, "learning_rate": 2.8884516280159286e-05, "loss": 0.0232, "step": 9500 }, { "epoch": 6.676739283204498, "grad_norm": 0.283993124961853, "learning_rate": 2.888404778636683e-05, "loss": 0.0373, "step": 9501 }, { "epoch": 6.6774420238931835, "grad_norm": 0.16477371752262115, "learning_rate": 2.8883579292574374e-05, "loss": 0.0207, "step": 9502 }, { "epoch": 6.6781447645818695, "grad_norm": 0.4945085346698761, "learning_rate": 2.8883110798781917e-05, "loss": 0.0414, "step": 9503 }, { "epoch": 6.678847505270555, "grad_norm": 0.3321438133716583, "learning_rate": 2.888264230498946e-05, "loss": 0.0453, "step": 9504 }, { "epoch": 6.679550245959241, "grad_norm": 0.23587870597839355, "learning_rate": 2.8882173811197e-05, "loss": 0.0174, "step": 9505 }, { "epoch": 6.680252986647927, "grad_norm": 0.24094976484775543, "learning_rate": 2.8881705317404545e-05, "loss": 0.0414, "step": 9506 }, { "epoch": 6.680955727336613, "grad_norm": 0.2876430153846741, "learning_rate": 2.888123682361209e-05, "loss": 0.0351, "step": 9507 }, { "epoch": 6.681658468025299, "grad_norm": 0.3594822585582733, "learning_rate": 2.8880768329819633e-05, "loss": 0.0722, "step": 9508 }, { "epoch": 6.682361208713985, "grad_norm": 0.38194143772125244, "learning_rate": 2.8880299836027173e-05, "loss": 0.1039, "step": 9509 }, { "epoch": 6.683063949402671, "grad_norm": 0.49685972929000854, "learning_rate": 2.8879831342234717e-05, "loss": 0.1733, "step": 9510 }, { "epoch": 6.683766690091356, "grad_norm": 1.3563613891601562, "learning_rate": 2.887936284844226e-05, "loss": 0.1818, "step": 9511 }, { "epoch": 6.684469430780042, "grad_norm": 2.294994592666626, "learning_rate": 2.8878894354649804e-05, "loss": 0.2273, "step": 9512 }, { "epoch": 6.685172171468728, "grad_norm": 1.2205166816711426, "learning_rate": 2.8878425860857344e-05, "loss": 0.3146, "step": 9513 }, { "epoch": 6.685874912157414, "grad_norm": 0.2581140398979187, "learning_rate": 2.8877957367064885e-05, "loss": 0.0706, "step": 9514 }, { "epoch": 6.6865776528461, "grad_norm": 0.1969408243894577, "learning_rate": 2.887748887327243e-05, "loss": 0.0387, "step": 9515 }, { "epoch": 6.6872803935347855, "grad_norm": 0.14361527562141418, "learning_rate": 2.8877020379479972e-05, "loss": 0.0197, "step": 9516 }, { "epoch": 6.6879831342234715, "grad_norm": 0.19115567207336426, "learning_rate": 2.8876551885687516e-05, "loss": 0.0259, "step": 9517 }, { "epoch": 6.688685874912157, "grad_norm": 0.13245117664337158, "learning_rate": 2.8876083391895056e-05, "loss": 0.0159, "step": 9518 }, { "epoch": 6.689388615600843, "grad_norm": 0.1818915754556656, "learning_rate": 2.88756148981026e-05, "loss": 0.0225, "step": 9519 }, { "epoch": 6.690091356289529, "grad_norm": 0.21065691113471985, "learning_rate": 2.8875146404310144e-05, "loss": 0.0163, "step": 9520 }, { "epoch": 6.690794096978215, "grad_norm": 0.22168651223182678, "learning_rate": 2.8874677910517687e-05, "loss": 0.0312, "step": 9521 }, { "epoch": 6.691496837666901, "grad_norm": 0.2124532014131546, "learning_rate": 2.8874209416725228e-05, "loss": 0.0254, "step": 9522 }, { "epoch": 6.692199578355587, "grad_norm": 0.2153923213481903, "learning_rate": 2.887374092293277e-05, "loss": 0.0208, "step": 9523 }, { "epoch": 6.692902319044273, "grad_norm": 0.98543381690979, "learning_rate": 2.8873272429140315e-05, "loss": 0.0338, "step": 9524 }, { "epoch": 6.693605059732959, "grad_norm": 0.1724800318479538, "learning_rate": 2.887280393534786e-05, "loss": 0.014, "step": 9525 }, { "epoch": 6.694307800421645, "grad_norm": 0.20008154213428497, "learning_rate": 2.88723354415554e-05, "loss": 0.0308, "step": 9526 }, { "epoch": 6.695010541110331, "grad_norm": 0.19979698956012726, "learning_rate": 2.8871866947762943e-05, "loss": 0.0285, "step": 9527 }, { "epoch": 6.695713281799016, "grad_norm": 0.19107818603515625, "learning_rate": 2.8871398453970487e-05, "loss": 0.0189, "step": 9528 }, { "epoch": 6.696416022487702, "grad_norm": 0.20033682882785797, "learning_rate": 2.887092996017803e-05, "loss": 0.0252, "step": 9529 }, { "epoch": 6.6971187631763875, "grad_norm": 0.16572503745555878, "learning_rate": 2.8870461466385574e-05, "loss": 0.0351, "step": 9530 }, { "epoch": 6.6978215038650735, "grad_norm": 0.22441436350345612, "learning_rate": 2.886999297259311e-05, "loss": 0.0291, "step": 9531 }, { "epoch": 6.698524244553759, "grad_norm": 0.5848424434661865, "learning_rate": 2.8869524478800655e-05, "loss": 0.0526, "step": 9532 }, { "epoch": 6.699226985242445, "grad_norm": 0.5176540613174438, "learning_rate": 2.88690559850082e-05, "loss": 0.0743, "step": 9533 }, { "epoch": 6.699929725931131, "grad_norm": 0.33384940028190613, "learning_rate": 2.8868587491215742e-05, "loss": 0.0749, "step": 9534 }, { "epoch": 6.700632466619817, "grad_norm": 0.6762542128562927, "learning_rate": 2.8868118997423283e-05, "loss": 0.0968, "step": 9535 }, { "epoch": 6.701335207308503, "grad_norm": 0.5361218452453613, "learning_rate": 2.8867650503630826e-05, "loss": 0.1652, "step": 9536 }, { "epoch": 6.702037947997189, "grad_norm": 1.1085619926452637, "learning_rate": 2.886718200983837e-05, "loss": 0.1873, "step": 9537 }, { "epoch": 6.702740688685875, "grad_norm": 1.38768470287323, "learning_rate": 2.8866713516045914e-05, "loss": 0.2608, "step": 9538 }, { "epoch": 6.703443429374561, "grad_norm": 0.263205885887146, "learning_rate": 2.8866245022253454e-05, "loss": 0.0718, "step": 9539 }, { "epoch": 6.704146170063247, "grad_norm": 0.1414055973291397, "learning_rate": 2.8865776528460998e-05, "loss": 0.0306, "step": 9540 }, { "epoch": 6.704848910751933, "grad_norm": 0.7476301789283752, "learning_rate": 2.886530803466854e-05, "loss": 0.0316, "step": 9541 }, { "epoch": 6.705551651440619, "grad_norm": 0.21057961881160736, "learning_rate": 2.8864839540876085e-05, "loss": 0.0187, "step": 9542 }, { "epoch": 6.7062543921293045, "grad_norm": 0.0987665206193924, "learning_rate": 2.886437104708363e-05, "loss": 0.0228, "step": 9543 }, { "epoch": 6.70695713281799, "grad_norm": 0.0992768332362175, "learning_rate": 2.886390255329117e-05, "loss": 0.0137, "step": 9544 }, { "epoch": 6.707659873506676, "grad_norm": 0.21293821930885315, "learning_rate": 2.8863434059498713e-05, "loss": 0.0371, "step": 9545 }, { "epoch": 6.708362614195362, "grad_norm": 0.2200496792793274, "learning_rate": 2.8862965565706257e-05, "loss": 0.026, "step": 9546 }, { "epoch": 6.709065354884048, "grad_norm": 0.12623418867588043, "learning_rate": 2.88624970719138e-05, "loss": 0.0155, "step": 9547 }, { "epoch": 6.709768095572734, "grad_norm": 0.20081397891044617, "learning_rate": 2.886202857812134e-05, "loss": 0.0205, "step": 9548 }, { "epoch": 6.710470836261419, "grad_norm": 0.34459832310676575, "learning_rate": 2.886156008432888e-05, "loss": 0.031, "step": 9549 }, { "epoch": 6.711173576950105, "grad_norm": 0.1689562201499939, "learning_rate": 2.8861091590536425e-05, "loss": 0.0152, "step": 9550 }, { "epoch": 6.711876317638791, "grad_norm": 0.21286101639270782, "learning_rate": 2.886062309674397e-05, "loss": 0.0303, "step": 9551 }, { "epoch": 6.712579058327477, "grad_norm": 0.19283427298069, "learning_rate": 2.886015460295151e-05, "loss": 0.0316, "step": 9552 }, { "epoch": 6.713281799016163, "grad_norm": 0.1532091647386551, "learning_rate": 2.8859686109159053e-05, "loss": 0.032, "step": 9553 }, { "epoch": 6.713984539704849, "grad_norm": 0.6736863851547241, "learning_rate": 2.8859217615366597e-05, "loss": 0.0286, "step": 9554 }, { "epoch": 6.714687280393535, "grad_norm": 0.12470236420631409, "learning_rate": 2.885874912157414e-05, "loss": 0.0157, "step": 9555 }, { "epoch": 6.715390021082221, "grad_norm": 0.27193447947502136, "learning_rate": 2.8858280627781684e-05, "loss": 0.0526, "step": 9556 }, { "epoch": 6.7160927617709065, "grad_norm": 0.3202257752418518, "learning_rate": 2.8857812133989224e-05, "loss": 0.0496, "step": 9557 }, { "epoch": 6.7167955024595924, "grad_norm": 0.3012125492095947, "learning_rate": 2.8857343640196768e-05, "loss": 0.0588, "step": 9558 }, { "epoch": 6.717498243148278, "grad_norm": 0.4872449040412903, "learning_rate": 2.8856875146404312e-05, "loss": 0.0858, "step": 9559 }, { "epoch": 6.718200983836964, "grad_norm": 0.6016753315925598, "learning_rate": 2.8856406652611855e-05, "loss": 0.1482, "step": 9560 }, { "epoch": 6.71890372452565, "grad_norm": 0.7019188404083252, "learning_rate": 2.8855938158819396e-05, "loss": 0.1589, "step": 9561 }, { "epoch": 6.719606465214336, "grad_norm": 0.9459525942802429, "learning_rate": 2.885546966502694e-05, "loss": 0.1978, "step": 9562 }, { "epoch": 6.720309205903022, "grad_norm": 1.4779444932937622, "learning_rate": 2.8855001171234483e-05, "loss": 0.2264, "step": 9563 }, { "epoch": 6.721011946591708, "grad_norm": 0.30390873551368713, "learning_rate": 2.8854532677442027e-05, "loss": 0.0953, "step": 9564 }, { "epoch": 6.721714687280394, "grad_norm": 0.24072174727916718, "learning_rate": 2.8854064183649567e-05, "loss": 0.0555, "step": 9565 }, { "epoch": 6.722417427969079, "grad_norm": 0.14916542172431946, "learning_rate": 2.8853595689857108e-05, "loss": 0.0267, "step": 9566 }, { "epoch": 6.723120168657765, "grad_norm": 0.1516915261745453, "learning_rate": 2.885312719606465e-05, "loss": 0.0229, "step": 9567 }, { "epoch": 6.723822909346451, "grad_norm": 0.16888393461704254, "learning_rate": 2.8852658702272195e-05, "loss": 0.0199, "step": 9568 }, { "epoch": 6.724525650035137, "grad_norm": 0.15931178629398346, "learning_rate": 2.885219020847974e-05, "loss": 0.0129, "step": 9569 }, { "epoch": 6.725228390723823, "grad_norm": 0.13790962100028992, "learning_rate": 2.885172171468728e-05, "loss": 0.0154, "step": 9570 }, { "epoch": 6.7259311314125085, "grad_norm": 0.1736249327659607, "learning_rate": 2.8851253220894823e-05, "loss": 0.0288, "step": 9571 }, { "epoch": 6.7266338721011945, "grad_norm": 0.1457759588956833, "learning_rate": 2.8850784727102367e-05, "loss": 0.0213, "step": 9572 }, { "epoch": 6.72733661278988, "grad_norm": 0.20495954155921936, "learning_rate": 2.885031623330991e-05, "loss": 0.0223, "step": 9573 }, { "epoch": 6.728039353478566, "grad_norm": 0.21788464486598969, "learning_rate": 2.884984773951745e-05, "loss": 0.0283, "step": 9574 }, { "epoch": 6.728742094167252, "grad_norm": 0.13742905855178833, "learning_rate": 2.8849379245724994e-05, "loss": 0.0179, "step": 9575 }, { "epoch": 6.729444834855938, "grad_norm": 0.35942885279655457, "learning_rate": 2.8848910751932538e-05, "loss": 0.0501, "step": 9576 }, { "epoch": 6.730147575544624, "grad_norm": 0.11436717957258224, "learning_rate": 2.8848442258140082e-05, "loss": 0.0144, "step": 9577 }, { "epoch": 6.73085031623331, "grad_norm": 0.27274468541145325, "learning_rate": 2.8847973764347622e-05, "loss": 0.0233, "step": 9578 }, { "epoch": 6.731553056921996, "grad_norm": 0.2710745334625244, "learning_rate": 2.8847505270555166e-05, "loss": 0.0519, "step": 9579 }, { "epoch": 6.732255797610682, "grad_norm": 0.4669567346572876, "learning_rate": 2.884703677676271e-05, "loss": 0.026, "step": 9580 }, { "epoch": 6.732958538299368, "grad_norm": 0.20807811617851257, "learning_rate": 2.8846568282970253e-05, "loss": 0.0576, "step": 9581 }, { "epoch": 6.733661278988054, "grad_norm": 0.17588277161121368, "learning_rate": 2.8846099789177797e-05, "loss": 0.0274, "step": 9582 }, { "epoch": 6.73436401967674, "grad_norm": 0.47078585624694824, "learning_rate": 2.8845631295385337e-05, "loss": 0.042, "step": 9583 }, { "epoch": 6.7350667603654255, "grad_norm": 0.46610620617866516, "learning_rate": 2.8845162801592878e-05, "loss": 0.092, "step": 9584 }, { "epoch": 6.735769501054111, "grad_norm": 0.7377228140830994, "learning_rate": 2.884469430780042e-05, "loss": 0.1096, "step": 9585 }, { "epoch": 6.736472241742797, "grad_norm": 0.7377826571464539, "learning_rate": 2.8844225814007965e-05, "loss": 0.1474, "step": 9586 }, { "epoch": 6.737174982431483, "grad_norm": 0.6772422790527344, "learning_rate": 2.8843757320215506e-05, "loss": 0.1793, "step": 9587 }, { "epoch": 6.737877723120168, "grad_norm": 1.4421591758728027, "learning_rate": 2.884328882642305e-05, "loss": 0.2842, "step": 9588 }, { "epoch": 6.738580463808854, "grad_norm": 0.49203938245773315, "learning_rate": 2.8842820332630593e-05, "loss": 0.1235, "step": 9589 }, { "epoch": 6.73928320449754, "grad_norm": 0.2012375146150589, "learning_rate": 2.8842351838838137e-05, "loss": 0.0278, "step": 9590 }, { "epoch": 6.739985945186226, "grad_norm": 0.12519031763076782, "learning_rate": 2.8841883345045677e-05, "loss": 0.0286, "step": 9591 }, { "epoch": 6.740688685874912, "grad_norm": 0.3699638545513153, "learning_rate": 2.884141485125322e-05, "loss": 0.0456, "step": 9592 }, { "epoch": 6.741391426563598, "grad_norm": 0.1833682358264923, "learning_rate": 2.8840946357460765e-05, "loss": 0.0213, "step": 9593 }, { "epoch": 6.742094167252284, "grad_norm": 0.16939552128314972, "learning_rate": 2.8840477863668308e-05, "loss": 0.0207, "step": 9594 }, { "epoch": 6.74279690794097, "grad_norm": 0.197504922747612, "learning_rate": 2.8840009369875852e-05, "loss": 0.0256, "step": 9595 }, { "epoch": 6.743499648629656, "grad_norm": 0.2662532925605774, "learning_rate": 2.8839540876083392e-05, "loss": 0.0312, "step": 9596 }, { "epoch": 6.744202389318342, "grad_norm": 0.1929301619529724, "learning_rate": 2.8839072382290936e-05, "loss": 0.0493, "step": 9597 }, { "epoch": 6.7449051300070275, "grad_norm": 0.13385213911533356, "learning_rate": 2.883860388849848e-05, "loss": 0.0145, "step": 9598 }, { "epoch": 6.745607870695713, "grad_norm": 0.269972026348114, "learning_rate": 2.8838135394706023e-05, "loss": 0.0456, "step": 9599 }, { "epoch": 6.746310611384399, "grad_norm": 0.1528766006231308, "learning_rate": 2.8837666900913564e-05, "loss": 0.0218, "step": 9600 }, { "epoch": 6.747013352073085, "grad_norm": 0.13651929795742035, "learning_rate": 2.8837198407121104e-05, "loss": 0.0328, "step": 9601 }, { "epoch": 6.747716092761771, "grad_norm": 0.17703144252300262, "learning_rate": 2.8836729913328648e-05, "loss": 0.0114, "step": 9602 }, { "epoch": 6.748418833450457, "grad_norm": 0.3757634460926056, "learning_rate": 2.883626141953619e-05, "loss": 0.0299, "step": 9603 }, { "epoch": 6.749121574139143, "grad_norm": 0.17968250811100006, "learning_rate": 2.8835792925743732e-05, "loss": 0.0338, "step": 9604 }, { "epoch": 6.749824314827828, "grad_norm": 0.31521543860435486, "learning_rate": 2.8835324431951276e-05, "loss": 0.0323, "step": 9605 }, { "epoch": 6.750527055516514, "grad_norm": 0.2650817334651947, "learning_rate": 2.883485593815882e-05, "loss": 0.0416, "step": 9606 }, { "epoch": 6.7512297962052, "grad_norm": 0.32796692848205566, "learning_rate": 2.8834387444366363e-05, "loss": 0.0628, "step": 9607 }, { "epoch": 6.751932536893886, "grad_norm": 0.27340009808540344, "learning_rate": 2.8833918950573907e-05, "loss": 0.0446, "step": 9608 }, { "epoch": 6.752635277582572, "grad_norm": 0.3083624243736267, "learning_rate": 2.8833450456781447e-05, "loss": 0.0701, "step": 9609 }, { "epoch": 6.753338018271258, "grad_norm": 0.48012804985046387, "learning_rate": 2.883298196298899e-05, "loss": 0.1472, "step": 9610 }, { "epoch": 6.754040758959944, "grad_norm": 0.6369045972824097, "learning_rate": 2.8832513469196535e-05, "loss": 0.1528, "step": 9611 }, { "epoch": 6.7547434996486295, "grad_norm": 0.752223551273346, "learning_rate": 2.883204497540408e-05, "loss": 0.1933, "step": 9612 }, { "epoch": 6.7554462403373154, "grad_norm": 2.0820109844207764, "learning_rate": 2.883157648161162e-05, "loss": 0.2569, "step": 9613 }, { "epoch": 6.756148981026001, "grad_norm": 0.3717661499977112, "learning_rate": 2.8831107987819162e-05, "loss": 0.0741, "step": 9614 }, { "epoch": 6.756851721714687, "grad_norm": 0.281486839056015, "learning_rate": 2.8830639494026706e-05, "loss": 0.0301, "step": 9615 }, { "epoch": 6.757554462403373, "grad_norm": 0.18585103750228882, "learning_rate": 2.883017100023425e-05, "loss": 0.0246, "step": 9616 }, { "epoch": 6.758257203092059, "grad_norm": 0.15056215226650238, "learning_rate": 2.882970250644179e-05, "loss": 0.0194, "step": 9617 }, { "epoch": 6.758959943780745, "grad_norm": 0.09837869554758072, "learning_rate": 2.8829234012649334e-05, "loss": 0.0184, "step": 9618 }, { "epoch": 6.759662684469431, "grad_norm": 0.17001888155937195, "learning_rate": 2.8828765518856874e-05, "loss": 0.0219, "step": 9619 }, { "epoch": 6.760365425158117, "grad_norm": 0.18336977064609528, "learning_rate": 2.8828297025064418e-05, "loss": 0.0187, "step": 9620 }, { "epoch": 6.761068165846803, "grad_norm": 0.15287895500659943, "learning_rate": 2.8827828531271962e-05, "loss": 0.0192, "step": 9621 }, { "epoch": 6.761770906535489, "grad_norm": 0.1587020754814148, "learning_rate": 2.8827360037479502e-05, "loss": 0.0327, "step": 9622 }, { "epoch": 6.762473647224175, "grad_norm": 0.26634711027145386, "learning_rate": 2.8826891543687046e-05, "loss": 0.0172, "step": 9623 }, { "epoch": 6.763176387912861, "grad_norm": 0.24684278666973114, "learning_rate": 2.882642304989459e-05, "loss": 0.0255, "step": 9624 }, { "epoch": 6.7638791286015465, "grad_norm": 0.3173573911190033, "learning_rate": 2.8825954556102133e-05, "loss": 0.0239, "step": 9625 }, { "epoch": 6.7645818692902315, "grad_norm": 0.17237766087055206, "learning_rate": 2.8825486062309674e-05, "loss": 0.0235, "step": 9626 }, { "epoch": 6.7652846099789175, "grad_norm": 0.171481192111969, "learning_rate": 2.8825017568517217e-05, "loss": 0.016, "step": 9627 }, { "epoch": 6.765987350667603, "grad_norm": 0.4748392701148987, "learning_rate": 2.882454907472476e-05, "loss": 0.0367, "step": 9628 }, { "epoch": 6.766690091356289, "grad_norm": 0.18890303373336792, "learning_rate": 2.8824080580932305e-05, "loss": 0.0462, "step": 9629 }, { "epoch": 6.767392832044975, "grad_norm": 0.1219087466597557, "learning_rate": 2.8823612087139845e-05, "loss": 0.0144, "step": 9630 }, { "epoch": 6.768095572733661, "grad_norm": 0.34864163398742676, "learning_rate": 2.882314359334739e-05, "loss": 0.0405, "step": 9631 }, { "epoch": 6.768798313422347, "grad_norm": 0.6238731145858765, "learning_rate": 2.8822675099554933e-05, "loss": 0.0644, "step": 9632 }, { "epoch": 6.769501054111033, "grad_norm": 0.3749718964099884, "learning_rate": 2.8822206605762476e-05, "loss": 0.0397, "step": 9633 }, { "epoch": 6.770203794799719, "grad_norm": 0.5963505506515503, "learning_rate": 2.882173811197002e-05, "loss": 0.1168, "step": 9634 }, { "epoch": 6.770906535488405, "grad_norm": 1.0118635892868042, "learning_rate": 2.882126961817756e-05, "loss": 0.1449, "step": 9635 }, { "epoch": 6.771609276177091, "grad_norm": 0.5704433917999268, "learning_rate": 2.88208011243851e-05, "loss": 0.1809, "step": 9636 }, { "epoch": 6.772312016865777, "grad_norm": 0.8269198536872864, "learning_rate": 2.8820332630592644e-05, "loss": 0.2251, "step": 9637 }, { "epoch": 6.773014757554463, "grad_norm": 1.843976616859436, "learning_rate": 2.8819864136800188e-05, "loss": 0.226, "step": 9638 }, { "epoch": 6.7737174982431485, "grad_norm": 0.22807589173316956, "learning_rate": 2.881939564300773e-05, "loss": 0.078, "step": 9639 }, { "epoch": 6.774420238931834, "grad_norm": 0.19571605324745178, "learning_rate": 2.8818927149215272e-05, "loss": 0.0329, "step": 9640 }, { "epoch": 6.77512297962052, "grad_norm": 0.12880635261535645, "learning_rate": 2.8818458655422816e-05, "loss": 0.0244, "step": 9641 }, { "epoch": 6.775825720309206, "grad_norm": 0.38631951808929443, "learning_rate": 2.881799016163036e-05, "loss": 0.0199, "step": 9642 }, { "epoch": 6.776528460997891, "grad_norm": 0.15890303254127502, "learning_rate": 2.88175216678379e-05, "loss": 0.0228, "step": 9643 }, { "epoch": 6.777231201686577, "grad_norm": 0.2591906487941742, "learning_rate": 2.8817053174045444e-05, "loss": 0.0198, "step": 9644 }, { "epoch": 6.777933942375263, "grad_norm": 0.3230529725551605, "learning_rate": 2.8816584680252987e-05, "loss": 0.0326, "step": 9645 }, { "epoch": 6.778636683063949, "grad_norm": 0.15630696713924408, "learning_rate": 2.881611618646053e-05, "loss": 0.029, "step": 9646 }, { "epoch": 6.779339423752635, "grad_norm": 0.13127721846103668, "learning_rate": 2.8815647692668075e-05, "loss": 0.0339, "step": 9647 }, { "epoch": 6.780042164441321, "grad_norm": 0.12014685571193695, "learning_rate": 2.8815179198875615e-05, "loss": 0.016, "step": 9648 }, { "epoch": 6.780744905130007, "grad_norm": 0.16961917281150818, "learning_rate": 2.881471070508316e-05, "loss": 0.0388, "step": 9649 }, { "epoch": 6.781447645818693, "grad_norm": 0.16997790336608887, "learning_rate": 2.8814242211290703e-05, "loss": 0.0161, "step": 9650 }, { "epoch": 6.782150386507379, "grad_norm": 0.23846465349197388, "learning_rate": 2.8813773717498246e-05, "loss": 0.0399, "step": 9651 }, { "epoch": 6.782853127196065, "grad_norm": 0.13381795585155487, "learning_rate": 2.8813305223705787e-05, "loss": 0.0165, "step": 9652 }, { "epoch": 6.7835558678847505, "grad_norm": 0.1603379100561142, "learning_rate": 2.8812836729913327e-05, "loss": 0.0444, "step": 9653 }, { "epoch": 6.784258608573436, "grad_norm": 0.2013673335313797, "learning_rate": 2.881236823612087e-05, "loss": 0.0475, "step": 9654 }, { "epoch": 6.784961349262122, "grad_norm": 0.30031824111938477, "learning_rate": 2.8811899742328415e-05, "loss": 0.0231, "step": 9655 }, { "epoch": 6.785664089950808, "grad_norm": 0.1962524652481079, "learning_rate": 2.8811431248535958e-05, "loss": 0.0292, "step": 9656 }, { "epoch": 6.786366830639494, "grad_norm": 0.2016938328742981, "learning_rate": 2.88109627547435e-05, "loss": 0.0278, "step": 9657 }, { "epoch": 6.78706957132818, "grad_norm": 0.26713958382606506, "learning_rate": 2.8810494260951042e-05, "loss": 0.0636, "step": 9658 }, { "epoch": 6.787772312016866, "grad_norm": 0.33154013752937317, "learning_rate": 2.8810025767158586e-05, "loss": 0.0803, "step": 9659 }, { "epoch": 6.788475052705552, "grad_norm": 0.5483110547065735, "learning_rate": 2.880955727336613e-05, "loss": 0.1025, "step": 9660 }, { "epoch": 6.789177793394238, "grad_norm": 1.2785807847976685, "learning_rate": 2.880908877957367e-05, "loss": 0.169, "step": 9661 }, { "epoch": 6.789880534082924, "grad_norm": 1.3452118635177612, "learning_rate": 2.8808620285781214e-05, "loss": 0.2154, "step": 9662 }, { "epoch": 6.79058327477161, "grad_norm": 4.512238025665283, "learning_rate": 2.8808151791988758e-05, "loss": 0.2802, "step": 9663 }, { "epoch": 6.791286015460296, "grad_norm": 0.2675861120223999, "learning_rate": 2.88076832981963e-05, "loss": 0.0782, "step": 9664 }, { "epoch": 6.791988756148981, "grad_norm": 0.24475830793380737, "learning_rate": 2.880721480440384e-05, "loss": 0.0406, "step": 9665 }, { "epoch": 6.792691496837667, "grad_norm": 0.2799290716648102, "learning_rate": 2.8806746310611385e-05, "loss": 0.0393, "step": 9666 }, { "epoch": 6.7933942375263525, "grad_norm": 0.20718522369861603, "learning_rate": 2.880627781681893e-05, "loss": 0.0268, "step": 9667 }, { "epoch": 6.794096978215038, "grad_norm": 0.15136098861694336, "learning_rate": 2.8805809323026473e-05, "loss": 0.0293, "step": 9668 }, { "epoch": 6.794799718903724, "grad_norm": 0.22101865708827972, "learning_rate": 2.8805340829234013e-05, "loss": 0.0144, "step": 9669 }, { "epoch": 6.79550245959241, "grad_norm": 0.13961240649223328, "learning_rate": 2.8804872335441557e-05, "loss": 0.0186, "step": 9670 }, { "epoch": 6.796205200281096, "grad_norm": 0.4101000130176544, "learning_rate": 2.8804403841649097e-05, "loss": 0.0255, "step": 9671 }, { "epoch": 6.796907940969782, "grad_norm": 0.16209624707698822, "learning_rate": 2.880393534785664e-05, "loss": 0.0272, "step": 9672 }, { "epoch": 6.797610681658468, "grad_norm": 0.14650337398052216, "learning_rate": 2.8803466854064185e-05, "loss": 0.0148, "step": 9673 }, { "epoch": 6.798313422347154, "grad_norm": 0.5203820466995239, "learning_rate": 2.8802998360271725e-05, "loss": 0.0288, "step": 9674 }, { "epoch": 6.79901616303584, "grad_norm": 0.23030886054039001, "learning_rate": 2.880252986647927e-05, "loss": 0.0285, "step": 9675 }, { "epoch": 6.799718903724526, "grad_norm": 0.24843476712703705, "learning_rate": 2.8802061372686812e-05, "loss": 0.0358, "step": 9676 }, { "epoch": 6.800421644413212, "grad_norm": 0.2073165476322174, "learning_rate": 2.8801592878894356e-05, "loss": 0.0171, "step": 9677 }, { "epoch": 6.801124385101898, "grad_norm": 0.28078728914260864, "learning_rate": 2.8801124385101896e-05, "loss": 0.0348, "step": 9678 }, { "epoch": 6.801827125790584, "grad_norm": 0.3125776946544647, "learning_rate": 2.880065589130944e-05, "loss": 0.0382, "step": 9679 }, { "epoch": 6.8025298664792695, "grad_norm": 0.20134077966213226, "learning_rate": 2.8800187397516984e-05, "loss": 0.0302, "step": 9680 }, { "epoch": 6.8032326071679545, "grad_norm": 0.36880165338516235, "learning_rate": 2.8799718903724528e-05, "loss": 0.031, "step": 9681 }, { "epoch": 6.8039353478566404, "grad_norm": 0.4630239009857178, "learning_rate": 2.879925040993207e-05, "loss": 0.0518, "step": 9682 }, { "epoch": 6.804638088545326, "grad_norm": 1.0639369487762451, "learning_rate": 2.8798781916139612e-05, "loss": 0.0512, "step": 9683 }, { "epoch": 6.805340829234012, "grad_norm": 0.4394376575946808, "learning_rate": 2.8798313422347155e-05, "loss": 0.1065, "step": 9684 }, { "epoch": 6.806043569922698, "grad_norm": 2.145711898803711, "learning_rate": 2.87978449285547e-05, "loss": 0.1574, "step": 9685 }, { "epoch": 6.806746310611384, "grad_norm": 0.727323591709137, "learning_rate": 2.8797376434762243e-05, "loss": 0.136, "step": 9686 }, { "epoch": 6.80744905130007, "grad_norm": 1.232437252998352, "learning_rate": 2.8796907940969783e-05, "loss": 0.2159, "step": 9687 }, { "epoch": 6.808151791988756, "grad_norm": 1.3009209632873535, "learning_rate": 2.8796439447177324e-05, "loss": 0.2593, "step": 9688 }, { "epoch": 6.808854532677442, "grad_norm": 0.2890832722187042, "learning_rate": 2.8795970953384867e-05, "loss": 0.1167, "step": 9689 }, { "epoch": 6.809557273366128, "grad_norm": 0.12811224162578583, "learning_rate": 2.879550245959241e-05, "loss": 0.0238, "step": 9690 }, { "epoch": 6.810260014054814, "grad_norm": 0.2048344910144806, "learning_rate": 2.879503396579995e-05, "loss": 0.0346, "step": 9691 }, { "epoch": 6.8109627547435, "grad_norm": 0.23334453999996185, "learning_rate": 2.8794565472007495e-05, "loss": 0.0414, "step": 9692 }, { "epoch": 6.811665495432186, "grad_norm": 0.09026803821325302, "learning_rate": 2.879409697821504e-05, "loss": 0.0172, "step": 9693 }, { "epoch": 6.8123682361208715, "grad_norm": 0.17655049264431, "learning_rate": 2.8793628484422583e-05, "loss": 0.0189, "step": 9694 }, { "epoch": 6.813070976809557, "grad_norm": 0.16547292470932007, "learning_rate": 2.8793159990630126e-05, "loss": 0.0221, "step": 9695 }, { "epoch": 6.813773717498243, "grad_norm": 0.12697266042232513, "learning_rate": 2.8792691496837667e-05, "loss": 0.0233, "step": 9696 }, { "epoch": 6.814476458186929, "grad_norm": 0.2407216876745224, "learning_rate": 2.879222300304521e-05, "loss": 0.0169, "step": 9697 }, { "epoch": 6.815179198875615, "grad_norm": 0.3050327003002167, "learning_rate": 2.8791754509252754e-05, "loss": 0.0162, "step": 9698 }, { "epoch": 6.815881939564301, "grad_norm": 0.20566076040267944, "learning_rate": 2.8791286015460298e-05, "loss": 0.0217, "step": 9699 }, { "epoch": 6.816584680252987, "grad_norm": 0.17556516826152802, "learning_rate": 2.8790817521667838e-05, "loss": 0.0269, "step": 9700 }, { "epoch": 6.817287420941673, "grad_norm": 0.39109766483306885, "learning_rate": 2.8790349027875382e-05, "loss": 0.024, "step": 9701 }, { "epoch": 6.817990161630359, "grad_norm": 0.16526181995868683, "learning_rate": 2.8789880534082926e-05, "loss": 0.0149, "step": 9702 }, { "epoch": 6.818692902319044, "grad_norm": 0.33145731687545776, "learning_rate": 2.878941204029047e-05, "loss": 0.0461, "step": 9703 }, { "epoch": 6.81939564300773, "grad_norm": 0.2332880198955536, "learning_rate": 2.878894354649801e-05, "loss": 0.04, "step": 9704 }, { "epoch": 6.820098383696416, "grad_norm": 0.45470404624938965, "learning_rate": 2.8788475052705553e-05, "loss": 0.0354, "step": 9705 }, { "epoch": 6.820801124385102, "grad_norm": 0.4394845962524414, "learning_rate": 2.8788006558913094e-05, "loss": 0.0372, "step": 9706 }, { "epoch": 6.821503865073788, "grad_norm": 0.24448442459106445, "learning_rate": 2.8787538065120637e-05, "loss": 0.0302, "step": 9707 }, { "epoch": 6.8222066057624735, "grad_norm": 0.26338502764701843, "learning_rate": 2.878706957132818e-05, "loss": 0.0718, "step": 9708 }, { "epoch": 6.822909346451159, "grad_norm": 0.4427189826965332, "learning_rate": 2.878660107753572e-05, "loss": 0.0643, "step": 9709 }, { "epoch": 6.823612087139845, "grad_norm": 0.5022890567779541, "learning_rate": 2.8786132583743265e-05, "loss": 0.1348, "step": 9710 }, { "epoch": 6.824314827828531, "grad_norm": 0.8154176473617554, "learning_rate": 2.878566408995081e-05, "loss": 0.1623, "step": 9711 }, { "epoch": 6.825017568517217, "grad_norm": 1.3721277713775635, "learning_rate": 2.8785195596158353e-05, "loss": 0.204, "step": 9712 }, { "epoch": 6.825720309205903, "grad_norm": 3.418489456176758, "learning_rate": 2.8784727102365893e-05, "loss": 0.2541, "step": 9713 }, { "epoch": 6.826423049894589, "grad_norm": 0.2950901389122009, "learning_rate": 2.8784258608573437e-05, "loss": 0.0655, "step": 9714 }, { "epoch": 6.827125790583275, "grad_norm": 0.1262485682964325, "learning_rate": 2.878379011478098e-05, "loss": 0.0255, "step": 9715 }, { "epoch": 6.827828531271961, "grad_norm": 0.19696597754955292, "learning_rate": 2.8783321620988524e-05, "loss": 0.0308, "step": 9716 }, { "epoch": 6.828531271960647, "grad_norm": 0.14530940353870392, "learning_rate": 2.8782853127196064e-05, "loss": 0.0203, "step": 9717 }, { "epoch": 6.829234012649333, "grad_norm": 0.16009092330932617, "learning_rate": 2.8782384633403608e-05, "loss": 0.0215, "step": 9718 }, { "epoch": 6.829936753338019, "grad_norm": 0.5343015193939209, "learning_rate": 2.8781916139611152e-05, "loss": 0.0179, "step": 9719 }, { "epoch": 6.830639494026704, "grad_norm": 0.212691992521286, "learning_rate": 2.8781447645818696e-05, "loss": 0.0218, "step": 9720 }, { "epoch": 6.83134223471539, "grad_norm": 0.1527569442987442, "learning_rate": 2.878097915202624e-05, "loss": 0.0166, "step": 9721 }, { "epoch": 6.8320449754040755, "grad_norm": 0.2104436457157135, "learning_rate": 2.878051065823378e-05, "loss": 0.0351, "step": 9722 }, { "epoch": 6.832747716092761, "grad_norm": 0.4306468963623047, "learning_rate": 2.878004216444132e-05, "loss": 0.023, "step": 9723 }, { "epoch": 6.833450456781447, "grad_norm": 0.25705769658088684, "learning_rate": 2.8779573670648864e-05, "loss": 0.0314, "step": 9724 }, { "epoch": 6.834153197470133, "grad_norm": 0.13645273447036743, "learning_rate": 2.8779105176856408e-05, "loss": 0.0154, "step": 9725 }, { "epoch": 6.834855938158819, "grad_norm": 0.18052296340465546, "learning_rate": 2.8778636683063948e-05, "loss": 0.0334, "step": 9726 }, { "epoch": 6.835558678847505, "grad_norm": 0.12127856910228729, "learning_rate": 2.877816818927149e-05, "loss": 0.016, "step": 9727 }, { "epoch": 6.836261419536191, "grad_norm": 0.1858801543712616, "learning_rate": 2.8777699695479035e-05, "loss": 0.0288, "step": 9728 }, { "epoch": 6.836964160224877, "grad_norm": 0.39998260140419006, "learning_rate": 2.877723120168658e-05, "loss": 0.0376, "step": 9729 }, { "epoch": 6.837666900913563, "grad_norm": 0.4237743318080902, "learning_rate": 2.877676270789412e-05, "loss": 0.0345, "step": 9730 }, { "epoch": 6.838369641602249, "grad_norm": 0.23721523582935333, "learning_rate": 2.8776294214101663e-05, "loss": 0.0457, "step": 9731 }, { "epoch": 6.839072382290935, "grad_norm": 0.16518372297286987, "learning_rate": 2.8775825720309207e-05, "loss": 0.0279, "step": 9732 }, { "epoch": 6.839775122979621, "grad_norm": 0.34104984998703003, "learning_rate": 2.877535722651675e-05, "loss": 0.0572, "step": 9733 }, { "epoch": 6.840477863668307, "grad_norm": 0.42165565490722656, "learning_rate": 2.8774888732724294e-05, "loss": 0.0784, "step": 9734 }, { "epoch": 6.8411806043569925, "grad_norm": 0.48501574993133545, "learning_rate": 2.8774420238931835e-05, "loss": 0.0968, "step": 9735 }, { "epoch": 6.841883345045678, "grad_norm": 0.9847889542579651, "learning_rate": 2.877395174513938e-05, "loss": 0.1743, "step": 9736 }, { "epoch": 6.842586085734364, "grad_norm": 1.3012791872024536, "learning_rate": 2.8773483251346922e-05, "loss": 0.1701, "step": 9737 }, { "epoch": 6.84328882642305, "grad_norm": 1.5913163423538208, "learning_rate": 2.8773014757554466e-05, "loss": 0.2323, "step": 9738 }, { "epoch": 6.843991567111736, "grad_norm": 0.31508031487464905, "learning_rate": 2.8772546263762006e-05, "loss": 0.0717, "step": 9739 }, { "epoch": 6.844694307800422, "grad_norm": 0.21583259105682373, "learning_rate": 2.8772077769969546e-05, "loss": 0.0266, "step": 9740 }, { "epoch": 6.845397048489108, "grad_norm": 0.12062423676252365, "learning_rate": 2.877160927617709e-05, "loss": 0.0216, "step": 9741 }, { "epoch": 6.846099789177793, "grad_norm": 0.11984805017709732, "learning_rate": 2.8771140782384634e-05, "loss": 0.0239, "step": 9742 }, { "epoch": 6.846802529866479, "grad_norm": 0.1318276971578598, "learning_rate": 2.8770672288592174e-05, "loss": 0.0229, "step": 9743 }, { "epoch": 6.847505270555165, "grad_norm": 0.08158121258020401, "learning_rate": 2.8770203794799718e-05, "loss": 0.0072, "step": 9744 }, { "epoch": 6.848208011243851, "grad_norm": 0.2806905508041382, "learning_rate": 2.876973530100726e-05, "loss": 0.0217, "step": 9745 }, { "epoch": 6.848910751932537, "grad_norm": 0.21787568926811218, "learning_rate": 2.8769266807214805e-05, "loss": 0.0148, "step": 9746 }, { "epoch": 6.849613492621223, "grad_norm": 0.21794722974300385, "learning_rate": 2.876879831342235e-05, "loss": 0.027, "step": 9747 }, { "epoch": 6.850316233309909, "grad_norm": 0.1920253187417984, "learning_rate": 2.876832981962989e-05, "loss": 0.0158, "step": 9748 }, { "epoch": 6.8510189739985945, "grad_norm": 0.31318309903144836, "learning_rate": 2.8767861325837433e-05, "loss": 0.0389, "step": 9749 }, { "epoch": 6.85172171468728, "grad_norm": 0.12461326271295547, "learning_rate": 2.8767392832044977e-05, "loss": 0.0108, "step": 9750 }, { "epoch": 6.852424455375966, "grad_norm": 0.19535043835639954, "learning_rate": 2.876692433825252e-05, "loss": 0.035, "step": 9751 }, { "epoch": 6.853127196064652, "grad_norm": 0.09742770344018936, "learning_rate": 2.876645584446006e-05, "loss": 0.0129, "step": 9752 }, { "epoch": 6.853829936753338, "grad_norm": 0.34027254581451416, "learning_rate": 2.8765987350667605e-05, "loss": 0.0246, "step": 9753 }, { "epoch": 6.854532677442024, "grad_norm": 1.8801393508911133, "learning_rate": 2.876551885687515e-05, "loss": 0.0317, "step": 9754 }, { "epoch": 6.85523541813071, "grad_norm": 0.6294184923171997, "learning_rate": 2.8765050363082692e-05, "loss": 0.032, "step": 9755 }, { "epoch": 6.855938158819396, "grad_norm": 0.4232763648033142, "learning_rate": 2.8764581869290232e-05, "loss": 0.0376, "step": 9756 }, { "epoch": 6.856640899508082, "grad_norm": 0.4414132833480835, "learning_rate": 2.8764113375497776e-05, "loss": 0.0823, "step": 9757 }, { "epoch": 6.857343640196767, "grad_norm": 0.33277440071105957, "learning_rate": 2.8763644881705317e-05, "loss": 0.0776, "step": 9758 }, { "epoch": 6.858046380885453, "grad_norm": 0.2794857323169708, "learning_rate": 2.876317638791286e-05, "loss": 0.0707, "step": 9759 }, { "epoch": 6.858749121574139, "grad_norm": 0.7430970668792725, "learning_rate": 2.8762707894120404e-05, "loss": 0.1444, "step": 9760 }, { "epoch": 6.859451862262825, "grad_norm": 0.6002799272537231, "learning_rate": 2.8762239400327944e-05, "loss": 0.1852, "step": 9761 }, { "epoch": 6.860154602951511, "grad_norm": 0.8693047761917114, "learning_rate": 2.8761770906535488e-05, "loss": 0.1892, "step": 9762 }, { "epoch": 6.8608573436401965, "grad_norm": 1.3533867597579956, "learning_rate": 2.8761302412743032e-05, "loss": 0.2829, "step": 9763 }, { "epoch": 6.861560084328882, "grad_norm": 0.48362040519714355, "learning_rate": 2.8760833918950576e-05, "loss": 0.0762, "step": 9764 }, { "epoch": 6.862262825017568, "grad_norm": 0.21352194249629974, "learning_rate": 2.8760365425158116e-05, "loss": 0.0353, "step": 9765 }, { "epoch": 6.862965565706254, "grad_norm": 0.13254889845848083, "learning_rate": 2.875989693136566e-05, "loss": 0.0206, "step": 9766 }, { "epoch": 6.86366830639494, "grad_norm": 0.0966777428984642, "learning_rate": 2.8759428437573203e-05, "loss": 0.0184, "step": 9767 }, { "epoch": 6.864371047083626, "grad_norm": 0.2790270745754242, "learning_rate": 2.8758959943780747e-05, "loss": 0.0334, "step": 9768 }, { "epoch": 6.865073787772312, "grad_norm": 0.17862115800380707, "learning_rate": 2.8758491449988287e-05, "loss": 0.0129, "step": 9769 }, { "epoch": 6.865776528460998, "grad_norm": 0.19887802004814148, "learning_rate": 2.875802295619583e-05, "loss": 0.0283, "step": 9770 }, { "epoch": 6.866479269149684, "grad_norm": 0.09234331548213959, "learning_rate": 2.8757554462403375e-05, "loss": 0.0116, "step": 9771 }, { "epoch": 6.86718200983837, "grad_norm": 0.18977127969264984, "learning_rate": 2.875708596861092e-05, "loss": 0.0154, "step": 9772 }, { "epoch": 6.867884750527056, "grad_norm": 0.11128304898738861, "learning_rate": 2.8756617474818462e-05, "loss": 0.0162, "step": 9773 }, { "epoch": 6.868587491215742, "grad_norm": 0.2849650979042053, "learning_rate": 2.8756148981026003e-05, "loss": 0.0399, "step": 9774 }, { "epoch": 6.869290231904428, "grad_norm": 0.3195715546607971, "learning_rate": 2.8755680487233543e-05, "loss": 0.029, "step": 9775 }, { "epoch": 6.8699929725931135, "grad_norm": 0.18852752447128296, "learning_rate": 2.8755211993441087e-05, "loss": 0.0258, "step": 9776 }, { "epoch": 6.870695713281799, "grad_norm": 0.4075023829936981, "learning_rate": 2.875474349964863e-05, "loss": 0.0403, "step": 9777 }, { "epoch": 6.871398453970485, "grad_norm": 0.3441878855228424, "learning_rate": 2.875427500585617e-05, "loss": 0.0261, "step": 9778 }, { "epoch": 6.872101194659171, "grad_norm": 0.18130168318748474, "learning_rate": 2.8753806512063714e-05, "loss": 0.0415, "step": 9779 }, { "epoch": 6.872803935347856, "grad_norm": 0.1413647085428238, "learning_rate": 2.8753338018271258e-05, "loss": 0.024, "step": 9780 }, { "epoch": 6.873506676036542, "grad_norm": 0.25190484523773193, "learning_rate": 2.8752869524478802e-05, "loss": 0.0348, "step": 9781 }, { "epoch": 6.874209416725228, "grad_norm": 0.25015291571617126, "learning_rate": 2.8752401030686342e-05, "loss": 0.0364, "step": 9782 }, { "epoch": 6.874912157413914, "grad_norm": 0.30428487062454224, "learning_rate": 2.8751932536893886e-05, "loss": 0.0842, "step": 9783 }, { "epoch": 6.8756148981026, "grad_norm": 0.3505842685699463, "learning_rate": 2.875146404310143e-05, "loss": 0.0959, "step": 9784 }, { "epoch": 6.876317638791286, "grad_norm": 0.351646363735199, "learning_rate": 2.8750995549308973e-05, "loss": 0.0915, "step": 9785 }, { "epoch": 6.877020379479972, "grad_norm": 1.2362630367279053, "learning_rate": 2.8750527055516517e-05, "loss": 0.1358, "step": 9786 }, { "epoch": 6.877723120168658, "grad_norm": 0.9412571787834167, "learning_rate": 2.8750058561724057e-05, "loss": 0.2011, "step": 9787 }, { "epoch": 6.878425860857344, "grad_norm": 1.6385480165481567, "learning_rate": 2.87495900679316e-05, "loss": 0.2272, "step": 9788 }, { "epoch": 6.87912860154603, "grad_norm": 0.2644965350627899, "learning_rate": 2.8749121574139145e-05, "loss": 0.0547, "step": 9789 }, { "epoch": 6.8798313422347155, "grad_norm": 0.18044202029705048, "learning_rate": 2.874865308034669e-05, "loss": 0.0296, "step": 9790 }, { "epoch": 6.880534082923401, "grad_norm": 0.2806285619735718, "learning_rate": 2.874818458655423e-05, "loss": 0.0271, "step": 9791 }, { "epoch": 6.881236823612087, "grad_norm": 0.14142385125160217, "learning_rate": 2.8747716092761773e-05, "loss": 0.0165, "step": 9792 }, { "epoch": 6.881939564300773, "grad_norm": 0.28702178597450256, "learning_rate": 2.8747247598969313e-05, "loss": 0.0205, "step": 9793 }, { "epoch": 6.882642304989459, "grad_norm": 0.09464030712842941, "learning_rate": 2.8746779105176857e-05, "loss": 0.017, "step": 9794 }, { "epoch": 6.883345045678145, "grad_norm": 0.11219792813062668, "learning_rate": 2.8746310611384397e-05, "loss": 0.017, "step": 9795 }, { "epoch": 6.884047786366831, "grad_norm": 0.2310839742422104, "learning_rate": 2.874584211759194e-05, "loss": 0.0313, "step": 9796 }, { "epoch": 6.884750527055516, "grad_norm": 0.16804973781108856, "learning_rate": 2.8745373623799485e-05, "loss": 0.0282, "step": 9797 }, { "epoch": 6.885453267744202, "grad_norm": 0.12939557433128357, "learning_rate": 2.8744905130007028e-05, "loss": 0.0156, "step": 9798 }, { "epoch": 6.886156008432888, "grad_norm": 0.2508302927017212, "learning_rate": 2.8744436636214572e-05, "loss": 0.0267, "step": 9799 }, { "epoch": 6.886858749121574, "grad_norm": 0.10390889644622803, "learning_rate": 2.8743968142422112e-05, "loss": 0.015, "step": 9800 }, { "epoch": 6.88756148981026, "grad_norm": 0.11814921349287033, "learning_rate": 2.8743499648629656e-05, "loss": 0.0197, "step": 9801 }, { "epoch": 6.888264230498946, "grad_norm": 0.15328514575958252, "learning_rate": 2.87430311548372e-05, "loss": 0.0161, "step": 9802 }, { "epoch": 6.888966971187632, "grad_norm": 0.32679885625839233, "learning_rate": 2.8742562661044744e-05, "loss": 0.0326, "step": 9803 }, { "epoch": 6.8896697118763175, "grad_norm": 0.3623190224170685, "learning_rate": 2.8742094167252284e-05, "loss": 0.0315, "step": 9804 }, { "epoch": 6.890372452565003, "grad_norm": 0.25791993737220764, "learning_rate": 2.8741625673459828e-05, "loss": 0.0246, "step": 9805 }, { "epoch": 6.891075193253689, "grad_norm": 0.30832064151763916, "learning_rate": 2.874115717966737e-05, "loss": 0.0479, "step": 9806 }, { "epoch": 6.891777933942375, "grad_norm": 0.30044621229171753, "learning_rate": 2.8740688685874915e-05, "loss": 0.0481, "step": 9807 }, { "epoch": 6.892480674631061, "grad_norm": 0.8943959474563599, "learning_rate": 2.8740220192082455e-05, "loss": 0.0708, "step": 9808 }, { "epoch": 6.893183415319747, "grad_norm": 0.511308491230011, "learning_rate": 2.873975169829e-05, "loss": 0.0977, "step": 9809 }, { "epoch": 6.893886156008433, "grad_norm": 0.4678712785243988, "learning_rate": 2.873928320449754e-05, "loss": 0.1197, "step": 9810 }, { "epoch": 6.894588896697119, "grad_norm": 0.7126848101615906, "learning_rate": 2.8738814710705083e-05, "loss": 0.1541, "step": 9811 }, { "epoch": 6.895291637385805, "grad_norm": 0.8861286640167236, "learning_rate": 2.8738346216912627e-05, "loss": 0.2218, "step": 9812 }, { "epoch": 6.895994378074491, "grad_norm": 1.2859079837799072, "learning_rate": 2.8737877723120167e-05, "loss": 0.2347, "step": 9813 }, { "epoch": 6.896697118763177, "grad_norm": 0.3921319246292114, "learning_rate": 2.873740922932771e-05, "loss": 0.0847, "step": 9814 }, { "epoch": 6.897399859451863, "grad_norm": 0.2084449827671051, "learning_rate": 2.8736940735535255e-05, "loss": 0.0293, "step": 9815 }, { "epoch": 6.8981026001405485, "grad_norm": 0.1404559463262558, "learning_rate": 2.87364722417428e-05, "loss": 0.0243, "step": 9816 }, { "epoch": 6.8988053408292345, "grad_norm": 0.1500425636768341, "learning_rate": 2.873600374795034e-05, "loss": 0.0213, "step": 9817 }, { "epoch": 6.8995080815179195, "grad_norm": 0.3155231177806854, "learning_rate": 2.8735535254157882e-05, "loss": 0.0187, "step": 9818 }, { "epoch": 6.900210822206605, "grad_norm": 0.11129837483167648, "learning_rate": 2.8735066760365426e-05, "loss": 0.0169, "step": 9819 }, { "epoch": 6.900913562895291, "grad_norm": 0.13751690089702606, "learning_rate": 2.873459826657297e-05, "loss": 0.0197, "step": 9820 }, { "epoch": 6.901616303583977, "grad_norm": 0.14550624787807465, "learning_rate": 2.873412977278051e-05, "loss": 0.019, "step": 9821 }, { "epoch": 6.902319044272663, "grad_norm": 0.24736957252025604, "learning_rate": 2.8733661278988054e-05, "loss": 0.0414, "step": 9822 }, { "epoch": 6.903021784961349, "grad_norm": 0.14033527672290802, "learning_rate": 2.8733192785195598e-05, "loss": 0.0105, "step": 9823 }, { "epoch": 6.903724525650035, "grad_norm": 0.41240349411964417, "learning_rate": 2.873272429140314e-05, "loss": 0.0339, "step": 9824 }, { "epoch": 6.904427266338721, "grad_norm": 0.17234276235103607, "learning_rate": 2.8732255797610685e-05, "loss": 0.0218, "step": 9825 }, { "epoch": 6.905130007027407, "grad_norm": 0.2395392507314682, "learning_rate": 2.8731787303818225e-05, "loss": 0.0343, "step": 9826 }, { "epoch": 6.905832747716093, "grad_norm": 0.16582927107810974, "learning_rate": 2.873131881002577e-05, "loss": 0.0255, "step": 9827 }, { "epoch": 6.906535488404779, "grad_norm": 0.44396913051605225, "learning_rate": 2.873085031623331e-05, "loss": 0.0227, "step": 9828 }, { "epoch": 6.907238229093465, "grad_norm": 0.22217202186584473, "learning_rate": 2.8730381822440853e-05, "loss": 0.0442, "step": 9829 }, { "epoch": 6.9079409697821506, "grad_norm": 0.16169168055057526, "learning_rate": 2.8729913328648394e-05, "loss": 0.0223, "step": 9830 }, { "epoch": 6.9086437104708365, "grad_norm": 0.2120649665594101, "learning_rate": 2.8729444834855937e-05, "loss": 0.0391, "step": 9831 }, { "epoch": 6.909346451159522, "grad_norm": 0.16410286724567413, "learning_rate": 2.872897634106348e-05, "loss": 0.0347, "step": 9832 }, { "epoch": 6.910049191848208, "grad_norm": 0.3380783498287201, "learning_rate": 2.8728507847271025e-05, "loss": 0.0793, "step": 9833 }, { "epoch": 6.910751932536894, "grad_norm": 0.375072181224823, "learning_rate": 2.8728039353478565e-05, "loss": 0.0682, "step": 9834 }, { "epoch": 6.911454673225579, "grad_norm": 0.4922373294830322, "learning_rate": 2.872757085968611e-05, "loss": 0.1265, "step": 9835 }, { "epoch": 6.912157413914265, "grad_norm": 0.874966025352478, "learning_rate": 2.8727102365893653e-05, "loss": 0.1669, "step": 9836 }, { "epoch": 6.912860154602951, "grad_norm": 2.8896825313568115, "learning_rate": 2.8726633872101196e-05, "loss": 0.2464, "step": 9837 }, { "epoch": 6.913562895291637, "grad_norm": 1.9196785688400269, "learning_rate": 2.872616537830874e-05, "loss": 0.2934, "step": 9838 }, { "epoch": 6.914265635980323, "grad_norm": 0.31982842087745667, "learning_rate": 2.872569688451628e-05, "loss": 0.0834, "step": 9839 }, { "epoch": 6.914968376669009, "grad_norm": 0.16470934450626373, "learning_rate": 2.8725228390723824e-05, "loss": 0.0332, "step": 9840 }, { "epoch": 6.915671117357695, "grad_norm": 0.20570160448551178, "learning_rate": 2.8724759896931368e-05, "loss": 0.0283, "step": 9841 }, { "epoch": 6.916373858046381, "grad_norm": 0.18026751279830933, "learning_rate": 2.872429140313891e-05, "loss": 0.0207, "step": 9842 }, { "epoch": 6.917076598735067, "grad_norm": 0.1402878761291504, "learning_rate": 2.8723822909346452e-05, "loss": 0.015, "step": 9843 }, { "epoch": 6.917779339423753, "grad_norm": 0.1735346019268036, "learning_rate": 2.8723354415553996e-05, "loss": 0.0201, "step": 9844 }, { "epoch": 6.9184820801124385, "grad_norm": 0.20777034759521484, "learning_rate": 2.8722885921761536e-05, "loss": 0.0297, "step": 9845 }, { "epoch": 6.919184820801124, "grad_norm": 0.14110764861106873, "learning_rate": 2.872241742796908e-05, "loss": 0.0184, "step": 9846 }, { "epoch": 6.91988756148981, "grad_norm": 0.17303994297981262, "learning_rate": 2.8721948934176623e-05, "loss": 0.0327, "step": 9847 }, { "epoch": 6.920590302178496, "grad_norm": 0.13574257493019104, "learning_rate": 2.8721480440384164e-05, "loss": 0.0135, "step": 9848 }, { "epoch": 6.921293042867182, "grad_norm": 0.19913695752620697, "learning_rate": 2.8721011946591707e-05, "loss": 0.0271, "step": 9849 }, { "epoch": 6.921995783555868, "grad_norm": 0.2551692724227905, "learning_rate": 2.872054345279925e-05, "loss": 0.0233, "step": 9850 }, { "epoch": 6.922698524244554, "grad_norm": 0.2394515722990036, "learning_rate": 2.8720074959006795e-05, "loss": 0.0292, "step": 9851 }, { "epoch": 6.92340126493324, "grad_norm": 0.37105992436408997, "learning_rate": 2.8719606465214335e-05, "loss": 0.0198, "step": 9852 }, { "epoch": 6.924104005621926, "grad_norm": 0.47242170572280884, "learning_rate": 2.871913797142188e-05, "loss": 0.037, "step": 9853 }, { "epoch": 6.924806746310612, "grad_norm": 0.2054852694272995, "learning_rate": 2.8718669477629423e-05, "loss": 0.0317, "step": 9854 }, { "epoch": 6.925509486999298, "grad_norm": 0.3409704267978668, "learning_rate": 2.8718200983836966e-05, "loss": 0.0271, "step": 9855 }, { "epoch": 6.926212227687984, "grad_norm": 0.2961689531803131, "learning_rate": 2.8717732490044507e-05, "loss": 0.0268, "step": 9856 }, { "epoch": 6.926914968376669, "grad_norm": 0.44175130128860474, "learning_rate": 2.871726399625205e-05, "loss": 0.0727, "step": 9857 }, { "epoch": 6.927617709065355, "grad_norm": 0.3645339012145996, "learning_rate": 2.8716795502459594e-05, "loss": 0.0474, "step": 9858 }, { "epoch": 6.9283204497540405, "grad_norm": 0.4615429639816284, "learning_rate": 2.8716327008667138e-05, "loss": 0.1061, "step": 9859 }, { "epoch": 6.929023190442726, "grad_norm": 0.5243494510650635, "learning_rate": 2.8715858514874678e-05, "loss": 0.1448, "step": 9860 }, { "epoch": 6.929725931131412, "grad_norm": 0.6919334530830383, "learning_rate": 2.8715390021082222e-05, "loss": 0.1604, "step": 9861 }, { "epoch": 6.930428671820098, "grad_norm": 0.9248089790344238, "learning_rate": 2.8714921527289762e-05, "loss": 0.2036, "step": 9862 }, { "epoch": 6.931131412508784, "grad_norm": 1.3130321502685547, "learning_rate": 2.8714453033497306e-05, "loss": 0.2204, "step": 9863 }, { "epoch": 6.93183415319747, "grad_norm": 0.306286096572876, "learning_rate": 2.871398453970485e-05, "loss": 0.0793, "step": 9864 }, { "epoch": 6.932536893886156, "grad_norm": 0.4132477939128876, "learning_rate": 2.871351604591239e-05, "loss": 0.034, "step": 9865 }, { "epoch": 6.933239634574842, "grad_norm": 0.0946970209479332, "learning_rate": 2.8713047552119934e-05, "loss": 0.0135, "step": 9866 }, { "epoch": 6.933942375263528, "grad_norm": 0.15350431203842163, "learning_rate": 2.8712579058327478e-05, "loss": 0.0167, "step": 9867 }, { "epoch": 6.934645115952214, "grad_norm": 0.1278267800807953, "learning_rate": 2.871211056453502e-05, "loss": 0.0245, "step": 9868 }, { "epoch": 6.9353478566409, "grad_norm": 0.17490383982658386, "learning_rate": 2.871164207074256e-05, "loss": 0.025, "step": 9869 }, { "epoch": 6.936050597329586, "grad_norm": 0.2879830300807953, "learning_rate": 2.8711173576950105e-05, "loss": 0.0251, "step": 9870 }, { "epoch": 6.9367533380182715, "grad_norm": 0.2368282675743103, "learning_rate": 2.871070508315765e-05, "loss": 0.0277, "step": 9871 }, { "epoch": 6.9374560787069575, "grad_norm": 0.14667430520057678, "learning_rate": 2.8710236589365193e-05, "loss": 0.0204, "step": 9872 }, { "epoch": 6.938158819395643, "grad_norm": 0.11400190740823746, "learning_rate": 2.8709768095572737e-05, "loss": 0.0144, "step": 9873 }, { "epoch": 6.938861560084328, "grad_norm": 0.15790626406669617, "learning_rate": 2.8709299601780277e-05, "loss": 0.0219, "step": 9874 }, { "epoch": 6.939564300773014, "grad_norm": 0.13493889570236206, "learning_rate": 2.870883110798782e-05, "loss": 0.017, "step": 9875 }, { "epoch": 6.9402670414617, "grad_norm": 0.2057834416627884, "learning_rate": 2.8708362614195364e-05, "loss": 0.0297, "step": 9876 }, { "epoch": 6.940969782150386, "grad_norm": 0.17743422091007233, "learning_rate": 2.8707894120402908e-05, "loss": 0.0138, "step": 9877 }, { "epoch": 6.941672522839072, "grad_norm": 0.23605938255786896, "learning_rate": 2.870742562661045e-05, "loss": 0.0441, "step": 9878 }, { "epoch": 6.942375263527758, "grad_norm": 0.13469980657100677, "learning_rate": 2.8706957132817992e-05, "loss": 0.0279, "step": 9879 }, { "epoch": 6.943078004216444, "grad_norm": 0.10476472973823547, "learning_rate": 2.8706488639025532e-05, "loss": 0.0105, "step": 9880 }, { "epoch": 6.94378074490513, "grad_norm": 0.15115074813365936, "learning_rate": 2.8706020145233076e-05, "loss": 0.0388, "step": 9881 }, { "epoch": 6.944483485593816, "grad_norm": 0.18579071760177612, "learning_rate": 2.8705551651440617e-05, "loss": 0.0388, "step": 9882 }, { "epoch": 6.945186226282502, "grad_norm": 0.2936340868473053, "learning_rate": 2.870508315764816e-05, "loss": 0.0841, "step": 9883 }, { "epoch": 6.945888966971188, "grad_norm": 0.32872873544692993, "learning_rate": 2.8704614663855704e-05, "loss": 0.0732, "step": 9884 }, { "epoch": 6.9465917076598735, "grad_norm": 0.43554115295410156, "learning_rate": 2.8704146170063248e-05, "loss": 0.1251, "step": 9885 }, { "epoch": 6.9472944483485595, "grad_norm": 0.5122206807136536, "learning_rate": 2.870367767627079e-05, "loss": 0.1693, "step": 9886 }, { "epoch": 6.947997189037245, "grad_norm": 0.6961690187454224, "learning_rate": 2.8703209182478332e-05, "loss": 0.1757, "step": 9887 }, { "epoch": 6.948699929725931, "grad_norm": 0.8516910076141357, "learning_rate": 2.8702740688685875e-05, "loss": 0.2126, "step": 9888 }, { "epoch": 6.949402670414617, "grad_norm": 0.3310295045375824, "learning_rate": 2.870227219489342e-05, "loss": 0.0834, "step": 9889 }, { "epoch": 6.950105411103303, "grad_norm": 0.0909588560461998, "learning_rate": 2.8701803701100963e-05, "loss": 0.0192, "step": 9890 }, { "epoch": 6.950808151791989, "grad_norm": 0.10589215904474258, "learning_rate": 2.8701335207308503e-05, "loss": 0.0204, "step": 9891 }, { "epoch": 6.951510892480675, "grad_norm": 0.16059957444667816, "learning_rate": 2.8700866713516047e-05, "loss": 0.0239, "step": 9892 }, { "epoch": 6.952213633169361, "grad_norm": 0.20125171542167664, "learning_rate": 2.870039821972359e-05, "loss": 0.0398, "step": 9893 }, { "epoch": 6.952916373858047, "grad_norm": 0.15300676226615906, "learning_rate": 2.8699929725931134e-05, "loss": 0.0123, "step": 9894 }, { "epoch": 6.953619114546732, "grad_norm": 0.15032798051834106, "learning_rate": 2.8699461232138675e-05, "loss": 0.0311, "step": 9895 }, { "epoch": 6.954321855235418, "grad_norm": 0.1417025774717331, "learning_rate": 2.869899273834622e-05, "loss": 0.0196, "step": 9896 }, { "epoch": 6.955024595924104, "grad_norm": 0.14405736327171326, "learning_rate": 2.869852424455376e-05, "loss": 0.0203, "step": 9897 }, { "epoch": 6.95572733661279, "grad_norm": 0.24993523955345154, "learning_rate": 2.8698055750761303e-05, "loss": 0.0231, "step": 9898 }, { "epoch": 6.956430077301476, "grad_norm": 0.269111305475235, "learning_rate": 2.8697587256968846e-05, "loss": 0.0241, "step": 9899 }, { "epoch": 6.9571328179901615, "grad_norm": 0.2754494547843933, "learning_rate": 2.8697118763176387e-05, "loss": 0.0177, "step": 9900 }, { "epoch": 6.957835558678847, "grad_norm": 0.1960372030735016, "learning_rate": 2.869665026938393e-05, "loss": 0.0255, "step": 9901 }, { "epoch": 6.958538299367533, "grad_norm": 0.2196757197380066, "learning_rate": 2.8696181775591474e-05, "loss": 0.012, "step": 9902 }, { "epoch": 6.959241040056219, "grad_norm": 0.33422979712486267, "learning_rate": 2.8695713281799018e-05, "loss": 0.0278, "step": 9903 }, { "epoch": 6.959943780744905, "grad_norm": 0.34153279662132263, "learning_rate": 2.8695244788006558e-05, "loss": 0.0312, "step": 9904 }, { "epoch": 6.960646521433591, "grad_norm": 0.35560283064842224, "learning_rate": 2.8694776294214102e-05, "loss": 0.0317, "step": 9905 }, { "epoch": 6.961349262122277, "grad_norm": 0.15960747003555298, "learning_rate": 2.8694307800421646e-05, "loss": 0.0253, "step": 9906 }, { "epoch": 6.962052002810963, "grad_norm": 0.3790944218635559, "learning_rate": 2.869383930662919e-05, "loss": 0.0753, "step": 9907 }, { "epoch": 6.962754743499649, "grad_norm": 0.3341270089149475, "learning_rate": 2.869337081283673e-05, "loss": 0.0667, "step": 9908 }, { "epoch": 6.963457484188335, "grad_norm": 0.6321704387664795, "learning_rate": 2.8692902319044273e-05, "loss": 0.1035, "step": 9909 }, { "epoch": 6.964160224877021, "grad_norm": 0.7839663028717041, "learning_rate": 2.8692433825251817e-05, "loss": 0.1208, "step": 9910 }, { "epoch": 6.964862965565707, "grad_norm": 0.7646161913871765, "learning_rate": 2.869196533145936e-05, "loss": 0.1986, "step": 9911 }, { "epoch": 6.965565706254392, "grad_norm": 1.0667742490768433, "learning_rate": 2.8691496837666905e-05, "loss": 0.2286, "step": 9912 }, { "epoch": 6.966268446943078, "grad_norm": 1.6357630491256714, "learning_rate": 2.8691028343874445e-05, "loss": 0.2361, "step": 9913 }, { "epoch": 6.9669711876317635, "grad_norm": 0.27829334139823914, "learning_rate": 2.869055985008199e-05, "loss": 0.0974, "step": 9914 }, { "epoch": 6.967673928320449, "grad_norm": 0.11964063346385956, "learning_rate": 2.869009135628953e-05, "loss": 0.0247, "step": 9915 }, { "epoch": 6.968376669009135, "grad_norm": 0.10321906954050064, "learning_rate": 2.8689622862497073e-05, "loss": 0.0169, "step": 9916 }, { "epoch": 6.969079409697821, "grad_norm": 0.10011882334947586, "learning_rate": 2.8689154368704613e-05, "loss": 0.0165, "step": 9917 }, { "epoch": 6.969782150386507, "grad_norm": 0.23980610072612762, "learning_rate": 2.8688685874912157e-05, "loss": 0.0221, "step": 9918 }, { "epoch": 6.970484891075193, "grad_norm": 0.09318172931671143, "learning_rate": 2.86882173811197e-05, "loss": 0.0096, "step": 9919 }, { "epoch": 6.971187631763879, "grad_norm": 0.251723051071167, "learning_rate": 2.8687748887327244e-05, "loss": 0.0361, "step": 9920 }, { "epoch": 6.971890372452565, "grad_norm": 0.1539393812417984, "learning_rate": 2.8687280393534785e-05, "loss": 0.0231, "step": 9921 }, { "epoch": 6.972593113141251, "grad_norm": 0.2363385409116745, "learning_rate": 2.8686811899742328e-05, "loss": 0.0195, "step": 9922 }, { "epoch": 6.973295853829937, "grad_norm": 0.20544081926345825, "learning_rate": 2.8686343405949872e-05, "loss": 0.0228, "step": 9923 }, { "epoch": 6.973998594518623, "grad_norm": 0.35924258828163147, "learning_rate": 2.8685874912157416e-05, "loss": 0.0252, "step": 9924 }, { "epoch": 6.974701335207309, "grad_norm": 0.17966997623443604, "learning_rate": 2.868540641836496e-05, "loss": 0.015, "step": 9925 }, { "epoch": 6.9754040758959945, "grad_norm": 0.17651230096817017, "learning_rate": 2.86849379245725e-05, "loss": 0.0295, "step": 9926 }, { "epoch": 6.9761068165846805, "grad_norm": 0.10556455701589584, "learning_rate": 2.8684469430780043e-05, "loss": 0.0184, "step": 9927 }, { "epoch": 6.976809557273366, "grad_norm": 0.16008685529232025, "learning_rate": 2.8684000936987587e-05, "loss": 0.0281, "step": 9928 }, { "epoch": 6.977512297962052, "grad_norm": 0.41532468795776367, "learning_rate": 2.868353244319513e-05, "loss": 0.0469, "step": 9929 }, { "epoch": 6.978215038650738, "grad_norm": 0.15031923353672028, "learning_rate": 2.868306394940267e-05, "loss": 0.021, "step": 9930 }, { "epoch": 6.978917779339424, "grad_norm": 0.23669901490211487, "learning_rate": 2.8682595455610215e-05, "loss": 0.0446, "step": 9931 }, { "epoch": 6.97962052002811, "grad_norm": 0.23459690809249878, "learning_rate": 2.8682126961817755e-05, "loss": 0.037, "step": 9932 }, { "epoch": 6.980323260716796, "grad_norm": 0.4881935119628906, "learning_rate": 2.86816584680253e-05, "loss": 0.1219, "step": 9933 }, { "epoch": 6.981026001405481, "grad_norm": 0.3634776175022125, "learning_rate": 2.868118997423284e-05, "loss": 0.083, "step": 9934 }, { "epoch": 6.981728742094167, "grad_norm": 0.5312870144844055, "learning_rate": 2.8680721480440383e-05, "loss": 0.1361, "step": 9935 }, { "epoch": 6.982431482782853, "grad_norm": 0.6158539056777954, "learning_rate": 2.8680252986647927e-05, "loss": 0.1582, "step": 9936 }, { "epoch": 6.983134223471539, "grad_norm": 1.007311224937439, "learning_rate": 2.867978449285547e-05, "loss": 0.1931, "step": 9937 }, { "epoch": 6.983836964160225, "grad_norm": 3.2549033164978027, "learning_rate": 2.8679315999063014e-05, "loss": 0.2873, "step": 9938 }, { "epoch": 6.984539704848911, "grad_norm": 0.20739425718784332, "learning_rate": 2.8678847505270555e-05, "loss": 0.0732, "step": 9939 }, { "epoch": 6.9852424455375965, "grad_norm": 0.15597955882549286, "learning_rate": 2.86783790114781e-05, "loss": 0.0333, "step": 9940 }, { "epoch": 6.9859451862262825, "grad_norm": 0.24215061962604523, "learning_rate": 2.8677910517685642e-05, "loss": 0.0277, "step": 9941 }, { "epoch": 6.986647926914968, "grad_norm": 0.11160491406917572, "learning_rate": 2.8677442023893186e-05, "loss": 0.0115, "step": 9942 }, { "epoch": 6.987350667603654, "grad_norm": 0.1600489318370819, "learning_rate": 2.8676973530100726e-05, "loss": 0.0243, "step": 9943 }, { "epoch": 6.98805340829234, "grad_norm": 0.1758985072374344, "learning_rate": 2.867650503630827e-05, "loss": 0.0213, "step": 9944 }, { "epoch": 6.988756148981026, "grad_norm": 0.1891636699438095, "learning_rate": 2.8676036542515814e-05, "loss": 0.0178, "step": 9945 }, { "epoch": 6.989458889669712, "grad_norm": 0.23565815389156342, "learning_rate": 2.8675568048723357e-05, "loss": 0.0354, "step": 9946 }, { "epoch": 6.990161630358398, "grad_norm": 0.1593700647354126, "learning_rate": 2.8675099554930898e-05, "loss": 0.0151, "step": 9947 }, { "epoch": 6.990864371047084, "grad_norm": 0.12197154760360718, "learning_rate": 2.867463106113844e-05, "loss": 0.0212, "step": 9948 }, { "epoch": 6.99156711173577, "grad_norm": 0.13954682648181915, "learning_rate": 2.8674162567345982e-05, "loss": 0.0099, "step": 9949 }, { "epoch": 6.992269852424456, "grad_norm": 0.24465970695018768, "learning_rate": 2.8673694073553525e-05, "loss": 0.036, "step": 9950 }, { "epoch": 6.992972593113141, "grad_norm": 0.1795680671930313, "learning_rate": 2.867322557976107e-05, "loss": 0.013, "step": 9951 }, { "epoch": 6.993675333801827, "grad_norm": 0.16076265275478363, "learning_rate": 2.867275708596861e-05, "loss": 0.0225, "step": 9952 }, { "epoch": 6.994378074490513, "grad_norm": 0.36713144183158875, "learning_rate": 2.8672288592176153e-05, "loss": 0.0455, "step": 9953 }, { "epoch": 6.9950808151791986, "grad_norm": 0.413688063621521, "learning_rate": 2.8671820098383697e-05, "loss": 0.0523, "step": 9954 }, { "epoch": 6.9957835558678845, "grad_norm": 0.2894760072231293, "learning_rate": 2.867135160459124e-05, "loss": 0.0513, "step": 9955 }, { "epoch": 6.99648629655657, "grad_norm": 0.4205327033996582, "learning_rate": 2.867088311079878e-05, "loss": 0.101, "step": 9956 }, { "epoch": 6.997189037245256, "grad_norm": 0.3762342035770416, "learning_rate": 2.8670414617006325e-05, "loss": 0.0743, "step": 9957 }, { "epoch": 6.997891777933942, "grad_norm": 0.6136941313743591, "learning_rate": 2.866994612321387e-05, "loss": 0.1281, "step": 9958 }, { "epoch": 6.998594518622628, "grad_norm": 0.7826639413833618, "learning_rate": 2.8669477629421412e-05, "loss": 0.1957, "step": 9959 }, { "epoch": 6.999297259311314, "grad_norm": 1.1103953123092651, "learning_rate": 2.8669009135628953e-05, "loss": 0.2027, "step": 9960 }, { "epoch": 7.0, "grad_norm": 0.7347806096076965, "learning_rate": 2.8668540641836496e-05, "loss": 0.1388, "step": 9961 }, { "epoch": 7.000702740688686, "grad_norm": 0.19062267243862152, "learning_rate": 2.866807214804404e-05, "loss": 0.0711, "step": 9962 }, { "epoch": 7.001405481377372, "grad_norm": 0.4291844964027405, "learning_rate": 2.8667603654251584e-05, "loss": 0.0347, "step": 9963 }, { "epoch": 7.002108222066058, "grad_norm": 0.1458965390920639, "learning_rate": 2.8667135160459127e-05, "loss": 0.0245, "step": 9964 }, { "epoch": 7.002810962754744, "grad_norm": 0.14755591750144958, "learning_rate": 2.8666666666666668e-05, "loss": 0.0175, "step": 9965 }, { "epoch": 7.00351370344343, "grad_norm": 0.19661839306354523, "learning_rate": 2.866619817287421e-05, "loss": 0.0309, "step": 9966 }, { "epoch": 7.0042164441321155, "grad_norm": 0.10860171914100647, "learning_rate": 2.8665729679081752e-05, "loss": 0.0153, "step": 9967 }, { "epoch": 7.0049191848208014, "grad_norm": 0.36992260813713074, "learning_rate": 2.8665261185289296e-05, "loss": 0.0175, "step": 9968 }, { "epoch": 7.005621925509487, "grad_norm": 0.21514740586280823, "learning_rate": 2.8664792691496836e-05, "loss": 0.0446, "step": 9969 }, { "epoch": 7.006324666198173, "grad_norm": 0.3110111653804779, "learning_rate": 2.866432419770438e-05, "loss": 0.0314, "step": 9970 }, { "epoch": 7.007027406886858, "grad_norm": 0.15938332676887512, "learning_rate": 2.8663855703911923e-05, "loss": 0.0167, "step": 9971 }, { "epoch": 7.007730147575544, "grad_norm": 0.19596756994724274, "learning_rate": 2.8663387210119467e-05, "loss": 0.02, "step": 9972 }, { "epoch": 7.00843288826423, "grad_norm": 0.18484744429588318, "learning_rate": 2.8662918716327007e-05, "loss": 0.0189, "step": 9973 }, { "epoch": 7.009135628952916, "grad_norm": 0.5832149386405945, "learning_rate": 2.866245022253455e-05, "loss": 0.0367, "step": 9974 }, { "epoch": 7.009838369641602, "grad_norm": 0.15828149020671844, "learning_rate": 2.8661981728742095e-05, "loss": 0.0161, "step": 9975 }, { "epoch": 7.010541110330288, "grad_norm": 0.21606042981147766, "learning_rate": 2.866151323494964e-05, "loss": 0.0329, "step": 9976 }, { "epoch": 7.011243851018974, "grad_norm": 0.309749573469162, "learning_rate": 2.8661044741157182e-05, "loss": 0.02, "step": 9977 }, { "epoch": 7.01194659170766, "grad_norm": 0.405041366815567, "learning_rate": 2.8660576247364723e-05, "loss": 0.035, "step": 9978 }, { "epoch": 7.012649332396346, "grad_norm": 0.10824228078126907, "learning_rate": 2.8660107753572266e-05, "loss": 0.0153, "step": 9979 }, { "epoch": 7.013352073085032, "grad_norm": 0.402679443359375, "learning_rate": 2.865963925977981e-05, "loss": 0.0649, "step": 9980 }, { "epoch": 7.0140548137737175, "grad_norm": 0.5355734825134277, "learning_rate": 2.8659170765987354e-05, "loss": 0.0489, "step": 9981 }, { "epoch": 7.0147575544624035, "grad_norm": 0.41294336318969727, "learning_rate": 2.8658702272194894e-05, "loss": 0.0913, "step": 9982 }, { "epoch": 7.015460295151089, "grad_norm": 0.4476640522480011, "learning_rate": 2.8658233778402438e-05, "loss": 0.1009, "step": 9983 }, { "epoch": 7.016163035839775, "grad_norm": 0.4859808385372162, "learning_rate": 2.8657765284609978e-05, "loss": 0.1416, "step": 9984 }, { "epoch": 7.016865776528461, "grad_norm": 0.7465022206306458, "learning_rate": 2.8657296790817522e-05, "loss": 0.1742, "step": 9985 }, { "epoch": 7.017568517217147, "grad_norm": Infinity, "learning_rate": 2.8657296790817522e-05, "loss": 0.2069, "step": 9986 }, { "epoch": 7.018271257905833, "grad_norm": 0.23267830908298492, "learning_rate": 2.8656828297025062e-05, "loss": 0.065, "step": 9987 }, { "epoch": 7.018973998594519, "grad_norm": 1.0138298273086548, "learning_rate": 2.8656359803232606e-05, "loss": 0.0389, "step": 9988 }, { "epoch": 7.019676739283205, "grad_norm": 0.16416612267494202, "learning_rate": 2.865589130944015e-05, "loss": 0.0305, "step": 9989 }, { "epoch": 7.02037947997189, "grad_norm": 0.14004461467266083, "learning_rate": 2.8655422815647693e-05, "loss": 0.0196, "step": 9990 }, { "epoch": 7.021082220660576, "grad_norm": 0.2692158818244934, "learning_rate": 2.8654954321855237e-05, "loss": 0.0203, "step": 9991 }, { "epoch": 7.021784961349262, "grad_norm": 0.12020023167133331, "learning_rate": 2.8654485828062778e-05, "loss": 0.0135, "step": 9992 }, { "epoch": 7.022487702037948, "grad_norm": 0.26630038022994995, "learning_rate": 2.865401733427032e-05, "loss": 0.0208, "step": 9993 }, { "epoch": 7.023190442726634, "grad_norm": 0.15086700022220612, "learning_rate": 2.8653548840477865e-05, "loss": 0.0182, "step": 9994 }, { "epoch": 7.0238931834153195, "grad_norm": 0.27400103211402893, "learning_rate": 2.865308034668541e-05, "loss": 0.0364, "step": 9995 }, { "epoch": 7.0245959241040055, "grad_norm": 0.1747036725282669, "learning_rate": 2.865261185289295e-05, "loss": 0.0274, "step": 9996 }, { "epoch": 7.025298664792691, "grad_norm": 0.2719075679779053, "learning_rate": 2.8652143359100493e-05, "loss": 0.0427, "step": 9997 }, { "epoch": 7.026001405481377, "grad_norm": 0.13386796414852142, "learning_rate": 2.8651674865308036e-05, "loss": 0.0149, "step": 9998 }, { "epoch": 7.026704146170063, "grad_norm": 0.17085500061511993, "learning_rate": 2.865120637151558e-05, "loss": 0.0237, "step": 9999 }, { "epoch": 7.027406886858749, "grad_norm": 0.3204750418663025, "learning_rate": 2.865073787772312e-05, "loss": 0.03, "step": 10000 }, { "epoch": 7.027406886858749, "eval_cer": 0.19584192552344168, "eval_loss": 0.2848150432109833, "eval_runtime": 18.2407, "eval_samples_per_second": 248.785, "eval_steps_per_second": 0.822, "eval_wer": 0.35214100164799866, "step": 10000 }, { "epoch": 7.028109627547435, "grad_norm": 0.16825780272483826, "learning_rate": 2.8650269383930664e-05, "loss": 0.0248, "step": 10001 }, { "epoch": 7.028812368236121, "grad_norm": 0.1723303198814392, "learning_rate": 2.8649800890138208e-05, "loss": 0.0329, "step": 10002 }, { "epoch": 7.029515108924807, "grad_norm": 0.21581576764583588, "learning_rate": 2.864933239634575e-05, "loss": 0.0519, "step": 10003 }, { "epoch": 7.030217849613493, "grad_norm": 0.5200768709182739, "learning_rate": 2.8648863902553292e-05, "loss": 0.0387, "step": 10004 }, { "epoch": 7.030920590302179, "grad_norm": 0.20804516971111298, "learning_rate": 2.8648395408760832e-05, "loss": 0.0353, "step": 10005 }, { "epoch": 7.031623330990865, "grad_norm": 0.19477210938930511, "learning_rate": 2.8647926914968376e-05, "loss": 0.0442, "step": 10006 }, { "epoch": 7.032326071679551, "grad_norm": 0.3237081468105316, "learning_rate": 2.864745842117592e-05, "loss": 0.0852, "step": 10007 }, { "epoch": 7.0330288123682365, "grad_norm": 0.6221606731414795, "learning_rate": 2.8646989927383464e-05, "loss": 0.1255, "step": 10008 }, { "epoch": 7.033731553056922, "grad_norm": 0.6617633104324341, "learning_rate": 2.8646521433591004e-05, "loss": 0.1618, "step": 10009 }, { "epoch": 7.0344342937456075, "grad_norm": 0.9681493043899536, "learning_rate": 2.8646052939798548e-05, "loss": 0.1623, "step": 10010 }, { "epoch": 7.035137034434293, "grad_norm": 0.9718230366706848, "learning_rate": 2.864558444600609e-05, "loss": 0.2415, "step": 10011 }, { "epoch": 7.035839775122979, "grad_norm": 0.41000330448150635, "learning_rate": 2.8645115952213635e-05, "loss": 0.0862, "step": 10012 }, { "epoch": 7.036542515811665, "grad_norm": 0.16739793121814728, "learning_rate": 2.8644647458421175e-05, "loss": 0.0381, "step": 10013 }, { "epoch": 7.037245256500351, "grad_norm": 0.11053162068128586, "learning_rate": 2.864417896462872e-05, "loss": 0.0216, "step": 10014 }, { "epoch": 7.037947997189037, "grad_norm": 0.19304496049880981, "learning_rate": 2.8643710470836263e-05, "loss": 0.0134, "step": 10015 }, { "epoch": 7.038650737877723, "grad_norm": 0.1755339801311493, "learning_rate": 2.8643241977043807e-05, "loss": 0.018, "step": 10016 }, { "epoch": 7.039353478566409, "grad_norm": 0.17989881336688995, "learning_rate": 2.864277348325135e-05, "loss": 0.0134, "step": 10017 }, { "epoch": 7.040056219255095, "grad_norm": 0.2081337720155716, "learning_rate": 2.864230498945889e-05, "loss": 0.0377, "step": 10018 }, { "epoch": 7.040758959943781, "grad_norm": 0.1860736459493637, "learning_rate": 2.8641836495666434e-05, "loss": 0.0246, "step": 10019 }, { "epoch": 7.041461700632467, "grad_norm": 0.14410631358623505, "learning_rate": 2.8641368001873975e-05, "loss": 0.0201, "step": 10020 }, { "epoch": 7.042164441321153, "grad_norm": 0.1757688820362091, "learning_rate": 2.864089950808152e-05, "loss": 0.027, "step": 10021 }, { "epoch": 7.0428671820098385, "grad_norm": 0.3858817219734192, "learning_rate": 2.864043101428906e-05, "loss": 0.0256, "step": 10022 }, { "epoch": 7.043569922698524, "grad_norm": 0.1321883499622345, "learning_rate": 2.8639962520496603e-05, "loss": 0.0122, "step": 10023 }, { "epoch": 7.04427266338721, "grad_norm": 0.24818457663059235, "learning_rate": 2.8639494026704146e-05, "loss": 0.0525, "step": 10024 }, { "epoch": 7.044975404075896, "grad_norm": 0.17377406358718872, "learning_rate": 2.863902553291169e-05, "loss": 0.0193, "step": 10025 }, { "epoch": 7.045678144764582, "grad_norm": 0.14599527418613434, "learning_rate": 2.863855703911923e-05, "loss": 0.0311, "step": 10026 }, { "epoch": 7.046380885453268, "grad_norm": 0.27824997901916504, "learning_rate": 2.8638088545326774e-05, "loss": 0.0493, "step": 10027 }, { "epoch": 7.047083626141954, "grad_norm": 0.19019433856010437, "learning_rate": 2.8637620051534318e-05, "loss": 0.0311, "step": 10028 }, { "epoch": 7.047786366830639, "grad_norm": 0.3361760675907135, "learning_rate": 2.863715155774186e-05, "loss": 0.04, "step": 10029 }, { "epoch": 7.048489107519325, "grad_norm": 0.20881111919879913, "learning_rate": 2.8636683063949405e-05, "loss": 0.0408, "step": 10030 }, { "epoch": 7.049191848208011, "grad_norm": 0.37283438444137573, "learning_rate": 2.8636214570156946e-05, "loss": 0.0574, "step": 10031 }, { "epoch": 7.049894588896697, "grad_norm": 0.3706407845020294, "learning_rate": 2.863574607636449e-05, "loss": 0.0574, "step": 10032 }, { "epoch": 7.050597329585383, "grad_norm": 1.0233486890792847, "learning_rate": 2.8635277582572033e-05, "loss": 0.1312, "step": 10033 }, { "epoch": 7.051300070274069, "grad_norm": 0.6554246544837952, "learning_rate": 2.8634809088779577e-05, "loss": 0.1595, "step": 10034 }, { "epoch": 7.052002810962755, "grad_norm": 1.0770576000213623, "learning_rate": 2.8634340594987117e-05, "loss": 0.228, "step": 10035 }, { "epoch": 7.0527055516514405, "grad_norm": 1.0070685148239136, "learning_rate": 2.863387210119466e-05, "loss": 0.2243, "step": 10036 }, { "epoch": 7.0534082923401265, "grad_norm": 0.1617760956287384, "learning_rate": 2.8633403607402204e-05, "loss": 0.0537, "step": 10037 }, { "epoch": 7.054111033028812, "grad_norm": 0.17755278944969177, "learning_rate": 2.8632935113609745e-05, "loss": 0.0283, "step": 10038 }, { "epoch": 7.054813773717498, "grad_norm": 0.1152239739894867, "learning_rate": 2.8632466619817285e-05, "loss": 0.0213, "step": 10039 }, { "epoch": 7.055516514406184, "grad_norm": 0.1801222413778305, "learning_rate": 2.863199812602483e-05, "loss": 0.0199, "step": 10040 }, { "epoch": 7.05621925509487, "grad_norm": 0.12920090556144714, "learning_rate": 2.8631529632232373e-05, "loss": 0.0169, "step": 10041 }, { "epoch": 7.056921995783556, "grad_norm": 0.11528950184583664, "learning_rate": 2.8631061138439916e-05, "loss": 0.0189, "step": 10042 }, { "epoch": 7.057624736472242, "grad_norm": 0.11597209423780441, "learning_rate": 2.863059264464746e-05, "loss": 0.017, "step": 10043 }, { "epoch": 7.058327477160928, "grad_norm": 0.4458935856819153, "learning_rate": 2.8630124150855e-05, "loss": 0.021, "step": 10044 }, { "epoch": 7.059030217849614, "grad_norm": 0.20431429147720337, "learning_rate": 2.8629655657062544e-05, "loss": 0.0226, "step": 10045 }, { "epoch": 7.0597329585383, "grad_norm": 0.14109644293785095, "learning_rate": 2.8629187163270088e-05, "loss": 0.0227, "step": 10046 }, { "epoch": 7.060435699226986, "grad_norm": 0.22888197004795074, "learning_rate": 2.862871866947763e-05, "loss": 0.0203, "step": 10047 }, { "epoch": 7.061138439915671, "grad_norm": 0.23403412103652954, "learning_rate": 2.8628250175685172e-05, "loss": 0.0227, "step": 10048 }, { "epoch": 7.061841180604357, "grad_norm": 0.43344584107398987, "learning_rate": 2.8627781681892716e-05, "loss": 0.0234, "step": 10049 }, { "epoch": 7.0625439212930425, "grad_norm": 0.1589985489845276, "learning_rate": 2.862731318810026e-05, "loss": 0.0172, "step": 10050 }, { "epoch": 7.0632466619817285, "grad_norm": 0.17618992924690247, "learning_rate": 2.8626844694307803e-05, "loss": 0.0227, "step": 10051 }, { "epoch": 7.063949402670414, "grad_norm": 0.17957943677902222, "learning_rate": 2.8626376200515343e-05, "loss": 0.0265, "step": 10052 }, { "epoch": 7.0646521433591, "grad_norm": 0.24501526355743408, "learning_rate": 2.8625907706722887e-05, "loss": 0.02, "step": 10053 }, { "epoch": 7.065354884047786, "grad_norm": 0.23227949440479279, "learning_rate": 2.862543921293043e-05, "loss": 0.0519, "step": 10054 }, { "epoch": 7.066057624736472, "grad_norm": 0.3593292534351349, "learning_rate": 2.862497071913797e-05, "loss": 0.0463, "step": 10055 }, { "epoch": 7.066760365425158, "grad_norm": 0.4185114800930023, "learning_rate": 2.8624502225345515e-05, "loss": 0.0724, "step": 10056 }, { "epoch": 7.067463106113844, "grad_norm": 0.3239460289478302, "learning_rate": 2.8624033731553055e-05, "loss": 0.0584, "step": 10057 }, { "epoch": 7.06816584680253, "grad_norm": 0.7220169901847839, "learning_rate": 2.86235652377606e-05, "loss": 0.1146, "step": 10058 }, { "epoch": 7.068868587491216, "grad_norm": 0.8224928975105286, "learning_rate": 2.8623096743968143e-05, "loss": 0.1731, "step": 10059 }, { "epoch": 7.069571328179902, "grad_norm": 0.795225977897644, "learning_rate": 2.8622628250175686e-05, "loss": 0.2028, "step": 10060 }, { "epoch": 7.070274068868588, "grad_norm": 2.0447843074798584, "learning_rate": 2.8622159756383227e-05, "loss": 0.2514, "step": 10061 }, { "epoch": 7.070976809557274, "grad_norm": 0.48495444655418396, "learning_rate": 2.862169126259077e-05, "loss": 0.0812, "step": 10062 }, { "epoch": 7.0716795502459595, "grad_norm": 0.17520853877067566, "learning_rate": 2.8621222768798314e-05, "loss": 0.0294, "step": 10063 }, { "epoch": 7.072382290934645, "grad_norm": 0.13269132375717163, "learning_rate": 2.8620754275005858e-05, "loss": 0.0195, "step": 10064 }, { "epoch": 7.073085031623331, "grad_norm": 0.1564185470342636, "learning_rate": 2.8620285781213402e-05, "loss": 0.0275, "step": 10065 }, { "epoch": 7.073787772312017, "grad_norm": 0.1358979195356369, "learning_rate": 2.8619817287420942e-05, "loss": 0.0245, "step": 10066 }, { "epoch": 7.074490513000702, "grad_norm": 0.15843893587589264, "learning_rate": 2.8619348793628486e-05, "loss": 0.0177, "step": 10067 }, { "epoch": 7.075193253689388, "grad_norm": 0.20903527736663818, "learning_rate": 2.861888029983603e-05, "loss": 0.0162, "step": 10068 }, { "epoch": 7.075895994378074, "grad_norm": 0.13196033239364624, "learning_rate": 2.8618411806043573e-05, "loss": 0.0243, "step": 10069 }, { "epoch": 7.07659873506676, "grad_norm": 0.11907488852739334, "learning_rate": 2.8617943312251114e-05, "loss": 0.0237, "step": 10070 }, { "epoch": 7.077301475755446, "grad_norm": 0.24227748811244965, "learning_rate": 2.8617474818458657e-05, "loss": 0.0172, "step": 10071 }, { "epoch": 7.078004216444132, "grad_norm": 0.13542893528938293, "learning_rate": 2.8617006324666198e-05, "loss": 0.0291, "step": 10072 }, { "epoch": 7.078706957132818, "grad_norm": 0.11391223967075348, "learning_rate": 2.861653783087374e-05, "loss": 0.0136, "step": 10073 }, { "epoch": 7.079409697821504, "grad_norm": 0.3153080344200134, "learning_rate": 2.861606933708128e-05, "loss": 0.0316, "step": 10074 }, { "epoch": 7.08011243851019, "grad_norm": 0.36352694034576416, "learning_rate": 2.8615600843288825e-05, "loss": 0.0141, "step": 10075 }, { "epoch": 7.080815179198876, "grad_norm": 0.26160937547683716, "learning_rate": 2.861513234949637e-05, "loss": 0.0409, "step": 10076 }, { "epoch": 7.0815179198875615, "grad_norm": 0.35037750005722046, "learning_rate": 2.8614663855703913e-05, "loss": 0.035, "step": 10077 }, { "epoch": 7.082220660576247, "grad_norm": 0.26860401034355164, "learning_rate": 2.8614195361911457e-05, "loss": 0.0348, "step": 10078 }, { "epoch": 7.082923401264933, "grad_norm": 0.2577749788761139, "learning_rate": 2.8613726868118997e-05, "loss": 0.0725, "step": 10079 }, { "epoch": 7.083626141953619, "grad_norm": 0.4220573902130127, "learning_rate": 2.861325837432654e-05, "loss": 0.0436, "step": 10080 }, { "epoch": 7.084328882642305, "grad_norm": 0.21684084832668304, "learning_rate": 2.8612789880534084e-05, "loss": 0.046, "step": 10081 }, { "epoch": 7.085031623330991, "grad_norm": 0.2865205705165863, "learning_rate": 2.8612321386741628e-05, "loss": 0.1007, "step": 10082 }, { "epoch": 7.085734364019677, "grad_norm": 0.6793549060821533, "learning_rate": 2.861185289294917e-05, "loss": 0.119, "step": 10083 }, { "epoch": 7.086437104708363, "grad_norm": 0.463923841714859, "learning_rate": 2.8611384399156712e-05, "loss": 0.1227, "step": 10084 }, { "epoch": 7.087139845397049, "grad_norm": 0.977183997631073, "learning_rate": 2.8610915905364256e-05, "loss": 0.2169, "step": 10085 }, { "epoch": 7.087842586085735, "grad_norm": 0.8348146080970764, "learning_rate": 2.86104474115718e-05, "loss": 0.2376, "step": 10086 }, { "epoch": 7.08854532677442, "grad_norm": 0.27068543434143066, "learning_rate": 2.860997891777934e-05, "loss": 0.0686, "step": 10087 }, { "epoch": 7.089248067463106, "grad_norm": 0.19279491901397705, "learning_rate": 2.8609510423986884e-05, "loss": 0.0276, "step": 10088 }, { "epoch": 7.089950808151792, "grad_norm": 0.13684023916721344, "learning_rate": 2.8609041930194427e-05, "loss": 0.0251, "step": 10089 }, { "epoch": 7.090653548840478, "grad_norm": 0.14285418391227722, "learning_rate": 2.8608573436401968e-05, "loss": 0.0216, "step": 10090 }, { "epoch": 7.0913562895291635, "grad_norm": 0.16464117169380188, "learning_rate": 2.860810494260951e-05, "loss": 0.0275, "step": 10091 }, { "epoch": 7.0920590302178494, "grad_norm": 0.1308138072490692, "learning_rate": 2.8607636448817052e-05, "loss": 0.0155, "step": 10092 }, { "epoch": 7.092761770906535, "grad_norm": 0.12643413245677948, "learning_rate": 2.8607167955024596e-05, "loss": 0.0233, "step": 10093 }, { "epoch": 7.093464511595221, "grad_norm": 0.2624111771583557, "learning_rate": 2.860669946123214e-05, "loss": 0.0275, "step": 10094 }, { "epoch": 7.094167252283907, "grad_norm": 0.15969033539295197, "learning_rate": 2.8606230967439683e-05, "loss": 0.0187, "step": 10095 }, { "epoch": 7.094869992972593, "grad_norm": 0.11879447847604752, "learning_rate": 2.8605762473647223e-05, "loss": 0.0148, "step": 10096 }, { "epoch": 7.095572733661279, "grad_norm": 0.2792961001396179, "learning_rate": 2.8605293979854767e-05, "loss": 0.0343, "step": 10097 }, { "epoch": 7.096275474349965, "grad_norm": 0.17175044119358063, "learning_rate": 2.860482548606231e-05, "loss": 0.0182, "step": 10098 }, { "epoch": 7.096978215038651, "grad_norm": 0.25708627700805664, "learning_rate": 2.8604356992269854e-05, "loss": 0.0245, "step": 10099 }, { "epoch": 7.097680955727337, "grad_norm": 0.19456709921360016, "learning_rate": 2.8603888498477395e-05, "loss": 0.0131, "step": 10100 }, { "epoch": 7.098383696416023, "grad_norm": 0.17131400108337402, "learning_rate": 2.860342000468494e-05, "loss": 0.0309, "step": 10101 }, { "epoch": 7.099086437104709, "grad_norm": 0.2054385244846344, "learning_rate": 2.8602951510892482e-05, "loss": 0.0299, "step": 10102 }, { "epoch": 7.099789177793395, "grad_norm": 0.13094550371170044, "learning_rate": 2.8602483017100026e-05, "loss": 0.0166, "step": 10103 }, { "epoch": 7.1004919184820805, "grad_norm": 0.1976046860218048, "learning_rate": 2.860201452330757e-05, "loss": 0.0424, "step": 10104 }, { "epoch": 7.101194659170766, "grad_norm": 0.25727519392967224, "learning_rate": 2.860154602951511e-05, "loss": 0.0494, "step": 10105 }, { "epoch": 7.1018973998594515, "grad_norm": 0.3392011821269989, "learning_rate": 2.8601077535722654e-05, "loss": 0.0666, "step": 10106 }, { "epoch": 7.102600140548137, "grad_norm": 0.3390904664993286, "learning_rate": 2.8600609041930194e-05, "loss": 0.0693, "step": 10107 }, { "epoch": 7.103302881236823, "grad_norm": 1.5568345785140991, "learning_rate": 2.8600140548137738e-05, "loss": 0.0955, "step": 10108 }, { "epoch": 7.104005621925509, "grad_norm": 0.8578689098358154, "learning_rate": 2.8599672054345278e-05, "loss": 0.1532, "step": 10109 }, { "epoch": 7.104708362614195, "grad_norm": 1.133326530456543, "learning_rate": 2.8599203560552822e-05, "loss": 0.1855, "step": 10110 }, { "epoch": 7.105411103302881, "grad_norm": 0.8030821084976196, "learning_rate": 2.8598735066760366e-05, "loss": 0.204, "step": 10111 }, { "epoch": 7.106113843991567, "grad_norm": 0.3066180646419525, "learning_rate": 2.859826657296791e-05, "loss": 0.0729, "step": 10112 }, { "epoch": 7.106816584680253, "grad_norm": 0.13829277455806732, "learning_rate": 2.859779807917545e-05, "loss": 0.0257, "step": 10113 }, { "epoch": 7.107519325368939, "grad_norm": 0.1856527179479599, "learning_rate": 2.8597329585382993e-05, "loss": 0.0232, "step": 10114 }, { "epoch": 7.108222066057625, "grad_norm": 0.16350814700126648, "learning_rate": 2.8596861091590537e-05, "loss": 0.0247, "step": 10115 }, { "epoch": 7.108924806746311, "grad_norm": 0.23546618223190308, "learning_rate": 2.859639259779808e-05, "loss": 0.0184, "step": 10116 }, { "epoch": 7.109627547434997, "grad_norm": 0.09968579560518265, "learning_rate": 2.8595924104005625e-05, "loss": 0.0105, "step": 10117 }, { "epoch": 7.1103302881236825, "grad_norm": 0.3758750855922699, "learning_rate": 2.8595455610213165e-05, "loss": 0.0141, "step": 10118 }, { "epoch": 7.111033028812368, "grad_norm": 0.19374699890613556, "learning_rate": 2.859498711642071e-05, "loss": 0.018, "step": 10119 }, { "epoch": 7.111735769501054, "grad_norm": 0.16918472945690155, "learning_rate": 2.8594518622628252e-05, "loss": 0.0188, "step": 10120 }, { "epoch": 7.11243851018974, "grad_norm": 0.2419942170381546, "learning_rate": 2.8594050128835796e-05, "loss": 0.0279, "step": 10121 }, { "epoch": 7.113141250878426, "grad_norm": 0.16514571011066437, "learning_rate": 2.8593581635043336e-05, "loss": 0.0235, "step": 10122 }, { "epoch": 7.113843991567112, "grad_norm": 0.1413009613752365, "learning_rate": 2.859311314125088e-05, "loss": 0.0119, "step": 10123 }, { "epoch": 7.114546732255798, "grad_norm": 0.1247975304722786, "learning_rate": 2.8592644647458424e-05, "loss": 0.0326, "step": 10124 }, { "epoch": 7.115249472944483, "grad_norm": 0.16592593491077423, "learning_rate": 2.8592176153665964e-05, "loss": 0.0207, "step": 10125 }, { "epoch": 7.115952213633169, "grad_norm": 0.349016934633255, "learning_rate": 2.8591707659873505e-05, "loss": 0.0303, "step": 10126 }, { "epoch": 7.116654954321855, "grad_norm": 0.16887810826301575, "learning_rate": 2.8591239166081048e-05, "loss": 0.0302, "step": 10127 }, { "epoch": 7.117357695010541, "grad_norm": 0.1427963674068451, "learning_rate": 2.8590770672288592e-05, "loss": 0.0193, "step": 10128 }, { "epoch": 7.118060435699227, "grad_norm": 0.18291369080543518, "learning_rate": 2.8590302178496136e-05, "loss": 0.0237, "step": 10129 }, { "epoch": 7.118763176387913, "grad_norm": 0.31880322098731995, "learning_rate": 2.858983368470368e-05, "loss": 0.0332, "step": 10130 }, { "epoch": 7.119465917076599, "grad_norm": 0.29254013299942017, "learning_rate": 2.858936519091122e-05, "loss": 0.0748, "step": 10131 }, { "epoch": 7.1201686577652845, "grad_norm": 0.47995230555534363, "learning_rate": 2.8588896697118764e-05, "loss": 0.0702, "step": 10132 }, { "epoch": 7.12087139845397, "grad_norm": 2.149388313293457, "learning_rate": 2.8588428203326307e-05, "loss": 0.1061, "step": 10133 }, { "epoch": 7.121574139142656, "grad_norm": 0.4953728914260864, "learning_rate": 2.858795970953385e-05, "loss": 0.1644, "step": 10134 }, { "epoch": 7.122276879831342, "grad_norm": 0.8161885142326355, "learning_rate": 2.858749121574139e-05, "loss": 0.1962, "step": 10135 }, { "epoch": 7.122979620520028, "grad_norm": 1.4276328086853027, "learning_rate": 2.8587022721948935e-05, "loss": 0.2607, "step": 10136 }, { "epoch": 7.123682361208714, "grad_norm": 0.2908376455307007, "learning_rate": 2.858655422815648e-05, "loss": 0.0595, "step": 10137 }, { "epoch": 7.1243851018974, "grad_norm": 0.18807195127010345, "learning_rate": 2.8586085734364022e-05, "loss": 0.0361, "step": 10138 }, { "epoch": 7.125087842586086, "grad_norm": 0.15630725026130676, "learning_rate": 2.8585617240571563e-05, "loss": 0.0279, "step": 10139 }, { "epoch": 7.125790583274772, "grad_norm": 0.12236148864030838, "learning_rate": 2.8585148746779107e-05, "loss": 0.02, "step": 10140 }, { "epoch": 7.126493323963458, "grad_norm": 0.23751811683177948, "learning_rate": 2.858468025298665e-05, "loss": 0.0188, "step": 10141 }, { "epoch": 7.127196064652144, "grad_norm": 0.09777884185314178, "learning_rate": 2.858421175919419e-05, "loss": 0.0085, "step": 10142 }, { "epoch": 7.12789880534083, "grad_norm": 0.14110858738422394, "learning_rate": 2.8583743265401734e-05, "loss": 0.0199, "step": 10143 }, { "epoch": 7.128601546029515, "grad_norm": 0.2136925607919693, "learning_rate": 2.8583274771609275e-05, "loss": 0.0265, "step": 10144 }, { "epoch": 7.129304286718201, "grad_norm": 0.12511707842350006, "learning_rate": 2.858280627781682e-05, "loss": 0.0212, "step": 10145 }, { "epoch": 7.1300070274068865, "grad_norm": 0.1176486387848854, "learning_rate": 2.8582337784024362e-05, "loss": 0.0182, "step": 10146 }, { "epoch": 7.130709768095572, "grad_norm": 0.1876746267080307, "learning_rate": 2.8581869290231906e-05, "loss": 0.0263, "step": 10147 }, { "epoch": 7.131412508784258, "grad_norm": 0.14533640444278717, "learning_rate": 2.8581400796439446e-05, "loss": 0.0159, "step": 10148 }, { "epoch": 7.132115249472944, "grad_norm": 0.2422095239162445, "learning_rate": 2.858093230264699e-05, "loss": 0.0288, "step": 10149 }, { "epoch": 7.13281799016163, "grad_norm": 0.2893047332763672, "learning_rate": 2.8580463808854534e-05, "loss": 0.0198, "step": 10150 }, { "epoch": 7.133520730850316, "grad_norm": 0.19625969231128693, "learning_rate": 2.8579995315062077e-05, "loss": 0.0242, "step": 10151 }, { "epoch": 7.134223471539002, "grad_norm": 0.23234182596206665, "learning_rate": 2.8579526821269618e-05, "loss": 0.0296, "step": 10152 }, { "epoch": 7.134926212227688, "grad_norm": 0.26217222213745117, "learning_rate": 2.857905832747716e-05, "loss": 0.029, "step": 10153 }, { "epoch": 7.135628952916374, "grad_norm": 0.3250102996826172, "learning_rate": 2.8578589833684705e-05, "loss": 0.0298, "step": 10154 }, { "epoch": 7.13633169360506, "grad_norm": 0.4214407503604889, "learning_rate": 2.857812133989225e-05, "loss": 0.0329, "step": 10155 }, { "epoch": 7.137034434293746, "grad_norm": 0.3195265233516693, "learning_rate": 2.8577652846099793e-05, "loss": 0.0624, "step": 10156 }, { "epoch": 7.137737174982432, "grad_norm": 0.5648598074913025, "learning_rate": 2.8577184352307333e-05, "loss": 0.0793, "step": 10157 }, { "epoch": 7.138439915671118, "grad_norm": 0.46787452697753906, "learning_rate": 2.8576715858514877e-05, "loss": 0.1168, "step": 10158 }, { "epoch": 7.1391426563598035, "grad_norm": 0.7602788209915161, "learning_rate": 2.8576247364722417e-05, "loss": 0.1341, "step": 10159 }, { "epoch": 7.139845397048489, "grad_norm": 2.395801305770874, "learning_rate": 2.857577887092996e-05, "loss": 0.2016, "step": 10160 }, { "epoch": 7.140548137737175, "grad_norm": 1.4821202754974365, "learning_rate": 2.85753103771375e-05, "loss": 0.2106, "step": 10161 }, { "epoch": 7.141250878425861, "grad_norm": 0.23632262647151947, "learning_rate": 2.8574841883345045e-05, "loss": 0.0754, "step": 10162 }, { "epoch": 7.141953619114547, "grad_norm": 0.12573322653770447, "learning_rate": 2.857437338955259e-05, "loss": 0.0266, "step": 10163 }, { "epoch": 7.142656359803232, "grad_norm": 0.16899733245372772, "learning_rate": 2.8573904895760132e-05, "loss": 0.025, "step": 10164 }, { "epoch": 7.143359100491918, "grad_norm": 0.11810819804668427, "learning_rate": 2.8573436401967673e-05, "loss": 0.0191, "step": 10165 }, { "epoch": 7.144061841180604, "grad_norm": 0.13251526653766632, "learning_rate": 2.8572967908175216e-05, "loss": 0.0191, "step": 10166 }, { "epoch": 7.14476458186929, "grad_norm": 0.1133100762963295, "learning_rate": 2.857249941438276e-05, "loss": 0.0192, "step": 10167 }, { "epoch": 7.145467322557976, "grad_norm": 0.17009331285953522, "learning_rate": 2.8572030920590304e-05, "loss": 0.0266, "step": 10168 }, { "epoch": 7.146170063246662, "grad_norm": 0.27664920687675476, "learning_rate": 2.8571562426797847e-05, "loss": 0.0185, "step": 10169 }, { "epoch": 7.146872803935348, "grad_norm": 0.30344247817993164, "learning_rate": 2.8571093933005388e-05, "loss": 0.0207, "step": 10170 }, { "epoch": 7.147575544624034, "grad_norm": 0.4547751843929291, "learning_rate": 2.857062543921293e-05, "loss": 0.0153, "step": 10171 }, { "epoch": 7.14827828531272, "grad_norm": 0.5409401059150696, "learning_rate": 2.8570156945420475e-05, "loss": 0.0178, "step": 10172 }, { "epoch": 7.1489810260014055, "grad_norm": 0.15423104166984558, "learning_rate": 2.856968845162802e-05, "loss": 0.0178, "step": 10173 }, { "epoch": 7.149683766690091, "grad_norm": 0.123049296438694, "learning_rate": 2.856921995783556e-05, "loss": 0.0267, "step": 10174 }, { "epoch": 7.150386507378777, "grad_norm": 0.27312877774238586, "learning_rate": 2.8568751464043103e-05, "loss": 0.0173, "step": 10175 }, { "epoch": 7.151089248067463, "grad_norm": 0.1668555587530136, "learning_rate": 2.8568282970250647e-05, "loss": 0.0311, "step": 10176 }, { "epoch": 7.151791988756149, "grad_norm": 0.2441021203994751, "learning_rate": 2.8567814476458187e-05, "loss": 0.0355, "step": 10177 }, { "epoch": 7.152494729444835, "grad_norm": 0.33546480536460876, "learning_rate": 2.8567345982665727e-05, "loss": 0.0219, "step": 10178 }, { "epoch": 7.153197470133521, "grad_norm": 0.2317405343055725, "learning_rate": 2.856687748887327e-05, "loss": 0.036, "step": 10179 }, { "epoch": 7.153900210822207, "grad_norm": 0.2111099362373352, "learning_rate": 2.8566408995080815e-05, "loss": 0.0311, "step": 10180 }, { "epoch": 7.154602951510893, "grad_norm": 0.40529125928878784, "learning_rate": 2.856594050128836e-05, "loss": 0.0543, "step": 10181 }, { "epoch": 7.155305692199578, "grad_norm": 0.3004523515701294, "learning_rate": 2.8565472007495902e-05, "loss": 0.069, "step": 10182 }, { "epoch": 7.156008432888264, "grad_norm": 0.5777682065963745, "learning_rate": 2.8565003513703443e-05, "loss": 0.1282, "step": 10183 }, { "epoch": 7.15671117357695, "grad_norm": 0.9143274426460266, "learning_rate": 2.8564535019910986e-05, "loss": 0.2064, "step": 10184 }, { "epoch": 7.157413914265636, "grad_norm": 1.252526879310608, "learning_rate": 2.856406652611853e-05, "loss": 0.2564, "step": 10185 }, { "epoch": 7.158116654954322, "grad_norm": 1.0529550313949585, "learning_rate": 2.8563598032326074e-05, "loss": 0.2189, "step": 10186 }, { "epoch": 7.1588193956430075, "grad_norm": 0.23764589428901672, "learning_rate": 2.8563129538533614e-05, "loss": 0.0641, "step": 10187 }, { "epoch": 7.159522136331693, "grad_norm": 0.27778366208076477, "learning_rate": 2.8562661044741158e-05, "loss": 0.0323, "step": 10188 }, { "epoch": 7.160224877020379, "grad_norm": 0.16764743626117706, "learning_rate": 2.85621925509487e-05, "loss": 0.0255, "step": 10189 }, { "epoch": 7.160927617709065, "grad_norm": 0.11150939017534256, "learning_rate": 2.8561724057156245e-05, "loss": 0.0201, "step": 10190 }, { "epoch": 7.161630358397751, "grad_norm": 0.19681480526924133, "learning_rate": 2.8561255563363786e-05, "loss": 0.0223, "step": 10191 }, { "epoch": 7.162333099086437, "grad_norm": 0.2998061180114746, "learning_rate": 2.856078706957133e-05, "loss": 0.0132, "step": 10192 }, { "epoch": 7.163035839775123, "grad_norm": 0.22987030446529388, "learning_rate": 2.8560318575778873e-05, "loss": 0.0246, "step": 10193 }, { "epoch": 7.163738580463809, "grad_norm": 0.15856383740901947, "learning_rate": 2.8559850081986413e-05, "loss": 0.0117, "step": 10194 }, { "epoch": 7.164441321152495, "grad_norm": 0.15213638544082642, "learning_rate": 2.8559381588193957e-05, "loss": 0.0259, "step": 10195 }, { "epoch": 7.165144061841181, "grad_norm": 0.21624772250652313, "learning_rate": 2.8558913094401498e-05, "loss": 0.015, "step": 10196 }, { "epoch": 7.165846802529867, "grad_norm": 0.5387992858886719, "learning_rate": 2.855844460060904e-05, "loss": 0.0365, "step": 10197 }, { "epoch": 7.166549543218553, "grad_norm": 0.3997077941894531, "learning_rate": 2.8557976106816585e-05, "loss": 0.0096, "step": 10198 }, { "epoch": 7.167252283907239, "grad_norm": 0.18819363415241241, "learning_rate": 2.855750761302413e-05, "loss": 0.0314, "step": 10199 }, { "epoch": 7.1679550245959245, "grad_norm": 0.2684866487979889, "learning_rate": 2.855703911923167e-05, "loss": 0.0288, "step": 10200 }, { "epoch": 7.16865776528461, "grad_norm": 0.16911788284778595, "learning_rate": 2.8556570625439213e-05, "loss": 0.0251, "step": 10201 }, { "epoch": 7.169360505973295, "grad_norm": 0.25381848216056824, "learning_rate": 2.8556102131646757e-05, "loss": 0.0336, "step": 10202 }, { "epoch": 7.170063246661981, "grad_norm": 0.3191494047641754, "learning_rate": 2.85556336378543e-05, "loss": 0.0348, "step": 10203 }, { "epoch": 7.170765987350667, "grad_norm": 0.21907782554626465, "learning_rate": 2.855516514406184e-05, "loss": 0.045, "step": 10204 }, { "epoch": 7.171468728039353, "grad_norm": 0.2721291482448578, "learning_rate": 2.8554696650269384e-05, "loss": 0.0333, "step": 10205 }, { "epoch": 7.172171468728039, "grad_norm": 0.3114466965198517, "learning_rate": 2.8554228156476928e-05, "loss": 0.0554, "step": 10206 }, { "epoch": 7.172874209416725, "grad_norm": 0.39609530568122864, "learning_rate": 2.8553759662684472e-05, "loss": 0.1113, "step": 10207 }, { "epoch": 7.173576950105411, "grad_norm": 0.47998014092445374, "learning_rate": 2.8553291168892015e-05, "loss": 0.1249, "step": 10208 }, { "epoch": 7.174279690794097, "grad_norm": 0.6078605651855469, "learning_rate": 2.8552822675099556e-05, "loss": 0.1887, "step": 10209 }, { "epoch": 7.174982431482783, "grad_norm": 0.820345938205719, "learning_rate": 2.85523541813071e-05, "loss": 0.2079, "step": 10210 }, { "epoch": 7.175685172171469, "grad_norm": 0.9509679079055786, "learning_rate": 2.8551885687514643e-05, "loss": 0.2048, "step": 10211 }, { "epoch": 7.176387912860155, "grad_norm": 0.3492265045642853, "learning_rate": 2.8551417193722184e-05, "loss": 0.074, "step": 10212 }, { "epoch": 7.177090653548841, "grad_norm": 0.14214344322681427, "learning_rate": 2.8550948699929724e-05, "loss": 0.0284, "step": 10213 }, { "epoch": 7.1777933942375265, "grad_norm": 0.25608178973197937, "learning_rate": 2.8550480206137268e-05, "loss": 0.0217, "step": 10214 }, { "epoch": 7.178496134926212, "grad_norm": 0.09270798414945602, "learning_rate": 2.855001171234481e-05, "loss": 0.0158, "step": 10215 }, { "epoch": 7.179198875614898, "grad_norm": 0.12952062487602234, "learning_rate": 2.8549543218552355e-05, "loss": 0.0099, "step": 10216 }, { "epoch": 7.179901616303584, "grad_norm": 0.1527094542980194, "learning_rate": 2.8549074724759895e-05, "loss": 0.0156, "step": 10217 }, { "epoch": 7.18060435699227, "grad_norm": 0.1334049105644226, "learning_rate": 2.854860623096744e-05, "loss": 0.0137, "step": 10218 }, { "epoch": 7.181307097680956, "grad_norm": 0.2547875642776489, "learning_rate": 2.8548137737174983e-05, "loss": 0.0087, "step": 10219 }, { "epoch": 7.182009838369642, "grad_norm": 0.2413707971572876, "learning_rate": 2.8547669243382527e-05, "loss": 0.0262, "step": 10220 }, { "epoch": 7.182712579058327, "grad_norm": 0.3481943607330322, "learning_rate": 2.854720074959007e-05, "loss": 0.0149, "step": 10221 }, { "epoch": 7.183415319747013, "grad_norm": 0.11380240321159363, "learning_rate": 2.854673225579761e-05, "loss": 0.0159, "step": 10222 }, { "epoch": 7.184118060435699, "grad_norm": 0.32091024518013, "learning_rate": 2.8546263762005154e-05, "loss": 0.0172, "step": 10223 }, { "epoch": 7.184820801124385, "grad_norm": 0.32872241735458374, "learning_rate": 2.8545795268212698e-05, "loss": 0.0284, "step": 10224 }, { "epoch": 7.185523541813071, "grad_norm": 0.24769607186317444, "learning_rate": 2.8545326774420242e-05, "loss": 0.0328, "step": 10225 }, { "epoch": 7.186226282501757, "grad_norm": 0.29678109288215637, "learning_rate": 2.8544858280627782e-05, "loss": 0.0328, "step": 10226 }, { "epoch": 7.186929023190443, "grad_norm": 0.14913757145404816, "learning_rate": 2.8544389786835326e-05, "loss": 0.0297, "step": 10227 }, { "epoch": 7.1876317638791285, "grad_norm": 0.19414037466049194, "learning_rate": 2.854392129304287e-05, "loss": 0.0336, "step": 10228 }, { "epoch": 7.188334504567814, "grad_norm": 0.304997056722641, "learning_rate": 2.854345279925041e-05, "loss": 0.079, "step": 10229 }, { "epoch": 7.1890372452565, "grad_norm": 0.25853022933006287, "learning_rate": 2.854298430545795e-05, "loss": 0.0442, "step": 10230 }, { "epoch": 7.189739985945186, "grad_norm": 0.6109940409660339, "learning_rate": 2.8542515811665494e-05, "loss": 0.0468, "step": 10231 }, { "epoch": 7.190442726633872, "grad_norm": 0.34757736325263977, "learning_rate": 2.8542047317873038e-05, "loss": 0.0692, "step": 10232 }, { "epoch": 7.191145467322558, "grad_norm": 0.36909008026123047, "learning_rate": 2.854157882408058e-05, "loss": 0.1092, "step": 10233 }, { "epoch": 7.191848208011244, "grad_norm": 0.8091253638267517, "learning_rate": 2.8541110330288125e-05, "loss": 0.1447, "step": 10234 }, { "epoch": 7.19255094869993, "grad_norm": 0.7732283473014832, "learning_rate": 2.8540641836495666e-05, "loss": 0.1899, "step": 10235 }, { "epoch": 7.193253689388616, "grad_norm": 1.6791417598724365, "learning_rate": 2.854017334270321e-05, "loss": 0.2202, "step": 10236 }, { "epoch": 7.193956430077302, "grad_norm": 0.27857550978660583, "learning_rate": 2.8539704848910753e-05, "loss": 0.0786, "step": 10237 }, { "epoch": 7.194659170765988, "grad_norm": 0.1252336949110031, "learning_rate": 2.8539236355118297e-05, "loss": 0.024, "step": 10238 }, { "epoch": 7.195361911454674, "grad_norm": 0.3486173152923584, "learning_rate": 2.8538767861325837e-05, "loss": 0.0249, "step": 10239 }, { "epoch": 7.1960646521433596, "grad_norm": 0.10986652225255966, "learning_rate": 2.853829936753338e-05, "loss": 0.019, "step": 10240 }, { "epoch": 7.196767392832045, "grad_norm": 0.11141145974397659, "learning_rate": 2.8537830873740925e-05, "loss": 0.0177, "step": 10241 }, { "epoch": 7.1974701335207305, "grad_norm": 0.18326082825660706, "learning_rate": 2.8537362379948468e-05, "loss": 0.0188, "step": 10242 }, { "epoch": 7.198172874209416, "grad_norm": 0.1080092340707779, "learning_rate": 2.853689388615601e-05, "loss": 0.0099, "step": 10243 }, { "epoch": 7.198875614898102, "grad_norm": 0.1285722553730011, "learning_rate": 2.8536425392363552e-05, "loss": 0.0114, "step": 10244 }, { "epoch": 7.199578355586788, "grad_norm": 0.1594901829957962, "learning_rate": 2.8535956898571096e-05, "loss": 0.0399, "step": 10245 }, { "epoch": 7.200281096275474, "grad_norm": 0.10483086854219437, "learning_rate": 2.8535488404778636e-05, "loss": 0.0148, "step": 10246 }, { "epoch": 7.20098383696416, "grad_norm": 0.2697985768318176, "learning_rate": 2.853501991098618e-05, "loss": 0.0472, "step": 10247 }, { "epoch": 7.201686577652846, "grad_norm": 0.3288112282752991, "learning_rate": 2.853455141719372e-05, "loss": 0.0187, "step": 10248 }, { "epoch": 7.202389318341532, "grad_norm": 0.32084712386131287, "learning_rate": 2.8534082923401264e-05, "loss": 0.0292, "step": 10249 }, { "epoch": 7.203092059030218, "grad_norm": 0.18897058069705963, "learning_rate": 2.8533614429608808e-05, "loss": 0.0227, "step": 10250 }, { "epoch": 7.203794799718904, "grad_norm": 0.15327641367912292, "learning_rate": 2.853314593581635e-05, "loss": 0.0319, "step": 10251 }, { "epoch": 7.20449754040759, "grad_norm": 0.1434965282678604, "learning_rate": 2.8532677442023892e-05, "loss": 0.0234, "step": 10252 }, { "epoch": 7.205200281096276, "grad_norm": 0.3542994558811188, "learning_rate": 2.8532208948231436e-05, "loss": 0.0439, "step": 10253 }, { "epoch": 7.205903021784962, "grad_norm": 2.5716421604156494, "learning_rate": 2.853174045443898e-05, "loss": 0.0578, "step": 10254 }, { "epoch": 7.2066057624736475, "grad_norm": 0.21026571094989777, "learning_rate": 2.8531271960646523e-05, "loss": 0.0414, "step": 10255 }, { "epoch": 7.207308503162333, "grad_norm": 0.22160819172859192, "learning_rate": 2.8530803466854063e-05, "loss": 0.0389, "step": 10256 }, { "epoch": 7.208011243851019, "grad_norm": 0.4351951777935028, "learning_rate": 2.8530334973061607e-05, "loss": 0.096, "step": 10257 }, { "epoch": 7.208713984539705, "grad_norm": 1.0251717567443848, "learning_rate": 2.852986647926915e-05, "loss": 0.1046, "step": 10258 }, { "epoch": 7.20941672522839, "grad_norm": 0.4992195665836334, "learning_rate": 2.8529397985476695e-05, "loss": 0.1864, "step": 10259 }, { "epoch": 7.210119465917076, "grad_norm": 1.6881484985351562, "learning_rate": 2.852892949168424e-05, "loss": 0.1948, "step": 10260 }, { "epoch": 7.210822206605762, "grad_norm": 1.1406725645065308, "learning_rate": 2.852846099789178e-05, "loss": 0.243, "step": 10261 }, { "epoch": 7.211524947294448, "grad_norm": 0.2856580317020416, "learning_rate": 2.8527992504099322e-05, "loss": 0.0686, "step": 10262 }, { "epoch": 7.212227687983134, "grad_norm": 0.19931462407112122, "learning_rate": 2.8527524010306866e-05, "loss": 0.0208, "step": 10263 }, { "epoch": 7.21293042867182, "grad_norm": 0.3550182580947876, "learning_rate": 2.8527055516514406e-05, "loss": 0.0369, "step": 10264 }, { "epoch": 7.213633169360506, "grad_norm": 0.17302246391773224, "learning_rate": 2.8526587022721947e-05, "loss": 0.0272, "step": 10265 }, { "epoch": 7.214335910049192, "grad_norm": 0.13865232467651367, "learning_rate": 2.852611852892949e-05, "loss": 0.0225, "step": 10266 }, { "epoch": 7.215038650737878, "grad_norm": 0.1648714691400528, "learning_rate": 2.8525650035137034e-05, "loss": 0.0263, "step": 10267 }, { "epoch": 7.215741391426564, "grad_norm": 0.09360802173614502, "learning_rate": 2.8525181541344578e-05, "loss": 0.008, "step": 10268 }, { "epoch": 7.2164441321152495, "grad_norm": 0.12513944506645203, "learning_rate": 2.8524713047552122e-05, "loss": 0.0277, "step": 10269 }, { "epoch": 7.217146872803935, "grad_norm": 0.28357958793640137, "learning_rate": 2.8524244553759662e-05, "loss": 0.0269, "step": 10270 }, { "epoch": 7.217849613492621, "grad_norm": 0.1366243213415146, "learning_rate": 2.8523776059967206e-05, "loss": 0.0207, "step": 10271 }, { "epoch": 7.218552354181307, "grad_norm": 0.1564950942993164, "learning_rate": 2.852330756617475e-05, "loss": 0.0295, "step": 10272 }, { "epoch": 7.219255094869993, "grad_norm": 0.1429765373468399, "learning_rate": 2.8522839072382293e-05, "loss": 0.0189, "step": 10273 }, { "epoch": 7.219957835558679, "grad_norm": 0.2499867081642151, "learning_rate": 2.8522370578589834e-05, "loss": 0.0284, "step": 10274 }, { "epoch": 7.220660576247365, "grad_norm": 0.23082762956619263, "learning_rate": 2.8521902084797377e-05, "loss": 0.0195, "step": 10275 }, { "epoch": 7.221363316936051, "grad_norm": 0.26708701252937317, "learning_rate": 2.852143359100492e-05, "loss": 0.045, "step": 10276 }, { "epoch": 7.222066057624737, "grad_norm": 0.2792053818702698, "learning_rate": 2.8520965097212465e-05, "loss": 0.0318, "step": 10277 }, { "epoch": 7.222768798313423, "grad_norm": 0.25259965658187866, "learning_rate": 2.8520496603420005e-05, "loss": 0.0213, "step": 10278 }, { "epoch": 7.223471539002108, "grad_norm": 1.1060000658035278, "learning_rate": 2.852002810962755e-05, "loss": 0.0436, "step": 10279 }, { "epoch": 7.224174279690794, "grad_norm": 0.5065612196922302, "learning_rate": 2.8519559615835093e-05, "loss": 0.0383, "step": 10280 }, { "epoch": 7.22487702037948, "grad_norm": 0.47776541113853455, "learning_rate": 2.8519091122042633e-05, "loss": 0.0623, "step": 10281 }, { "epoch": 7.225579761068166, "grad_norm": 0.6321958899497986, "learning_rate": 2.8518622628250177e-05, "loss": 0.0902, "step": 10282 }, { "epoch": 7.2262825017568515, "grad_norm": 0.9729534983634949, "learning_rate": 2.8518154134457717e-05, "loss": 0.136, "step": 10283 }, { "epoch": 7.226985242445537, "grad_norm": 0.6829997897148132, "learning_rate": 2.851768564066526e-05, "loss": 0.1359, "step": 10284 }, { "epoch": 7.227687983134223, "grad_norm": 0.6882984042167664, "learning_rate": 2.8517217146872804e-05, "loss": 0.1654, "step": 10285 }, { "epoch": 7.228390723822909, "grad_norm": 1.6272637844085693, "learning_rate": 2.8516748653080348e-05, "loss": 0.2451, "step": 10286 }, { "epoch": 7.229093464511595, "grad_norm": 0.36419960856437683, "learning_rate": 2.851628015928789e-05, "loss": 0.0871, "step": 10287 }, { "epoch": 7.229796205200281, "grad_norm": 0.13561080396175385, "learning_rate": 2.8515811665495432e-05, "loss": 0.0205, "step": 10288 }, { "epoch": 7.230498945888967, "grad_norm": 0.14468660950660706, "learning_rate": 2.8515343171702976e-05, "loss": 0.0289, "step": 10289 }, { "epoch": 7.231201686577653, "grad_norm": 0.17493852972984314, "learning_rate": 2.851487467791052e-05, "loss": 0.0364, "step": 10290 }, { "epoch": 7.231904427266339, "grad_norm": 0.16301265358924866, "learning_rate": 2.851440618411806e-05, "loss": 0.0144, "step": 10291 }, { "epoch": 7.232607167955025, "grad_norm": 0.2231975644826889, "learning_rate": 2.8513937690325604e-05, "loss": 0.0287, "step": 10292 }, { "epoch": 7.233309908643711, "grad_norm": 0.27214279770851135, "learning_rate": 2.8513469196533147e-05, "loss": 0.0245, "step": 10293 }, { "epoch": 7.234012649332397, "grad_norm": 0.09416429698467255, "learning_rate": 2.851300070274069e-05, "loss": 0.0136, "step": 10294 }, { "epoch": 7.2347153900210825, "grad_norm": 0.2306353598833084, "learning_rate": 2.8512532208948235e-05, "loss": 0.021, "step": 10295 }, { "epoch": 7.2354181307097685, "grad_norm": 0.12029793113470078, "learning_rate": 2.8512063715155775e-05, "loss": 0.0145, "step": 10296 }, { "epoch": 7.236120871398454, "grad_norm": 0.2912665009498596, "learning_rate": 2.851159522136332e-05, "loss": 0.0344, "step": 10297 }, { "epoch": 7.236823612087139, "grad_norm": 0.2325769066810608, "learning_rate": 2.8511126727570863e-05, "loss": 0.0146, "step": 10298 }, { "epoch": 7.237526352775825, "grad_norm": 0.22863753139972687, "learning_rate": 2.8510658233778403e-05, "loss": 0.0604, "step": 10299 }, { "epoch": 7.238229093464511, "grad_norm": 0.17557984590530396, "learning_rate": 2.8510189739985943e-05, "loss": 0.0177, "step": 10300 }, { "epoch": 7.238931834153197, "grad_norm": 0.9744418859481812, "learning_rate": 2.8509721246193487e-05, "loss": 0.0264, "step": 10301 }, { "epoch": 7.239634574841883, "grad_norm": 0.12234855443239212, "learning_rate": 2.850925275240103e-05, "loss": 0.0253, "step": 10302 }, { "epoch": 7.240337315530569, "grad_norm": 0.16430984437465668, "learning_rate": 2.8508784258608575e-05, "loss": 0.0184, "step": 10303 }, { "epoch": 7.241040056219255, "grad_norm": 0.35078901052474976, "learning_rate": 2.8508315764816115e-05, "loss": 0.0372, "step": 10304 }, { "epoch": 7.241742796907941, "grad_norm": 0.2321939766407013, "learning_rate": 2.850784727102366e-05, "loss": 0.0424, "step": 10305 }, { "epoch": 7.242445537596627, "grad_norm": 0.37762629985809326, "learning_rate": 2.8507378777231202e-05, "loss": 0.0631, "step": 10306 }, { "epoch": 7.243148278285313, "grad_norm": 0.332434743642807, "learning_rate": 2.8506910283438746e-05, "loss": 0.0878, "step": 10307 }, { "epoch": 7.243851018973999, "grad_norm": 0.6782646775245667, "learning_rate": 2.850644178964629e-05, "loss": 0.1529, "step": 10308 }, { "epoch": 7.2445537596626846, "grad_norm": 0.6274649500846863, "learning_rate": 2.850597329585383e-05, "loss": 0.1407, "step": 10309 }, { "epoch": 7.2452565003513705, "grad_norm": 1.2348347902297974, "learning_rate": 2.8505504802061374e-05, "loss": 0.2074, "step": 10310 }, { "epoch": 7.245959241040056, "grad_norm": 1.4151883125305176, "learning_rate": 2.8505036308268918e-05, "loss": 0.2418, "step": 10311 }, { "epoch": 7.246661981728742, "grad_norm": 0.20694303512573242, "learning_rate": 2.850456781447646e-05, "loss": 0.061, "step": 10312 }, { "epoch": 7.247364722417428, "grad_norm": 0.2714124023914337, "learning_rate": 2.8504099320684e-05, "loss": 0.0488, "step": 10313 }, { "epoch": 7.248067463106114, "grad_norm": 0.4520392417907715, "learning_rate": 2.8503630826891545e-05, "loss": 0.0378, "step": 10314 }, { "epoch": 7.2487702037948, "grad_norm": 0.0924195647239685, "learning_rate": 2.850316233309909e-05, "loss": 0.0199, "step": 10315 }, { "epoch": 7.249472944483486, "grad_norm": 0.1298542022705078, "learning_rate": 2.850269383930663e-05, "loss": 0.0145, "step": 10316 }, { "epoch": 7.250175685172172, "grad_norm": 0.11263538151979446, "learning_rate": 2.850222534551417e-05, "loss": 0.0194, "step": 10317 }, { "epoch": 7.250878425860857, "grad_norm": 0.3544664680957794, "learning_rate": 2.8501756851721713e-05, "loss": 0.0219, "step": 10318 }, { "epoch": 7.251581166549543, "grad_norm": 0.1171259805560112, "learning_rate": 2.8501288357929257e-05, "loss": 0.0182, "step": 10319 }, { "epoch": 7.252283907238229, "grad_norm": 0.17066800594329834, "learning_rate": 2.85008198641368e-05, "loss": 0.0272, "step": 10320 }, { "epoch": 7.252986647926915, "grad_norm": 0.1493062525987625, "learning_rate": 2.8500351370344345e-05, "loss": 0.0201, "step": 10321 }, { "epoch": 7.253689388615601, "grad_norm": 0.1766098141670227, "learning_rate": 2.8499882876551885e-05, "loss": 0.0302, "step": 10322 }, { "epoch": 7.254392129304287, "grad_norm": 0.16030563414096832, "learning_rate": 2.849941438275943e-05, "loss": 0.0136, "step": 10323 }, { "epoch": 7.2550948699929725, "grad_norm": 0.18429824709892273, "learning_rate": 2.8498945888966972e-05, "loss": 0.024, "step": 10324 }, { "epoch": 7.255797610681658, "grad_norm": 0.134675532579422, "learning_rate": 2.8498477395174516e-05, "loss": 0.017, "step": 10325 }, { "epoch": 7.256500351370344, "grad_norm": 0.17428089678287506, "learning_rate": 2.8498008901382056e-05, "loss": 0.0195, "step": 10326 }, { "epoch": 7.25720309205903, "grad_norm": 0.22683987021446228, "learning_rate": 2.84975404075896e-05, "loss": 0.0311, "step": 10327 }, { "epoch": 7.257905832747716, "grad_norm": 0.18900929391384125, "learning_rate": 2.8497071913797144e-05, "loss": 0.0225, "step": 10328 }, { "epoch": 7.258608573436402, "grad_norm": 0.22217869758605957, "learning_rate": 2.8496603420004688e-05, "loss": 0.0281, "step": 10329 }, { "epoch": 7.259311314125088, "grad_norm": 0.3300914466381073, "learning_rate": 2.8496134926212228e-05, "loss": 0.0606, "step": 10330 }, { "epoch": 7.260014054813774, "grad_norm": 0.3087947964668274, "learning_rate": 2.8495666432419772e-05, "loss": 0.0447, "step": 10331 }, { "epoch": 7.26071679550246, "grad_norm": 0.6203629374504089, "learning_rate": 2.8495197938627315e-05, "loss": 0.0887, "step": 10332 }, { "epoch": 7.261419536191146, "grad_norm": 0.49109920859336853, "learning_rate": 2.849472944483486e-05, "loss": 0.1088, "step": 10333 }, { "epoch": 7.262122276879832, "grad_norm": 0.9639217853546143, "learning_rate": 2.84942609510424e-05, "loss": 0.166, "step": 10334 }, { "epoch": 7.262825017568518, "grad_norm": 1.5175729990005493, "learning_rate": 2.849379245724994e-05, "loss": 0.1893, "step": 10335 }, { "epoch": 7.263527758257203, "grad_norm": 2.330139398574829, "learning_rate": 2.8493323963457484e-05, "loss": 0.2445, "step": 10336 }, { "epoch": 7.264230498945889, "grad_norm": 0.16940627992153168, "learning_rate": 2.8492855469665027e-05, "loss": 0.0653, "step": 10337 }, { "epoch": 7.2649332396345745, "grad_norm": 0.14737801253795624, "learning_rate": 2.849238697587257e-05, "loss": 0.03, "step": 10338 }, { "epoch": 7.26563598032326, "grad_norm": 0.13761994242668152, "learning_rate": 2.849191848208011e-05, "loss": 0.0221, "step": 10339 }, { "epoch": 7.266338721011946, "grad_norm": 0.17808987200260162, "learning_rate": 2.8491449988287655e-05, "loss": 0.0184, "step": 10340 }, { "epoch": 7.267041461700632, "grad_norm": 0.1494632363319397, "learning_rate": 2.84909814944952e-05, "loss": 0.0235, "step": 10341 }, { "epoch": 7.267744202389318, "grad_norm": 0.13249599933624268, "learning_rate": 2.8490513000702743e-05, "loss": 0.0156, "step": 10342 }, { "epoch": 7.268446943078004, "grad_norm": 0.17698417603969574, "learning_rate": 2.8490044506910283e-05, "loss": 0.0155, "step": 10343 }, { "epoch": 7.26914968376669, "grad_norm": 0.1979515254497528, "learning_rate": 2.8489576013117827e-05, "loss": 0.0311, "step": 10344 }, { "epoch": 7.269852424455376, "grad_norm": 0.22893185913562775, "learning_rate": 2.848910751932537e-05, "loss": 0.0212, "step": 10345 }, { "epoch": 7.270555165144062, "grad_norm": 0.09988623112440109, "learning_rate": 2.8488639025532914e-05, "loss": 0.0076, "step": 10346 }, { "epoch": 7.271257905832748, "grad_norm": 0.13591383397579193, "learning_rate": 2.8488170531740458e-05, "loss": 0.0344, "step": 10347 }, { "epoch": 7.271960646521434, "grad_norm": 0.16994917392730713, "learning_rate": 2.8487702037947998e-05, "loss": 0.0189, "step": 10348 }, { "epoch": 7.27266338721012, "grad_norm": 0.2642715871334076, "learning_rate": 2.8487233544155542e-05, "loss": 0.0299, "step": 10349 }, { "epoch": 7.2733661278988055, "grad_norm": 0.1713116466999054, "learning_rate": 2.8486765050363086e-05, "loss": 0.0164, "step": 10350 }, { "epoch": 7.2740688685874915, "grad_norm": 0.21155306696891785, "learning_rate": 2.8486296556570626e-05, "loss": 0.028, "step": 10351 }, { "epoch": 7.274771609276177, "grad_norm": 0.31445974111557007, "learning_rate": 2.8485828062778166e-05, "loss": 0.0376, "step": 10352 }, { "epoch": 7.275474349964863, "grad_norm": 0.17005281150341034, "learning_rate": 2.848535956898571e-05, "loss": 0.0171, "step": 10353 }, { "epoch": 7.276177090653549, "grad_norm": 0.27626538276672363, "learning_rate": 2.8484891075193254e-05, "loss": 0.0493, "step": 10354 }, { "epoch": 7.276879831342235, "grad_norm": 0.24024821817874908, "learning_rate": 2.8484422581400797e-05, "loss": 0.031, "step": 10355 }, { "epoch": 7.27758257203092, "grad_norm": 0.6003785133361816, "learning_rate": 2.8483954087608338e-05, "loss": 0.058, "step": 10356 }, { "epoch": 7.278285312719606, "grad_norm": 0.4444182515144348, "learning_rate": 2.848348559381588e-05, "loss": 0.0719, "step": 10357 }, { "epoch": 7.278988053408292, "grad_norm": 0.6167373061180115, "learning_rate": 2.8483017100023425e-05, "loss": 0.1186, "step": 10358 }, { "epoch": 7.279690794096978, "grad_norm": 0.8001025319099426, "learning_rate": 2.848254860623097e-05, "loss": 0.1584, "step": 10359 }, { "epoch": 7.280393534785664, "grad_norm": 0.6468865275382996, "learning_rate": 2.8482080112438513e-05, "loss": 0.1695, "step": 10360 }, { "epoch": 7.28109627547435, "grad_norm": 1.1256662607192993, "learning_rate": 2.8481611618646053e-05, "loss": 0.2173, "step": 10361 }, { "epoch": 7.281799016163036, "grad_norm": 0.28214147686958313, "learning_rate": 2.8481143124853597e-05, "loss": 0.0691, "step": 10362 }, { "epoch": 7.282501756851722, "grad_norm": 0.15920858085155487, "learning_rate": 2.848067463106114e-05, "loss": 0.0232, "step": 10363 }, { "epoch": 7.2832044975404076, "grad_norm": 0.20006294548511505, "learning_rate": 2.8480206137268684e-05, "loss": 0.0283, "step": 10364 }, { "epoch": 7.2839072382290935, "grad_norm": 0.42874813079833984, "learning_rate": 2.8479737643476224e-05, "loss": 0.0239, "step": 10365 }, { "epoch": 7.284609978917779, "grad_norm": 0.15485365688800812, "learning_rate": 2.8479269149683768e-05, "loss": 0.0129, "step": 10366 }, { "epoch": 7.285312719606465, "grad_norm": 0.1676255464553833, "learning_rate": 2.8478800655891312e-05, "loss": 0.0192, "step": 10367 }, { "epoch": 7.286015460295151, "grad_norm": 0.10650492459535599, "learning_rate": 2.8478332162098852e-05, "loss": 0.0193, "step": 10368 }, { "epoch": 7.286718200983837, "grad_norm": 0.27482712268829346, "learning_rate": 2.8477863668306393e-05, "loss": 0.0197, "step": 10369 }, { "epoch": 7.287420941672523, "grad_norm": 0.22629332542419434, "learning_rate": 2.8477395174513936e-05, "loss": 0.0491, "step": 10370 }, { "epoch": 7.288123682361209, "grad_norm": 0.16915306448936462, "learning_rate": 2.847692668072148e-05, "loss": 0.0315, "step": 10371 }, { "epoch": 7.288826423049895, "grad_norm": 0.561231791973114, "learning_rate": 2.8476458186929024e-05, "loss": 0.0312, "step": 10372 }, { "epoch": 7.289529163738581, "grad_norm": 0.23933129012584686, "learning_rate": 2.8475989693136568e-05, "loss": 0.0151, "step": 10373 }, { "epoch": 7.290231904427266, "grad_norm": 0.27414438128471375, "learning_rate": 2.8475521199344108e-05, "loss": 0.0388, "step": 10374 }, { "epoch": 7.290934645115952, "grad_norm": 0.20048178732395172, "learning_rate": 2.847505270555165e-05, "loss": 0.0251, "step": 10375 }, { "epoch": 7.291637385804638, "grad_norm": 0.23961147665977478, "learning_rate": 2.8474584211759195e-05, "loss": 0.0225, "step": 10376 }, { "epoch": 7.292340126493324, "grad_norm": 0.19407658278942108, "learning_rate": 2.847411571796674e-05, "loss": 0.0484, "step": 10377 }, { "epoch": 7.29304286718201, "grad_norm": 0.2739998400211334, "learning_rate": 2.847364722417428e-05, "loss": 0.0212, "step": 10378 }, { "epoch": 7.2937456078706955, "grad_norm": 0.17252039909362793, "learning_rate": 2.8473178730381823e-05, "loss": 0.0287, "step": 10379 }, { "epoch": 7.294448348559381, "grad_norm": 0.30976852774620056, "learning_rate": 2.8472710236589367e-05, "loss": 0.0362, "step": 10380 }, { "epoch": 7.295151089248067, "grad_norm": 0.2535390555858612, "learning_rate": 2.847224174279691e-05, "loss": 0.0507, "step": 10381 }, { "epoch": 7.295853829936753, "grad_norm": 0.596549391746521, "learning_rate": 2.847177324900445e-05, "loss": 0.0698, "step": 10382 }, { "epoch": 7.296556570625439, "grad_norm": 0.4932982921600342, "learning_rate": 2.8471304755211995e-05, "loss": 0.1191, "step": 10383 }, { "epoch": 7.297259311314125, "grad_norm": 0.584754228591919, "learning_rate": 2.847083626141954e-05, "loss": 0.1667, "step": 10384 }, { "epoch": 7.297962052002811, "grad_norm": 0.9123458862304688, "learning_rate": 2.8470367767627082e-05, "loss": 0.199, "step": 10385 }, { "epoch": 7.298664792691497, "grad_norm": 1.1075067520141602, "learning_rate": 2.8469899273834622e-05, "loss": 0.2265, "step": 10386 }, { "epoch": 7.299367533380183, "grad_norm": 0.4391394555568695, "learning_rate": 2.8469430780042163e-05, "loss": 0.0625, "step": 10387 }, { "epoch": 7.300070274068869, "grad_norm": 0.15349258482456207, "learning_rate": 2.8468962286249706e-05, "loss": 0.0284, "step": 10388 }, { "epoch": 7.300773014757555, "grad_norm": 0.5178960561752319, "learning_rate": 2.846849379245725e-05, "loss": 0.0201, "step": 10389 }, { "epoch": 7.301475755446241, "grad_norm": 0.19062484800815582, "learning_rate": 2.8468025298664794e-05, "loss": 0.0204, "step": 10390 }, { "epoch": 7.3021784961349265, "grad_norm": 0.18384404480457306, "learning_rate": 2.8467556804872334e-05, "loss": 0.0393, "step": 10391 }, { "epoch": 7.3028812368236125, "grad_norm": 0.1592058539390564, "learning_rate": 2.8467088311079878e-05, "loss": 0.0189, "step": 10392 }, { "epoch": 7.303583977512298, "grad_norm": 0.1342972218990326, "learning_rate": 2.846661981728742e-05, "loss": 0.0239, "step": 10393 }, { "epoch": 7.304286718200984, "grad_norm": 0.1623118370771408, "learning_rate": 2.8466151323494965e-05, "loss": 0.0234, "step": 10394 }, { "epoch": 7.304989458889669, "grad_norm": 0.20503197610378265, "learning_rate": 2.8465682829702506e-05, "loss": 0.0317, "step": 10395 }, { "epoch": 7.305692199578355, "grad_norm": 0.16726870834827423, "learning_rate": 2.846521433591005e-05, "loss": 0.0372, "step": 10396 }, { "epoch": 7.306394940267041, "grad_norm": 0.19541022181510925, "learning_rate": 2.8464745842117593e-05, "loss": 0.0233, "step": 10397 }, { "epoch": 7.307097680955727, "grad_norm": 0.20428824424743652, "learning_rate": 2.8464277348325137e-05, "loss": 0.0192, "step": 10398 }, { "epoch": 7.307800421644413, "grad_norm": 0.14790138602256775, "learning_rate": 2.846380885453268e-05, "loss": 0.0229, "step": 10399 }, { "epoch": 7.308503162333099, "grad_norm": 0.17800500988960266, "learning_rate": 2.846334036074022e-05, "loss": 0.0194, "step": 10400 }, { "epoch": 7.309205903021785, "grad_norm": 0.3868916630744934, "learning_rate": 2.8462871866947765e-05, "loss": 0.042, "step": 10401 }, { "epoch": 7.309908643710471, "grad_norm": 0.27041420340538025, "learning_rate": 2.846240337315531e-05, "loss": 0.0426, "step": 10402 }, { "epoch": 7.310611384399157, "grad_norm": 0.2313060611486435, "learning_rate": 2.846193487936285e-05, "loss": 0.0143, "step": 10403 }, { "epoch": 7.311314125087843, "grad_norm": 0.35430532693862915, "learning_rate": 2.846146638557039e-05, "loss": 0.0431, "step": 10404 }, { "epoch": 7.3120168657765285, "grad_norm": 0.2583432197570801, "learning_rate": 2.8460997891777933e-05, "loss": 0.0521, "step": 10405 }, { "epoch": 7.3127196064652145, "grad_norm": 0.2959102988243103, "learning_rate": 2.8460529397985477e-05, "loss": 0.0594, "step": 10406 }, { "epoch": 7.3134223471539, "grad_norm": 0.36543363332748413, "learning_rate": 2.846006090419302e-05, "loss": 0.0993, "step": 10407 }, { "epoch": 7.314125087842586, "grad_norm": 0.7257431149482727, "learning_rate": 2.845959241040056e-05, "loss": 0.1376, "step": 10408 }, { "epoch": 7.314827828531272, "grad_norm": 0.5509680509567261, "learning_rate": 2.8459123916608104e-05, "loss": 0.1391, "step": 10409 }, { "epoch": 7.315530569219958, "grad_norm": 7.324347496032715, "learning_rate": 2.8458655422815648e-05, "loss": 0.1784, "step": 10410 }, { "epoch": 7.316233309908644, "grad_norm": 0.9884722828865051, "learning_rate": 2.8458186929023192e-05, "loss": 0.2362, "step": 10411 }, { "epoch": 7.31693605059733, "grad_norm": 0.27390021085739136, "learning_rate": 2.8457718435230736e-05, "loss": 0.0613, "step": 10412 }, { "epoch": 7.317638791286015, "grad_norm": 0.18463802337646484, "learning_rate": 2.8457249941438276e-05, "loss": 0.0376, "step": 10413 }, { "epoch": 7.318341531974701, "grad_norm": 0.15919458866119385, "learning_rate": 2.845678144764582e-05, "loss": 0.037, "step": 10414 }, { "epoch": 7.319044272663387, "grad_norm": 0.11907456070184708, "learning_rate": 2.8456312953853363e-05, "loss": 0.0169, "step": 10415 }, { "epoch": 7.319747013352073, "grad_norm": 0.1804346889257431, "learning_rate": 2.8455844460060907e-05, "loss": 0.0342, "step": 10416 }, { "epoch": 7.320449754040759, "grad_norm": 0.19085757434368134, "learning_rate": 2.8455375966268447e-05, "loss": 0.0187, "step": 10417 }, { "epoch": 7.321152494729445, "grad_norm": 0.11340057104825974, "learning_rate": 2.845490747247599e-05, "loss": 0.0171, "step": 10418 }, { "epoch": 7.3218552354181305, "grad_norm": 0.11471317708492279, "learning_rate": 2.8454438978683535e-05, "loss": 0.0194, "step": 10419 }, { "epoch": 7.3225579761068165, "grad_norm": 0.22009965777397156, "learning_rate": 2.845397048489108e-05, "loss": 0.0223, "step": 10420 }, { "epoch": 7.323260716795502, "grad_norm": 0.16760312020778656, "learning_rate": 2.8453501991098615e-05, "loss": 0.0254, "step": 10421 }, { "epoch": 7.323963457484188, "grad_norm": 0.15125016868114471, "learning_rate": 2.845303349730616e-05, "loss": 0.0341, "step": 10422 }, { "epoch": 7.324666198172874, "grad_norm": 0.17375117540359497, "learning_rate": 2.8452565003513703e-05, "loss": 0.0192, "step": 10423 }, { "epoch": 7.32536893886156, "grad_norm": 0.16467566788196564, "learning_rate": 2.8452096509721247e-05, "loss": 0.0336, "step": 10424 }, { "epoch": 7.326071679550246, "grad_norm": 0.26694604754447937, "learning_rate": 2.845162801592879e-05, "loss": 0.0222, "step": 10425 }, { "epoch": 7.326774420238932, "grad_norm": 0.20642630755901337, "learning_rate": 2.845115952213633e-05, "loss": 0.0239, "step": 10426 }, { "epoch": 7.327477160927618, "grad_norm": 0.40643855929374695, "learning_rate": 2.8450691028343874e-05, "loss": 0.0409, "step": 10427 }, { "epoch": 7.328179901616304, "grad_norm": 0.13728871941566467, "learning_rate": 2.8450222534551418e-05, "loss": 0.0195, "step": 10428 }, { "epoch": 7.32888264230499, "grad_norm": 0.39342474937438965, "learning_rate": 2.8449754040758962e-05, "loss": 0.0499, "step": 10429 }, { "epoch": 7.329585382993676, "grad_norm": 0.3275671601295471, "learning_rate": 2.8449285546966502e-05, "loss": 0.0376, "step": 10430 }, { "epoch": 7.330288123682362, "grad_norm": 0.489454984664917, "learning_rate": 2.8448817053174046e-05, "loss": 0.0675, "step": 10431 }, { "epoch": 7.3309908643710475, "grad_norm": 0.6827239394187927, "learning_rate": 2.844834855938159e-05, "loss": 0.0684, "step": 10432 }, { "epoch": 7.3316936050597326, "grad_norm": 1.0112680196762085, "learning_rate": 2.8447880065589133e-05, "loss": 0.132, "step": 10433 }, { "epoch": 7.3323963457484185, "grad_norm": 0.7853127121925354, "learning_rate": 2.8447411571796674e-05, "loss": 0.1421, "step": 10434 }, { "epoch": 7.333099086437104, "grad_norm": 0.6934992074966431, "learning_rate": 2.8446943078004217e-05, "loss": 0.1952, "step": 10435 }, { "epoch": 7.33380182712579, "grad_norm": 0.9677119851112366, "learning_rate": 2.844647458421176e-05, "loss": 0.2328, "step": 10436 }, { "epoch": 7.334504567814476, "grad_norm": 0.19413159787654877, "learning_rate": 2.8446006090419305e-05, "loss": 0.0524, "step": 10437 }, { "epoch": 7.335207308503162, "grad_norm": 0.15914732217788696, "learning_rate": 2.8445537596626845e-05, "loss": 0.0269, "step": 10438 }, { "epoch": 7.335910049191848, "grad_norm": 0.1236315667629242, "learning_rate": 2.8445069102834386e-05, "loss": 0.0216, "step": 10439 }, { "epoch": 7.336612789880534, "grad_norm": 0.15729321539402008, "learning_rate": 2.844460060904193e-05, "loss": 0.0205, "step": 10440 }, { "epoch": 7.33731553056922, "grad_norm": 0.1517043262720108, "learning_rate": 2.8444132115249473e-05, "loss": 0.022, "step": 10441 }, { "epoch": 7.338018271257906, "grad_norm": 0.07353862375020981, "learning_rate": 2.8443663621457017e-05, "loss": 0.0065, "step": 10442 }, { "epoch": 7.338721011946592, "grad_norm": 0.1447957307100296, "learning_rate": 2.8443195127664557e-05, "loss": 0.0186, "step": 10443 }, { "epoch": 7.339423752635278, "grad_norm": 0.09559643268585205, "learning_rate": 2.84427266338721e-05, "loss": 0.0146, "step": 10444 }, { "epoch": 7.340126493323964, "grad_norm": 0.12981081008911133, "learning_rate": 2.8442258140079645e-05, "loss": 0.0196, "step": 10445 }, { "epoch": 7.3408292340126495, "grad_norm": 0.25640490651130676, "learning_rate": 2.8441789646287188e-05, "loss": 0.0158, "step": 10446 }, { "epoch": 7.3415319747013355, "grad_norm": 0.24355754256248474, "learning_rate": 2.844132115249473e-05, "loss": 0.0263, "step": 10447 }, { "epoch": 7.342234715390021, "grad_norm": 0.118146151304245, "learning_rate": 2.8440852658702272e-05, "loss": 0.0092, "step": 10448 }, { "epoch": 7.342937456078707, "grad_norm": 0.2259831428527832, "learning_rate": 2.8440384164909816e-05, "loss": 0.0345, "step": 10449 }, { "epoch": 7.343640196767393, "grad_norm": 0.1666029393672943, "learning_rate": 2.843991567111736e-05, "loss": 0.0182, "step": 10450 }, { "epoch": 7.344342937456078, "grad_norm": 0.343281090259552, "learning_rate": 2.8439447177324904e-05, "loss": 0.0317, "step": 10451 }, { "epoch": 7.345045678144764, "grad_norm": 0.26514601707458496, "learning_rate": 2.8438978683532444e-05, "loss": 0.0365, "step": 10452 }, { "epoch": 7.34574841883345, "grad_norm": 0.1713542938232422, "learning_rate": 2.8438510189739988e-05, "loss": 0.0196, "step": 10453 }, { "epoch": 7.346451159522136, "grad_norm": 0.26672330498695374, "learning_rate": 2.843804169594753e-05, "loss": 0.0402, "step": 10454 }, { "epoch": 7.347153900210822, "grad_norm": 0.19605450332164764, "learning_rate": 2.843757320215507e-05, "loss": 0.0393, "step": 10455 }, { "epoch": 7.347856640899508, "grad_norm": 0.3171564042568207, "learning_rate": 2.8437104708362612e-05, "loss": 0.0663, "step": 10456 }, { "epoch": 7.348559381588194, "grad_norm": 0.5458561182022095, "learning_rate": 2.8436636214570156e-05, "loss": 0.085, "step": 10457 }, { "epoch": 7.34926212227688, "grad_norm": 1.7240244150161743, "learning_rate": 2.84361677207777e-05, "loss": 0.1051, "step": 10458 }, { "epoch": 7.349964862965566, "grad_norm": 0.7960001230239868, "learning_rate": 2.8435699226985243e-05, "loss": 0.2132, "step": 10459 }, { "epoch": 7.3506676036542515, "grad_norm": 0.7211527228355408, "learning_rate": 2.8435230733192787e-05, "loss": 0.2228, "step": 10460 }, { "epoch": 7.3513703443429375, "grad_norm": 1.157808780670166, "learning_rate": 2.8434762239400327e-05, "loss": 0.2229, "step": 10461 }, { "epoch": 7.352073085031623, "grad_norm": 0.27953049540519714, "learning_rate": 2.843429374560787e-05, "loss": 0.0654, "step": 10462 }, { "epoch": 7.352775825720309, "grad_norm": 0.5273613929748535, "learning_rate": 2.8433825251815415e-05, "loss": 0.036, "step": 10463 }, { "epoch": 7.353478566408995, "grad_norm": 0.3574502170085907, "learning_rate": 2.843335675802296e-05, "loss": 0.0257, "step": 10464 }, { "epoch": 7.354181307097681, "grad_norm": 0.13401782512664795, "learning_rate": 2.84328882642305e-05, "loss": 0.027, "step": 10465 }, { "epoch": 7.354884047786367, "grad_norm": 0.23847301304340363, "learning_rate": 2.8432419770438042e-05, "loss": 0.0285, "step": 10466 }, { "epoch": 7.355586788475053, "grad_norm": 0.23563116788864136, "learning_rate": 2.8431951276645586e-05, "loss": 0.0179, "step": 10467 }, { "epoch": 7.356289529163739, "grad_norm": 0.1962539553642273, "learning_rate": 2.843148278285313e-05, "loss": 0.0216, "step": 10468 }, { "epoch": 7.356992269852425, "grad_norm": 0.3375108242034912, "learning_rate": 2.843101428906067e-05, "loss": 0.0202, "step": 10469 }, { "epoch": 7.357695010541111, "grad_norm": 0.1782933920621872, "learning_rate": 2.8430545795268214e-05, "loss": 0.0233, "step": 10470 }, { "epoch": 7.358397751229797, "grad_norm": 0.1545080542564392, "learning_rate": 2.8430077301475758e-05, "loss": 0.0238, "step": 10471 }, { "epoch": 7.359100491918482, "grad_norm": 0.2395140826702118, "learning_rate": 2.84296088076833e-05, "loss": 0.0424, "step": 10472 }, { "epoch": 7.359803232607168, "grad_norm": 0.07319673895835876, "learning_rate": 2.8429140313890842e-05, "loss": 0.006, "step": 10473 }, { "epoch": 7.3605059732958535, "grad_norm": 0.34980326890945435, "learning_rate": 2.8428671820098382e-05, "loss": 0.0504, "step": 10474 }, { "epoch": 7.3612087139845395, "grad_norm": 0.24999083578586578, "learning_rate": 2.8428203326305926e-05, "loss": 0.0141, "step": 10475 }, { "epoch": 7.361911454673225, "grad_norm": 0.2103220671415329, "learning_rate": 2.842773483251347e-05, "loss": 0.0322, "step": 10476 }, { "epoch": 7.362614195361911, "grad_norm": 0.19323067367076874, "learning_rate": 2.8427266338721013e-05, "loss": 0.0445, "step": 10477 }, { "epoch": 7.363316936050597, "grad_norm": 0.20985902845859528, "learning_rate": 2.8426797844928554e-05, "loss": 0.0141, "step": 10478 }, { "epoch": 7.364019676739283, "grad_norm": 0.22468793392181396, "learning_rate": 2.8426329351136097e-05, "loss": 0.0441, "step": 10479 }, { "epoch": 7.364722417427969, "grad_norm": 0.3281170129776001, "learning_rate": 2.842586085734364e-05, "loss": 0.0526, "step": 10480 }, { "epoch": 7.365425158116655, "grad_norm": 0.5101807713508606, "learning_rate": 2.8425392363551185e-05, "loss": 0.0478, "step": 10481 }, { "epoch": 7.366127898805341, "grad_norm": 0.38307905197143555, "learning_rate": 2.8424923869758725e-05, "loss": 0.0707, "step": 10482 }, { "epoch": 7.366830639494027, "grad_norm": 0.47086235880851746, "learning_rate": 2.842445537596627e-05, "loss": 0.1069, "step": 10483 }, { "epoch": 7.367533380182713, "grad_norm": 0.585972011089325, "learning_rate": 2.8423986882173813e-05, "loss": 0.1563, "step": 10484 }, { "epoch": 7.368236120871399, "grad_norm": 0.983415424823761, "learning_rate": 2.8423518388381356e-05, "loss": 0.1651, "step": 10485 }, { "epoch": 7.368938861560085, "grad_norm": 3.3123984336853027, "learning_rate": 2.84230498945889e-05, "loss": 0.2489, "step": 10486 }, { "epoch": 7.3696416022487705, "grad_norm": 0.2934236228466034, "learning_rate": 2.842258140079644e-05, "loss": 0.0508, "step": 10487 }, { "epoch": 7.370344342937456, "grad_norm": 0.1326909065246582, "learning_rate": 2.8422112907003984e-05, "loss": 0.0216, "step": 10488 }, { "epoch": 7.371047083626142, "grad_norm": 0.24555827677249908, "learning_rate": 2.8421644413211528e-05, "loss": 0.0241, "step": 10489 }, { "epoch": 7.371749824314827, "grad_norm": 0.2059486210346222, "learning_rate": 2.8421175919419068e-05, "loss": 0.0201, "step": 10490 }, { "epoch": 7.372452565003513, "grad_norm": 0.18533802032470703, "learning_rate": 2.842070742562661e-05, "loss": 0.0293, "step": 10491 }, { "epoch": 7.373155305692199, "grad_norm": 0.08496317267417908, "learning_rate": 2.8420238931834152e-05, "loss": 0.0075, "step": 10492 }, { "epoch": 7.373858046380885, "grad_norm": 0.2709095776081085, "learning_rate": 2.8419770438041696e-05, "loss": 0.0434, "step": 10493 }, { "epoch": 7.374560787069571, "grad_norm": 0.12856853008270264, "learning_rate": 2.841930194424924e-05, "loss": 0.0209, "step": 10494 }, { "epoch": 7.375263527758257, "grad_norm": 0.22988146543502808, "learning_rate": 2.841883345045678e-05, "loss": 0.0213, "step": 10495 }, { "epoch": 7.375966268446943, "grad_norm": 0.14610132575035095, "learning_rate": 2.8418364956664324e-05, "loss": 0.0224, "step": 10496 }, { "epoch": 7.376669009135629, "grad_norm": 0.28792551159858704, "learning_rate": 2.8417896462871867e-05, "loss": 0.0326, "step": 10497 }, { "epoch": 7.377371749824315, "grad_norm": 0.30379000306129456, "learning_rate": 2.841742796907941e-05, "loss": 0.0177, "step": 10498 }, { "epoch": 7.378074490513001, "grad_norm": 0.25649094581604004, "learning_rate": 2.8416959475286955e-05, "loss": 0.02, "step": 10499 }, { "epoch": 7.378777231201687, "grad_norm": 0.12725698947906494, "learning_rate": 2.8416490981494495e-05, "loss": 0.0179, "step": 10500 }, { "epoch": 7.3794799718903725, "grad_norm": 0.2620183527469635, "learning_rate": 2.841602248770204e-05, "loss": 0.0414, "step": 10501 }, { "epoch": 7.3801827125790584, "grad_norm": 0.20270845293998718, "learning_rate": 2.8415553993909583e-05, "loss": 0.0379, "step": 10502 }, { "epoch": 7.380885453267744, "grad_norm": 0.19780506193637848, "learning_rate": 2.8415085500117126e-05, "loss": 0.0304, "step": 10503 }, { "epoch": 7.38158819395643, "grad_norm": 0.17212386429309845, "learning_rate": 2.8414617006324667e-05, "loss": 0.0331, "step": 10504 }, { "epoch": 7.382290934645116, "grad_norm": 0.4438331425189972, "learning_rate": 2.841414851253221e-05, "loss": 0.0474, "step": 10505 }, { "epoch": 7.382993675333802, "grad_norm": 0.2356790155172348, "learning_rate": 2.8413680018739754e-05, "loss": 0.0752, "step": 10506 }, { "epoch": 7.383696416022488, "grad_norm": 0.4439500868320465, "learning_rate": 2.8413211524947298e-05, "loss": 0.0698, "step": 10507 }, { "epoch": 7.384399156711174, "grad_norm": 0.4779139757156372, "learning_rate": 2.8412743031154835e-05, "loss": 0.1183, "step": 10508 }, { "epoch": 7.38510189739986, "grad_norm": 0.7290963530540466, "learning_rate": 2.841227453736238e-05, "loss": 0.1596, "step": 10509 }, { "epoch": 7.385804638088545, "grad_norm": 2.1408159732818604, "learning_rate": 2.8411806043569922e-05, "loss": 0.1926, "step": 10510 }, { "epoch": 7.386507378777231, "grad_norm": 1.3024338483810425, "learning_rate": 2.8411337549777466e-05, "loss": 0.2271, "step": 10511 }, { "epoch": 7.387210119465917, "grad_norm": 0.19773553311824799, "learning_rate": 2.841086905598501e-05, "loss": 0.0767, "step": 10512 }, { "epoch": 7.387912860154603, "grad_norm": 0.2115238606929779, "learning_rate": 2.841040056219255e-05, "loss": 0.0294, "step": 10513 }, { "epoch": 7.388615600843289, "grad_norm": 0.13151119649410248, "learning_rate": 2.8409932068400094e-05, "loss": 0.0238, "step": 10514 }, { "epoch": 7.3893183415319745, "grad_norm": 0.07701414823532104, "learning_rate": 2.8409463574607638e-05, "loss": 0.0128, "step": 10515 }, { "epoch": 7.3900210822206605, "grad_norm": 0.16367045044898987, "learning_rate": 2.840899508081518e-05, "loss": 0.0172, "step": 10516 }, { "epoch": 7.390723822909346, "grad_norm": 0.09563209116458893, "learning_rate": 2.840852658702272e-05, "loss": 0.0136, "step": 10517 }, { "epoch": 7.391426563598032, "grad_norm": 0.18029619753360748, "learning_rate": 2.8408058093230265e-05, "loss": 0.0204, "step": 10518 }, { "epoch": 7.392129304286718, "grad_norm": 0.22265611588954926, "learning_rate": 2.840758959943781e-05, "loss": 0.0204, "step": 10519 }, { "epoch": 7.392832044975404, "grad_norm": 0.8001222610473633, "learning_rate": 2.8407121105645353e-05, "loss": 0.0194, "step": 10520 }, { "epoch": 7.39353478566409, "grad_norm": 0.13820567727088928, "learning_rate": 2.8406652611852893e-05, "loss": 0.0162, "step": 10521 }, { "epoch": 7.394237526352776, "grad_norm": 0.2336248904466629, "learning_rate": 2.8406184118060437e-05, "loss": 0.0307, "step": 10522 }, { "epoch": 7.394940267041462, "grad_norm": 0.14529676735401154, "learning_rate": 2.840571562426798e-05, "loss": 0.0207, "step": 10523 }, { "epoch": 7.395643007730148, "grad_norm": 0.25968992710113525, "learning_rate": 2.8405247130475524e-05, "loss": 0.0538, "step": 10524 }, { "epoch": 7.396345748418834, "grad_norm": 0.1588040143251419, "learning_rate": 2.8404778636683065e-05, "loss": 0.0155, "step": 10525 }, { "epoch": 7.39704848910752, "grad_norm": 0.18006129562854767, "learning_rate": 2.8404310142890605e-05, "loss": 0.0308, "step": 10526 }, { "epoch": 7.397751229796206, "grad_norm": 0.31836530566215515, "learning_rate": 2.840384164909815e-05, "loss": 0.0399, "step": 10527 }, { "epoch": 7.398453970484891, "grad_norm": 0.2257545441389084, "learning_rate": 2.8403373155305692e-05, "loss": 0.0181, "step": 10528 }, { "epoch": 7.3991567111735765, "grad_norm": 0.16853788495063782, "learning_rate": 2.8402904661513236e-05, "loss": 0.0423, "step": 10529 }, { "epoch": 7.3998594518622625, "grad_norm": 0.2430584877729416, "learning_rate": 2.8402436167720777e-05, "loss": 0.0454, "step": 10530 }, { "epoch": 7.400562192550948, "grad_norm": 0.3546387851238251, "learning_rate": 2.840196767392832e-05, "loss": 0.0569, "step": 10531 }, { "epoch": 7.401264933239634, "grad_norm": 0.4635317325592041, "learning_rate": 2.8401499180135864e-05, "loss": 0.1246, "step": 10532 }, { "epoch": 7.40196767392832, "grad_norm": 0.5230059027671814, "learning_rate": 2.8401030686343408e-05, "loss": 0.15, "step": 10533 }, { "epoch": 7.402670414617006, "grad_norm": 1.508215069770813, "learning_rate": 2.8400562192550948e-05, "loss": 0.1856, "step": 10534 }, { "epoch": 7.403373155305692, "grad_norm": 0.5879708528518677, "learning_rate": 2.8400093698758492e-05, "loss": 0.1702, "step": 10535 }, { "epoch": 7.404075895994378, "grad_norm": 3.590294599533081, "learning_rate": 2.8399625204966035e-05, "loss": 0.246, "step": 10536 }, { "epoch": 7.404778636683064, "grad_norm": 0.28275638818740845, "learning_rate": 2.839915671117358e-05, "loss": 0.0767, "step": 10537 }, { "epoch": 7.40548137737175, "grad_norm": 0.4277175962924957, "learning_rate": 2.8398688217381123e-05, "loss": 0.0267, "step": 10538 }, { "epoch": 7.406184118060436, "grad_norm": 0.2163076400756836, "learning_rate": 2.8398219723588663e-05, "loss": 0.0182, "step": 10539 }, { "epoch": 7.406886858749122, "grad_norm": 0.15596003830432892, "learning_rate": 2.8397751229796207e-05, "loss": 0.0273, "step": 10540 }, { "epoch": 7.407589599437808, "grad_norm": 0.1374027580022812, "learning_rate": 2.839728273600375e-05, "loss": 0.0245, "step": 10541 }, { "epoch": 7.4082923401264935, "grad_norm": 0.19182781875133514, "learning_rate": 2.8396814242211294e-05, "loss": 0.012, "step": 10542 }, { "epoch": 7.408995080815179, "grad_norm": 0.5283179879188538, "learning_rate": 2.839634574841883e-05, "loss": 0.0264, "step": 10543 }, { "epoch": 7.409697821503865, "grad_norm": 0.16743826866149902, "learning_rate": 2.8395877254626375e-05, "loss": 0.026, "step": 10544 }, { "epoch": 7.410400562192551, "grad_norm": 0.18652905523777008, "learning_rate": 2.839540876083392e-05, "loss": 0.0272, "step": 10545 }, { "epoch": 7.411103302881237, "grad_norm": 0.11258243769407272, "learning_rate": 2.8394940267041463e-05, "loss": 0.0114, "step": 10546 }, { "epoch": 7.411806043569923, "grad_norm": 0.1319417953491211, "learning_rate": 2.8394471773249003e-05, "loss": 0.0232, "step": 10547 }, { "epoch": 7.412508784258609, "grad_norm": 0.13264116644859314, "learning_rate": 2.8394003279456547e-05, "loss": 0.0175, "step": 10548 }, { "epoch": 7.413211524947294, "grad_norm": 0.24843256175518036, "learning_rate": 2.839353478566409e-05, "loss": 0.0311, "step": 10549 }, { "epoch": 7.41391426563598, "grad_norm": 0.1389116644859314, "learning_rate": 2.8393066291871634e-05, "loss": 0.0169, "step": 10550 }, { "epoch": 7.414617006324666, "grad_norm": 0.4417627155780792, "learning_rate": 2.8392597798079178e-05, "loss": 0.0285, "step": 10551 }, { "epoch": 7.415319747013352, "grad_norm": 0.2646951675415039, "learning_rate": 2.8392129304286718e-05, "loss": 0.0451, "step": 10552 }, { "epoch": 7.416022487702038, "grad_norm": 0.15992045402526855, "learning_rate": 2.8391660810494262e-05, "loss": 0.0229, "step": 10553 }, { "epoch": 7.416725228390724, "grad_norm": 0.31168702244758606, "learning_rate": 2.8391192316701806e-05, "loss": 0.0491, "step": 10554 }, { "epoch": 7.41742796907941, "grad_norm": 0.36502623558044434, "learning_rate": 2.839072382290935e-05, "loss": 0.0424, "step": 10555 }, { "epoch": 7.4181307097680955, "grad_norm": 0.4073335826396942, "learning_rate": 2.839025532911689e-05, "loss": 0.0682, "step": 10556 }, { "epoch": 7.418833450456781, "grad_norm": 0.39457178115844727, "learning_rate": 2.8389786835324433e-05, "loss": 0.079, "step": 10557 }, { "epoch": 7.419536191145467, "grad_norm": 0.5093479156494141, "learning_rate": 2.8389318341531977e-05, "loss": 0.1145, "step": 10558 }, { "epoch": 7.420238931834153, "grad_norm": 1.0628330707550049, "learning_rate": 2.838884984773952e-05, "loss": 0.1597, "step": 10559 }, { "epoch": 7.420941672522839, "grad_norm": 0.8776503801345825, "learning_rate": 2.8388381353947058e-05, "loss": 0.195, "step": 10560 }, { "epoch": 7.421644413211525, "grad_norm": 1.2045971155166626, "learning_rate": 2.83879128601546e-05, "loss": 0.2045, "step": 10561 }, { "epoch": 7.422347153900211, "grad_norm": 0.2441740334033966, "learning_rate": 2.8387444366362145e-05, "loss": 0.0805, "step": 10562 }, { "epoch": 7.423049894588897, "grad_norm": 0.12697765231132507, "learning_rate": 2.838697587256969e-05, "loss": 0.0302, "step": 10563 }, { "epoch": 7.423752635277583, "grad_norm": 0.24887560307979584, "learning_rate": 2.8386507378777233e-05, "loss": 0.0189, "step": 10564 }, { "epoch": 7.424455375966269, "grad_norm": 0.13425330817699432, "learning_rate": 2.8386038884984773e-05, "loss": 0.0222, "step": 10565 }, { "epoch": 7.425158116654955, "grad_norm": 0.20837251842021942, "learning_rate": 2.8385570391192317e-05, "loss": 0.0247, "step": 10566 }, { "epoch": 7.42586085734364, "grad_norm": 0.12188433855772018, "learning_rate": 2.838510189739986e-05, "loss": 0.0137, "step": 10567 }, { "epoch": 7.426563598032326, "grad_norm": 0.21820102632045746, "learning_rate": 2.8384633403607404e-05, "loss": 0.016, "step": 10568 }, { "epoch": 7.427266338721012, "grad_norm": 0.17718496918678284, "learning_rate": 2.8384164909814945e-05, "loss": 0.027, "step": 10569 }, { "epoch": 7.4279690794096975, "grad_norm": 0.14795762300491333, "learning_rate": 2.8383696416022488e-05, "loss": 0.025, "step": 10570 }, { "epoch": 7.4286718200983834, "grad_norm": 0.17415690422058105, "learning_rate": 2.8383227922230032e-05, "loss": 0.021, "step": 10571 }, { "epoch": 7.429374560787069, "grad_norm": 0.14748606085777283, "learning_rate": 2.8382759428437576e-05, "loss": 0.0233, "step": 10572 }, { "epoch": 7.430077301475755, "grad_norm": 0.18567106127738953, "learning_rate": 2.8382290934645116e-05, "loss": 0.0161, "step": 10573 }, { "epoch": 7.430780042164441, "grad_norm": 0.1722511202096939, "learning_rate": 2.838182244085266e-05, "loss": 0.0357, "step": 10574 }, { "epoch": 7.431482782853127, "grad_norm": 0.24613983929157257, "learning_rate": 2.8381353947060203e-05, "loss": 0.0233, "step": 10575 }, { "epoch": 7.432185523541813, "grad_norm": 0.41490495204925537, "learning_rate": 2.8380885453267747e-05, "loss": 0.0606, "step": 10576 }, { "epoch": 7.432888264230499, "grad_norm": 0.17667391896247864, "learning_rate": 2.8380416959475288e-05, "loss": 0.0407, "step": 10577 }, { "epoch": 7.433591004919185, "grad_norm": 0.14221744239330292, "learning_rate": 2.8379948465682828e-05, "loss": 0.0252, "step": 10578 }, { "epoch": 7.434293745607871, "grad_norm": 0.24761223793029785, "learning_rate": 2.837947997189037e-05, "loss": 0.0333, "step": 10579 }, { "epoch": 7.434996486296557, "grad_norm": 0.2839229106903076, "learning_rate": 2.8379011478097915e-05, "loss": 0.0501, "step": 10580 }, { "epoch": 7.435699226985243, "grad_norm": 0.3814791738986969, "learning_rate": 2.837854298430546e-05, "loss": 0.0732, "step": 10581 }, { "epoch": 7.436401967673929, "grad_norm": 0.4086576998233795, "learning_rate": 2.8378074490513e-05, "loss": 0.0888, "step": 10582 }, { "epoch": 7.4371047083626145, "grad_norm": 0.6423905491828918, "learning_rate": 2.8377605996720543e-05, "loss": 0.1036, "step": 10583 }, { "epoch": 7.4378074490513, "grad_norm": 0.6485064625740051, "learning_rate": 2.8377137502928087e-05, "loss": 0.1842, "step": 10584 }, { "epoch": 7.438510189739986, "grad_norm": 0.8507484793663025, "learning_rate": 2.837666900913563e-05, "loss": 0.2183, "step": 10585 }, { "epoch": 7.439212930428672, "grad_norm": 1.324892520904541, "learning_rate": 2.837620051534317e-05, "loss": 0.2203, "step": 10586 }, { "epoch": 7.439915671117357, "grad_norm": 0.30113333463668823, "learning_rate": 2.8375732021550715e-05, "loss": 0.072, "step": 10587 }, { "epoch": 7.440618411806043, "grad_norm": 0.11682070046663284, "learning_rate": 2.837526352775826e-05, "loss": 0.0178, "step": 10588 }, { "epoch": 7.441321152494729, "grad_norm": 0.11559148877859116, "learning_rate": 2.8374795033965802e-05, "loss": 0.0163, "step": 10589 }, { "epoch": 7.442023893183415, "grad_norm": 0.14030639827251434, "learning_rate": 2.8374326540173346e-05, "loss": 0.0165, "step": 10590 }, { "epoch": 7.442726633872101, "grad_norm": 0.15410377085208893, "learning_rate": 2.8373858046380886e-05, "loss": 0.0183, "step": 10591 }, { "epoch": 7.443429374560787, "grad_norm": 0.09970470517873764, "learning_rate": 2.837338955258843e-05, "loss": 0.0122, "step": 10592 }, { "epoch": 7.444132115249473, "grad_norm": 0.17397212982177734, "learning_rate": 2.8372921058795974e-05, "loss": 0.0188, "step": 10593 }, { "epoch": 7.444834855938159, "grad_norm": 0.2665573060512543, "learning_rate": 2.8372452565003517e-05, "loss": 0.031, "step": 10594 }, { "epoch": 7.445537596626845, "grad_norm": 0.13058005273342133, "learning_rate": 2.8371984071211054e-05, "loss": 0.0206, "step": 10595 }, { "epoch": 7.446240337315531, "grad_norm": 0.15117895603179932, "learning_rate": 2.8371515577418598e-05, "loss": 0.0164, "step": 10596 }, { "epoch": 7.4469430780042165, "grad_norm": 0.3606702983379364, "learning_rate": 2.8371047083626142e-05, "loss": 0.0284, "step": 10597 }, { "epoch": 7.447645818692902, "grad_norm": 0.14950788021087646, "learning_rate": 2.8370578589833685e-05, "loss": 0.0177, "step": 10598 }, { "epoch": 7.448348559381588, "grad_norm": 0.19800828397274017, "learning_rate": 2.8370110096041226e-05, "loss": 0.0371, "step": 10599 }, { "epoch": 7.449051300070274, "grad_norm": 0.1694648712873459, "learning_rate": 2.836964160224877e-05, "loss": 0.0185, "step": 10600 }, { "epoch": 7.44975404075896, "grad_norm": 0.3975076973438263, "learning_rate": 2.8369173108456313e-05, "loss": 0.0357, "step": 10601 }, { "epoch": 7.450456781447646, "grad_norm": 0.22653664648532867, "learning_rate": 2.8368704614663857e-05, "loss": 0.0313, "step": 10602 }, { "epoch": 7.451159522136332, "grad_norm": 0.3020986318588257, "learning_rate": 2.83682361208714e-05, "loss": 0.0241, "step": 10603 }, { "epoch": 7.451862262825018, "grad_norm": 0.4413236975669861, "learning_rate": 2.836776762707894e-05, "loss": 0.0386, "step": 10604 }, { "epoch": 7.452565003513703, "grad_norm": 0.2668340802192688, "learning_rate": 2.8367299133286485e-05, "loss": 0.0496, "step": 10605 }, { "epoch": 7.453267744202389, "grad_norm": 1.5601617097854614, "learning_rate": 2.836683063949403e-05, "loss": 0.0583, "step": 10606 }, { "epoch": 7.453970484891075, "grad_norm": 1.0509473085403442, "learning_rate": 2.8366362145701572e-05, "loss": 0.0935, "step": 10607 }, { "epoch": 7.454673225579761, "grad_norm": 0.6408141851425171, "learning_rate": 2.8365893651909113e-05, "loss": 0.1059, "step": 10608 }, { "epoch": 7.455375966268447, "grad_norm": 1.0212446451187134, "learning_rate": 2.8365425158116656e-05, "loss": 0.1692, "step": 10609 }, { "epoch": 7.456078706957133, "grad_norm": 0.9130793213844299, "learning_rate": 2.83649566643242e-05, "loss": 0.1992, "step": 10610 }, { "epoch": 7.4567814476458185, "grad_norm": 1.4474031925201416, "learning_rate": 2.8364488170531744e-05, "loss": 0.2335, "step": 10611 }, { "epoch": 7.457484188334504, "grad_norm": 0.5880023241043091, "learning_rate": 2.836401967673928e-05, "loss": 0.0692, "step": 10612 }, { "epoch": 7.45818692902319, "grad_norm": 0.18484948575496674, "learning_rate": 2.8363551182946824e-05, "loss": 0.0363, "step": 10613 }, { "epoch": 7.458889669711876, "grad_norm": 0.1209612786769867, "learning_rate": 2.8363082689154368e-05, "loss": 0.0221, "step": 10614 }, { "epoch": 7.459592410400562, "grad_norm": 0.11206749826669693, "learning_rate": 2.8362614195361912e-05, "loss": 0.0229, "step": 10615 }, { "epoch": 7.460295151089248, "grad_norm": 0.20072774589061737, "learning_rate": 2.8362145701569456e-05, "loss": 0.0252, "step": 10616 }, { "epoch": 7.460997891777934, "grad_norm": 0.17713138461112976, "learning_rate": 2.8361677207776996e-05, "loss": 0.0144, "step": 10617 }, { "epoch": 7.46170063246662, "grad_norm": 0.38906341791152954, "learning_rate": 2.836120871398454e-05, "loss": 0.0209, "step": 10618 }, { "epoch": 7.462403373155306, "grad_norm": 0.27872174978256226, "learning_rate": 2.8360740220192083e-05, "loss": 0.0305, "step": 10619 }, { "epoch": 7.463106113843992, "grad_norm": 0.0964062288403511, "learning_rate": 2.8360271726399627e-05, "loss": 0.0219, "step": 10620 }, { "epoch": 7.463808854532678, "grad_norm": 0.19678758084774017, "learning_rate": 2.8359803232607167e-05, "loss": 0.0348, "step": 10621 }, { "epoch": 7.464511595221364, "grad_norm": 0.13471576571464539, "learning_rate": 2.835933473881471e-05, "loss": 0.024, "step": 10622 }, { "epoch": 7.46521433591005, "grad_norm": 0.18811187148094177, "learning_rate": 2.8358866245022255e-05, "loss": 0.0132, "step": 10623 }, { "epoch": 7.4659170765987355, "grad_norm": 0.28778398036956787, "learning_rate": 2.83583977512298e-05, "loss": 0.0312, "step": 10624 }, { "epoch": 7.466619817287421, "grad_norm": 0.12143117189407349, "learning_rate": 2.835792925743734e-05, "loss": 0.0165, "step": 10625 }, { "epoch": 7.4673225579761064, "grad_norm": 0.17773276567459106, "learning_rate": 2.8357460763644883e-05, "loss": 0.0332, "step": 10626 }, { "epoch": 7.468025298664792, "grad_norm": 0.2489572912454605, "learning_rate": 2.8356992269852426e-05, "loss": 0.0311, "step": 10627 }, { "epoch": 7.468728039353478, "grad_norm": 0.1792534440755844, "learning_rate": 2.835652377605997e-05, "loss": 0.0209, "step": 10628 }, { "epoch": 7.469430780042164, "grad_norm": 0.4841915965080261, "learning_rate": 2.8356055282267514e-05, "loss": 0.0415, "step": 10629 }, { "epoch": 7.47013352073085, "grad_norm": 0.4914470613002777, "learning_rate": 2.835558678847505e-05, "loss": 0.0414, "step": 10630 }, { "epoch": 7.470836261419536, "grad_norm": 0.21250887215137482, "learning_rate": 2.8355118294682594e-05, "loss": 0.0558, "step": 10631 }, { "epoch": 7.471539002108222, "grad_norm": 0.44283729791641235, "learning_rate": 2.8354649800890138e-05, "loss": 0.0842, "step": 10632 }, { "epoch": 7.472241742796908, "grad_norm": 0.4207106828689575, "learning_rate": 2.8354181307097682e-05, "loss": 0.1187, "step": 10633 }, { "epoch": 7.472944483485594, "grad_norm": 1.054202914237976, "learning_rate": 2.8353712813305222e-05, "loss": 0.1668, "step": 10634 }, { "epoch": 7.47364722417428, "grad_norm": 1.0300121307373047, "learning_rate": 2.8353244319512766e-05, "loss": 0.2037, "step": 10635 }, { "epoch": 7.474349964862966, "grad_norm": 1.2358908653259277, "learning_rate": 2.835277582572031e-05, "loss": 0.2749, "step": 10636 }, { "epoch": 7.475052705551652, "grad_norm": 0.31794044375419617, "learning_rate": 2.8352307331927853e-05, "loss": 0.0995, "step": 10637 }, { "epoch": 7.4757554462403375, "grad_norm": 0.13416437804698944, "learning_rate": 2.8351838838135394e-05, "loss": 0.0288, "step": 10638 }, { "epoch": 7.476458186929023, "grad_norm": 0.1766093224287033, "learning_rate": 2.8351370344342938e-05, "loss": 0.0239, "step": 10639 }, { "epoch": 7.477160927617709, "grad_norm": 0.16348446905612946, "learning_rate": 2.835090185055048e-05, "loss": 0.0167, "step": 10640 }, { "epoch": 7.477863668306395, "grad_norm": 0.2329167276620865, "learning_rate": 2.8350433356758025e-05, "loss": 0.0239, "step": 10641 }, { "epoch": 7.478566408995081, "grad_norm": 0.09506922215223312, "learning_rate": 2.834996486296557e-05, "loss": 0.0095, "step": 10642 }, { "epoch": 7.479269149683767, "grad_norm": 0.2192562371492386, "learning_rate": 2.834949636917311e-05, "loss": 0.0267, "step": 10643 }, { "epoch": 7.479971890372452, "grad_norm": 0.14238616824150085, "learning_rate": 2.8349027875380653e-05, "loss": 0.0263, "step": 10644 }, { "epoch": 7.480674631061138, "grad_norm": 0.20545785129070282, "learning_rate": 2.8348559381588196e-05, "loss": 0.0245, "step": 10645 }, { "epoch": 7.481377371749824, "grad_norm": 0.14105238020420074, "learning_rate": 2.834809088779574e-05, "loss": 0.014, "step": 10646 }, { "epoch": 7.48208011243851, "grad_norm": 0.313215434551239, "learning_rate": 2.8347622394003277e-05, "loss": 0.0342, "step": 10647 }, { "epoch": 7.482782853127196, "grad_norm": 0.1561059057712555, "learning_rate": 2.834715390021082e-05, "loss": 0.015, "step": 10648 }, { "epoch": 7.483485593815882, "grad_norm": 0.3599403202533722, "learning_rate": 2.8346685406418365e-05, "loss": 0.04, "step": 10649 }, { "epoch": 7.484188334504568, "grad_norm": 0.1628771722316742, "learning_rate": 2.834621691262591e-05, "loss": 0.0141, "step": 10650 }, { "epoch": 7.484891075193254, "grad_norm": 0.2045862078666687, "learning_rate": 2.834574841883345e-05, "loss": 0.0352, "step": 10651 }, { "epoch": 7.4855938158819395, "grad_norm": 0.226817786693573, "learning_rate": 2.8345279925040992e-05, "loss": 0.0445, "step": 10652 }, { "epoch": 7.486296556570625, "grad_norm": 0.3865256905555725, "learning_rate": 2.8344811431248536e-05, "loss": 0.0411, "step": 10653 }, { "epoch": 7.486999297259311, "grad_norm": 0.39821138978004456, "learning_rate": 2.834434293745608e-05, "loss": 0.0284, "step": 10654 }, { "epoch": 7.487702037947997, "grad_norm": 0.3296501636505127, "learning_rate": 2.8343874443663624e-05, "loss": 0.0734, "step": 10655 }, { "epoch": 7.488404778636683, "grad_norm": 2.7139339447021484, "learning_rate": 2.8343405949871164e-05, "loss": 0.0631, "step": 10656 }, { "epoch": 7.489107519325369, "grad_norm": 0.4213346838951111, "learning_rate": 2.8342937456078708e-05, "loss": 0.0787, "step": 10657 }, { "epoch": 7.489810260014055, "grad_norm": 0.7409932017326355, "learning_rate": 2.834246896228625e-05, "loss": 0.1015, "step": 10658 }, { "epoch": 7.490513000702741, "grad_norm": 0.6038417220115662, "learning_rate": 2.8342000468493795e-05, "loss": 0.1659, "step": 10659 }, { "epoch": 7.491215741391427, "grad_norm": 1.5768237113952637, "learning_rate": 2.8341531974701335e-05, "loss": 0.2243, "step": 10660 }, { "epoch": 7.491918482080113, "grad_norm": 1.4545433521270752, "learning_rate": 2.834106348090888e-05, "loss": 0.2478, "step": 10661 }, { "epoch": 7.492621222768799, "grad_norm": 0.2824713885784149, "learning_rate": 2.8340594987116423e-05, "loss": 0.0722, "step": 10662 }, { "epoch": 7.493323963457485, "grad_norm": 0.17180027067661285, "learning_rate": 2.8340126493323967e-05, "loss": 0.0219, "step": 10663 }, { "epoch": 7.49402670414617, "grad_norm": 0.21919317543506622, "learning_rate": 2.8339657999531507e-05, "loss": 0.0376, "step": 10664 }, { "epoch": 7.494729444834856, "grad_norm": 0.22748374938964844, "learning_rate": 2.8339189505739047e-05, "loss": 0.0203, "step": 10665 }, { "epoch": 7.4954321855235415, "grad_norm": 0.24662034213542938, "learning_rate": 2.833872101194659e-05, "loss": 0.0201, "step": 10666 }, { "epoch": 7.496134926212227, "grad_norm": 0.22386488318443298, "learning_rate": 2.8338252518154135e-05, "loss": 0.0206, "step": 10667 }, { "epoch": 7.496837666900913, "grad_norm": 0.15499936044216156, "learning_rate": 2.833778402436168e-05, "loss": 0.0171, "step": 10668 }, { "epoch": 7.497540407589599, "grad_norm": 0.19838346540927887, "learning_rate": 2.833731553056922e-05, "loss": 0.0221, "step": 10669 }, { "epoch": 7.498243148278285, "grad_norm": 0.1712058037519455, "learning_rate": 2.8336847036776763e-05, "loss": 0.0267, "step": 10670 }, { "epoch": 7.498945888966971, "grad_norm": 0.16609561443328857, "learning_rate": 2.8336378542984306e-05, "loss": 0.0188, "step": 10671 }, { "epoch": 7.499648629655657, "grad_norm": 0.3011547327041626, "learning_rate": 2.833591004919185e-05, "loss": 0.0308, "step": 10672 }, { "epoch": 7.500351370344343, "grad_norm": 0.12463817745447159, "learning_rate": 2.833544155539939e-05, "loss": 0.0142, "step": 10673 }, { "epoch": 7.501054111033029, "grad_norm": 0.13234646618366241, "learning_rate": 2.8334973061606934e-05, "loss": 0.0187, "step": 10674 }, { "epoch": 7.501756851721715, "grad_norm": 0.10232234746217728, "learning_rate": 2.8334504567814478e-05, "loss": 0.0141, "step": 10675 }, { "epoch": 7.502459592410401, "grad_norm": 0.25026074051856995, "learning_rate": 2.833403607402202e-05, "loss": 0.0292, "step": 10676 }, { "epoch": 7.503162333099087, "grad_norm": 0.21424022316932678, "learning_rate": 2.8333567580229562e-05, "loss": 0.042, "step": 10677 }, { "epoch": 7.503865073787773, "grad_norm": 0.21093806624412537, "learning_rate": 2.8333099086437106e-05, "loss": 0.0144, "step": 10678 }, { "epoch": 7.5045678144764585, "grad_norm": 0.31423652172088623, "learning_rate": 2.833263059264465e-05, "loss": 0.0446, "step": 10679 }, { "epoch": 7.505270555165144, "grad_norm": 0.3037637770175934, "learning_rate": 2.8332162098852193e-05, "loss": 0.0333, "step": 10680 }, { "epoch": 7.505973295853829, "grad_norm": 0.35030969977378845, "learning_rate": 2.8331693605059737e-05, "loss": 0.0641, "step": 10681 }, { "epoch": 7.506676036542515, "grad_norm": 0.6258600354194641, "learning_rate": 2.8331225111267274e-05, "loss": 0.0822, "step": 10682 }, { "epoch": 7.507378777231201, "grad_norm": 0.9611529111862183, "learning_rate": 2.8330756617474817e-05, "loss": 0.1139, "step": 10683 }, { "epoch": 7.508081517919887, "grad_norm": 1.0978703498840332, "learning_rate": 2.833028812368236e-05, "loss": 0.156, "step": 10684 }, { "epoch": 7.508784258608573, "grad_norm": 0.9545159339904785, "learning_rate": 2.8329819629889905e-05, "loss": 0.1976, "step": 10685 }, { "epoch": 7.509486999297259, "grad_norm": 1.8281903266906738, "learning_rate": 2.8329351136097445e-05, "loss": 0.2357, "step": 10686 }, { "epoch": 7.510189739985945, "grad_norm": 0.33966514468193054, "learning_rate": 2.832888264230499e-05, "loss": 0.114, "step": 10687 }, { "epoch": 7.510892480674631, "grad_norm": 0.19123859703540802, "learning_rate": 2.8328414148512533e-05, "loss": 0.0295, "step": 10688 }, { "epoch": 7.511595221363317, "grad_norm": 0.13880367577075958, "learning_rate": 2.8327945654720076e-05, "loss": 0.0276, "step": 10689 }, { "epoch": 7.512297962052003, "grad_norm": 0.1521795392036438, "learning_rate": 2.832747716092762e-05, "loss": 0.0182, "step": 10690 }, { "epoch": 7.513000702740689, "grad_norm": 0.13979403674602509, "learning_rate": 2.832700866713516e-05, "loss": 0.0184, "step": 10691 }, { "epoch": 7.513703443429375, "grad_norm": 0.1297173649072647, "learning_rate": 2.8326540173342704e-05, "loss": 0.0135, "step": 10692 }, { "epoch": 7.5144061841180605, "grad_norm": 0.10235393792390823, "learning_rate": 2.8326071679550248e-05, "loss": 0.0164, "step": 10693 }, { "epoch": 7.515108924806746, "grad_norm": 0.15360234677791595, "learning_rate": 2.832560318575779e-05, "loss": 0.0262, "step": 10694 }, { "epoch": 7.515811665495432, "grad_norm": 0.15548314154148102, "learning_rate": 2.8325134691965332e-05, "loss": 0.0312, "step": 10695 }, { "epoch": 7.516514406184118, "grad_norm": 0.13605989515781403, "learning_rate": 2.8324666198172876e-05, "loss": 0.0161, "step": 10696 }, { "epoch": 7.517217146872804, "grad_norm": 0.2232738435268402, "learning_rate": 2.832419770438042e-05, "loss": 0.0218, "step": 10697 }, { "epoch": 7.51791988756149, "grad_norm": 0.11694754660129547, "learning_rate": 2.8323729210587963e-05, "loss": 0.0121, "step": 10698 }, { "epoch": 7.518622628250176, "grad_norm": 0.44324690103530884, "learning_rate": 2.83232607167955e-05, "loss": 0.0435, "step": 10699 }, { "epoch": 7.519325368938862, "grad_norm": 0.25046506524086, "learning_rate": 2.8322792223003044e-05, "loss": 0.021, "step": 10700 }, { "epoch": 7.520028109627548, "grad_norm": 0.27220815420150757, "learning_rate": 2.8322323729210587e-05, "loss": 0.0312, "step": 10701 }, { "epoch": 7.520730850316234, "grad_norm": 0.20058336853981018, "learning_rate": 2.832185523541813e-05, "loss": 0.0382, "step": 10702 }, { "epoch": 7.521433591004919, "grad_norm": 0.27930834889411926, "learning_rate": 2.8321386741625675e-05, "loss": 0.0189, "step": 10703 }, { "epoch": 7.522136331693605, "grad_norm": 0.3942670524120331, "learning_rate": 2.8320918247833215e-05, "loss": 0.0519, "step": 10704 }, { "epoch": 7.522839072382291, "grad_norm": 0.4411008656024933, "learning_rate": 2.832044975404076e-05, "loss": 0.0396, "step": 10705 }, { "epoch": 7.523541813070977, "grad_norm": 0.7483949661254883, "learning_rate": 2.8319981260248303e-05, "loss": 0.0731, "step": 10706 }, { "epoch": 7.5242445537596625, "grad_norm": 1.5792996883392334, "learning_rate": 2.8319512766455846e-05, "loss": 0.1061, "step": 10707 }, { "epoch": 7.524947294448348, "grad_norm": 1.6767168045043945, "learning_rate": 2.8319044272663387e-05, "loss": 0.1256, "step": 10708 }, { "epoch": 7.525650035137034, "grad_norm": 1.8950622081756592, "learning_rate": 2.831857577887093e-05, "loss": 0.1711, "step": 10709 }, { "epoch": 7.52635277582572, "grad_norm": 1.6766992807388306, "learning_rate": 2.8318107285078474e-05, "loss": 0.1909, "step": 10710 }, { "epoch": 7.527055516514406, "grad_norm": 2.1602234840393066, "learning_rate": 2.8317638791286018e-05, "loss": 0.2554, "step": 10711 }, { "epoch": 7.527758257203092, "grad_norm": 0.7852967381477356, "learning_rate": 2.831717029749356e-05, "loss": 0.0872, "step": 10712 }, { "epoch": 7.528460997891778, "grad_norm": 0.282871276140213, "learning_rate": 2.8316701803701102e-05, "loss": 0.0402, "step": 10713 }, { "epoch": 7.529163738580464, "grad_norm": 0.1794709414243698, "learning_rate": 2.8316233309908646e-05, "loss": 0.0227, "step": 10714 }, { "epoch": 7.52986647926915, "grad_norm": 0.10614708811044693, "learning_rate": 2.831576481611619e-05, "loss": 0.0189, "step": 10715 }, { "epoch": 7.530569219957836, "grad_norm": 0.2841579020023346, "learning_rate": 2.8315296322323733e-05, "loss": 0.0224, "step": 10716 }, { "epoch": 7.531271960646522, "grad_norm": 0.20742274820804596, "learning_rate": 2.831482782853127e-05, "loss": 0.0149, "step": 10717 }, { "epoch": 7.531974701335208, "grad_norm": 0.1653134971857071, "learning_rate": 2.8314359334738814e-05, "loss": 0.0194, "step": 10718 }, { "epoch": 7.5326774420238936, "grad_norm": 1.27114737033844, "learning_rate": 2.8313890840946358e-05, "loss": 0.0443, "step": 10719 }, { "epoch": 7.533380182712579, "grad_norm": 0.1918966919183731, "learning_rate": 2.83134223471539e-05, "loss": 0.0224, "step": 10720 }, { "epoch": 7.5340829234012645, "grad_norm": 0.10330352932214737, "learning_rate": 2.831295385336144e-05, "loss": 0.0114, "step": 10721 }, { "epoch": 7.53478566408995, "grad_norm": 0.2464042752981186, "learning_rate": 2.8312485359568985e-05, "loss": 0.0317, "step": 10722 }, { "epoch": 7.535488404778636, "grad_norm": 0.13444951176643372, "learning_rate": 2.831201686577653e-05, "loss": 0.0267, "step": 10723 }, { "epoch": 7.536191145467322, "grad_norm": 0.21018241345882416, "learning_rate": 2.8311548371984073e-05, "loss": 0.0179, "step": 10724 }, { "epoch": 7.536893886156008, "grad_norm": 0.16699564456939697, "learning_rate": 2.8311079878191613e-05, "loss": 0.0141, "step": 10725 }, { "epoch": 7.537596626844694, "grad_norm": 0.21653589606285095, "learning_rate": 2.8310611384399157e-05, "loss": 0.0319, "step": 10726 }, { "epoch": 7.53829936753338, "grad_norm": 0.20772913098335266, "learning_rate": 2.83101428906067e-05, "loss": 0.0377, "step": 10727 }, { "epoch": 7.539002108222066, "grad_norm": 0.3277468979358673, "learning_rate": 2.8309674396814244e-05, "loss": 0.0256, "step": 10728 }, { "epoch": 7.539704848910752, "grad_norm": 0.19271548092365265, "learning_rate": 2.8309205903021788e-05, "loss": 0.0344, "step": 10729 }, { "epoch": 7.540407589599438, "grad_norm": 0.22056202590465546, "learning_rate": 2.830873740922933e-05, "loss": 0.0375, "step": 10730 }, { "epoch": 7.541110330288124, "grad_norm": 0.40429598093032837, "learning_rate": 2.8308268915436872e-05, "loss": 0.0667, "step": 10731 }, { "epoch": 7.54181307097681, "grad_norm": 0.23644845187664032, "learning_rate": 2.8307800421644416e-05, "loss": 0.0687, "step": 10732 }, { "epoch": 7.542515811665496, "grad_norm": 1.0590063333511353, "learning_rate": 2.830733192785196e-05, "loss": 0.1107, "step": 10733 }, { "epoch": 7.5432185523541815, "grad_norm": 1.1598362922668457, "learning_rate": 2.8306863434059497e-05, "loss": 0.16, "step": 10734 }, { "epoch": 7.543921293042867, "grad_norm": 0.7575961947441101, "learning_rate": 2.830639494026704e-05, "loss": 0.2776, "step": 10735 }, { "epoch": 7.544624033731553, "grad_norm": 1.6586041450500488, "learning_rate": 2.8305926446474584e-05, "loss": 0.2446, "step": 10736 }, { "epoch": 7.545326774420239, "grad_norm": 0.22244423627853394, "learning_rate": 2.8305457952682128e-05, "loss": 0.0698, "step": 10737 }, { "epoch": 7.546029515108925, "grad_norm": 0.19080880284309387, "learning_rate": 2.8304989458889668e-05, "loss": 0.0539, "step": 10738 }, { "epoch": 7.546732255797611, "grad_norm": 0.2726988196372986, "learning_rate": 2.8304520965097212e-05, "loss": 0.026, "step": 10739 }, { "epoch": 7.547434996486297, "grad_norm": 0.26453953981399536, "learning_rate": 2.8304052471304756e-05, "loss": 0.0411, "step": 10740 }, { "epoch": 7.548137737174983, "grad_norm": 0.13778461515903473, "learning_rate": 2.83035839775123e-05, "loss": 0.0201, "step": 10741 }, { "epoch": 7.548840477863668, "grad_norm": 0.12355674058198929, "learning_rate": 2.8303115483719843e-05, "loss": 0.0111, "step": 10742 }, { "epoch": 7.549543218552354, "grad_norm": 0.09585722535848618, "learning_rate": 2.8302646989927383e-05, "loss": 0.022, "step": 10743 }, { "epoch": 7.55024595924104, "grad_norm": 0.09562642127275467, "learning_rate": 2.8302178496134927e-05, "loss": 0.0114, "step": 10744 }, { "epoch": 7.550948699929726, "grad_norm": 0.1737682819366455, "learning_rate": 2.830171000234247e-05, "loss": 0.022, "step": 10745 }, { "epoch": 7.551651440618412, "grad_norm": 0.07139898091554642, "learning_rate": 2.8301241508550014e-05, "loss": 0.0084, "step": 10746 }, { "epoch": 7.552354181307098, "grad_norm": 0.15286242961883545, "learning_rate": 2.8300773014757555e-05, "loss": 0.0277, "step": 10747 }, { "epoch": 7.5530569219957835, "grad_norm": 0.1931137889623642, "learning_rate": 2.83003045209651e-05, "loss": 0.0223, "step": 10748 }, { "epoch": 7.553759662684469, "grad_norm": 0.48158836364746094, "learning_rate": 2.8299836027172642e-05, "loss": 0.0283, "step": 10749 }, { "epoch": 7.554462403373155, "grad_norm": 0.21861720085144043, "learning_rate": 2.8299367533380186e-05, "loss": 0.0216, "step": 10750 }, { "epoch": 7.555165144061841, "grad_norm": 0.19763682782649994, "learning_rate": 2.8298899039587726e-05, "loss": 0.0255, "step": 10751 }, { "epoch": 7.555867884750527, "grad_norm": 0.18810251355171204, "learning_rate": 2.8298430545795267e-05, "loss": 0.0343, "step": 10752 }, { "epoch": 7.556570625439213, "grad_norm": 0.24171161651611328, "learning_rate": 2.829796205200281e-05, "loss": 0.0424, "step": 10753 }, { "epoch": 7.557273366127899, "grad_norm": 0.29954347014427185, "learning_rate": 2.8297493558210354e-05, "loss": 0.0398, "step": 10754 }, { "epoch": 7.557976106816585, "grad_norm": 0.2616119980812073, "learning_rate": 2.8297025064417898e-05, "loss": 0.0356, "step": 10755 }, { "epoch": 7.558678847505271, "grad_norm": 0.32762888073921204, "learning_rate": 2.8296556570625438e-05, "loss": 0.0896, "step": 10756 }, { "epoch": 7.559381588193957, "grad_norm": 0.319198340177536, "learning_rate": 2.8296088076832982e-05, "loss": 0.087, "step": 10757 }, { "epoch": 7.560084328882642, "grad_norm": 0.5661336183547974, "learning_rate": 2.8295619583040526e-05, "loss": 0.1207, "step": 10758 }, { "epoch": 7.560787069571328, "grad_norm": 0.8191341757774353, "learning_rate": 2.829515108924807e-05, "loss": 0.1553, "step": 10759 }, { "epoch": 7.561489810260014, "grad_norm": 1.1055148839950562, "learning_rate": 2.829468259545561e-05, "loss": 0.2008, "step": 10760 }, { "epoch": 7.5621925509487, "grad_norm": 1.1429834365844727, "learning_rate": 2.8294214101663153e-05, "loss": 0.2435, "step": 10761 }, { "epoch": 7.5628952916373855, "grad_norm": 0.34870994091033936, "learning_rate": 2.8293745607870697e-05, "loss": 0.0706, "step": 10762 }, { "epoch": 7.563598032326071, "grad_norm": 0.22975225746631622, "learning_rate": 2.829327711407824e-05, "loss": 0.0359, "step": 10763 }, { "epoch": 7.564300773014757, "grad_norm": 0.13113299012184143, "learning_rate": 2.829280862028578e-05, "loss": 0.0211, "step": 10764 }, { "epoch": 7.565003513703443, "grad_norm": 0.36693960428237915, "learning_rate": 2.8292340126493325e-05, "loss": 0.0256, "step": 10765 }, { "epoch": 7.565706254392129, "grad_norm": 0.09334585070610046, "learning_rate": 2.829187163270087e-05, "loss": 0.0223, "step": 10766 }, { "epoch": 7.566408995080815, "grad_norm": 0.11906454712152481, "learning_rate": 2.8291403138908412e-05, "loss": 0.0213, "step": 10767 }, { "epoch": 7.567111735769501, "grad_norm": 0.19684453308582306, "learning_rate": 2.8290934645115956e-05, "loss": 0.0187, "step": 10768 }, { "epoch": 7.567814476458187, "grad_norm": 0.18516972661018372, "learning_rate": 2.8290466151323493e-05, "loss": 0.0187, "step": 10769 }, { "epoch": 7.568517217146873, "grad_norm": 0.257442831993103, "learning_rate": 2.8289997657531037e-05, "loss": 0.0236, "step": 10770 }, { "epoch": 7.569219957835559, "grad_norm": 0.15806253254413605, "learning_rate": 2.828952916373858e-05, "loss": 0.0292, "step": 10771 }, { "epoch": 7.569922698524245, "grad_norm": 0.8859333992004395, "learning_rate": 2.8289060669946124e-05, "loss": 0.0271, "step": 10772 }, { "epoch": 7.570625439212931, "grad_norm": 0.2095138132572174, "learning_rate": 2.8288592176153665e-05, "loss": 0.0114, "step": 10773 }, { "epoch": 7.5713281799016166, "grad_norm": 0.36159107089042664, "learning_rate": 2.8288123682361208e-05, "loss": 0.0426, "step": 10774 }, { "epoch": 7.5720309205903025, "grad_norm": 0.64947909116745, "learning_rate": 2.8287655188568752e-05, "loss": 0.0359, "step": 10775 }, { "epoch": 7.572733661278988, "grad_norm": 0.4859391152858734, "learning_rate": 2.8287186694776296e-05, "loss": 0.059, "step": 10776 }, { "epoch": 7.573436401967674, "grad_norm": 0.17076528072357178, "learning_rate": 2.8286718200983836e-05, "loss": 0.047, "step": 10777 }, { "epoch": 7.57413914265636, "grad_norm": 0.18109938502311707, "learning_rate": 2.828624970719138e-05, "loss": 0.0222, "step": 10778 }, { "epoch": 7.574841883345046, "grad_norm": 0.2497410625219345, "learning_rate": 2.8285781213398924e-05, "loss": 0.0534, "step": 10779 }, { "epoch": 7.575544624033731, "grad_norm": 0.4528485834598541, "learning_rate": 2.8285312719606467e-05, "loss": 0.0385, "step": 10780 }, { "epoch": 7.576247364722417, "grad_norm": 0.4347855746746063, "learning_rate": 2.828484422581401e-05, "loss": 0.0551, "step": 10781 }, { "epoch": 7.576950105411103, "grad_norm": 0.3641187250614166, "learning_rate": 2.828437573202155e-05, "loss": 0.0882, "step": 10782 }, { "epoch": 7.577652846099789, "grad_norm": 0.9770341515541077, "learning_rate": 2.8283907238229095e-05, "loss": 0.1334, "step": 10783 }, { "epoch": 7.578355586788475, "grad_norm": 0.5341147780418396, "learning_rate": 2.828343874443664e-05, "loss": 0.1784, "step": 10784 }, { "epoch": 7.579058327477161, "grad_norm": 0.9632569551467896, "learning_rate": 2.8282970250644182e-05, "loss": 0.1731, "step": 10785 }, { "epoch": 7.579761068165847, "grad_norm": 1.002151370048523, "learning_rate": 2.828250175685172e-05, "loss": 0.2302, "step": 10786 }, { "epoch": 7.580463808854533, "grad_norm": 0.34267657995224, "learning_rate": 2.8282033263059263e-05, "loss": 0.1081, "step": 10787 }, { "epoch": 7.581166549543219, "grad_norm": 0.24801971018314362, "learning_rate": 2.8281564769266807e-05, "loss": 0.0223, "step": 10788 }, { "epoch": 7.5818692902319045, "grad_norm": 0.1874133199453354, "learning_rate": 2.828109627547435e-05, "loss": 0.0365, "step": 10789 }, { "epoch": 7.58257203092059, "grad_norm": 0.23867028951644897, "learning_rate": 2.828062778168189e-05, "loss": 0.0194, "step": 10790 }, { "epoch": 7.583274771609276, "grad_norm": 0.19958992302417755, "learning_rate": 2.8280159287889435e-05, "loss": 0.0195, "step": 10791 }, { "epoch": 7.583977512297962, "grad_norm": 0.13952460885047913, "learning_rate": 2.827969079409698e-05, "loss": 0.0138, "step": 10792 }, { "epoch": 7.584680252986648, "grad_norm": 0.15137521922588348, "learning_rate": 2.8279222300304522e-05, "loss": 0.0167, "step": 10793 }, { "epoch": 7.585382993675334, "grad_norm": 0.10360247641801834, "learning_rate": 2.8278753806512066e-05, "loss": 0.0174, "step": 10794 }, { "epoch": 7.58608573436402, "grad_norm": 0.15554174780845642, "learning_rate": 2.8278285312719606e-05, "loss": 0.0171, "step": 10795 }, { "epoch": 7.586788475052706, "grad_norm": 0.14925043284893036, "learning_rate": 2.827781681892715e-05, "loss": 0.0149, "step": 10796 }, { "epoch": 7.587491215741391, "grad_norm": 0.2190113514661789, "learning_rate": 2.8277348325134694e-05, "loss": 0.0262, "step": 10797 }, { "epoch": 7.588193956430077, "grad_norm": 0.19190461933612823, "learning_rate": 2.8276879831342237e-05, "loss": 0.0369, "step": 10798 }, { "epoch": 7.588896697118763, "grad_norm": 0.21939511597156525, "learning_rate": 2.8276411337549778e-05, "loss": 0.0196, "step": 10799 }, { "epoch": 7.589599437807449, "grad_norm": 0.25985121726989746, "learning_rate": 2.827594284375732e-05, "loss": 0.0354, "step": 10800 }, { "epoch": 7.590302178496135, "grad_norm": 0.19660606980323792, "learning_rate": 2.8275474349964865e-05, "loss": 0.0458, "step": 10801 }, { "epoch": 7.591004919184821, "grad_norm": 0.207490935921669, "learning_rate": 2.827500585617241e-05, "loss": 0.0364, "step": 10802 }, { "epoch": 7.5917076598735065, "grad_norm": 0.19272169470787048, "learning_rate": 2.827453736237995e-05, "loss": 0.0407, "step": 10803 }, { "epoch": 7.592410400562192, "grad_norm": 0.28271013498306274, "learning_rate": 2.827406886858749e-05, "loss": 0.036, "step": 10804 }, { "epoch": 7.593113141250878, "grad_norm": 0.24151428043842316, "learning_rate": 2.8273600374795033e-05, "loss": 0.0565, "step": 10805 }, { "epoch": 7.593815881939564, "grad_norm": 0.39034363627433777, "learning_rate": 2.8273131881002577e-05, "loss": 0.0699, "step": 10806 }, { "epoch": 7.59451862262825, "grad_norm": 0.44332173466682434, "learning_rate": 2.827266338721012e-05, "loss": 0.0746, "step": 10807 }, { "epoch": 7.595221363316936, "grad_norm": 0.42000192403793335, "learning_rate": 2.827219489341766e-05, "loss": 0.1254, "step": 10808 }, { "epoch": 7.595924104005622, "grad_norm": 0.7759125828742981, "learning_rate": 2.8271726399625205e-05, "loss": 0.1618, "step": 10809 }, { "epoch": 7.596626844694308, "grad_norm": 1.122571349143982, "learning_rate": 2.827125790583275e-05, "loss": 0.255, "step": 10810 }, { "epoch": 7.597329585382994, "grad_norm": 0.7810690999031067, "learning_rate": 2.8270789412040292e-05, "loss": 0.2373, "step": 10811 }, { "epoch": 7.59803232607168, "grad_norm": 0.26431623101234436, "learning_rate": 2.8270320918247833e-05, "loss": 0.0753, "step": 10812 }, { "epoch": 7.598735066760366, "grad_norm": 0.1987958401441574, "learning_rate": 2.8269852424455376e-05, "loss": 0.0263, "step": 10813 }, { "epoch": 7.599437807449052, "grad_norm": 0.11319063603878021, "learning_rate": 2.826938393066292e-05, "loss": 0.0141, "step": 10814 }, { "epoch": 7.6001405481377375, "grad_norm": 0.22669945657253265, "learning_rate": 2.8268915436870464e-05, "loss": 0.0268, "step": 10815 }, { "epoch": 7.6008432888264235, "grad_norm": 0.17265553772449493, "learning_rate": 2.8268446943078004e-05, "loss": 0.0251, "step": 10816 }, { "epoch": 7.601546029515109, "grad_norm": 0.11201256513595581, "learning_rate": 2.8267978449285548e-05, "loss": 0.021, "step": 10817 }, { "epoch": 7.602248770203794, "grad_norm": 0.20099811255931854, "learning_rate": 2.826750995549309e-05, "loss": 0.0123, "step": 10818 }, { "epoch": 7.60295151089248, "grad_norm": 0.17893220484256744, "learning_rate": 2.8267041461700635e-05, "loss": 0.0225, "step": 10819 }, { "epoch": 7.603654251581166, "grad_norm": 0.1508701890707016, "learning_rate": 2.826657296790818e-05, "loss": 0.0236, "step": 10820 }, { "epoch": 7.604356992269852, "grad_norm": 0.1107516810297966, "learning_rate": 2.8266104474115716e-05, "loss": 0.0139, "step": 10821 }, { "epoch": 7.605059732958538, "grad_norm": 0.22968420386314392, "learning_rate": 2.826563598032326e-05, "loss": 0.048, "step": 10822 }, { "epoch": 7.605762473647224, "grad_norm": 0.18767333030700684, "learning_rate": 2.8265167486530803e-05, "loss": 0.0193, "step": 10823 }, { "epoch": 7.60646521433591, "grad_norm": 0.23809632658958435, "learning_rate": 2.8264698992738347e-05, "loss": 0.0411, "step": 10824 }, { "epoch": 7.607167955024596, "grad_norm": 0.16981585323810577, "learning_rate": 2.8264230498945887e-05, "loss": 0.023, "step": 10825 }, { "epoch": 7.607870695713282, "grad_norm": 0.18015792965888977, "learning_rate": 2.826376200515343e-05, "loss": 0.0237, "step": 10826 }, { "epoch": 7.608573436401968, "grad_norm": 0.30836722254753113, "learning_rate": 2.8263293511360975e-05, "loss": 0.0454, "step": 10827 }, { "epoch": 7.609276177090654, "grad_norm": 0.1726853996515274, "learning_rate": 2.826282501756852e-05, "loss": 0.028, "step": 10828 }, { "epoch": 7.6099789177793395, "grad_norm": 0.921586811542511, "learning_rate": 2.826235652377606e-05, "loss": 0.0408, "step": 10829 }, { "epoch": 7.6106816584680255, "grad_norm": 0.32617250084877014, "learning_rate": 2.8261888029983603e-05, "loss": 0.0685, "step": 10830 }, { "epoch": 7.611384399156711, "grad_norm": 0.4140160083770752, "learning_rate": 2.8261419536191146e-05, "loss": 0.0698, "step": 10831 }, { "epoch": 7.612087139845397, "grad_norm": 0.4601594805717468, "learning_rate": 2.826095104239869e-05, "loss": 0.0755, "step": 10832 }, { "epoch": 7.612789880534083, "grad_norm": 0.4418909549713135, "learning_rate": 2.8260482548606234e-05, "loss": 0.1192, "step": 10833 }, { "epoch": 7.613492621222769, "grad_norm": 0.9467178583145142, "learning_rate": 2.8260014054813774e-05, "loss": 0.1914, "step": 10834 }, { "epoch": 7.614195361911454, "grad_norm": 0.8182012438774109, "learning_rate": 2.8259545561021318e-05, "loss": 0.1971, "step": 10835 }, { "epoch": 7.61489810260014, "grad_norm": 0.6210089921951294, "learning_rate": 2.825907706722886e-05, "loss": 0.2186, "step": 10836 }, { "epoch": 7.615600843288826, "grad_norm": 0.29839831590652466, "learning_rate": 2.8258608573436405e-05, "loss": 0.0728, "step": 10837 }, { "epoch": 7.616303583977512, "grad_norm": 0.26708781719207764, "learning_rate": 2.8258140079643946e-05, "loss": 0.0311, "step": 10838 }, { "epoch": 7.617006324666198, "grad_norm": 0.23325778543949127, "learning_rate": 2.8257671585851486e-05, "loss": 0.0486, "step": 10839 }, { "epoch": 7.617709065354884, "grad_norm": 0.12286189198493958, "learning_rate": 2.825720309205903e-05, "loss": 0.017, "step": 10840 }, { "epoch": 7.61841180604357, "grad_norm": 0.14212694764137268, "learning_rate": 2.8256734598266574e-05, "loss": 0.019, "step": 10841 }, { "epoch": 7.619114546732256, "grad_norm": 0.14289990067481995, "learning_rate": 2.8256266104474114e-05, "loss": 0.015, "step": 10842 }, { "epoch": 7.6198172874209416, "grad_norm": 0.18127256631851196, "learning_rate": 2.8255797610681658e-05, "loss": 0.0206, "step": 10843 }, { "epoch": 7.6205200281096275, "grad_norm": 0.10319894552230835, "learning_rate": 2.82553291168892e-05, "loss": 0.0184, "step": 10844 }, { "epoch": 7.621222768798313, "grad_norm": 0.17845728993415833, "learning_rate": 2.8254860623096745e-05, "loss": 0.0249, "step": 10845 }, { "epoch": 7.621925509486999, "grad_norm": 0.11012209951877594, "learning_rate": 2.825439212930429e-05, "loss": 0.0115, "step": 10846 }, { "epoch": 7.622628250175685, "grad_norm": 0.1357268989086151, "learning_rate": 2.825392363551183e-05, "loss": 0.0181, "step": 10847 }, { "epoch": 7.623330990864371, "grad_norm": 0.18338648974895477, "learning_rate": 2.8253455141719373e-05, "loss": 0.0254, "step": 10848 }, { "epoch": 7.624033731553057, "grad_norm": 0.20355001091957092, "learning_rate": 2.8252986647926917e-05, "loss": 0.0366, "step": 10849 }, { "epoch": 7.624736472241743, "grad_norm": 0.12668006122112274, "learning_rate": 2.825251815413446e-05, "loss": 0.029, "step": 10850 }, { "epoch": 7.625439212930429, "grad_norm": 0.4038730561733246, "learning_rate": 2.8252049660342e-05, "loss": 0.0384, "step": 10851 }, { "epoch": 7.626141953619115, "grad_norm": 0.4794606864452362, "learning_rate": 2.8251581166549544e-05, "loss": 0.0363, "step": 10852 }, { "epoch": 7.626844694307801, "grad_norm": 0.19653300940990448, "learning_rate": 2.8251112672757088e-05, "loss": 0.0292, "step": 10853 }, { "epoch": 7.627547434996487, "grad_norm": 0.2582058310508728, "learning_rate": 2.8250644178964632e-05, "loss": 0.0314, "step": 10854 }, { "epoch": 7.628250175685173, "grad_norm": 0.48717835545539856, "learning_rate": 2.8250175685172172e-05, "loss": 0.0451, "step": 10855 }, { "epoch": 7.6289529163738585, "grad_norm": 0.2694604694843292, "learning_rate": 2.8249707191379712e-05, "loss": 0.0623, "step": 10856 }, { "epoch": 7.629655657062544, "grad_norm": 0.29232504963874817, "learning_rate": 2.8249238697587256e-05, "loss": 0.0762, "step": 10857 }, { "epoch": 7.6303583977512295, "grad_norm": 0.5865505337715149, "learning_rate": 2.82487702037948e-05, "loss": 0.1599, "step": 10858 }, { "epoch": 7.631061138439915, "grad_norm": 0.5123229622840881, "learning_rate": 2.8248301710002344e-05, "loss": 0.2114, "step": 10859 }, { "epoch": 7.631763879128601, "grad_norm": 1.3272191286087036, "learning_rate": 2.8247833216209884e-05, "loss": 0.2122, "step": 10860 }, { "epoch": 7.632466619817287, "grad_norm": 1.806230902671814, "learning_rate": 2.8247364722417428e-05, "loss": 0.2221, "step": 10861 }, { "epoch": 7.633169360505973, "grad_norm": 0.22920534014701843, "learning_rate": 2.824689622862497e-05, "loss": 0.07, "step": 10862 }, { "epoch": 7.633872101194659, "grad_norm": 0.13213419914245605, "learning_rate": 2.8246427734832515e-05, "loss": 0.0238, "step": 10863 }, { "epoch": 7.634574841883345, "grad_norm": 0.15717588365077972, "learning_rate": 2.8245959241040055e-05, "loss": 0.0246, "step": 10864 }, { "epoch": 7.635277582572031, "grad_norm": 0.10201577842235565, "learning_rate": 2.82454907472476e-05, "loss": 0.0187, "step": 10865 }, { "epoch": 7.635980323260717, "grad_norm": 0.24482935667037964, "learning_rate": 2.8245022253455143e-05, "loss": 0.0233, "step": 10866 }, { "epoch": 7.636683063949403, "grad_norm": 0.1199704259634018, "learning_rate": 2.8244553759662687e-05, "loss": 0.0123, "step": 10867 }, { "epoch": 7.637385804638089, "grad_norm": 0.1104881539940834, "learning_rate": 2.8244085265870227e-05, "loss": 0.0191, "step": 10868 }, { "epoch": 7.638088545326775, "grad_norm": 0.12074466049671173, "learning_rate": 2.824361677207777e-05, "loss": 0.0234, "step": 10869 }, { "epoch": 7.6387912860154605, "grad_norm": 0.47834280133247375, "learning_rate": 2.8243148278285314e-05, "loss": 0.0214, "step": 10870 }, { "epoch": 7.6394940267041465, "grad_norm": 0.10791703313589096, "learning_rate": 2.8242679784492858e-05, "loss": 0.0164, "step": 10871 }, { "epoch": 7.640196767392832, "grad_norm": 0.10438999533653259, "learning_rate": 2.8242211290700402e-05, "loss": 0.0198, "step": 10872 }, { "epoch": 7.640899508081518, "grad_norm": 0.23837202787399292, "learning_rate": 2.824174279690794e-05, "loss": 0.0144, "step": 10873 }, { "epoch": 7.641602248770203, "grad_norm": 0.2776528596878052, "learning_rate": 2.8241274303115483e-05, "loss": 0.047, "step": 10874 }, { "epoch": 7.642304989458889, "grad_norm": 0.1418510228395462, "learning_rate": 2.8240805809323026e-05, "loss": 0.0215, "step": 10875 }, { "epoch": 7.643007730147575, "grad_norm": 0.27831801772117615, "learning_rate": 2.824033731553057e-05, "loss": 0.041, "step": 10876 }, { "epoch": 7.643710470836261, "grad_norm": 0.18511657416820526, "learning_rate": 2.823986882173811e-05, "loss": 0.0441, "step": 10877 }, { "epoch": 7.644413211524947, "grad_norm": 0.14439216256141663, "learning_rate": 2.8239400327945654e-05, "loss": 0.0213, "step": 10878 }, { "epoch": 7.645115952213633, "grad_norm": 0.1998128443956375, "learning_rate": 2.8238931834153198e-05, "loss": 0.0425, "step": 10879 }, { "epoch": 7.645818692902319, "grad_norm": 0.222604438662529, "learning_rate": 2.823846334036074e-05, "loss": 0.0731, "step": 10880 }, { "epoch": 7.646521433591005, "grad_norm": 0.30399179458618164, "learning_rate": 2.8237994846568285e-05, "loss": 0.0528, "step": 10881 }, { "epoch": 7.647224174279691, "grad_norm": 0.5789604783058167, "learning_rate": 2.8237526352775826e-05, "loss": 0.1028, "step": 10882 }, { "epoch": 7.647926914968377, "grad_norm": 0.39337044954299927, "learning_rate": 2.823705785898337e-05, "loss": 0.1363, "step": 10883 }, { "epoch": 7.6486296556570625, "grad_norm": 0.5151695609092712, "learning_rate": 2.8236589365190913e-05, "loss": 0.1568, "step": 10884 }, { "epoch": 7.6493323963457485, "grad_norm": 0.9614532589912415, "learning_rate": 2.8236120871398457e-05, "loss": 0.234, "step": 10885 }, { "epoch": 7.650035137034434, "grad_norm": 1.9747283458709717, "learning_rate": 2.8235652377605997e-05, "loss": 0.295, "step": 10886 }, { "epoch": 7.65073787772312, "grad_norm": 0.21474087238311768, "learning_rate": 2.823518388381354e-05, "loss": 0.0658, "step": 10887 }, { "epoch": 7.651440618411806, "grad_norm": 0.11033285409212112, "learning_rate": 2.8234715390021085e-05, "loss": 0.0345, "step": 10888 }, { "epoch": 7.652143359100492, "grad_norm": 0.12691904604434967, "learning_rate": 2.8234246896228628e-05, "loss": 0.0199, "step": 10889 }, { "epoch": 7.652846099789178, "grad_norm": 0.4668242335319519, "learning_rate": 2.823377840243617e-05, "loss": 0.0265, "step": 10890 }, { "epoch": 7.653548840477864, "grad_norm": 0.12631641328334808, "learning_rate": 2.823330990864371e-05, "loss": 0.0203, "step": 10891 }, { "epoch": 7.65425158116655, "grad_norm": 0.1493341475725174, "learning_rate": 2.8232841414851253e-05, "loss": 0.0187, "step": 10892 }, { "epoch": 7.654954321855236, "grad_norm": 0.08123579621315002, "learning_rate": 2.8232372921058796e-05, "loss": 0.0121, "step": 10893 }, { "epoch": 7.655657062543922, "grad_norm": 0.246215358376503, "learning_rate": 2.823190442726634e-05, "loss": 0.0239, "step": 10894 }, { "epoch": 7.656359803232607, "grad_norm": 0.23538537323474884, "learning_rate": 2.823143593347388e-05, "loss": 0.021, "step": 10895 }, { "epoch": 7.657062543921293, "grad_norm": 0.11043446511030197, "learning_rate": 2.8230967439681424e-05, "loss": 0.0117, "step": 10896 }, { "epoch": 7.657765284609979, "grad_norm": 0.21798793971538544, "learning_rate": 2.8230498945888968e-05, "loss": 0.0232, "step": 10897 }, { "epoch": 7.6584680252986645, "grad_norm": 0.10645458847284317, "learning_rate": 2.823003045209651e-05, "loss": 0.018, "step": 10898 }, { "epoch": 7.6591707659873505, "grad_norm": 0.16828419268131256, "learning_rate": 2.8229561958304052e-05, "loss": 0.0235, "step": 10899 }, { "epoch": 7.659873506676036, "grad_norm": 0.147715762257576, "learning_rate": 2.8229093464511596e-05, "loss": 0.0211, "step": 10900 }, { "epoch": 7.660576247364722, "grad_norm": 0.7535310387611389, "learning_rate": 2.822862497071914e-05, "loss": 0.0323, "step": 10901 }, { "epoch": 7.661278988053408, "grad_norm": 0.17083153128623962, "learning_rate": 2.8228156476926683e-05, "loss": 0.0313, "step": 10902 }, { "epoch": 7.661981728742094, "grad_norm": 0.10888982564210892, "learning_rate": 2.8227687983134223e-05, "loss": 0.0245, "step": 10903 }, { "epoch": 7.66268446943078, "grad_norm": 0.1753917932510376, "learning_rate": 2.8227219489341767e-05, "loss": 0.0424, "step": 10904 }, { "epoch": 7.663387210119466, "grad_norm": 0.23597823083400726, "learning_rate": 2.822675099554931e-05, "loss": 0.0584, "step": 10905 }, { "epoch": 7.664089950808152, "grad_norm": 0.29032623767852783, "learning_rate": 2.8226282501756855e-05, "loss": 0.031, "step": 10906 }, { "epoch": 7.664792691496838, "grad_norm": 0.26979920268058777, "learning_rate": 2.82258140079644e-05, "loss": 0.071, "step": 10907 }, { "epoch": 7.665495432185524, "grad_norm": 0.6942504644393921, "learning_rate": 2.8225345514171935e-05, "loss": 0.1123, "step": 10908 }, { "epoch": 7.66619817287421, "grad_norm": 0.5654909610748291, "learning_rate": 2.822487702037948e-05, "loss": 0.1641, "step": 10909 }, { "epoch": 7.666900913562896, "grad_norm": 0.595115065574646, "learning_rate": 2.8224408526587023e-05, "loss": 0.1869, "step": 10910 }, { "epoch": 7.6676036542515815, "grad_norm": 1.3836982250213623, "learning_rate": 2.8223940032794567e-05, "loss": 0.2097, "step": 10911 }, { "epoch": 7.668306394940267, "grad_norm": 0.29119351506233215, "learning_rate": 2.8223471539002107e-05, "loss": 0.064, "step": 10912 }, { "epoch": 7.6690091356289525, "grad_norm": 0.11129038780927658, "learning_rate": 2.822300304520965e-05, "loss": 0.0187, "step": 10913 }, { "epoch": 7.669711876317638, "grad_norm": 0.2405015081167221, "learning_rate": 2.8222534551417194e-05, "loss": 0.031, "step": 10914 }, { "epoch": 7.670414617006324, "grad_norm": 0.25254499912261963, "learning_rate": 2.8222066057624738e-05, "loss": 0.0202, "step": 10915 }, { "epoch": 7.67111735769501, "grad_norm": 0.1408524364233017, "learning_rate": 2.822159756383228e-05, "loss": 0.0249, "step": 10916 }, { "epoch": 7.671820098383696, "grad_norm": 0.0963728129863739, "learning_rate": 2.8221129070039822e-05, "loss": 0.0062, "step": 10917 }, { "epoch": 7.672522839072382, "grad_norm": 0.11173689365386963, "learning_rate": 2.8220660576247366e-05, "loss": 0.0182, "step": 10918 }, { "epoch": 7.673225579761068, "grad_norm": 0.14365895092487335, "learning_rate": 2.822019208245491e-05, "loss": 0.0182, "step": 10919 }, { "epoch": 7.673928320449754, "grad_norm": 0.1747010350227356, "learning_rate": 2.8219723588662453e-05, "loss": 0.0375, "step": 10920 }, { "epoch": 7.67463106113844, "grad_norm": 0.2586846351623535, "learning_rate": 2.8219255094869994e-05, "loss": 0.0178, "step": 10921 }, { "epoch": 7.675333801827126, "grad_norm": 0.3299073874950409, "learning_rate": 2.8218786601077537e-05, "loss": 0.019, "step": 10922 }, { "epoch": 7.676036542515812, "grad_norm": 0.11909591406583786, "learning_rate": 2.821831810728508e-05, "loss": 0.0236, "step": 10923 }, { "epoch": 7.676739283204498, "grad_norm": 0.22234977781772614, "learning_rate": 2.8217849613492625e-05, "loss": 0.032, "step": 10924 }, { "epoch": 7.6774420238931835, "grad_norm": 0.12222029268741608, "learning_rate": 2.8217381119700165e-05, "loss": 0.0108, "step": 10925 }, { "epoch": 7.6781447645818695, "grad_norm": 0.144514799118042, "learning_rate": 2.8216912625907705e-05, "loss": 0.0225, "step": 10926 }, { "epoch": 7.678847505270555, "grad_norm": 0.22889065742492676, "learning_rate": 2.821644413211525e-05, "loss": 0.0313, "step": 10927 }, { "epoch": 7.679550245959241, "grad_norm": 0.2685621678829193, "learning_rate": 2.8215975638322793e-05, "loss": 0.038, "step": 10928 }, { "epoch": 7.680252986647927, "grad_norm": 0.16075731813907623, "learning_rate": 2.8215507144530333e-05, "loss": 0.0314, "step": 10929 }, { "epoch": 7.680955727336613, "grad_norm": 0.23933418095111847, "learning_rate": 2.8215038650737877e-05, "loss": 0.0476, "step": 10930 }, { "epoch": 7.681658468025299, "grad_norm": 0.19151556491851807, "learning_rate": 2.821457015694542e-05, "loss": 0.0411, "step": 10931 }, { "epoch": 7.682361208713985, "grad_norm": 0.3734782636165619, "learning_rate": 2.8214101663152964e-05, "loss": 0.1242, "step": 10932 }, { "epoch": 7.683063949402671, "grad_norm": 0.7018726468086243, "learning_rate": 2.8213633169360508e-05, "loss": 0.1274, "step": 10933 }, { "epoch": 7.683766690091356, "grad_norm": 0.5150479674339294, "learning_rate": 2.821316467556805e-05, "loss": 0.1521, "step": 10934 }, { "epoch": 7.684469430780042, "grad_norm": 0.7013506293296814, "learning_rate": 2.8212696181775592e-05, "loss": 0.2327, "step": 10935 }, { "epoch": 7.685172171468728, "grad_norm": 1.130753755569458, "learning_rate": 2.8212227687983136e-05, "loss": 0.2424, "step": 10936 }, { "epoch": 7.685874912157414, "grad_norm": 0.3311194181442261, "learning_rate": 2.821175919419068e-05, "loss": 0.0632, "step": 10937 }, { "epoch": 7.6865776528461, "grad_norm": 0.5433008670806885, "learning_rate": 2.821129070039822e-05, "loss": 0.0425, "step": 10938 }, { "epoch": 7.6872803935347855, "grad_norm": 0.178428053855896, "learning_rate": 2.8210822206605764e-05, "loss": 0.0367, "step": 10939 }, { "epoch": 7.6879831342234715, "grad_norm": 0.24188028275966644, "learning_rate": 2.8210353712813307e-05, "loss": 0.0111, "step": 10940 }, { "epoch": 7.688685874912157, "grad_norm": 0.14017832279205322, "learning_rate": 2.820988521902085e-05, "loss": 0.0198, "step": 10941 }, { "epoch": 7.689388615600843, "grad_norm": 0.08263355493545532, "learning_rate": 2.820941672522839e-05, "loss": 0.0067, "step": 10942 }, { "epoch": 7.690091356289529, "grad_norm": 0.10215825587511063, "learning_rate": 2.8208948231435932e-05, "loss": 0.0182, "step": 10943 }, { "epoch": 7.690794096978215, "grad_norm": 0.15878897905349731, "learning_rate": 2.8208479737643476e-05, "loss": 0.0346, "step": 10944 }, { "epoch": 7.691496837666901, "grad_norm": 0.14612935483455658, "learning_rate": 2.820801124385102e-05, "loss": 0.0243, "step": 10945 }, { "epoch": 7.692199578355587, "grad_norm": 0.24286554753780365, "learning_rate": 2.8207542750058563e-05, "loss": 0.0143, "step": 10946 }, { "epoch": 7.692902319044273, "grad_norm": 0.27689996361732483, "learning_rate": 2.8207074256266103e-05, "loss": 0.0532, "step": 10947 }, { "epoch": 7.693605059732959, "grad_norm": 0.30105915665626526, "learning_rate": 2.8206605762473647e-05, "loss": 0.0234, "step": 10948 }, { "epoch": 7.694307800421645, "grad_norm": 0.09908697009086609, "learning_rate": 2.820613726868119e-05, "loss": 0.0127, "step": 10949 }, { "epoch": 7.695010541110331, "grad_norm": 0.13979481160640717, "learning_rate": 2.8205668774888735e-05, "loss": 0.0286, "step": 10950 }, { "epoch": 7.695713281799016, "grad_norm": 0.1958961933851242, "learning_rate": 2.8205200281096275e-05, "loss": 0.0414, "step": 10951 }, { "epoch": 7.696416022487702, "grad_norm": 0.26773902773857117, "learning_rate": 2.820473178730382e-05, "loss": 0.0269, "step": 10952 }, { "epoch": 7.6971187631763875, "grad_norm": 0.17520223557949066, "learning_rate": 2.8204263293511362e-05, "loss": 0.0408, "step": 10953 }, { "epoch": 7.6978215038650735, "grad_norm": 0.17437142133712769, "learning_rate": 2.8203794799718906e-05, "loss": 0.0335, "step": 10954 }, { "epoch": 7.698524244553759, "grad_norm": 0.18692836165428162, "learning_rate": 2.8203326305926446e-05, "loss": 0.0384, "step": 10955 }, { "epoch": 7.699226985242445, "grad_norm": 0.2673952281475067, "learning_rate": 2.820285781213399e-05, "loss": 0.0655, "step": 10956 }, { "epoch": 7.699929725931131, "grad_norm": 0.3742476999759674, "learning_rate": 2.8202389318341534e-05, "loss": 0.0683, "step": 10957 }, { "epoch": 7.700632466619817, "grad_norm": 0.449553906917572, "learning_rate": 2.8201920824549078e-05, "loss": 0.1217, "step": 10958 }, { "epoch": 7.701335207308503, "grad_norm": 0.4753788411617279, "learning_rate": 2.820145233075662e-05, "loss": 0.1363, "step": 10959 }, { "epoch": 7.702037947997189, "grad_norm": 0.7921175360679626, "learning_rate": 2.820098383696416e-05, "loss": 0.2125, "step": 10960 }, { "epoch": 7.702740688685875, "grad_norm": 1.046561360359192, "learning_rate": 2.8200515343171702e-05, "loss": 0.2181, "step": 10961 }, { "epoch": 7.703443429374561, "grad_norm": 0.2756248414516449, "learning_rate": 2.8200046849379246e-05, "loss": 0.0945, "step": 10962 }, { "epoch": 7.704146170063247, "grad_norm": 0.13358138501644135, "learning_rate": 2.819957835558679e-05, "loss": 0.0245, "step": 10963 }, { "epoch": 7.704848910751933, "grad_norm": 0.17611902952194214, "learning_rate": 2.819910986179433e-05, "loss": 0.0474, "step": 10964 }, { "epoch": 7.705551651440619, "grad_norm": 0.14677315950393677, "learning_rate": 2.8198641368001873e-05, "loss": 0.0235, "step": 10965 }, { "epoch": 7.7062543921293045, "grad_norm": 0.18189893662929535, "learning_rate": 2.8198172874209417e-05, "loss": 0.0312, "step": 10966 }, { "epoch": 7.70695713281799, "grad_norm": 0.14013628661632538, "learning_rate": 2.819770438041696e-05, "loss": 0.021, "step": 10967 }, { "epoch": 7.707659873506676, "grad_norm": 0.25426843762397766, "learning_rate": 2.81972358866245e-05, "loss": 0.0132, "step": 10968 }, { "epoch": 7.708362614195362, "grad_norm": 0.1582912653684616, "learning_rate": 2.8196767392832045e-05, "loss": 0.0193, "step": 10969 }, { "epoch": 7.709065354884048, "grad_norm": 0.1482819765806198, "learning_rate": 2.819629889903959e-05, "loss": 0.0244, "step": 10970 }, { "epoch": 7.709768095572734, "grad_norm": 0.14218096435070038, "learning_rate": 2.8195830405247132e-05, "loss": 0.029, "step": 10971 }, { "epoch": 7.710470836261419, "grad_norm": 0.12324971705675125, "learning_rate": 2.8195361911454676e-05, "loss": 0.0282, "step": 10972 }, { "epoch": 7.711173576950105, "grad_norm": 0.11470673978328705, "learning_rate": 2.8194893417662216e-05, "loss": 0.0151, "step": 10973 }, { "epoch": 7.711876317638791, "grad_norm": 0.18339422345161438, "learning_rate": 2.819442492386976e-05, "loss": 0.0253, "step": 10974 }, { "epoch": 7.712579058327477, "grad_norm": 0.0941457524895668, "learning_rate": 2.8193956430077304e-05, "loss": 0.0107, "step": 10975 }, { "epoch": 7.713281799016163, "grad_norm": 0.2814662456512451, "learning_rate": 2.8193487936284848e-05, "loss": 0.0469, "step": 10976 }, { "epoch": 7.713984539704849, "grad_norm": 0.20356538891792297, "learning_rate": 2.8193019442492388e-05, "loss": 0.0382, "step": 10977 }, { "epoch": 7.714687280393535, "grad_norm": 0.22199666500091553, "learning_rate": 2.819255094869993e-05, "loss": 0.0215, "step": 10978 }, { "epoch": 7.715390021082221, "grad_norm": 0.1590181440114975, "learning_rate": 2.8192082454907472e-05, "loss": 0.0394, "step": 10979 }, { "epoch": 7.7160927617709065, "grad_norm": 1.1168755292892456, "learning_rate": 2.8191613961115016e-05, "loss": 0.0718, "step": 10980 }, { "epoch": 7.7167955024595924, "grad_norm": 0.233510822057724, "learning_rate": 2.8191145467322556e-05, "loss": 0.0608, "step": 10981 }, { "epoch": 7.717498243148278, "grad_norm": 0.320778489112854, "learning_rate": 2.81906769735301e-05, "loss": 0.0749, "step": 10982 }, { "epoch": 7.718200983836964, "grad_norm": 0.5733261704444885, "learning_rate": 2.8190208479737644e-05, "loss": 0.1059, "step": 10983 }, { "epoch": 7.71890372452565, "grad_norm": 0.578029453754425, "learning_rate": 2.8189739985945187e-05, "loss": 0.1395, "step": 10984 }, { "epoch": 7.719606465214336, "grad_norm": 0.8958289623260498, "learning_rate": 2.818927149215273e-05, "loss": 0.1821, "step": 10985 }, { "epoch": 7.720309205903022, "grad_norm": 0.9925659894943237, "learning_rate": 2.818880299836027e-05, "loss": 0.2208, "step": 10986 }, { "epoch": 7.721011946591708, "grad_norm": 0.2544754445552826, "learning_rate": 2.8188334504567815e-05, "loss": 0.0619, "step": 10987 }, { "epoch": 7.721714687280394, "grad_norm": 0.12964865565299988, "learning_rate": 2.818786601077536e-05, "loss": 0.0288, "step": 10988 }, { "epoch": 7.722417427969079, "grad_norm": 0.4961196780204773, "learning_rate": 2.8187397516982903e-05, "loss": 0.026, "step": 10989 }, { "epoch": 7.723120168657765, "grad_norm": 0.13030312955379486, "learning_rate": 2.8186929023190443e-05, "loss": 0.0149, "step": 10990 }, { "epoch": 7.723822909346451, "grad_norm": 0.11416751891374588, "learning_rate": 2.8186460529397987e-05, "loss": 0.0188, "step": 10991 }, { "epoch": 7.724525650035137, "grad_norm": 0.0973980501294136, "learning_rate": 2.818599203560553e-05, "loss": 0.0139, "step": 10992 }, { "epoch": 7.725228390723823, "grad_norm": 0.13592848181724548, "learning_rate": 2.8185523541813074e-05, "loss": 0.0169, "step": 10993 }, { "epoch": 7.7259311314125085, "grad_norm": 0.3021695911884308, "learning_rate": 2.8185055048020614e-05, "loss": 0.0294, "step": 10994 }, { "epoch": 7.7266338721011945, "grad_norm": 0.21050423383712769, "learning_rate": 2.8184586554228155e-05, "loss": 0.0122, "step": 10995 }, { "epoch": 7.72733661278988, "grad_norm": 0.1436862349510193, "learning_rate": 2.81841180604357e-05, "loss": 0.0279, "step": 10996 }, { "epoch": 7.728039353478566, "grad_norm": 1.105301856994629, "learning_rate": 2.8183649566643242e-05, "loss": 0.0327, "step": 10997 }, { "epoch": 7.728742094167252, "grad_norm": 0.22556327283382416, "learning_rate": 2.8183181072850786e-05, "loss": 0.0146, "step": 10998 }, { "epoch": 7.729444834855938, "grad_norm": 0.1794922798871994, "learning_rate": 2.8182712579058326e-05, "loss": 0.036, "step": 10999 }, { "epoch": 7.730147575544624, "grad_norm": 0.17303252220153809, "learning_rate": 2.818224408526587e-05, "loss": 0.0158, "step": 11000 }, { "epoch": 7.730147575544624, "eval_cer": 0.1947235769696573, "eval_loss": 0.2805321514606476, "eval_runtime": 18.4391, "eval_samples_per_second": 246.108, "eval_steps_per_second": 0.813, "eval_wer": 0.3498226306527751, "step": 11000 }, { "epoch": 7.73085031623331, "grad_norm": 0.2378627359867096, "learning_rate": 2.8181775591473414e-05, "loss": 0.035, "step": 11001 }, { "epoch": 7.731553056921996, "grad_norm": 0.12820333242416382, "learning_rate": 2.8181307097680957e-05, "loss": 0.0234, "step": 11002 }, { "epoch": 7.732255797610682, "grad_norm": 0.20425142347812653, "learning_rate": 2.8180838603888498e-05, "loss": 0.0232, "step": 11003 }, { "epoch": 7.732958538299368, "grad_norm": 0.17573295533657074, "learning_rate": 2.818037011009604e-05, "loss": 0.0355, "step": 11004 }, { "epoch": 7.733661278988054, "grad_norm": 0.20722241699695587, "learning_rate": 2.8179901616303585e-05, "loss": 0.0445, "step": 11005 }, { "epoch": 7.73436401967674, "grad_norm": 0.4682607650756836, "learning_rate": 2.817943312251113e-05, "loss": 0.0585, "step": 11006 }, { "epoch": 7.7350667603654255, "grad_norm": 0.34039306640625, "learning_rate": 2.817896462871867e-05, "loss": 0.0743, "step": 11007 }, { "epoch": 7.735769501054111, "grad_norm": 0.6320792436599731, "learning_rate": 2.8178496134926213e-05, "loss": 0.1226, "step": 11008 }, { "epoch": 7.736472241742797, "grad_norm": 0.8542987704277039, "learning_rate": 2.8178027641133757e-05, "loss": 0.155, "step": 11009 }, { "epoch": 7.737174982431483, "grad_norm": 1.110617756843567, "learning_rate": 2.81775591473413e-05, "loss": 0.2277, "step": 11010 }, { "epoch": 7.737877723120168, "grad_norm": 0.8940987586975098, "learning_rate": 2.8177090653548844e-05, "loss": 0.2433, "step": 11011 }, { "epoch": 7.738580463808854, "grad_norm": 0.25305747985839844, "learning_rate": 2.8176622159756384e-05, "loss": 0.0627, "step": 11012 }, { "epoch": 7.73928320449754, "grad_norm": 0.129794642329216, "learning_rate": 2.8176153665963925e-05, "loss": 0.0312, "step": 11013 }, { "epoch": 7.739985945186226, "grad_norm": 0.25415390729904175, "learning_rate": 2.817568517217147e-05, "loss": 0.054, "step": 11014 }, { "epoch": 7.740688685874912, "grad_norm": 0.10966604202985764, "learning_rate": 2.8175216678379012e-05, "loss": 0.0147, "step": 11015 }, { "epoch": 7.741391426563598, "grad_norm": 0.26424935460090637, "learning_rate": 2.8174748184586553e-05, "loss": 0.0187, "step": 11016 }, { "epoch": 7.742094167252284, "grad_norm": 0.11034445464611053, "learning_rate": 2.8174279690794096e-05, "loss": 0.0128, "step": 11017 }, { "epoch": 7.74279690794097, "grad_norm": 0.12144314497709274, "learning_rate": 2.817381119700164e-05, "loss": 0.02, "step": 11018 }, { "epoch": 7.743499648629656, "grad_norm": 0.1504473239183426, "learning_rate": 2.8173342703209184e-05, "loss": 0.0173, "step": 11019 }, { "epoch": 7.744202389318342, "grad_norm": 0.34531521797180176, "learning_rate": 2.8172874209416724e-05, "loss": 0.0328, "step": 11020 }, { "epoch": 7.7449051300070275, "grad_norm": 0.11200707405805588, "learning_rate": 2.8172405715624268e-05, "loss": 0.0119, "step": 11021 }, { "epoch": 7.745607870695713, "grad_norm": 0.1801013946533203, "learning_rate": 2.817193722183181e-05, "loss": 0.0258, "step": 11022 }, { "epoch": 7.746310611384399, "grad_norm": 0.1427612602710724, "learning_rate": 2.8171468728039355e-05, "loss": 0.0165, "step": 11023 }, { "epoch": 7.747013352073085, "grad_norm": 0.16151711344718933, "learning_rate": 2.81710002342469e-05, "loss": 0.0266, "step": 11024 }, { "epoch": 7.747716092761771, "grad_norm": 0.2036992758512497, "learning_rate": 2.817053174045444e-05, "loss": 0.0243, "step": 11025 }, { "epoch": 7.748418833450457, "grad_norm": 0.20083163678646088, "learning_rate": 2.8170063246661983e-05, "loss": 0.0481, "step": 11026 }, { "epoch": 7.749121574139143, "grad_norm": 0.19897541403770447, "learning_rate": 2.8169594752869527e-05, "loss": 0.0261, "step": 11027 }, { "epoch": 7.749824314827828, "grad_norm": 0.2333768755197525, "learning_rate": 2.816912625907707e-05, "loss": 0.0187, "step": 11028 }, { "epoch": 7.750527055516514, "grad_norm": 0.22121615707874298, "learning_rate": 2.816865776528461e-05, "loss": 0.0404, "step": 11029 }, { "epoch": 7.7512297962052, "grad_norm": 0.7351882457733154, "learning_rate": 2.816818927149215e-05, "loss": 0.0459, "step": 11030 }, { "epoch": 7.751932536893886, "grad_norm": 0.1727423369884491, "learning_rate": 2.8167720777699695e-05, "loss": 0.041, "step": 11031 }, { "epoch": 7.752635277582572, "grad_norm": 0.46043992042541504, "learning_rate": 2.816725228390724e-05, "loss": 0.0956, "step": 11032 }, { "epoch": 7.753338018271258, "grad_norm": 0.755976140499115, "learning_rate": 2.816678379011478e-05, "loss": 0.1176, "step": 11033 }, { "epoch": 7.754040758959944, "grad_norm": 1.0431846380233765, "learning_rate": 2.8166315296322323e-05, "loss": 0.1397, "step": 11034 }, { "epoch": 7.7547434996486295, "grad_norm": 1.3424632549285889, "learning_rate": 2.8165846802529866e-05, "loss": 0.1881, "step": 11035 }, { "epoch": 7.7554462403373154, "grad_norm": 0.9748526215553284, "learning_rate": 2.816537830873741e-05, "loss": 0.3017, "step": 11036 }, { "epoch": 7.756148981026001, "grad_norm": 0.361689954996109, "learning_rate": 2.8164909814944954e-05, "loss": 0.0786, "step": 11037 }, { "epoch": 7.756851721714687, "grad_norm": 0.2158910483121872, "learning_rate": 2.8164441321152494e-05, "loss": 0.0228, "step": 11038 }, { "epoch": 7.757554462403373, "grad_norm": 0.11097247898578644, "learning_rate": 2.8163972827360038e-05, "loss": 0.0193, "step": 11039 }, { "epoch": 7.758257203092059, "grad_norm": 0.3271417021751404, "learning_rate": 2.816350433356758e-05, "loss": 0.025, "step": 11040 }, { "epoch": 7.758959943780745, "grad_norm": 0.10806777328252792, "learning_rate": 2.8163035839775125e-05, "loss": 0.0166, "step": 11041 }, { "epoch": 7.759662684469431, "grad_norm": 0.18760229647159576, "learning_rate": 2.8162567345982666e-05, "loss": 0.0171, "step": 11042 }, { "epoch": 7.760365425158117, "grad_norm": 0.15312422811985016, "learning_rate": 2.816209885219021e-05, "loss": 0.0136, "step": 11043 }, { "epoch": 7.761068165846803, "grad_norm": 0.2925340235233307, "learning_rate": 2.8161630358397753e-05, "loss": 0.0412, "step": 11044 }, { "epoch": 7.761770906535489, "grad_norm": 0.17903712391853333, "learning_rate": 2.8161161864605297e-05, "loss": 0.0277, "step": 11045 }, { "epoch": 7.762473647224175, "grad_norm": 0.13500872254371643, "learning_rate": 2.8160693370812837e-05, "loss": 0.017, "step": 11046 }, { "epoch": 7.763176387912861, "grad_norm": 0.3029116094112396, "learning_rate": 2.816022487702038e-05, "loss": 0.0349, "step": 11047 }, { "epoch": 7.7638791286015465, "grad_norm": 0.1881933957338333, "learning_rate": 2.815975638322792e-05, "loss": 0.0162, "step": 11048 }, { "epoch": 7.7645818692902315, "grad_norm": 0.7244691252708435, "learning_rate": 2.8159287889435465e-05, "loss": 0.0252, "step": 11049 }, { "epoch": 7.7652846099789175, "grad_norm": 0.2582504153251648, "learning_rate": 2.815881939564301e-05, "loss": 0.0271, "step": 11050 }, { "epoch": 7.765987350667603, "grad_norm": 0.5181587934494019, "learning_rate": 2.815835090185055e-05, "loss": 0.0271, "step": 11051 }, { "epoch": 7.766690091356289, "grad_norm": 0.20262792706489563, "learning_rate": 2.8157882408058093e-05, "loss": 0.0231, "step": 11052 }, { "epoch": 7.767392832044975, "grad_norm": 0.1619345098733902, "learning_rate": 2.8157413914265637e-05, "loss": 0.0261, "step": 11053 }, { "epoch": 7.768095572733661, "grad_norm": 0.25026264786720276, "learning_rate": 2.815694542047318e-05, "loss": 0.0405, "step": 11054 }, { "epoch": 7.768798313422347, "grad_norm": 0.7878456115722656, "learning_rate": 2.815647692668072e-05, "loss": 0.0372, "step": 11055 }, { "epoch": 7.769501054111033, "grad_norm": 0.5200526118278503, "learning_rate": 2.8156008432888264e-05, "loss": 0.0529, "step": 11056 }, { "epoch": 7.770203794799719, "grad_norm": 0.4173181354999542, "learning_rate": 2.8155539939095808e-05, "loss": 0.0746, "step": 11057 }, { "epoch": 7.770906535488405, "grad_norm": 0.47398027777671814, "learning_rate": 2.8155071445303352e-05, "loss": 0.1318, "step": 11058 }, { "epoch": 7.771609276177091, "grad_norm": 0.8783945441246033, "learning_rate": 2.8154602951510892e-05, "loss": 0.1776, "step": 11059 }, { "epoch": 7.772312016865777, "grad_norm": 1.0346769094467163, "learning_rate": 2.8154134457718436e-05, "loss": 0.2015, "step": 11060 }, { "epoch": 7.773014757554463, "grad_norm": 2.3148415088653564, "learning_rate": 2.815366596392598e-05, "loss": 0.2317, "step": 11061 }, { "epoch": 7.7737174982431485, "grad_norm": 0.2123379111289978, "learning_rate": 2.8153197470133523e-05, "loss": 0.06, "step": 11062 }, { "epoch": 7.774420238931834, "grad_norm": 0.115195631980896, "learning_rate": 2.8152728976341067e-05, "loss": 0.0259, "step": 11063 }, { "epoch": 7.77512297962052, "grad_norm": 0.15253835916519165, "learning_rate": 2.8152260482548607e-05, "loss": 0.0223, "step": 11064 }, { "epoch": 7.775825720309206, "grad_norm": 0.18121366202831268, "learning_rate": 2.8151791988756148e-05, "loss": 0.0296, "step": 11065 }, { "epoch": 7.776528460997891, "grad_norm": 0.12939715385437012, "learning_rate": 2.815132349496369e-05, "loss": 0.0193, "step": 11066 }, { "epoch": 7.777231201686577, "grad_norm": 0.15599818527698517, "learning_rate": 2.8150855001171235e-05, "loss": 0.0246, "step": 11067 }, { "epoch": 7.777933942375263, "grad_norm": 0.19786663353443146, "learning_rate": 2.8150386507378776e-05, "loss": 0.0241, "step": 11068 }, { "epoch": 7.778636683063949, "grad_norm": 0.21119241416454315, "learning_rate": 2.814991801358632e-05, "loss": 0.021, "step": 11069 }, { "epoch": 7.779339423752635, "grad_norm": 0.14698532223701477, "learning_rate": 2.8149449519793863e-05, "loss": 0.0235, "step": 11070 }, { "epoch": 7.780042164441321, "grad_norm": 0.16839195787906647, "learning_rate": 2.8148981026001407e-05, "loss": 0.0133, "step": 11071 }, { "epoch": 7.780744905130007, "grad_norm": 0.16638921201229095, "learning_rate": 2.814851253220895e-05, "loss": 0.0214, "step": 11072 }, { "epoch": 7.781447645818693, "grad_norm": 0.1668468415737152, "learning_rate": 2.814804403841649e-05, "loss": 0.0136, "step": 11073 }, { "epoch": 7.782150386507379, "grad_norm": 0.19671380519866943, "learning_rate": 2.8147575544624034e-05, "loss": 0.0377, "step": 11074 }, { "epoch": 7.782853127196065, "grad_norm": 0.1487283557653427, "learning_rate": 2.8147107050831578e-05, "loss": 0.0191, "step": 11075 }, { "epoch": 7.7835558678847505, "grad_norm": 0.21953894197940826, "learning_rate": 2.8146638557039122e-05, "loss": 0.052, "step": 11076 }, { "epoch": 7.784258608573436, "grad_norm": 0.24409306049346924, "learning_rate": 2.8146170063246662e-05, "loss": 0.0409, "step": 11077 }, { "epoch": 7.784961349262122, "grad_norm": 0.15446004271507263, "learning_rate": 2.8145701569454206e-05, "loss": 0.0151, "step": 11078 }, { "epoch": 7.785664089950808, "grad_norm": 0.573665201663971, "learning_rate": 2.814523307566175e-05, "loss": 0.044, "step": 11079 }, { "epoch": 7.786366830639494, "grad_norm": 0.17649969458580017, "learning_rate": 2.8144764581869293e-05, "loss": 0.04, "step": 11080 }, { "epoch": 7.78706957132818, "grad_norm": 0.49974772334098816, "learning_rate": 2.8144296088076834e-05, "loss": 0.0576, "step": 11081 }, { "epoch": 7.787772312016866, "grad_norm": 0.5097247362136841, "learning_rate": 2.8143827594284374e-05, "loss": 0.0601, "step": 11082 }, { "epoch": 7.788475052705552, "grad_norm": 0.30081912875175476, "learning_rate": 2.8143359100491918e-05, "loss": 0.1002, "step": 11083 }, { "epoch": 7.789177793394238, "grad_norm": 1.5230520963668823, "learning_rate": 2.814289060669946e-05, "loss": 0.1433, "step": 11084 }, { "epoch": 7.789880534082924, "grad_norm": 0.7182759642601013, "learning_rate": 2.8142422112907005e-05, "loss": 0.1872, "step": 11085 }, { "epoch": 7.79058327477161, "grad_norm": 1.070192575454712, "learning_rate": 2.8141953619114546e-05, "loss": 0.2356, "step": 11086 }, { "epoch": 7.791286015460296, "grad_norm": 0.31565749645233154, "learning_rate": 2.814148512532209e-05, "loss": 0.0648, "step": 11087 }, { "epoch": 7.791988756148981, "grad_norm": 0.13858026266098022, "learning_rate": 2.8141016631529633e-05, "loss": 0.0253, "step": 11088 }, { "epoch": 7.792691496837667, "grad_norm": 0.12411579489707947, "learning_rate": 2.8140548137737177e-05, "loss": 0.021, "step": 11089 }, { "epoch": 7.7933942375263525, "grad_norm": 0.1144464984536171, "learning_rate": 2.8140079643944717e-05, "loss": 0.019, "step": 11090 }, { "epoch": 7.794096978215038, "grad_norm": 1.055800199508667, "learning_rate": 2.813961115015226e-05, "loss": 0.0181, "step": 11091 }, { "epoch": 7.794799718903724, "grad_norm": 0.2950217127799988, "learning_rate": 2.8139142656359805e-05, "loss": 0.0182, "step": 11092 }, { "epoch": 7.79550245959241, "grad_norm": 0.16226427257061005, "learning_rate": 2.8138674162567348e-05, "loss": 0.0264, "step": 11093 }, { "epoch": 7.796205200281096, "grad_norm": 0.15012140572071075, "learning_rate": 2.813820566877489e-05, "loss": 0.0188, "step": 11094 }, { "epoch": 7.796907940969782, "grad_norm": 0.2625048756599426, "learning_rate": 2.8137737174982432e-05, "loss": 0.0387, "step": 11095 }, { "epoch": 7.797610681658468, "grad_norm": 0.1489834487438202, "learning_rate": 2.8137268681189976e-05, "loss": 0.0143, "step": 11096 }, { "epoch": 7.798313422347154, "grad_norm": 0.21605926752090454, "learning_rate": 2.813680018739752e-05, "loss": 0.0225, "step": 11097 }, { "epoch": 7.79901616303584, "grad_norm": 0.16425000131130219, "learning_rate": 2.8136331693605064e-05, "loss": 0.017, "step": 11098 }, { "epoch": 7.799718903724526, "grad_norm": 0.3111160397529602, "learning_rate": 2.8135863199812604e-05, "loss": 0.0264, "step": 11099 }, { "epoch": 7.800421644413212, "grad_norm": 0.168186217546463, "learning_rate": 2.8135394706020144e-05, "loss": 0.0131, "step": 11100 }, { "epoch": 7.801124385101898, "grad_norm": 0.2867015600204468, "learning_rate": 2.8134926212227688e-05, "loss": 0.0249, "step": 11101 }, { "epoch": 7.801827125790584, "grad_norm": 0.3193269968032837, "learning_rate": 2.813445771843523e-05, "loss": 0.0476, "step": 11102 }, { "epoch": 7.8025298664792695, "grad_norm": 0.2056986689567566, "learning_rate": 2.8133989224642772e-05, "loss": 0.0455, "step": 11103 }, { "epoch": 7.8032326071679545, "grad_norm": 0.19372133910655975, "learning_rate": 2.8133520730850316e-05, "loss": 0.0271, "step": 11104 }, { "epoch": 7.8039353478566404, "grad_norm": 0.33316537737846375, "learning_rate": 2.813305223705786e-05, "loss": 0.0618, "step": 11105 }, { "epoch": 7.804638088545326, "grad_norm": 0.26099586486816406, "learning_rate": 2.8132583743265403e-05, "loss": 0.0507, "step": 11106 }, { "epoch": 7.805340829234012, "grad_norm": 0.9210512638092041, "learning_rate": 2.8132115249472944e-05, "loss": 0.1085, "step": 11107 }, { "epoch": 7.806043569922698, "grad_norm": 0.6826267838478088, "learning_rate": 2.8131646755680487e-05, "loss": 0.124, "step": 11108 }, { "epoch": 7.806746310611384, "grad_norm": 0.7953299880027771, "learning_rate": 2.813117826188803e-05, "loss": 0.1788, "step": 11109 }, { "epoch": 7.80744905130007, "grad_norm": 0.7883527874946594, "learning_rate": 2.8130709768095575e-05, "loss": 0.2123, "step": 11110 }, { "epoch": 7.808151791988756, "grad_norm": 1.216875433921814, "learning_rate": 2.813024127430312e-05, "loss": 0.2622, "step": 11111 }, { "epoch": 7.808854532677442, "grad_norm": 0.26107168197631836, "learning_rate": 2.812977278051066e-05, "loss": 0.0749, "step": 11112 }, { "epoch": 7.809557273366128, "grad_norm": 0.15419863164424896, "learning_rate": 2.8129304286718202e-05, "loss": 0.0345, "step": 11113 }, { "epoch": 7.810260014054814, "grad_norm": 0.15278059244155884, "learning_rate": 2.8128835792925746e-05, "loss": 0.0266, "step": 11114 }, { "epoch": 7.8109627547435, "grad_norm": 0.17548881471157074, "learning_rate": 2.812836729913329e-05, "loss": 0.029, "step": 11115 }, { "epoch": 7.811665495432186, "grad_norm": 0.12508073449134827, "learning_rate": 2.812789880534083e-05, "loss": 0.031, "step": 11116 }, { "epoch": 7.8123682361208715, "grad_norm": 0.12669378519058228, "learning_rate": 2.812743031154837e-05, "loss": 0.0199, "step": 11117 }, { "epoch": 7.813070976809557, "grad_norm": 0.17245237529277802, "learning_rate": 2.8126961817755914e-05, "loss": 0.0272, "step": 11118 }, { "epoch": 7.813773717498243, "grad_norm": 0.16697603464126587, "learning_rate": 2.8126493323963458e-05, "loss": 0.028, "step": 11119 }, { "epoch": 7.814476458186929, "grad_norm": 0.11419756710529327, "learning_rate": 2.8126024830171e-05, "loss": 0.0191, "step": 11120 }, { "epoch": 7.815179198875615, "grad_norm": 0.17499375343322754, "learning_rate": 2.8125556336378542e-05, "loss": 0.0156, "step": 11121 }, { "epoch": 7.815881939564301, "grad_norm": 0.1892242431640625, "learning_rate": 2.8125087842586086e-05, "loss": 0.0254, "step": 11122 }, { "epoch": 7.816584680252987, "grad_norm": 0.1082482561469078, "learning_rate": 2.812461934879363e-05, "loss": 0.0178, "step": 11123 }, { "epoch": 7.817287420941673, "grad_norm": 0.159809872508049, "learning_rate": 2.8124150855001173e-05, "loss": 0.0252, "step": 11124 }, { "epoch": 7.817990161630359, "grad_norm": 0.13954120874404907, "learning_rate": 2.8123682361208714e-05, "loss": 0.0229, "step": 11125 }, { "epoch": 7.818692902319044, "grad_norm": 0.15178684890270233, "learning_rate": 2.8123213867416257e-05, "loss": 0.0256, "step": 11126 }, { "epoch": 7.81939564300773, "grad_norm": 0.2645086348056793, "learning_rate": 2.81227453736238e-05, "loss": 0.0428, "step": 11127 }, { "epoch": 7.820098383696416, "grad_norm": 0.18267208337783813, "learning_rate": 2.8122276879831345e-05, "loss": 0.0225, "step": 11128 }, { "epoch": 7.820801124385102, "grad_norm": 0.17339389026165009, "learning_rate": 2.8121808386038885e-05, "loss": 0.0344, "step": 11129 }, { "epoch": 7.821503865073788, "grad_norm": 0.2194002866744995, "learning_rate": 2.812133989224643e-05, "loss": 0.045, "step": 11130 }, { "epoch": 7.8222066057624735, "grad_norm": 0.24636723101139069, "learning_rate": 2.8120871398453973e-05, "loss": 0.0739, "step": 11131 }, { "epoch": 7.822909346451159, "grad_norm": 0.28096821904182434, "learning_rate": 2.8120402904661516e-05, "loss": 0.0763, "step": 11132 }, { "epoch": 7.823612087139845, "grad_norm": 0.4163598120212555, "learning_rate": 2.8119934410869057e-05, "loss": 0.1032, "step": 11133 }, { "epoch": 7.824314827828531, "grad_norm": 0.4799940884113312, "learning_rate": 2.81194659170766e-05, "loss": 0.1583, "step": 11134 }, { "epoch": 7.825017568517217, "grad_norm": 0.809755802154541, "learning_rate": 2.811899742328414e-05, "loss": 0.2111, "step": 11135 }, { "epoch": 7.825720309205903, "grad_norm": 13.150553703308105, "learning_rate": 2.8118528929491684e-05, "loss": 0.2865, "step": 11136 }, { "epoch": 7.826423049894589, "grad_norm": 0.30448630452156067, "learning_rate": 2.8118060435699228e-05, "loss": 0.1247, "step": 11137 }, { "epoch": 7.827125790583275, "grad_norm": 0.48001593351364136, "learning_rate": 2.811759194190677e-05, "loss": 0.0259, "step": 11138 }, { "epoch": 7.827828531271961, "grad_norm": 0.1713745892047882, "learning_rate": 2.8117123448114312e-05, "loss": 0.0321, "step": 11139 }, { "epoch": 7.828531271960647, "grad_norm": 0.21482673287391663, "learning_rate": 2.8116654954321856e-05, "loss": 0.0203, "step": 11140 }, { "epoch": 7.829234012649333, "grad_norm": 0.103814035654068, "learning_rate": 2.81161864605294e-05, "loss": 0.0187, "step": 11141 }, { "epoch": 7.829936753338019, "grad_norm": 0.20968274772167206, "learning_rate": 2.811571796673694e-05, "loss": 0.0255, "step": 11142 }, { "epoch": 7.830639494026704, "grad_norm": 0.15339352190494537, "learning_rate": 2.8115249472944484e-05, "loss": 0.0276, "step": 11143 }, { "epoch": 7.83134223471539, "grad_norm": 0.19064383208751678, "learning_rate": 2.8114780979152027e-05, "loss": 0.0282, "step": 11144 }, { "epoch": 7.8320449754040755, "grad_norm": 0.1170353963971138, "learning_rate": 2.811431248535957e-05, "loss": 0.0266, "step": 11145 }, { "epoch": 7.832747716092761, "grad_norm": 0.13219568133354187, "learning_rate": 2.811384399156711e-05, "loss": 0.0266, "step": 11146 }, { "epoch": 7.833450456781447, "grad_norm": 0.11421836167573929, "learning_rate": 2.8113375497774655e-05, "loss": 0.0137, "step": 11147 }, { "epoch": 7.834153197470133, "grad_norm": 0.2555832266807556, "learning_rate": 2.81129070039822e-05, "loss": 0.0246, "step": 11148 }, { "epoch": 7.834855938158819, "grad_norm": 0.16287676990032196, "learning_rate": 2.8112438510189743e-05, "loss": 0.0268, "step": 11149 }, { "epoch": 7.835558678847505, "grad_norm": 0.27678048610687256, "learning_rate": 2.8111970016397286e-05, "loss": 0.0181, "step": 11150 }, { "epoch": 7.836261419536191, "grad_norm": 0.21496909856796265, "learning_rate": 2.8111501522604827e-05, "loss": 0.0252, "step": 11151 }, { "epoch": 7.836964160224877, "grad_norm": 0.14251011610031128, "learning_rate": 2.8111033028812367e-05, "loss": 0.0196, "step": 11152 }, { "epoch": 7.837666900913563, "grad_norm": 0.1771264225244522, "learning_rate": 2.811056453501991e-05, "loss": 0.0225, "step": 11153 }, { "epoch": 7.838369641602249, "grad_norm": 0.20682650804519653, "learning_rate": 2.8110096041227455e-05, "loss": 0.0294, "step": 11154 }, { "epoch": 7.839072382290935, "grad_norm": 0.5674413442611694, "learning_rate": 2.8109627547434995e-05, "loss": 0.0418, "step": 11155 }, { "epoch": 7.839775122979621, "grad_norm": 0.24849963188171387, "learning_rate": 2.810915905364254e-05, "loss": 0.0472, "step": 11156 }, { "epoch": 7.840477863668307, "grad_norm": 0.47609564661979675, "learning_rate": 2.8108690559850082e-05, "loss": 0.0833, "step": 11157 }, { "epoch": 7.8411806043569925, "grad_norm": 0.7159491777420044, "learning_rate": 2.8108222066057626e-05, "loss": 0.12, "step": 11158 }, { "epoch": 7.841883345045678, "grad_norm": 0.964439868927002, "learning_rate": 2.8107753572265166e-05, "loss": 0.1854, "step": 11159 }, { "epoch": 7.842586085734364, "grad_norm": 0.5935278534889221, "learning_rate": 2.810728507847271e-05, "loss": 0.1723, "step": 11160 }, { "epoch": 7.84328882642305, "grad_norm": 1.5215204954147339, "learning_rate": 2.8106816584680254e-05, "loss": 0.2067, "step": 11161 }, { "epoch": 7.843991567111736, "grad_norm": 0.3428206443786621, "learning_rate": 2.8106348090887798e-05, "loss": 0.0835, "step": 11162 }, { "epoch": 7.844694307800422, "grad_norm": 0.13142041862010956, "learning_rate": 2.810587959709534e-05, "loss": 0.0254, "step": 11163 }, { "epoch": 7.845397048489108, "grad_norm": 0.16859900951385498, "learning_rate": 2.810541110330288e-05, "loss": 0.0174, "step": 11164 }, { "epoch": 7.846099789177793, "grad_norm": 0.2786014974117279, "learning_rate": 2.8104942609510425e-05, "loss": 0.0187, "step": 11165 }, { "epoch": 7.846802529866479, "grad_norm": 0.32289785146713257, "learning_rate": 2.810447411571797e-05, "loss": 0.023, "step": 11166 }, { "epoch": 7.847505270555165, "grad_norm": 0.11921331286430359, "learning_rate": 2.8104005621925513e-05, "loss": 0.007, "step": 11167 }, { "epoch": 7.848208011243851, "grad_norm": 0.10863720625638962, "learning_rate": 2.8103537128133053e-05, "loss": 0.0271, "step": 11168 }, { "epoch": 7.848910751932537, "grad_norm": 0.17650839686393738, "learning_rate": 2.8103068634340593e-05, "loss": 0.0367, "step": 11169 }, { "epoch": 7.849613492621223, "grad_norm": 0.32839664816856384, "learning_rate": 2.8102600140548137e-05, "loss": 0.0212, "step": 11170 }, { "epoch": 7.850316233309909, "grad_norm": 0.12338722497224808, "learning_rate": 2.810213164675568e-05, "loss": 0.0118, "step": 11171 }, { "epoch": 7.8510189739985945, "grad_norm": 0.16254477202892303, "learning_rate": 2.810166315296322e-05, "loss": 0.0192, "step": 11172 }, { "epoch": 7.85172171468728, "grad_norm": 0.12170343846082687, "learning_rate": 2.8101194659170765e-05, "loss": 0.0229, "step": 11173 }, { "epoch": 7.852424455375966, "grad_norm": 0.11007452011108398, "learning_rate": 2.810072616537831e-05, "loss": 0.0226, "step": 11174 }, { "epoch": 7.853127196064652, "grad_norm": 0.08402994275093079, "learning_rate": 2.8100257671585852e-05, "loss": 0.012, "step": 11175 }, { "epoch": 7.853829936753338, "grad_norm": 0.27334412932395935, "learning_rate": 2.8099789177793396e-05, "loss": 0.0421, "step": 11176 }, { "epoch": 7.854532677442024, "grad_norm": 0.37284761667251587, "learning_rate": 2.8099320684000937e-05, "loss": 0.0419, "step": 11177 }, { "epoch": 7.85523541813071, "grad_norm": 0.1476069688796997, "learning_rate": 2.809885219020848e-05, "loss": 0.0171, "step": 11178 }, { "epoch": 7.855938158819396, "grad_norm": 0.23475854098796844, "learning_rate": 2.8098383696416024e-05, "loss": 0.0438, "step": 11179 }, { "epoch": 7.856640899508082, "grad_norm": 0.35333961248397827, "learning_rate": 2.8097915202623568e-05, "loss": 0.0597, "step": 11180 }, { "epoch": 7.857343640196767, "grad_norm": 0.27259117364883423, "learning_rate": 2.8097446708831108e-05, "loss": 0.0578, "step": 11181 }, { "epoch": 7.858046380885453, "grad_norm": 0.24345354735851288, "learning_rate": 2.8096978215038652e-05, "loss": 0.0659, "step": 11182 }, { "epoch": 7.858749121574139, "grad_norm": 0.6198297739028931, "learning_rate": 2.8096509721246195e-05, "loss": 0.1407, "step": 11183 }, { "epoch": 7.859451862262825, "grad_norm": 0.6092819571495056, "learning_rate": 2.809604122745374e-05, "loss": 0.1342, "step": 11184 }, { "epoch": 7.860154602951511, "grad_norm": 0.799241304397583, "learning_rate": 2.809557273366128e-05, "loss": 0.1865, "step": 11185 }, { "epoch": 7.8608573436401965, "grad_norm": 1.0158442258834839, "learning_rate": 2.8095104239868823e-05, "loss": 0.2366, "step": 11186 }, { "epoch": 7.861560084328882, "grad_norm": 0.2572207450866699, "learning_rate": 2.8094635746076364e-05, "loss": 0.0684, "step": 11187 }, { "epoch": 7.862262825017568, "grad_norm": 0.15229660272598267, "learning_rate": 2.8094167252283907e-05, "loss": 0.0293, "step": 11188 }, { "epoch": 7.862965565706254, "grad_norm": 0.16693070530891418, "learning_rate": 2.809369875849145e-05, "loss": 0.0245, "step": 11189 }, { "epoch": 7.86366830639494, "grad_norm": 0.2866445779800415, "learning_rate": 2.809323026469899e-05, "loss": 0.023, "step": 11190 }, { "epoch": 7.864371047083626, "grad_norm": 0.14383505284786224, "learning_rate": 2.8092761770906535e-05, "loss": 0.0253, "step": 11191 }, { "epoch": 7.865073787772312, "grad_norm": 0.13529394567012787, "learning_rate": 2.809229327711408e-05, "loss": 0.0162, "step": 11192 }, { "epoch": 7.865776528460998, "grad_norm": 0.13879328966140747, "learning_rate": 2.8091824783321623e-05, "loss": 0.021, "step": 11193 }, { "epoch": 7.866479269149684, "grad_norm": 0.20704320073127747, "learning_rate": 2.8091356289529163e-05, "loss": 0.039, "step": 11194 }, { "epoch": 7.86718200983837, "grad_norm": 0.18644501268863678, "learning_rate": 2.8090887795736707e-05, "loss": 0.0351, "step": 11195 }, { "epoch": 7.867884750527056, "grad_norm": 0.12344889342784882, "learning_rate": 2.809041930194425e-05, "loss": 0.0101, "step": 11196 }, { "epoch": 7.868587491215742, "grad_norm": 0.13281641900539398, "learning_rate": 2.8089950808151794e-05, "loss": 0.0206, "step": 11197 }, { "epoch": 7.869290231904428, "grad_norm": 0.19362692534923553, "learning_rate": 2.8089482314359334e-05, "loss": 0.0122, "step": 11198 }, { "epoch": 7.8699929725931135, "grad_norm": 0.14105583727359772, "learning_rate": 2.8089013820566878e-05, "loss": 0.0237, "step": 11199 }, { "epoch": 7.870695713281799, "grad_norm": 0.4733785390853882, "learning_rate": 2.8088545326774422e-05, "loss": 0.0302, "step": 11200 }, { "epoch": 7.871398453970485, "grad_norm": 0.4009442925453186, "learning_rate": 2.8088076832981966e-05, "loss": 0.0283, "step": 11201 }, { "epoch": 7.872101194659171, "grad_norm": 0.2518312335014343, "learning_rate": 2.808760833918951e-05, "loss": 0.0423, "step": 11202 }, { "epoch": 7.872803935347856, "grad_norm": 0.1346367448568344, "learning_rate": 2.808713984539705e-05, "loss": 0.0174, "step": 11203 }, { "epoch": 7.873506676036542, "grad_norm": 0.18389315903186798, "learning_rate": 2.808667135160459e-05, "loss": 0.0298, "step": 11204 }, { "epoch": 7.874209416725228, "grad_norm": 0.2700779139995575, "learning_rate": 2.8086202857812134e-05, "loss": 0.0343, "step": 11205 }, { "epoch": 7.874912157413914, "grad_norm": 0.40795353055000305, "learning_rate": 2.8085734364019677e-05, "loss": 0.0874, "step": 11206 }, { "epoch": 7.8756148981026, "grad_norm": 0.8015267252922058, "learning_rate": 2.8085265870227218e-05, "loss": 0.0868, "step": 11207 }, { "epoch": 7.876317638791286, "grad_norm": 0.8535261154174805, "learning_rate": 2.808479737643476e-05, "loss": 0.1293, "step": 11208 }, { "epoch": 7.877020379479972, "grad_norm": 0.84588223695755, "learning_rate": 2.8084328882642305e-05, "loss": 0.1273, "step": 11209 }, { "epoch": 7.877723120168658, "grad_norm": 1.4941242933273315, "learning_rate": 2.808386038884985e-05, "loss": 0.2179, "step": 11210 }, { "epoch": 7.878425860857344, "grad_norm": 1.6463358402252197, "learning_rate": 2.808339189505739e-05, "loss": 0.2464, "step": 11211 }, { "epoch": 7.87912860154603, "grad_norm": 0.2088988721370697, "learning_rate": 2.8082923401264933e-05, "loss": 0.0698, "step": 11212 }, { "epoch": 7.8798313422347155, "grad_norm": 0.1280967742204666, "learning_rate": 2.8082454907472477e-05, "loss": 0.0294, "step": 11213 }, { "epoch": 7.880534082923401, "grad_norm": 0.34316372871398926, "learning_rate": 2.808198641368002e-05, "loss": 0.0211, "step": 11214 }, { "epoch": 7.881236823612087, "grad_norm": 0.2772105932235718, "learning_rate": 2.8081517919887564e-05, "loss": 0.0225, "step": 11215 }, { "epoch": 7.881939564300773, "grad_norm": 0.14924870431423187, "learning_rate": 2.8081049426095105e-05, "loss": 0.0171, "step": 11216 }, { "epoch": 7.882642304989459, "grad_norm": 0.18605975806713104, "learning_rate": 2.8080580932302648e-05, "loss": 0.0145, "step": 11217 }, { "epoch": 7.883345045678145, "grad_norm": 0.2884882688522339, "learning_rate": 2.8080112438510192e-05, "loss": 0.0194, "step": 11218 }, { "epoch": 7.884047786366831, "grad_norm": 0.4185335040092468, "learning_rate": 2.8079643944717736e-05, "loss": 0.0233, "step": 11219 }, { "epoch": 7.884750527055516, "grad_norm": 0.1391218900680542, "learning_rate": 2.8079175450925276e-05, "loss": 0.0209, "step": 11220 }, { "epoch": 7.885453267744202, "grad_norm": 0.9051554203033447, "learning_rate": 2.807870695713282e-05, "loss": 0.0217, "step": 11221 }, { "epoch": 7.886156008432888, "grad_norm": 0.22439917922019958, "learning_rate": 2.807823846334036e-05, "loss": 0.0204, "step": 11222 }, { "epoch": 7.886858749121574, "grad_norm": 0.6535073518753052, "learning_rate": 2.8077769969547904e-05, "loss": 0.0147, "step": 11223 }, { "epoch": 7.88756148981026, "grad_norm": 0.23665228486061096, "learning_rate": 2.8077301475755444e-05, "loss": 0.0463, "step": 11224 }, { "epoch": 7.888264230498946, "grad_norm": 0.15132413804531097, "learning_rate": 2.8076832981962988e-05, "loss": 0.018, "step": 11225 }, { "epoch": 7.888966971187632, "grad_norm": 0.30059126019477844, "learning_rate": 2.807636448817053e-05, "loss": 0.0357, "step": 11226 }, { "epoch": 7.8896697118763175, "grad_norm": 0.20803450047969818, "learning_rate": 2.8075895994378075e-05, "loss": 0.0353, "step": 11227 }, { "epoch": 7.890372452565003, "grad_norm": 0.22190305590629578, "learning_rate": 2.807542750058562e-05, "loss": 0.0232, "step": 11228 }, { "epoch": 7.891075193253689, "grad_norm": 0.18280738592147827, "learning_rate": 2.807495900679316e-05, "loss": 0.0426, "step": 11229 }, { "epoch": 7.891777933942375, "grad_norm": 0.34983983635902405, "learning_rate": 2.8074490513000703e-05, "loss": 0.045, "step": 11230 }, { "epoch": 7.892480674631061, "grad_norm": 0.39468327164649963, "learning_rate": 2.8074022019208247e-05, "loss": 0.0682, "step": 11231 }, { "epoch": 7.893183415319747, "grad_norm": 0.33853933215141296, "learning_rate": 2.807355352541579e-05, "loss": 0.0682, "step": 11232 }, { "epoch": 7.893886156008433, "grad_norm": 0.39331430196762085, "learning_rate": 2.807308503162333e-05, "loss": 0.1059, "step": 11233 }, { "epoch": 7.894588896697119, "grad_norm": 0.9068189859390259, "learning_rate": 2.8072616537830875e-05, "loss": 0.1795, "step": 11234 }, { "epoch": 7.895291637385805, "grad_norm": 0.8709612488746643, "learning_rate": 2.807214804403842e-05, "loss": 0.2001, "step": 11235 }, { "epoch": 7.895994378074491, "grad_norm": 0.9237762093544006, "learning_rate": 2.8071679550245962e-05, "loss": 0.2209, "step": 11236 }, { "epoch": 7.896697118763177, "grad_norm": 0.3052862286567688, "learning_rate": 2.8071211056453502e-05, "loss": 0.0789, "step": 11237 }, { "epoch": 7.897399859451863, "grad_norm": 0.20376569032669067, "learning_rate": 2.8070742562661046e-05, "loss": 0.0258, "step": 11238 }, { "epoch": 7.8981026001405485, "grad_norm": 0.20451179146766663, "learning_rate": 2.8070274068868586e-05, "loss": 0.0207, "step": 11239 }, { "epoch": 7.8988053408292345, "grad_norm": 0.22885935008525848, "learning_rate": 2.806980557507613e-05, "loss": 0.0283, "step": 11240 }, { "epoch": 7.8995080815179195, "grad_norm": 0.25832194089889526, "learning_rate": 2.8069337081283674e-05, "loss": 0.0349, "step": 11241 }, { "epoch": 7.900210822206605, "grad_norm": 0.0897606909275055, "learning_rate": 2.8068868587491214e-05, "loss": 0.0195, "step": 11242 }, { "epoch": 7.900913562895291, "grad_norm": 0.11198287457227707, "learning_rate": 2.8068400093698758e-05, "loss": 0.0131, "step": 11243 }, { "epoch": 7.901616303583977, "grad_norm": 0.13375847041606903, "learning_rate": 2.8067931599906302e-05, "loss": 0.0299, "step": 11244 }, { "epoch": 7.902319044272663, "grad_norm": 0.14221058785915375, "learning_rate": 2.8067463106113845e-05, "loss": 0.0222, "step": 11245 }, { "epoch": 7.903021784961349, "grad_norm": 0.14318588376045227, "learning_rate": 2.8066994612321386e-05, "loss": 0.0182, "step": 11246 }, { "epoch": 7.903724525650035, "grad_norm": 0.1583537757396698, "learning_rate": 2.806652611852893e-05, "loss": 0.0232, "step": 11247 }, { "epoch": 7.904427266338721, "grad_norm": 0.11369886249303818, "learning_rate": 2.8066057624736473e-05, "loss": 0.0117, "step": 11248 }, { "epoch": 7.905130007027407, "grad_norm": 0.19458939135074615, "learning_rate": 2.8065589130944017e-05, "loss": 0.0351, "step": 11249 }, { "epoch": 7.905832747716093, "grad_norm": 0.2128392904996872, "learning_rate": 2.8065120637151557e-05, "loss": 0.0235, "step": 11250 }, { "epoch": 7.906535488404779, "grad_norm": 0.2702138125896454, "learning_rate": 2.80646521433591e-05, "loss": 0.026, "step": 11251 }, { "epoch": 7.907238229093465, "grad_norm": 0.19834017753601074, "learning_rate": 2.8064183649566645e-05, "loss": 0.026, "step": 11252 }, { "epoch": 7.9079409697821506, "grad_norm": 0.3519231379032135, "learning_rate": 2.806371515577419e-05, "loss": 0.0253, "step": 11253 }, { "epoch": 7.9086437104708365, "grad_norm": 0.29007941484451294, "learning_rate": 2.8063246661981732e-05, "loss": 0.0425, "step": 11254 }, { "epoch": 7.909346451159522, "grad_norm": 0.5724202394485474, "learning_rate": 2.8062778168189273e-05, "loss": 0.0524, "step": 11255 }, { "epoch": 7.910049191848208, "grad_norm": 0.2783881723880768, "learning_rate": 2.8062309674396816e-05, "loss": 0.0597, "step": 11256 }, { "epoch": 7.910751932536894, "grad_norm": 0.5256803035736084, "learning_rate": 2.8061841180604357e-05, "loss": 0.0831, "step": 11257 }, { "epoch": 7.911454673225579, "grad_norm": 0.6956315636634827, "learning_rate": 2.80613726868119e-05, "loss": 0.1133, "step": 11258 }, { "epoch": 7.912157413914265, "grad_norm": 0.6033449769020081, "learning_rate": 2.806090419301944e-05, "loss": 0.1728, "step": 11259 }, { "epoch": 7.912860154602951, "grad_norm": 0.9097537398338318, "learning_rate": 2.8060435699226984e-05, "loss": 0.1576, "step": 11260 }, { "epoch": 7.913562895291637, "grad_norm": 1.7414047718048096, "learning_rate": 2.8059967205434528e-05, "loss": 0.29, "step": 11261 }, { "epoch": 7.914265635980323, "grad_norm": 0.7295612096786499, "learning_rate": 2.8059498711642072e-05, "loss": 0.0828, "step": 11262 }, { "epoch": 7.914968376669009, "grad_norm": 0.2458271086215973, "learning_rate": 2.8059030217849612e-05, "loss": 0.0211, "step": 11263 }, { "epoch": 7.915671117357695, "grad_norm": 0.32681164145469666, "learning_rate": 2.8058561724057156e-05, "loss": 0.037, "step": 11264 }, { "epoch": 7.916373858046381, "grad_norm": 0.1636582762002945, "learning_rate": 2.80580932302647e-05, "loss": 0.0191, "step": 11265 }, { "epoch": 7.917076598735067, "grad_norm": 0.2645375430583954, "learning_rate": 2.8057624736472243e-05, "loss": 0.0209, "step": 11266 }, { "epoch": 7.917779339423753, "grad_norm": 0.1037532389163971, "learning_rate": 2.8057156242679787e-05, "loss": 0.0097, "step": 11267 }, { "epoch": 7.9184820801124385, "grad_norm": 0.27565449476242065, "learning_rate": 2.8056687748887327e-05, "loss": 0.0147, "step": 11268 }, { "epoch": 7.919184820801124, "grad_norm": 0.288687527179718, "learning_rate": 2.805621925509487e-05, "loss": 0.0274, "step": 11269 }, { "epoch": 7.91988756148981, "grad_norm": 0.15104202926158905, "learning_rate": 2.8055750761302415e-05, "loss": 0.0204, "step": 11270 }, { "epoch": 7.920590302178496, "grad_norm": 0.13153861463069916, "learning_rate": 2.805528226750996e-05, "loss": 0.0116, "step": 11271 }, { "epoch": 7.921293042867182, "grad_norm": 0.13043230772018433, "learning_rate": 2.80548137737175e-05, "loss": 0.0214, "step": 11272 }, { "epoch": 7.921995783555868, "grad_norm": 0.5703258514404297, "learning_rate": 2.8054345279925043e-05, "loss": 0.0184, "step": 11273 }, { "epoch": 7.922698524244554, "grad_norm": 0.12021931260824203, "learning_rate": 2.8053876786132583e-05, "loss": 0.0258, "step": 11274 }, { "epoch": 7.92340126493324, "grad_norm": 0.23374561965465546, "learning_rate": 2.8053408292340127e-05, "loss": 0.0158, "step": 11275 }, { "epoch": 7.924104005621926, "grad_norm": 0.2506572902202606, "learning_rate": 2.805293979854767e-05, "loss": 0.0445, "step": 11276 }, { "epoch": 7.924806746310612, "grad_norm": 0.2827605903148651, "learning_rate": 2.805247130475521e-05, "loss": 0.0448, "step": 11277 }, { "epoch": 7.925509486999298, "grad_norm": 0.14834082126617432, "learning_rate": 2.8052002810962755e-05, "loss": 0.0255, "step": 11278 }, { "epoch": 7.926212227687984, "grad_norm": 0.22457227110862732, "learning_rate": 2.8051534317170298e-05, "loss": 0.044, "step": 11279 }, { "epoch": 7.926914968376669, "grad_norm": 0.28725242614746094, "learning_rate": 2.8051065823377842e-05, "loss": 0.0532, "step": 11280 }, { "epoch": 7.927617709065355, "grad_norm": 0.3826650083065033, "learning_rate": 2.8050597329585382e-05, "loss": 0.0804, "step": 11281 }, { "epoch": 7.9283204497540405, "grad_norm": 0.827762246131897, "learning_rate": 2.8050128835792926e-05, "loss": 0.0658, "step": 11282 }, { "epoch": 7.929023190442726, "grad_norm": 0.7921033501625061, "learning_rate": 2.804966034200047e-05, "loss": 0.12, "step": 11283 }, { "epoch": 7.929725931131412, "grad_norm": 1.092176079750061, "learning_rate": 2.8049191848208013e-05, "loss": 0.1585, "step": 11284 }, { "epoch": 7.930428671820098, "grad_norm": 1.87347412109375, "learning_rate": 2.8048723354415554e-05, "loss": 0.2222, "step": 11285 }, { "epoch": 7.931131412508784, "grad_norm": 1.9887984991073608, "learning_rate": 2.8048254860623098e-05, "loss": 0.2381, "step": 11286 }, { "epoch": 7.93183415319747, "grad_norm": 0.3008216321468353, "learning_rate": 2.804778636683064e-05, "loss": 0.0723, "step": 11287 }, { "epoch": 7.932536893886156, "grad_norm": 0.1522047370672226, "learning_rate": 2.8047317873038185e-05, "loss": 0.0161, "step": 11288 }, { "epoch": 7.933239634574842, "grad_norm": 0.12303128093481064, "learning_rate": 2.8046849379245725e-05, "loss": 0.0321, "step": 11289 }, { "epoch": 7.933942375263528, "grad_norm": 0.2229485809803009, "learning_rate": 2.804638088545327e-05, "loss": 0.0207, "step": 11290 }, { "epoch": 7.934645115952214, "grad_norm": 0.1571967452764511, "learning_rate": 2.804591239166081e-05, "loss": 0.0252, "step": 11291 }, { "epoch": 7.9353478566409, "grad_norm": 0.12283863127231598, "learning_rate": 2.8045443897868353e-05, "loss": 0.0125, "step": 11292 }, { "epoch": 7.936050597329586, "grad_norm": 0.15343903005123138, "learning_rate": 2.8044975404075897e-05, "loss": 0.02, "step": 11293 }, { "epoch": 7.9367533380182715, "grad_norm": 0.1952916979789734, "learning_rate": 2.8044506910283437e-05, "loss": 0.0187, "step": 11294 }, { "epoch": 7.9374560787069575, "grad_norm": 0.2202315479516983, "learning_rate": 2.804403841649098e-05, "loss": 0.0204, "step": 11295 }, { "epoch": 7.938158819395643, "grad_norm": 0.11679386347532272, "learning_rate": 2.8043569922698525e-05, "loss": 0.0085, "step": 11296 }, { "epoch": 7.938861560084328, "grad_norm": 0.30607834458351135, "learning_rate": 2.804310142890607e-05, "loss": 0.0322, "step": 11297 }, { "epoch": 7.939564300773014, "grad_norm": 0.1499340534210205, "learning_rate": 2.804263293511361e-05, "loss": 0.0163, "step": 11298 }, { "epoch": 7.9402670414617, "grad_norm": 0.24753230810165405, "learning_rate": 2.8042164441321152e-05, "loss": 0.0291, "step": 11299 }, { "epoch": 7.940969782150386, "grad_norm": 0.1580180525779724, "learning_rate": 2.8041695947528696e-05, "loss": 0.0148, "step": 11300 }, { "epoch": 7.941672522839072, "grad_norm": 0.3225456476211548, "learning_rate": 2.804122745373624e-05, "loss": 0.0561, "step": 11301 }, { "epoch": 7.942375263527758, "grad_norm": 0.4353141784667969, "learning_rate": 2.8040758959943784e-05, "loss": 0.0664, "step": 11302 }, { "epoch": 7.943078004216444, "grad_norm": 0.11751475930213928, "learning_rate": 2.8040290466151324e-05, "loss": 0.0186, "step": 11303 }, { "epoch": 7.94378074490513, "grad_norm": 0.34932658076286316, "learning_rate": 2.8039821972358868e-05, "loss": 0.0291, "step": 11304 }, { "epoch": 7.944483485593816, "grad_norm": 0.2452460527420044, "learning_rate": 2.803935347856641e-05, "loss": 0.0416, "step": 11305 }, { "epoch": 7.945186226282502, "grad_norm": 0.343150794506073, "learning_rate": 2.8038884984773955e-05, "loss": 0.0659, "step": 11306 }, { "epoch": 7.945888966971188, "grad_norm": 0.35238581895828247, "learning_rate": 2.8038416490981495e-05, "loss": 0.1111, "step": 11307 }, { "epoch": 7.9465917076598735, "grad_norm": 0.4586244821548462, "learning_rate": 2.803794799718904e-05, "loss": 0.1047, "step": 11308 }, { "epoch": 7.9472944483485595, "grad_norm": 0.7176613807678223, "learning_rate": 2.803747950339658e-05, "loss": 0.1936, "step": 11309 }, { "epoch": 7.947997189037245, "grad_norm": 0.9140382409095764, "learning_rate": 2.8037011009604123e-05, "loss": 0.1884, "step": 11310 }, { "epoch": 7.948699929725931, "grad_norm": 1.2866711616516113, "learning_rate": 2.8036542515811664e-05, "loss": 0.2167, "step": 11311 }, { "epoch": 7.949402670414617, "grad_norm": 0.26868218183517456, "learning_rate": 2.8036074022019207e-05, "loss": 0.0732, "step": 11312 }, { "epoch": 7.950105411103303, "grad_norm": 0.352770060300827, "learning_rate": 2.803560552822675e-05, "loss": 0.0528, "step": 11313 }, { "epoch": 7.950808151791989, "grad_norm": 0.1615224927663803, "learning_rate": 2.8035137034434295e-05, "loss": 0.0194, "step": 11314 }, { "epoch": 7.951510892480675, "grad_norm": 0.13930033147335052, "learning_rate": 2.803466854064184e-05, "loss": 0.0193, "step": 11315 }, { "epoch": 7.952213633169361, "grad_norm": 0.13743636012077332, "learning_rate": 2.803420004684938e-05, "loss": 0.016, "step": 11316 }, { "epoch": 7.952916373858047, "grad_norm": 0.1977926641702652, "learning_rate": 2.8033731553056923e-05, "loss": 0.0375, "step": 11317 }, { "epoch": 7.953619114546732, "grad_norm": 0.19255369901657104, "learning_rate": 2.8033263059264466e-05, "loss": 0.0146, "step": 11318 }, { "epoch": 7.954321855235418, "grad_norm": 0.14411872625350952, "learning_rate": 2.803279456547201e-05, "loss": 0.0137, "step": 11319 }, { "epoch": 7.955024595924104, "grad_norm": 0.25972968339920044, "learning_rate": 2.803232607167955e-05, "loss": 0.022, "step": 11320 }, { "epoch": 7.95572733661279, "grad_norm": 0.17440003156661987, "learning_rate": 2.8031857577887094e-05, "loss": 0.0163, "step": 11321 }, { "epoch": 7.956430077301476, "grad_norm": 0.24896056950092316, "learning_rate": 2.8031389084094638e-05, "loss": 0.0214, "step": 11322 }, { "epoch": 7.9571328179901615, "grad_norm": 0.15051160752773285, "learning_rate": 2.803092059030218e-05, "loss": 0.0243, "step": 11323 }, { "epoch": 7.957835558678847, "grad_norm": 0.5899866819381714, "learning_rate": 2.8030452096509722e-05, "loss": 0.0271, "step": 11324 }, { "epoch": 7.958538299367533, "grad_norm": 0.3101128339767456, "learning_rate": 2.8029983602717266e-05, "loss": 0.0173, "step": 11325 }, { "epoch": 7.959241040056219, "grad_norm": 0.2567971646785736, "learning_rate": 2.8029515108924806e-05, "loss": 0.0322, "step": 11326 }, { "epoch": 7.959943780744905, "grad_norm": 0.17071932554244995, "learning_rate": 2.802904661513235e-05, "loss": 0.0308, "step": 11327 }, { "epoch": 7.960646521433591, "grad_norm": 0.19368813931941986, "learning_rate": 2.8028578121339893e-05, "loss": 0.0287, "step": 11328 }, { "epoch": 7.961349262122277, "grad_norm": 0.43035414814949036, "learning_rate": 2.8028109627547434e-05, "loss": 0.0583, "step": 11329 }, { "epoch": 7.962052002810963, "grad_norm": 0.2720687985420227, "learning_rate": 2.8027641133754977e-05, "loss": 0.0485, "step": 11330 }, { "epoch": 7.962754743499649, "grad_norm": 0.38876116275787354, "learning_rate": 2.802717263996252e-05, "loss": 0.0661, "step": 11331 }, { "epoch": 7.963457484188335, "grad_norm": 0.6186659336090088, "learning_rate": 2.8026704146170065e-05, "loss": 0.1041, "step": 11332 }, { "epoch": 7.964160224877021, "grad_norm": 0.6949255466461182, "learning_rate": 2.8026235652377605e-05, "loss": 0.0984, "step": 11333 }, { "epoch": 7.964862965565707, "grad_norm": 0.7257712483406067, "learning_rate": 2.802576715858515e-05, "loss": 0.1481, "step": 11334 }, { "epoch": 7.965565706254392, "grad_norm": 0.7515950202941895, "learning_rate": 2.8025298664792693e-05, "loss": 0.2018, "step": 11335 }, { "epoch": 7.966268446943078, "grad_norm": 0.9594013094902039, "learning_rate": 2.8024830171000236e-05, "loss": 0.2198, "step": 11336 }, { "epoch": 7.9669711876317635, "grad_norm": 0.611850380897522, "learning_rate": 2.8024361677207777e-05, "loss": 0.0976, "step": 11337 }, { "epoch": 7.967673928320449, "grad_norm": 0.10785282403230667, "learning_rate": 2.802389318341532e-05, "loss": 0.024, "step": 11338 }, { "epoch": 7.968376669009135, "grad_norm": 0.09717202186584473, "learning_rate": 2.8023424689622864e-05, "loss": 0.0131, "step": 11339 }, { "epoch": 7.969079409697821, "grad_norm": 0.25097334384918213, "learning_rate": 2.8022956195830408e-05, "loss": 0.0202, "step": 11340 }, { "epoch": 7.969782150386507, "grad_norm": 0.07280335575342178, "learning_rate": 2.802248770203795e-05, "loss": 0.0174, "step": 11341 }, { "epoch": 7.970484891075193, "grad_norm": 0.2016313672065735, "learning_rate": 2.8022019208245492e-05, "loss": 0.0197, "step": 11342 }, { "epoch": 7.971187631763879, "grad_norm": 0.23464658856391907, "learning_rate": 2.8021550714453036e-05, "loss": 0.0148, "step": 11343 }, { "epoch": 7.971890372452565, "grad_norm": 0.1557953804731369, "learning_rate": 2.8021082220660576e-05, "loss": 0.0149, "step": 11344 }, { "epoch": 7.972593113141251, "grad_norm": 0.21050423383712769, "learning_rate": 2.802061372686812e-05, "loss": 0.0392, "step": 11345 }, { "epoch": 7.973295853829937, "grad_norm": 0.29488062858581543, "learning_rate": 2.802014523307566e-05, "loss": 0.0109, "step": 11346 }, { "epoch": 7.973998594518623, "grad_norm": 0.17411287128925323, "learning_rate": 2.8019676739283204e-05, "loss": 0.0236, "step": 11347 }, { "epoch": 7.974701335207309, "grad_norm": 1.4143049716949463, "learning_rate": 2.8019208245490748e-05, "loss": 0.0161, "step": 11348 }, { "epoch": 7.9754040758959945, "grad_norm": 0.2792789041996002, "learning_rate": 2.801873975169829e-05, "loss": 0.0252, "step": 11349 }, { "epoch": 7.9761068165846805, "grad_norm": 0.10671818256378174, "learning_rate": 2.801827125790583e-05, "loss": 0.0177, "step": 11350 }, { "epoch": 7.976809557273366, "grad_norm": 0.41698798537254333, "learning_rate": 2.8017802764113375e-05, "loss": 0.0225, "step": 11351 }, { "epoch": 7.977512297962052, "grad_norm": 0.23289957642555237, "learning_rate": 2.801733427032092e-05, "loss": 0.0308, "step": 11352 }, { "epoch": 7.978215038650738, "grad_norm": 0.32429197430610657, "learning_rate": 2.8016865776528463e-05, "loss": 0.0476, "step": 11353 }, { "epoch": 7.978917779339424, "grad_norm": 0.15517915785312653, "learning_rate": 2.8016397282736006e-05, "loss": 0.025, "step": 11354 }, { "epoch": 7.97962052002811, "grad_norm": 1.4503426551818848, "learning_rate": 2.8015928788943547e-05, "loss": 0.0487, "step": 11355 }, { "epoch": 7.980323260716796, "grad_norm": 0.30731987953186035, "learning_rate": 2.801546029515109e-05, "loss": 0.0486, "step": 11356 }, { "epoch": 7.981026001405481, "grad_norm": 0.4758206903934479, "learning_rate": 2.8014991801358634e-05, "loss": 0.096, "step": 11357 }, { "epoch": 7.981728742094167, "grad_norm": 1.241390347480774, "learning_rate": 2.8014523307566178e-05, "loss": 0.112, "step": 11358 }, { "epoch": 7.982431482782853, "grad_norm": 0.5701252818107605, "learning_rate": 2.801405481377372e-05, "loss": 0.1798, "step": 11359 }, { "epoch": 7.983134223471539, "grad_norm": 1.2235661745071411, "learning_rate": 2.8013586319981262e-05, "loss": 0.2301, "step": 11360 }, { "epoch": 7.983836964160225, "grad_norm": 1.7470548152923584, "learning_rate": 2.8013117826188802e-05, "loss": 0.2815, "step": 11361 }, { "epoch": 7.984539704848911, "grad_norm": 0.19717589020729065, "learning_rate": 2.8012649332396346e-05, "loss": 0.0562, "step": 11362 }, { "epoch": 7.9852424455375965, "grad_norm": 0.4178701937198639, "learning_rate": 2.8012180838603886e-05, "loss": 0.0297, "step": 11363 }, { "epoch": 7.9859451862262825, "grad_norm": 0.12800279259681702, "learning_rate": 2.801171234481143e-05, "loss": 0.034, "step": 11364 }, { "epoch": 7.986647926914968, "grad_norm": 0.12486220896244049, "learning_rate": 2.8011243851018974e-05, "loss": 0.019, "step": 11365 }, { "epoch": 7.987350667603654, "grad_norm": 0.15790081024169922, "learning_rate": 2.8010775357226518e-05, "loss": 0.0174, "step": 11366 }, { "epoch": 7.98805340829234, "grad_norm": 0.22651441395282745, "learning_rate": 2.801030686343406e-05, "loss": 0.0238, "step": 11367 }, { "epoch": 7.988756148981026, "grad_norm": 0.2516651749610901, "learning_rate": 2.80098383696416e-05, "loss": 0.0144, "step": 11368 }, { "epoch": 7.989458889669712, "grad_norm": 0.24305197596549988, "learning_rate": 2.8009369875849145e-05, "loss": 0.0287, "step": 11369 }, { "epoch": 7.990161630358398, "grad_norm": 0.169418066740036, "learning_rate": 2.800890138205669e-05, "loss": 0.0251, "step": 11370 }, { "epoch": 7.990864371047084, "grad_norm": 0.24109576642513275, "learning_rate": 2.8008432888264233e-05, "loss": 0.0239, "step": 11371 }, { "epoch": 7.99156711173577, "grad_norm": 0.17238786816596985, "learning_rate": 2.8007964394471773e-05, "loss": 0.0247, "step": 11372 }, { "epoch": 7.992269852424456, "grad_norm": 0.7895814180374146, "learning_rate": 2.8007495900679317e-05, "loss": 0.0372, "step": 11373 }, { "epoch": 7.992972593113141, "grad_norm": 0.18049080669879913, "learning_rate": 2.800702740688686e-05, "loss": 0.0127, "step": 11374 }, { "epoch": 7.993675333801827, "grad_norm": 0.1366037279367447, "learning_rate": 2.8006558913094404e-05, "loss": 0.034, "step": 11375 }, { "epoch": 7.994378074490513, "grad_norm": 0.14851605892181396, "learning_rate": 2.8006090419301945e-05, "loss": 0.0359, "step": 11376 }, { "epoch": 7.9950808151791986, "grad_norm": 0.23565064370632172, "learning_rate": 2.800562192550949e-05, "loss": 0.0579, "step": 11377 }, { "epoch": 7.9957835558678845, "grad_norm": 0.34498271346092224, "learning_rate": 2.800515343171703e-05, "loss": 0.0541, "step": 11378 }, { "epoch": 7.99648629655657, "grad_norm": 0.3681020140647888, "learning_rate": 2.8004684937924572e-05, "loss": 0.0399, "step": 11379 }, { "epoch": 7.997189037245256, "grad_norm": 0.5720534324645996, "learning_rate": 2.8004216444132116e-05, "loss": 0.0911, "step": 11380 }, { "epoch": 7.997891777933942, "grad_norm": 0.749087929725647, "learning_rate": 2.8003747950339657e-05, "loss": 0.1229, "step": 11381 }, { "epoch": 7.998594518622628, "grad_norm": 7.0487871170043945, "learning_rate": 2.80032794565472e-05, "loss": 0.2072, "step": 11382 }, { "epoch": 7.999297259311314, "grad_norm": 0.9277560114860535, "learning_rate": 2.8002810962754744e-05, "loss": 0.2055, "step": 11383 }, { "epoch": 8.0, "grad_norm": 0.494368314743042, "learning_rate": 2.8002342468962288e-05, "loss": 0.1524, "step": 11384 }, { "epoch": 8.000702740688686, "grad_norm": 0.6796692609786987, "learning_rate": 2.8001873975169828e-05, "loss": 0.0795, "step": 11385 }, { "epoch": 8.001405481377372, "grad_norm": 0.2307494878768921, "learning_rate": 2.8001405481377372e-05, "loss": 0.06, "step": 11386 }, { "epoch": 8.002108222066058, "grad_norm": 0.14658240973949432, "learning_rate": 2.8000936987584916e-05, "loss": 0.0345, "step": 11387 }, { "epoch": 8.002810962754744, "grad_norm": 0.16426515579223633, "learning_rate": 2.800046849379246e-05, "loss": 0.0219, "step": 11388 }, { "epoch": 8.00351370344343, "grad_norm": 0.07979383319616318, "learning_rate": 2.8e-05, "loss": 0.0095, "step": 11389 }, { "epoch": 8.004216444132116, "grad_norm": 0.18026205897331238, "learning_rate": 2.7999531506207543e-05, "loss": 0.0194, "step": 11390 }, { "epoch": 8.004919184820801, "grad_norm": 0.2961854338645935, "learning_rate": 2.7999063012415087e-05, "loss": 0.0201, "step": 11391 }, { "epoch": 8.005621925509487, "grad_norm": 0.2691172957420349, "learning_rate": 2.799859451862263e-05, "loss": 0.049, "step": 11392 }, { "epoch": 8.006324666198173, "grad_norm": 0.22366629540920258, "learning_rate": 2.7998126024830174e-05, "loss": 0.0277, "step": 11393 }, { "epoch": 8.00702740688686, "grad_norm": 0.1879684031009674, "learning_rate": 2.7997657531037715e-05, "loss": 0.0127, "step": 11394 }, { "epoch": 8.007730147575545, "grad_norm": 0.13947175443172455, "learning_rate": 2.799718903724526e-05, "loss": 0.0214, "step": 11395 }, { "epoch": 8.008432888264231, "grad_norm": 0.12925153970718384, "learning_rate": 2.79967205434528e-05, "loss": 0.0146, "step": 11396 }, { "epoch": 8.009135628952917, "grad_norm": 0.15940269827842712, "learning_rate": 2.7996252049660343e-05, "loss": 0.0223, "step": 11397 }, { "epoch": 8.009838369641603, "grad_norm": 0.17733314633369446, "learning_rate": 2.7995783555867883e-05, "loss": 0.0216, "step": 11398 }, { "epoch": 8.010541110330289, "grad_norm": 0.4590957760810852, "learning_rate": 2.7995315062075427e-05, "loss": 0.0188, "step": 11399 }, { "epoch": 8.011243851018975, "grad_norm": 0.2312295287847519, "learning_rate": 2.799484656828297e-05, "loss": 0.0378, "step": 11400 }, { "epoch": 8.01194659170766, "grad_norm": 0.2272830605506897, "learning_rate": 2.7994378074490514e-05, "loss": 0.0189, "step": 11401 }, { "epoch": 8.012649332396347, "grad_norm": 0.22035571932792664, "learning_rate": 2.7993909580698054e-05, "loss": 0.0551, "step": 11402 }, { "epoch": 8.013352073085033, "grad_norm": 0.3930771052837372, "learning_rate": 2.7993441086905598e-05, "loss": 0.0352, "step": 11403 }, { "epoch": 8.014054813773717, "grad_norm": 0.18292182683944702, "learning_rate": 2.7992972593113142e-05, "loss": 0.0335, "step": 11404 }, { "epoch": 8.014757554462403, "grad_norm": 0.3606186807155609, "learning_rate": 2.7992504099320686e-05, "loss": 0.0734, "step": 11405 }, { "epoch": 8.015460295151088, "grad_norm": 1.0989086627960205, "learning_rate": 2.799203560552823e-05, "loss": 0.1175, "step": 11406 }, { "epoch": 8.016163035839774, "grad_norm": 0.46955204010009766, "learning_rate": 2.799156711173577e-05, "loss": 0.135, "step": 11407 }, { "epoch": 8.01686577652846, "grad_norm": 1.0240554809570312, "learning_rate": 2.7991098617943313e-05, "loss": 0.2019, "step": 11408 }, { "epoch": 8.017568517217146, "grad_norm": 1.2967609167099, "learning_rate": 2.7990630124150857e-05, "loss": 0.1949, "step": 11409 }, { "epoch": 8.018271257905832, "grad_norm": 0.8915266990661621, "learning_rate": 2.79901616303584e-05, "loss": 0.0704, "step": 11410 }, { "epoch": 8.018973998594518, "grad_norm": 0.28018781542778015, "learning_rate": 2.798969313656594e-05, "loss": 0.0367, "step": 11411 }, { "epoch": 8.019676739283204, "grad_norm": 0.12997937202453613, "learning_rate": 2.7989224642773485e-05, "loss": 0.0208, "step": 11412 }, { "epoch": 8.02037947997189, "grad_norm": 0.1354350745677948, "learning_rate": 2.7988756148981025e-05, "loss": 0.0213, "step": 11413 }, { "epoch": 8.021082220660576, "grad_norm": 0.09974400699138641, "learning_rate": 2.798828765518857e-05, "loss": 0.0167, "step": 11414 }, { "epoch": 8.021784961349262, "grad_norm": 0.12112115323543549, "learning_rate": 2.798781916139611e-05, "loss": 0.0215, "step": 11415 }, { "epoch": 8.022487702037948, "grad_norm": 0.12164943665266037, "learning_rate": 2.7987350667603653e-05, "loss": 0.02, "step": 11416 }, { "epoch": 8.023190442726634, "grad_norm": 0.13479679822921753, "learning_rate": 2.7986882173811197e-05, "loss": 0.0239, "step": 11417 }, { "epoch": 8.02389318341532, "grad_norm": 0.15043234825134277, "learning_rate": 2.798641368001874e-05, "loss": 0.0199, "step": 11418 }, { "epoch": 8.024595924104005, "grad_norm": 0.18716593086719513, "learning_rate": 2.7985945186226284e-05, "loss": 0.0402, "step": 11419 }, { "epoch": 8.025298664792691, "grad_norm": 0.14046381413936615, "learning_rate": 2.7985476692433825e-05, "loss": 0.0228, "step": 11420 }, { "epoch": 8.026001405481377, "grad_norm": 0.20314453542232513, "learning_rate": 2.7985008198641368e-05, "loss": 0.015, "step": 11421 }, { "epoch": 8.026704146170063, "grad_norm": 0.24366480112075806, "learning_rate": 2.7984539704848912e-05, "loss": 0.0685, "step": 11422 }, { "epoch": 8.02740688685875, "grad_norm": 0.19048719108104706, "learning_rate": 2.7984071211056456e-05, "loss": 0.0154, "step": 11423 }, { "epoch": 8.028109627547435, "grad_norm": 0.2429187148809433, "learning_rate": 2.7983602717263996e-05, "loss": 0.0271, "step": 11424 }, { "epoch": 8.028812368236121, "grad_norm": 0.1617676317691803, "learning_rate": 2.798313422347154e-05, "loss": 0.0302, "step": 11425 }, { "epoch": 8.029515108924807, "grad_norm": 0.18738240003585815, "learning_rate": 2.7982665729679084e-05, "loss": 0.0227, "step": 11426 }, { "epoch": 8.030217849613493, "grad_norm": 0.5820369124412537, "learning_rate": 2.7982197235886627e-05, "loss": 0.0442, "step": 11427 }, { "epoch": 8.030920590302179, "grad_norm": 0.2686006426811218, "learning_rate": 2.7981728742094168e-05, "loss": 0.0572, "step": 11428 }, { "epoch": 8.031623330990865, "grad_norm": 0.20700764656066895, "learning_rate": 2.798126024830171e-05, "loss": 0.0352, "step": 11429 }, { "epoch": 8.03232607167955, "grad_norm": 0.6477649807929993, "learning_rate": 2.7980791754509255e-05, "loss": 0.1049, "step": 11430 }, { "epoch": 8.033028812368237, "grad_norm": 1.21756911277771, "learning_rate": 2.7980323260716795e-05, "loss": 0.1, "step": 11431 }, { "epoch": 8.033731553056922, "grad_norm": 0.5479359030723572, "learning_rate": 2.797985476692434e-05, "loss": 0.1377, "step": 11432 }, { "epoch": 8.034434293745608, "grad_norm": 0.8317524194717407, "learning_rate": 2.797938627313188e-05, "loss": 0.1796, "step": 11433 }, { "epoch": 8.035137034434294, "grad_norm": 1.5797609090805054, "learning_rate": 2.7978917779339423e-05, "loss": 0.1878, "step": 11434 }, { "epoch": 8.03583977512298, "grad_norm": 0.2044316828250885, "learning_rate": 2.7978449285546967e-05, "loss": 0.0718, "step": 11435 }, { "epoch": 8.036542515811666, "grad_norm": 0.14508093893527985, "learning_rate": 2.797798079175451e-05, "loss": 0.0186, "step": 11436 }, { "epoch": 8.037245256500352, "grad_norm": 0.33570596575737, "learning_rate": 2.797751229796205e-05, "loss": 0.0298, "step": 11437 }, { "epoch": 8.037947997189038, "grad_norm": 0.15001891553401947, "learning_rate": 2.7977043804169595e-05, "loss": 0.0166, "step": 11438 }, { "epoch": 8.038650737877724, "grad_norm": 0.1967543065547943, "learning_rate": 2.797657531037714e-05, "loss": 0.0224, "step": 11439 }, { "epoch": 8.03935347856641, "grad_norm": 0.28654852509498596, "learning_rate": 2.7976106816584682e-05, "loss": 0.0202, "step": 11440 }, { "epoch": 8.040056219255096, "grad_norm": 0.503896951675415, "learning_rate": 2.7975638322792222e-05, "loss": 0.023, "step": 11441 }, { "epoch": 8.04075895994378, "grad_norm": 0.24406903982162476, "learning_rate": 2.7975169828999766e-05, "loss": 0.0309, "step": 11442 }, { "epoch": 8.041461700632466, "grad_norm": 0.10437776148319244, "learning_rate": 2.797470133520731e-05, "loss": 0.015, "step": 11443 }, { "epoch": 8.042164441321152, "grad_norm": 0.1409076303243637, "learning_rate": 2.7974232841414854e-05, "loss": 0.0176, "step": 11444 }, { "epoch": 8.042867182009838, "grad_norm": 0.21547351777553558, "learning_rate": 2.7973764347622397e-05, "loss": 0.0156, "step": 11445 }, { "epoch": 8.043569922698524, "grad_norm": 0.16715316474437714, "learning_rate": 2.7973295853829938e-05, "loss": 0.0154, "step": 11446 }, { "epoch": 8.04427266338721, "grad_norm": 0.1283947229385376, "learning_rate": 2.797282736003748e-05, "loss": 0.0259, "step": 11447 }, { "epoch": 8.044975404075895, "grad_norm": 0.1295379251241684, "learning_rate": 2.7972358866245022e-05, "loss": 0.0131, "step": 11448 }, { "epoch": 8.045678144764581, "grad_norm": 0.151309996843338, "learning_rate": 2.7971890372452565e-05, "loss": 0.0284, "step": 11449 }, { "epoch": 8.046380885453267, "grad_norm": 0.19575819373130798, "learning_rate": 2.7971421878660106e-05, "loss": 0.0347, "step": 11450 }, { "epoch": 8.047083626141953, "grad_norm": 0.1440170705318451, "learning_rate": 2.797095338486765e-05, "loss": 0.0256, "step": 11451 }, { "epoch": 8.047786366830639, "grad_norm": 0.27639105916023254, "learning_rate": 2.7970484891075193e-05, "loss": 0.0318, "step": 11452 }, { "epoch": 8.048489107519325, "grad_norm": 0.25681936740875244, "learning_rate": 2.7970016397282737e-05, "loss": 0.0295, "step": 11453 }, { "epoch": 8.049191848208011, "grad_norm": 0.3265092074871063, "learning_rate": 2.7969547903490277e-05, "loss": 0.053, "step": 11454 }, { "epoch": 8.049894588896697, "grad_norm": 0.3107292652130127, "learning_rate": 2.796907940969782e-05, "loss": 0.0811, "step": 11455 }, { "epoch": 8.050597329585383, "grad_norm": 0.4505515992641449, "learning_rate": 2.7968610915905365e-05, "loss": 0.115, "step": 11456 }, { "epoch": 8.051300070274069, "grad_norm": 0.976851224899292, "learning_rate": 2.796814242211291e-05, "loss": 0.1693, "step": 11457 }, { "epoch": 8.052002810962755, "grad_norm": 1.2015752792358398, "learning_rate": 2.7967673928320452e-05, "loss": 0.1808, "step": 11458 }, { "epoch": 8.05270555165144, "grad_norm": 1.6793090105056763, "learning_rate": 2.7967205434527993e-05, "loss": 0.2957, "step": 11459 }, { "epoch": 8.053408292340126, "grad_norm": 0.27025341987609863, "learning_rate": 2.7966736940735536e-05, "loss": 0.0784, "step": 11460 }, { "epoch": 8.054111033028812, "grad_norm": 0.2248089462518692, "learning_rate": 2.796626844694308e-05, "loss": 0.0363, "step": 11461 }, { "epoch": 8.054813773717498, "grad_norm": 0.3013382852077484, "learning_rate": 2.7965799953150624e-05, "loss": 0.0359, "step": 11462 }, { "epoch": 8.055516514406184, "grad_norm": 0.13751250505447388, "learning_rate": 2.7965331459358164e-05, "loss": 0.0156, "step": 11463 }, { "epoch": 8.05621925509487, "grad_norm": 0.14152581989765167, "learning_rate": 2.7964862965565708e-05, "loss": 0.0122, "step": 11464 }, { "epoch": 8.056921995783556, "grad_norm": 0.1255345642566681, "learning_rate": 2.796439447177325e-05, "loss": 0.0088, "step": 11465 }, { "epoch": 8.057624736472242, "grad_norm": 0.11608325690031052, "learning_rate": 2.7963925977980792e-05, "loss": 0.0127, "step": 11466 }, { "epoch": 8.058327477160928, "grad_norm": 0.12138267606496811, "learning_rate": 2.7963457484188336e-05, "loss": 0.0148, "step": 11467 }, { "epoch": 8.059030217849614, "grad_norm": 0.12676340341567993, "learning_rate": 2.7962988990395876e-05, "loss": 0.0173, "step": 11468 }, { "epoch": 8.0597329585383, "grad_norm": 0.14760667085647583, "learning_rate": 2.796252049660342e-05, "loss": 0.0117, "step": 11469 }, { "epoch": 8.060435699226986, "grad_norm": 0.26003095507621765, "learning_rate": 2.7962052002810963e-05, "loss": 0.0364, "step": 11470 }, { "epoch": 8.061138439915672, "grad_norm": 0.2058156430721283, "learning_rate": 2.7961583509018507e-05, "loss": 0.0234, "step": 11471 }, { "epoch": 8.061841180604358, "grad_norm": 0.21077178418636322, "learning_rate": 2.7961115015226047e-05, "loss": 0.0274, "step": 11472 }, { "epoch": 8.062543921293043, "grad_norm": 0.1775210201740265, "learning_rate": 2.796064652143359e-05, "loss": 0.0105, "step": 11473 }, { "epoch": 8.06324666198173, "grad_norm": 0.20552781224250793, "learning_rate": 2.7960178027641135e-05, "loss": 0.0274, "step": 11474 }, { "epoch": 8.063949402670415, "grad_norm": 0.26493147015571594, "learning_rate": 2.795970953384868e-05, "loss": 0.0319, "step": 11475 }, { "epoch": 8.064652143359101, "grad_norm": 0.14341257512569427, "learning_rate": 2.795924104005622e-05, "loss": 0.0177, "step": 11476 }, { "epoch": 8.065354884047787, "grad_norm": 0.29321742057800293, "learning_rate": 2.7958772546263763e-05, "loss": 0.0406, "step": 11477 }, { "epoch": 8.066057624736473, "grad_norm": 0.5096074938774109, "learning_rate": 2.7958304052471306e-05, "loss": 0.065, "step": 11478 }, { "epoch": 8.066760365425159, "grad_norm": 0.22815439105033875, "learning_rate": 2.795783555867885e-05, "loss": 0.0631, "step": 11479 }, { "epoch": 8.067463106113845, "grad_norm": 0.32879430055618286, "learning_rate": 2.795736706488639e-05, "loss": 0.0745, "step": 11480 }, { "epoch": 8.068165846802529, "grad_norm": 0.5063245892524719, "learning_rate": 2.7956898571093934e-05, "loss": 0.1276, "step": 11481 }, { "epoch": 8.068868587491215, "grad_norm": 0.8911059498786926, "learning_rate": 2.7956430077301478e-05, "loss": 0.1237, "step": 11482 }, { "epoch": 8.0695713281799, "grad_norm": 1.800390362739563, "learning_rate": 2.7955961583509018e-05, "loss": 0.1999, "step": 11483 }, { "epoch": 8.070274068868587, "grad_norm": 0.9592876434326172, "learning_rate": 2.7955493089716562e-05, "loss": 0.2545, "step": 11484 }, { "epoch": 8.070976809557273, "grad_norm": 0.2693748474121094, "learning_rate": 2.7955024595924102e-05, "loss": 0.0749, "step": 11485 }, { "epoch": 8.071679550245959, "grad_norm": 0.20525962114334106, "learning_rate": 2.7954556102131646e-05, "loss": 0.0332, "step": 11486 }, { "epoch": 8.072382290934645, "grad_norm": 0.12362919002771378, "learning_rate": 2.795408760833919e-05, "loss": 0.0173, "step": 11487 }, { "epoch": 8.07308503162333, "grad_norm": 0.18309177458286285, "learning_rate": 2.7953619114546734e-05, "loss": 0.0237, "step": 11488 }, { "epoch": 8.073787772312016, "grad_norm": 0.27360036969184875, "learning_rate": 2.7953150620754274e-05, "loss": 0.0279, "step": 11489 }, { "epoch": 8.074490513000702, "grad_norm": 0.13393427431583405, "learning_rate": 2.7952682126961818e-05, "loss": 0.0147, "step": 11490 }, { "epoch": 8.075193253689388, "grad_norm": 0.27107760310173035, "learning_rate": 2.795221363316936e-05, "loss": 0.0161, "step": 11491 }, { "epoch": 8.075895994378074, "grad_norm": 0.11380840837955475, "learning_rate": 2.7951745139376905e-05, "loss": 0.0121, "step": 11492 }, { "epoch": 8.07659873506676, "grad_norm": 0.28347042202949524, "learning_rate": 2.795127664558445e-05, "loss": 0.0335, "step": 11493 }, { "epoch": 8.077301475755446, "grad_norm": 0.1452491581439972, "learning_rate": 2.795080815179199e-05, "loss": 0.0327, "step": 11494 }, { "epoch": 8.078004216444132, "grad_norm": 0.24459153413772583, "learning_rate": 2.7950339657999533e-05, "loss": 0.0302, "step": 11495 }, { "epoch": 8.078706957132818, "grad_norm": 0.3163432776927948, "learning_rate": 2.7949871164207077e-05, "loss": 0.0262, "step": 11496 }, { "epoch": 8.079409697821504, "grad_norm": 0.18111173808574677, "learning_rate": 2.794940267041462e-05, "loss": 0.0251, "step": 11497 }, { "epoch": 8.08011243851019, "grad_norm": 0.14132912456989288, "learning_rate": 2.794893417662216e-05, "loss": 0.0158, "step": 11498 }, { "epoch": 8.080815179198876, "grad_norm": 0.17346496880054474, "learning_rate": 2.7948465682829704e-05, "loss": 0.027, "step": 11499 }, { "epoch": 8.081517919887562, "grad_norm": 0.16589166224002838, "learning_rate": 2.7947997189037245e-05, "loss": 0.0338, "step": 11500 }, { "epoch": 8.082220660576247, "grad_norm": 0.20941990613937378, "learning_rate": 2.794752869524479e-05, "loss": 0.0171, "step": 11501 }, { "epoch": 8.082923401264933, "grad_norm": 0.24709609150886536, "learning_rate": 2.794706020145233e-05, "loss": 0.0319, "step": 11502 }, { "epoch": 8.08362614195362, "grad_norm": 0.5070090293884277, "learning_rate": 2.7946591707659872e-05, "loss": 0.0246, "step": 11503 }, { "epoch": 8.084328882642305, "grad_norm": 0.2172584980726242, "learning_rate": 2.7946123213867416e-05, "loss": 0.0368, "step": 11504 }, { "epoch": 8.085031623330991, "grad_norm": 0.7671838402748108, "learning_rate": 2.794565472007496e-05, "loss": 0.0645, "step": 11505 }, { "epoch": 8.085734364019677, "grad_norm": 0.8071747422218323, "learning_rate": 2.7945186226282504e-05, "loss": 0.1116, "step": 11506 }, { "epoch": 8.086437104708363, "grad_norm": 0.6606447696685791, "learning_rate": 2.7944717732490044e-05, "loss": 0.1572, "step": 11507 }, { "epoch": 8.087139845397049, "grad_norm": 0.8878746628761292, "learning_rate": 2.7944249238697588e-05, "loss": 0.1986, "step": 11508 }, { "epoch": 8.087842586085735, "grad_norm": 1.1515045166015625, "learning_rate": 2.794378074490513e-05, "loss": 0.1801, "step": 11509 }, { "epoch": 8.08854532677442, "grad_norm": 0.3066137731075287, "learning_rate": 2.7943312251112675e-05, "loss": 0.0602, "step": 11510 }, { "epoch": 8.089248067463107, "grad_norm": 0.17428076267242432, "learning_rate": 2.7942843757320215e-05, "loss": 0.023, "step": 11511 }, { "epoch": 8.089950808151793, "grad_norm": 0.09531709551811218, "learning_rate": 2.794237526352776e-05, "loss": 0.0177, "step": 11512 }, { "epoch": 8.090653548840478, "grad_norm": 0.20893606543540955, "learning_rate": 2.7941906769735303e-05, "loss": 0.0261, "step": 11513 }, { "epoch": 8.091356289529164, "grad_norm": 0.12748830020427704, "learning_rate": 2.7941438275942847e-05, "loss": 0.0165, "step": 11514 }, { "epoch": 8.09205903021785, "grad_norm": 0.19894346594810486, "learning_rate": 2.7940969782150387e-05, "loss": 0.0267, "step": 11515 }, { "epoch": 8.092761770906536, "grad_norm": 0.12973271310329437, "learning_rate": 2.794050128835793e-05, "loss": 0.0189, "step": 11516 }, { "epoch": 8.093464511595222, "grad_norm": 0.5027895569801331, "learning_rate": 2.7940032794565474e-05, "loss": 0.0151, "step": 11517 }, { "epoch": 8.094167252283908, "grad_norm": 0.18727770447731018, "learning_rate": 2.7939564300773015e-05, "loss": 0.0177, "step": 11518 }, { "epoch": 8.094869992972592, "grad_norm": 0.14025269448757172, "learning_rate": 2.793909580698056e-05, "loss": 0.0162, "step": 11519 }, { "epoch": 8.095572733661278, "grad_norm": 0.2020655870437622, "learning_rate": 2.79386273131881e-05, "loss": 0.0301, "step": 11520 }, { "epoch": 8.096275474349964, "grad_norm": 0.20230655372142792, "learning_rate": 2.7938158819395643e-05, "loss": 0.0182, "step": 11521 }, { "epoch": 8.09697821503865, "grad_norm": 0.24656042456626892, "learning_rate": 2.7937690325603186e-05, "loss": 0.0178, "step": 11522 }, { "epoch": 8.097680955727336, "grad_norm": 0.24158023297786713, "learning_rate": 2.793722183181073e-05, "loss": 0.018, "step": 11523 }, { "epoch": 8.098383696416022, "grad_norm": 0.18995404243469238, "learning_rate": 2.793675333801827e-05, "loss": 0.0351, "step": 11524 }, { "epoch": 8.099086437104708, "grad_norm": 0.16095224022865295, "learning_rate": 2.7936284844225814e-05, "loss": 0.0362, "step": 11525 }, { "epoch": 8.099789177793394, "grad_norm": 0.1877245008945465, "learning_rate": 2.7935816350433358e-05, "loss": 0.0147, "step": 11526 }, { "epoch": 8.10049191848208, "grad_norm": 0.31149542331695557, "learning_rate": 2.79353478566409e-05, "loss": 0.0445, "step": 11527 }, { "epoch": 8.101194659170766, "grad_norm": 1.3885302543640137, "learning_rate": 2.7934879362848442e-05, "loss": 0.0375, "step": 11528 }, { "epoch": 8.101897399859451, "grad_norm": 0.44601914286613464, "learning_rate": 2.7934410869055986e-05, "loss": 0.0564, "step": 11529 }, { "epoch": 8.102600140548137, "grad_norm": 0.573316216468811, "learning_rate": 2.793394237526353e-05, "loss": 0.0821, "step": 11530 }, { "epoch": 8.103302881236823, "grad_norm": 0.4367632269859314, "learning_rate": 2.7933473881471073e-05, "loss": 0.0918, "step": 11531 }, { "epoch": 8.10400562192551, "grad_norm": 0.8250226378440857, "learning_rate": 2.7933005387678617e-05, "loss": 0.161, "step": 11532 }, { "epoch": 8.104708362614195, "grad_norm": 0.9411904811859131, "learning_rate": 2.7932536893886157e-05, "loss": 0.1835, "step": 11533 }, { "epoch": 8.105411103302881, "grad_norm": 1.1358143091201782, "learning_rate": 2.79320684000937e-05, "loss": 0.2216, "step": 11534 }, { "epoch": 8.106113843991567, "grad_norm": 0.6956408619880676, "learning_rate": 2.793159990630124e-05, "loss": 0.0945, "step": 11535 }, { "epoch": 8.106816584680253, "grad_norm": 0.38263770937919617, "learning_rate": 2.7931131412508785e-05, "loss": 0.0298, "step": 11536 }, { "epoch": 8.107519325368939, "grad_norm": 0.44251713156700134, "learning_rate": 2.7930662918716325e-05, "loss": 0.0327, "step": 11537 }, { "epoch": 8.108222066057625, "grad_norm": 0.20147274434566498, "learning_rate": 2.793019442492387e-05, "loss": 0.02, "step": 11538 }, { "epoch": 8.10892480674631, "grad_norm": 0.09635082632303238, "learning_rate": 2.7929725931131413e-05, "loss": 0.0121, "step": 11539 }, { "epoch": 8.109627547434997, "grad_norm": 0.13452211022377014, "learning_rate": 2.7929257437338956e-05, "loss": 0.0149, "step": 11540 }, { "epoch": 8.110330288123683, "grad_norm": 0.245136097073555, "learning_rate": 2.7928788943546497e-05, "loss": 0.0142, "step": 11541 }, { "epoch": 8.111033028812368, "grad_norm": 0.3461887836456299, "learning_rate": 2.792832044975404e-05, "loss": 0.0176, "step": 11542 }, { "epoch": 8.111735769501054, "grad_norm": 0.1944284737110138, "learning_rate": 2.7927851955961584e-05, "loss": 0.0197, "step": 11543 }, { "epoch": 8.11243851018974, "grad_norm": 0.14069265127182007, "learning_rate": 2.7927383462169128e-05, "loss": 0.0161, "step": 11544 }, { "epoch": 8.113141250878426, "grad_norm": 0.15435908734798431, "learning_rate": 2.792691496837667e-05, "loss": 0.0223, "step": 11545 }, { "epoch": 8.113843991567112, "grad_norm": 0.2683972120285034, "learning_rate": 2.7926446474584212e-05, "loss": 0.0123, "step": 11546 }, { "epoch": 8.114546732255798, "grad_norm": 0.5060185790061951, "learning_rate": 2.7925977980791756e-05, "loss": 0.0283, "step": 11547 }, { "epoch": 8.115249472944484, "grad_norm": 0.2966032028198242, "learning_rate": 2.79255094869993e-05, "loss": 0.0237, "step": 11548 }, { "epoch": 8.11595221363317, "grad_norm": 0.26916730403900146, "learning_rate": 2.7925040993206843e-05, "loss": 0.0209, "step": 11549 }, { "epoch": 8.116654954321856, "grad_norm": 0.17763908207416534, "learning_rate": 2.7924572499414383e-05, "loss": 0.0283, "step": 11550 }, { "epoch": 8.117357695010542, "grad_norm": 0.1592669039964676, "learning_rate": 2.7924104005621927e-05, "loss": 0.0146, "step": 11551 }, { "epoch": 8.118060435699228, "grad_norm": 0.16888833045959473, "learning_rate": 2.792363551182947e-05, "loss": 0.0341, "step": 11552 }, { "epoch": 8.118763176387914, "grad_norm": 0.30842509865760803, "learning_rate": 2.792316701803701e-05, "loss": 0.0453, "step": 11553 }, { "epoch": 8.1194659170766, "grad_norm": 0.22170571982860565, "learning_rate": 2.792269852424455e-05, "loss": 0.0513, "step": 11554 }, { "epoch": 8.120168657765285, "grad_norm": 0.40349307656288147, "learning_rate": 2.7922230030452095e-05, "loss": 0.0721, "step": 11555 }, { "epoch": 8.120871398453971, "grad_norm": 0.6305455565452576, "learning_rate": 2.792176153665964e-05, "loss": 0.1461, "step": 11556 }, { "epoch": 8.121574139142655, "grad_norm": 0.6297188997268677, "learning_rate": 2.7921293042867183e-05, "loss": 0.1606, "step": 11557 }, { "epoch": 8.122276879831341, "grad_norm": 1.1764459609985352, "learning_rate": 2.7920824549074727e-05, "loss": 0.2036, "step": 11558 }, { "epoch": 8.122979620520027, "grad_norm": 1.8612160682678223, "learning_rate": 2.7920356055282267e-05, "loss": 0.1959, "step": 11559 }, { "epoch": 8.123682361208713, "grad_norm": 0.20872116088867188, "learning_rate": 2.791988756148981e-05, "loss": 0.079, "step": 11560 }, { "epoch": 8.1243851018974, "grad_norm": 0.4159833490848541, "learning_rate": 2.7919419067697354e-05, "loss": 0.0273, "step": 11561 }, { "epoch": 8.125087842586085, "grad_norm": 0.18046148121356964, "learning_rate": 2.7918950573904898e-05, "loss": 0.0191, "step": 11562 }, { "epoch": 8.125790583274771, "grad_norm": 0.08941780775785446, "learning_rate": 2.791848208011244e-05, "loss": 0.014, "step": 11563 }, { "epoch": 8.126493323963457, "grad_norm": 0.32830217480659485, "learning_rate": 2.7918013586319982e-05, "loss": 0.0203, "step": 11564 }, { "epoch": 8.127196064652143, "grad_norm": 0.21995484828948975, "learning_rate": 2.7917545092527526e-05, "loss": 0.0224, "step": 11565 }, { "epoch": 8.127898805340829, "grad_norm": 0.07829933613538742, "learning_rate": 2.791707659873507e-05, "loss": 0.0125, "step": 11566 }, { "epoch": 8.128601546029515, "grad_norm": 0.15865321457386017, "learning_rate": 2.791660810494261e-05, "loss": 0.0229, "step": 11567 }, { "epoch": 8.1293042867182, "grad_norm": 0.21536268293857574, "learning_rate": 2.7916139611150154e-05, "loss": 0.0214, "step": 11568 }, { "epoch": 8.130007027406887, "grad_norm": 0.10984020680189133, "learning_rate": 2.7915671117357697e-05, "loss": 0.017, "step": 11569 }, { "epoch": 8.130709768095572, "grad_norm": 0.10651743412017822, "learning_rate": 2.7915202623565238e-05, "loss": 0.0154, "step": 11570 }, { "epoch": 8.131412508784258, "grad_norm": 0.1775500476360321, "learning_rate": 2.791473412977278e-05, "loss": 0.0189, "step": 11571 }, { "epoch": 8.132115249472944, "grad_norm": 0.2781200110912323, "learning_rate": 2.7914265635980322e-05, "loss": 0.0281, "step": 11572 }, { "epoch": 8.13281799016163, "grad_norm": 0.6726841926574707, "learning_rate": 2.7913797142187865e-05, "loss": 0.0207, "step": 11573 }, { "epoch": 8.133520730850316, "grad_norm": 0.22111044824123383, "learning_rate": 2.791332864839541e-05, "loss": 0.0305, "step": 11574 }, { "epoch": 8.134223471539002, "grad_norm": 0.31011447310447693, "learning_rate": 2.7912860154602953e-05, "loss": 0.0319, "step": 11575 }, { "epoch": 8.134926212227688, "grad_norm": 0.18671074509620667, "learning_rate": 2.7912391660810493e-05, "loss": 0.0265, "step": 11576 }, { "epoch": 8.135628952916374, "grad_norm": 0.16125057637691498, "learning_rate": 2.7911923167018037e-05, "loss": 0.0277, "step": 11577 }, { "epoch": 8.13633169360506, "grad_norm": 0.2656441926956177, "learning_rate": 2.791145467322558e-05, "loss": 0.0489, "step": 11578 }, { "epoch": 8.137034434293746, "grad_norm": 0.3710891306400299, "learning_rate": 2.7910986179433124e-05, "loss": 0.0651, "step": 11579 }, { "epoch": 8.137737174982432, "grad_norm": 0.557820737361908, "learning_rate": 2.7910517685640665e-05, "loss": 0.0682, "step": 11580 }, { "epoch": 8.138439915671118, "grad_norm": 0.4826844036579132, "learning_rate": 2.791004919184821e-05, "loss": 0.1354, "step": 11581 }, { "epoch": 8.139142656359803, "grad_norm": 1.5087171792984009, "learning_rate": 2.7909580698055752e-05, "loss": 0.1417, "step": 11582 }, { "epoch": 8.13984539704849, "grad_norm": 1.3393595218658447, "learning_rate": 2.7909112204263296e-05, "loss": 0.1551, "step": 11583 }, { "epoch": 8.140548137737175, "grad_norm": 1.269771695137024, "learning_rate": 2.790864371047084e-05, "loss": 0.2353, "step": 11584 }, { "epoch": 8.141250878425861, "grad_norm": 0.4909808039665222, "learning_rate": 2.790817521667838e-05, "loss": 0.06, "step": 11585 }, { "epoch": 8.141953619114547, "grad_norm": 0.1242656484246254, "learning_rate": 2.7907706722885924e-05, "loss": 0.0197, "step": 11586 }, { "epoch": 8.142656359803233, "grad_norm": 0.09865598380565643, "learning_rate": 2.7907238229093464e-05, "loss": 0.0183, "step": 11587 }, { "epoch": 8.143359100491919, "grad_norm": 0.19075100123882294, "learning_rate": 2.7906769735301008e-05, "loss": 0.0179, "step": 11588 }, { "epoch": 8.144061841180605, "grad_norm": 0.1279066801071167, "learning_rate": 2.7906301241508548e-05, "loss": 0.0166, "step": 11589 }, { "epoch": 8.14476458186929, "grad_norm": 0.3114076554775238, "learning_rate": 2.7905832747716092e-05, "loss": 0.0132, "step": 11590 }, { "epoch": 8.145467322557977, "grad_norm": 0.10890365391969681, "learning_rate": 2.7905364253923636e-05, "loss": 0.0177, "step": 11591 }, { "epoch": 8.146170063246663, "grad_norm": 0.1919460892677307, "learning_rate": 2.790489576013118e-05, "loss": 0.0319, "step": 11592 }, { "epoch": 8.146872803935349, "grad_norm": 0.15935513377189636, "learning_rate": 2.790442726633872e-05, "loss": 0.0211, "step": 11593 }, { "epoch": 8.147575544624035, "grad_norm": 0.08037934452295303, "learning_rate": 2.7903958772546263e-05, "loss": 0.0075, "step": 11594 }, { "epoch": 8.14827828531272, "grad_norm": 0.14594896137714386, "learning_rate": 2.7903490278753807e-05, "loss": 0.0187, "step": 11595 }, { "epoch": 8.148981026001405, "grad_norm": 0.1833915263414383, "learning_rate": 2.790302178496135e-05, "loss": 0.0162, "step": 11596 }, { "epoch": 8.14968376669009, "grad_norm": 0.20874933898448944, "learning_rate": 2.7902553291168895e-05, "loss": 0.0385, "step": 11597 }, { "epoch": 8.150386507378776, "grad_norm": 0.21217374503612518, "learning_rate": 2.7902084797376435e-05, "loss": 0.0255, "step": 11598 }, { "epoch": 8.151089248067462, "grad_norm": 0.2298794686794281, "learning_rate": 2.790161630358398e-05, "loss": 0.0274, "step": 11599 }, { "epoch": 8.151791988756148, "grad_norm": 0.2430797666311264, "learning_rate": 2.7901147809791522e-05, "loss": 0.0453, "step": 11600 }, { "epoch": 8.152494729444834, "grad_norm": 0.14947083592414856, "learning_rate": 2.7900679315999066e-05, "loss": 0.0209, "step": 11601 }, { "epoch": 8.15319747013352, "grad_norm": 0.18552415072917938, "learning_rate": 2.7900210822206606e-05, "loss": 0.0343, "step": 11602 }, { "epoch": 8.153900210822206, "grad_norm": 0.45588812232017517, "learning_rate": 2.789974232841415e-05, "loss": 0.0572, "step": 11603 }, { "epoch": 8.154602951510892, "grad_norm": 0.3445875942707062, "learning_rate": 2.7899273834621694e-05, "loss": 0.0458, "step": 11604 }, { "epoch": 8.155305692199578, "grad_norm": 0.8069422841072083, "learning_rate": 2.7898805340829234e-05, "loss": 0.0682, "step": 11605 }, { "epoch": 8.156008432888264, "grad_norm": 0.6625093817710876, "learning_rate": 2.7898336847036774e-05, "loss": 0.1352, "step": 11606 }, { "epoch": 8.15671117357695, "grad_norm": 1.4174087047576904, "learning_rate": 2.7897868353244318e-05, "loss": 0.1675, "step": 11607 }, { "epoch": 8.157413914265636, "grad_norm": 1.4493324756622314, "learning_rate": 2.7897399859451862e-05, "loss": 0.1958, "step": 11608 }, { "epoch": 8.158116654954322, "grad_norm": 0.6167373657226562, "learning_rate": 2.7896931365659406e-05, "loss": 0.1784, "step": 11609 }, { "epoch": 8.158819395643008, "grad_norm": 0.33197274804115295, "learning_rate": 2.789646287186695e-05, "loss": 0.0816, "step": 11610 }, { "epoch": 8.159522136331693, "grad_norm": 0.10827666521072388, "learning_rate": 2.789599437807449e-05, "loss": 0.0187, "step": 11611 }, { "epoch": 8.16022487702038, "grad_norm": 0.14751677215099335, "learning_rate": 2.7895525884282033e-05, "loss": 0.0255, "step": 11612 }, { "epoch": 8.160927617709065, "grad_norm": 0.11561665683984756, "learning_rate": 2.7895057390489577e-05, "loss": 0.0126, "step": 11613 }, { "epoch": 8.161630358397751, "grad_norm": 0.1519162952899933, "learning_rate": 2.789458889669712e-05, "loss": 0.0248, "step": 11614 }, { "epoch": 8.162333099086437, "grad_norm": 0.08687417954206467, "learning_rate": 2.789412040290466e-05, "loss": 0.011, "step": 11615 }, { "epoch": 8.163035839775123, "grad_norm": 0.12288597971200943, "learning_rate": 2.7893651909112205e-05, "loss": 0.0104, "step": 11616 }, { "epoch": 8.163738580463809, "grad_norm": 0.1553061157464981, "learning_rate": 2.789318341531975e-05, "loss": 0.0158, "step": 11617 }, { "epoch": 8.164441321152495, "grad_norm": 0.2101469784975052, "learning_rate": 2.7892714921527292e-05, "loss": 0.0263, "step": 11618 }, { "epoch": 8.16514406184118, "grad_norm": 0.13721658289432526, "learning_rate": 2.7892246427734833e-05, "loss": 0.022, "step": 11619 }, { "epoch": 8.165846802529867, "grad_norm": 0.23276089131832123, "learning_rate": 2.7891777933942376e-05, "loss": 0.0246, "step": 11620 }, { "epoch": 8.166549543218553, "grad_norm": 0.38424843549728394, "learning_rate": 2.789130944014992e-05, "loss": 0.0144, "step": 11621 }, { "epoch": 8.167252283907239, "grad_norm": 0.12344104796648026, "learning_rate": 2.789084094635746e-05, "loss": 0.0287, "step": 11622 }, { "epoch": 8.167955024595924, "grad_norm": 0.2856539189815521, "learning_rate": 2.7890372452565004e-05, "loss": 0.0181, "step": 11623 }, { "epoch": 8.16865776528461, "grad_norm": 0.18917180597782135, "learning_rate": 2.7889903958772545e-05, "loss": 0.0387, "step": 11624 }, { "epoch": 8.169360505973296, "grad_norm": 0.32357603311538696, "learning_rate": 2.788943546498009e-05, "loss": 0.0627, "step": 11625 }, { "epoch": 8.170063246661982, "grad_norm": 0.21810154616832733, "learning_rate": 2.7888966971187632e-05, "loss": 0.0311, "step": 11626 }, { "epoch": 8.170765987350668, "grad_norm": 0.2880038917064667, "learning_rate": 2.7888498477395176e-05, "loss": 0.0208, "step": 11627 }, { "epoch": 8.171468728039354, "grad_norm": 0.32305699586868286, "learning_rate": 2.7888029983602716e-05, "loss": 0.0589, "step": 11628 }, { "epoch": 8.17217146872804, "grad_norm": 0.3084862530231476, "learning_rate": 2.788756148981026e-05, "loss": 0.06, "step": 11629 }, { "epoch": 8.172874209416726, "grad_norm": 0.5158478617668152, "learning_rate": 2.7887092996017804e-05, "loss": 0.1025, "step": 11630 }, { "epoch": 8.173576950105412, "grad_norm": 2.3538365364074707, "learning_rate": 2.7886624502225347e-05, "loss": 0.1146, "step": 11631 }, { "epoch": 8.174279690794098, "grad_norm": 0.4831637740135193, "learning_rate": 2.7886156008432888e-05, "loss": 0.1716, "step": 11632 }, { "epoch": 8.174982431482784, "grad_norm": 0.7298352718353271, "learning_rate": 2.788568751464043e-05, "loss": 0.1729, "step": 11633 }, { "epoch": 8.17568517217147, "grad_norm": 1.8077912330627441, "learning_rate": 2.7885219020847975e-05, "loss": 0.2109, "step": 11634 }, { "epoch": 8.176387912860154, "grad_norm": 0.3759428560733795, "learning_rate": 2.788475052705552e-05, "loss": 0.0898, "step": 11635 }, { "epoch": 8.17709065354884, "grad_norm": 0.3966875672340393, "learning_rate": 2.7884282033263063e-05, "loss": 0.0236, "step": 11636 }, { "epoch": 8.177793394237526, "grad_norm": 0.10857966542243958, "learning_rate": 2.7883813539470603e-05, "loss": 0.0242, "step": 11637 }, { "epoch": 8.178496134926212, "grad_norm": 0.10032939910888672, "learning_rate": 2.7883345045678147e-05, "loss": 0.0126, "step": 11638 }, { "epoch": 8.179198875614897, "grad_norm": 0.1911817491054535, "learning_rate": 2.788287655188569e-05, "loss": 0.0182, "step": 11639 }, { "epoch": 8.179901616303583, "grad_norm": 0.2264518439769745, "learning_rate": 2.788240805809323e-05, "loss": 0.0241, "step": 11640 }, { "epoch": 8.18060435699227, "grad_norm": 0.17151720821857452, "learning_rate": 2.788193956430077e-05, "loss": 0.0203, "step": 11641 }, { "epoch": 8.181307097680955, "grad_norm": 0.15578067302703857, "learning_rate": 2.7881471070508315e-05, "loss": 0.0117, "step": 11642 }, { "epoch": 8.182009838369641, "grad_norm": 0.34819966554641724, "learning_rate": 2.788100257671586e-05, "loss": 0.0283, "step": 11643 }, { "epoch": 8.182712579058327, "grad_norm": 0.2284715175628662, "learning_rate": 2.7880534082923402e-05, "loss": 0.0175, "step": 11644 }, { "epoch": 8.183415319747013, "grad_norm": 0.8128102421760559, "learning_rate": 2.7880065589130943e-05, "loss": 0.0387, "step": 11645 }, { "epoch": 8.184118060435699, "grad_norm": 0.11006270349025726, "learning_rate": 2.7879597095338486e-05, "loss": 0.0122, "step": 11646 }, { "epoch": 8.184820801124385, "grad_norm": 0.20876720547676086, "learning_rate": 2.787912860154603e-05, "loss": 0.0267, "step": 11647 }, { "epoch": 8.18552354181307, "grad_norm": 0.4410187005996704, "learning_rate": 2.7878660107753574e-05, "loss": 0.0335, "step": 11648 }, { "epoch": 8.186226282501757, "grad_norm": 0.3234148621559143, "learning_rate": 2.7878191613961117e-05, "loss": 0.0348, "step": 11649 }, { "epoch": 8.186929023190443, "grad_norm": 0.18199506402015686, "learning_rate": 2.7877723120168658e-05, "loss": 0.0266, "step": 11650 }, { "epoch": 8.187631763879128, "grad_norm": 0.323959082365036, "learning_rate": 2.78772546263762e-05, "loss": 0.0284, "step": 11651 }, { "epoch": 8.188334504567814, "grad_norm": 0.2765852212905884, "learning_rate": 2.7876786132583745e-05, "loss": 0.0434, "step": 11652 }, { "epoch": 8.1890372452565, "grad_norm": 0.26669391989707947, "learning_rate": 2.787631763879129e-05, "loss": 0.0414, "step": 11653 }, { "epoch": 8.189739985945186, "grad_norm": 0.47778502106666565, "learning_rate": 2.787584914499883e-05, "loss": 0.0514, "step": 11654 }, { "epoch": 8.190442726633872, "grad_norm": 0.48254668712615967, "learning_rate": 2.7875380651206373e-05, "loss": 0.0933, "step": 11655 }, { "epoch": 8.191145467322558, "grad_norm": 0.4983602464199066, "learning_rate": 2.7874912157413917e-05, "loss": 0.1023, "step": 11656 }, { "epoch": 8.191848208011244, "grad_norm": 0.9296179413795471, "learning_rate": 2.7874443663621457e-05, "loss": 0.1557, "step": 11657 }, { "epoch": 8.19255094869993, "grad_norm": 0.9913389086723328, "learning_rate": 2.7873975169828997e-05, "loss": 0.2065, "step": 11658 }, { "epoch": 8.193253689388616, "grad_norm": 6.838597774505615, "learning_rate": 2.787350667603654e-05, "loss": 0.2212, "step": 11659 }, { "epoch": 8.193956430077302, "grad_norm": 0.269368439912796, "learning_rate": 2.7873038182244085e-05, "loss": 0.0655, "step": 11660 }, { "epoch": 8.194659170765988, "grad_norm": 0.2084011733531952, "learning_rate": 2.787256968845163e-05, "loss": 0.0234, "step": 11661 }, { "epoch": 8.195361911454674, "grad_norm": 0.13443419337272644, "learning_rate": 2.7872101194659172e-05, "loss": 0.0172, "step": 11662 }, { "epoch": 8.19606465214336, "grad_norm": 0.16410517692565918, "learning_rate": 2.7871632700866713e-05, "loss": 0.0264, "step": 11663 }, { "epoch": 8.196767392832045, "grad_norm": 0.13599395751953125, "learning_rate": 2.7871164207074256e-05, "loss": 0.0161, "step": 11664 }, { "epoch": 8.197470133520731, "grad_norm": 0.15338248014450073, "learning_rate": 2.78706957132818e-05, "loss": 0.0196, "step": 11665 }, { "epoch": 8.198172874209417, "grad_norm": 0.13762998580932617, "learning_rate": 2.7870227219489344e-05, "loss": 0.0144, "step": 11666 }, { "epoch": 8.198875614898103, "grad_norm": 0.1890782117843628, "learning_rate": 2.7869758725696884e-05, "loss": 0.0285, "step": 11667 }, { "epoch": 8.19957835558679, "grad_norm": 0.42580491304397583, "learning_rate": 2.7869290231904428e-05, "loss": 0.0221, "step": 11668 }, { "epoch": 8.200281096275475, "grad_norm": 0.12016790360212326, "learning_rate": 2.786882173811197e-05, "loss": 0.0152, "step": 11669 }, { "epoch": 8.200983836964161, "grad_norm": 0.36470547318458557, "learning_rate": 2.7868353244319515e-05, "loss": 0.0277, "step": 11670 }, { "epoch": 8.201686577652847, "grad_norm": 0.15952293574810028, "learning_rate": 2.7867884750527056e-05, "loss": 0.0167, "step": 11671 }, { "epoch": 8.202389318341533, "grad_norm": 0.4012382924556732, "learning_rate": 2.78674162567346e-05, "loss": 0.0279, "step": 11672 }, { "epoch": 8.203092059030217, "grad_norm": 0.2098003625869751, "learning_rate": 2.7866947762942143e-05, "loss": 0.0299, "step": 11673 }, { "epoch": 8.203794799718903, "grad_norm": 0.21644090116024017, "learning_rate": 2.7866479269149687e-05, "loss": 0.0272, "step": 11674 }, { "epoch": 8.204497540407589, "grad_norm": 0.1839359998703003, "learning_rate": 2.7866010775357227e-05, "loss": 0.0219, "step": 11675 }, { "epoch": 8.205200281096275, "grad_norm": 0.18968768417835236, "learning_rate": 2.7865542281564767e-05, "loss": 0.0211, "step": 11676 }, { "epoch": 8.20590302178496, "grad_norm": 0.4934930205345154, "learning_rate": 2.786507378777231e-05, "loss": 0.0469, "step": 11677 }, { "epoch": 8.206605762473647, "grad_norm": 0.4802268445491791, "learning_rate": 2.7864605293979855e-05, "loss": 0.0469, "step": 11678 }, { "epoch": 8.207308503162333, "grad_norm": 0.24701127409934998, "learning_rate": 2.78641368001874e-05, "loss": 0.0541, "step": 11679 }, { "epoch": 8.208011243851018, "grad_norm": 0.5908156037330627, "learning_rate": 2.786366830639494e-05, "loss": 0.0897, "step": 11680 }, { "epoch": 8.208713984539704, "grad_norm": 0.4169710576534271, "learning_rate": 2.7863199812602483e-05, "loss": 0.1097, "step": 11681 }, { "epoch": 8.20941672522839, "grad_norm": 0.8108794689178467, "learning_rate": 2.7862731318810026e-05, "loss": 0.1523, "step": 11682 }, { "epoch": 8.210119465917076, "grad_norm": 0.8574408292770386, "learning_rate": 2.786226282501757e-05, "loss": 0.1804, "step": 11683 }, { "epoch": 8.210822206605762, "grad_norm": 2.0513083934783936, "learning_rate": 2.7861794331225114e-05, "loss": 0.2194, "step": 11684 }, { "epoch": 8.211524947294448, "grad_norm": 0.2908778190612793, "learning_rate": 2.7861325837432654e-05, "loss": 0.0824, "step": 11685 }, { "epoch": 8.212227687983134, "grad_norm": 0.10681681334972382, "learning_rate": 2.7860857343640198e-05, "loss": 0.0184, "step": 11686 }, { "epoch": 8.21293042867182, "grad_norm": 0.17172574996948242, "learning_rate": 2.7860388849847742e-05, "loss": 0.0166, "step": 11687 }, { "epoch": 8.213633169360506, "grad_norm": 0.26070913672447205, "learning_rate": 2.7859920356055285e-05, "loss": 0.026, "step": 11688 }, { "epoch": 8.214335910049192, "grad_norm": 0.14918582141399384, "learning_rate": 2.7859451862262826e-05, "loss": 0.0149, "step": 11689 }, { "epoch": 8.215038650737878, "grad_norm": 0.1806170791387558, "learning_rate": 2.785898336847037e-05, "loss": 0.0246, "step": 11690 }, { "epoch": 8.215741391426564, "grad_norm": 0.13324260711669922, "learning_rate": 2.7858514874677913e-05, "loss": 0.0202, "step": 11691 }, { "epoch": 8.21644413211525, "grad_norm": 0.20256638526916504, "learning_rate": 2.7858046380885454e-05, "loss": 0.0276, "step": 11692 }, { "epoch": 8.217146872803935, "grad_norm": 0.20115971565246582, "learning_rate": 2.7857577887092994e-05, "loss": 0.019, "step": 11693 }, { "epoch": 8.217849613492621, "grad_norm": 0.25836965441703796, "learning_rate": 2.7857109393300538e-05, "loss": 0.0212, "step": 11694 }, { "epoch": 8.218552354181307, "grad_norm": 0.1260327398777008, "learning_rate": 2.785664089950808e-05, "loss": 0.015, "step": 11695 }, { "epoch": 8.219255094869993, "grad_norm": 0.12885242700576782, "learning_rate": 2.7856172405715625e-05, "loss": 0.0284, "step": 11696 }, { "epoch": 8.219957835558679, "grad_norm": 0.24971796572208405, "learning_rate": 2.785570391192317e-05, "loss": 0.0419, "step": 11697 }, { "epoch": 8.220660576247365, "grad_norm": 0.13995343446731567, "learning_rate": 2.785523541813071e-05, "loss": 0.0204, "step": 11698 }, { "epoch": 8.221363316936051, "grad_norm": 0.20654499530792236, "learning_rate": 2.7854766924338253e-05, "loss": 0.0165, "step": 11699 }, { "epoch": 8.222066057624737, "grad_norm": 0.1789635717868805, "learning_rate": 2.7854298430545797e-05, "loss": 0.0263, "step": 11700 }, { "epoch": 8.222768798313423, "grad_norm": 0.23417553305625916, "learning_rate": 2.785382993675334e-05, "loss": 0.0384, "step": 11701 }, { "epoch": 8.223471539002109, "grad_norm": 0.3472510874271393, "learning_rate": 2.785336144296088e-05, "loss": 0.0597, "step": 11702 }, { "epoch": 8.224174279690795, "grad_norm": 0.31219482421875, "learning_rate": 2.7852892949168424e-05, "loss": 0.0455, "step": 11703 }, { "epoch": 8.22487702037948, "grad_norm": 0.2572139799594879, "learning_rate": 2.7852424455375968e-05, "loss": 0.033, "step": 11704 }, { "epoch": 8.225579761068166, "grad_norm": 0.4079623520374298, "learning_rate": 2.7851955961583512e-05, "loss": 0.1047, "step": 11705 }, { "epoch": 8.226282501756852, "grad_norm": 0.6055755615234375, "learning_rate": 2.7851487467791052e-05, "loss": 0.139, "step": 11706 }, { "epoch": 8.226985242445538, "grad_norm": 0.47529837489128113, "learning_rate": 2.7851018973998596e-05, "loss": 0.1524, "step": 11707 }, { "epoch": 8.227687983134224, "grad_norm": 0.8849090933799744, "learning_rate": 2.785055048020614e-05, "loss": 0.2086, "step": 11708 }, { "epoch": 8.22839072382291, "grad_norm": 1.2985252141952515, "learning_rate": 2.785008198641368e-05, "loss": 0.2136, "step": 11709 }, { "epoch": 8.229093464511596, "grad_norm": 0.3934968411922455, "learning_rate": 2.7849613492621224e-05, "loss": 0.1007, "step": 11710 }, { "epoch": 8.22979620520028, "grad_norm": 0.301439106464386, "learning_rate": 2.7849144998828764e-05, "loss": 0.0465, "step": 11711 }, { "epoch": 8.230498945888966, "grad_norm": 0.14906489849090576, "learning_rate": 2.7848676505036308e-05, "loss": 0.0315, "step": 11712 }, { "epoch": 8.231201686577652, "grad_norm": 0.30606570839881897, "learning_rate": 2.784820801124385e-05, "loss": 0.0095, "step": 11713 }, { "epoch": 8.231904427266338, "grad_norm": 0.1731240302324295, "learning_rate": 2.7847739517451395e-05, "loss": 0.017, "step": 11714 }, { "epoch": 8.232607167955024, "grad_norm": 0.16558684408664703, "learning_rate": 2.7847271023658936e-05, "loss": 0.0134, "step": 11715 }, { "epoch": 8.23330990864371, "grad_norm": 0.5767574906349182, "learning_rate": 2.784680252986648e-05, "loss": 0.0272, "step": 11716 }, { "epoch": 8.234012649332396, "grad_norm": 0.45692792534828186, "learning_rate": 2.7846334036074023e-05, "loss": 0.0143, "step": 11717 }, { "epoch": 8.234715390021082, "grad_norm": 0.11351250112056732, "learning_rate": 2.7845865542281567e-05, "loss": 0.0181, "step": 11718 }, { "epoch": 8.235418130709768, "grad_norm": 0.12892143428325653, "learning_rate": 2.7845397048489107e-05, "loss": 0.012, "step": 11719 }, { "epoch": 8.236120871398454, "grad_norm": 0.3003205955028534, "learning_rate": 2.784492855469665e-05, "loss": 0.0321, "step": 11720 }, { "epoch": 8.23682361208714, "grad_norm": 0.12298636883497238, "learning_rate": 2.7844460060904194e-05, "loss": 0.0168, "step": 11721 }, { "epoch": 8.237526352775825, "grad_norm": 0.4820932149887085, "learning_rate": 2.7843991567111738e-05, "loss": 0.0272, "step": 11722 }, { "epoch": 8.238229093464511, "grad_norm": 0.12740278244018555, "learning_rate": 2.7843523073319282e-05, "loss": 0.0138, "step": 11723 }, { "epoch": 8.238931834153197, "grad_norm": 0.37576964497566223, "learning_rate": 2.7843054579526822e-05, "loss": 0.0331, "step": 11724 }, { "epoch": 8.239634574841883, "grad_norm": 0.19046756625175476, "learning_rate": 2.7842586085734366e-05, "loss": 0.0231, "step": 11725 }, { "epoch": 8.240337315530569, "grad_norm": 0.7929025888442993, "learning_rate": 2.784211759194191e-05, "loss": 0.0553, "step": 11726 }, { "epoch": 8.241040056219255, "grad_norm": 0.35854774713516235, "learning_rate": 2.784164909814945e-05, "loss": 0.0156, "step": 11727 }, { "epoch": 8.24174279690794, "grad_norm": 0.4813508987426758, "learning_rate": 2.784118060435699e-05, "loss": 0.0668, "step": 11728 }, { "epoch": 8.242445537596627, "grad_norm": 0.4090951681137085, "learning_rate": 2.7840712110564534e-05, "loss": 0.0677, "step": 11729 }, { "epoch": 8.243148278285313, "grad_norm": 0.6520717740058899, "learning_rate": 2.7840243616772078e-05, "loss": 0.0731, "step": 11730 }, { "epoch": 8.243851018973999, "grad_norm": 0.4516896903514862, "learning_rate": 2.783977512297962e-05, "loss": 0.0814, "step": 11731 }, { "epoch": 8.244553759662685, "grad_norm": 0.94002366065979, "learning_rate": 2.7839306629187162e-05, "loss": 0.1607, "step": 11732 }, { "epoch": 8.24525650035137, "grad_norm": 1.0173966884613037, "learning_rate": 2.7838838135394706e-05, "loss": 0.1857, "step": 11733 }, { "epoch": 8.245959241040056, "grad_norm": 1.290103554725647, "learning_rate": 2.783836964160225e-05, "loss": 0.2199, "step": 11734 }, { "epoch": 8.246661981728742, "grad_norm": 0.3306066691875458, "learning_rate": 2.7837901147809793e-05, "loss": 0.0555, "step": 11735 }, { "epoch": 8.247364722417428, "grad_norm": 0.5041919946670532, "learning_rate": 2.7837432654017337e-05, "loss": 0.037, "step": 11736 }, { "epoch": 8.248067463106114, "grad_norm": 0.1980142742395401, "learning_rate": 2.7836964160224877e-05, "loss": 0.0243, "step": 11737 }, { "epoch": 8.2487702037948, "grad_norm": 0.13302811980247498, "learning_rate": 2.783649566643242e-05, "loss": 0.0147, "step": 11738 }, { "epoch": 8.249472944483486, "grad_norm": 0.1433946043252945, "learning_rate": 2.7836027172639965e-05, "loss": 0.0362, "step": 11739 }, { "epoch": 8.250175685172172, "grad_norm": 0.1937677562236786, "learning_rate": 2.783555867884751e-05, "loss": 0.0149, "step": 11740 }, { "epoch": 8.250878425860858, "grad_norm": 0.07920946180820465, "learning_rate": 2.783509018505505e-05, "loss": 0.0103, "step": 11741 }, { "epoch": 8.251581166549544, "grad_norm": 0.1442558914422989, "learning_rate": 2.7834621691262592e-05, "loss": 0.0199, "step": 11742 }, { "epoch": 8.25228390723823, "grad_norm": 0.2847009599208832, "learning_rate": 2.7834153197470136e-05, "loss": 0.0139, "step": 11743 }, { "epoch": 8.252986647926916, "grad_norm": 0.260717511177063, "learning_rate": 2.7833684703677676e-05, "loss": 0.0221, "step": 11744 }, { "epoch": 8.253689388615602, "grad_norm": 0.29466715455055237, "learning_rate": 2.7833216209885217e-05, "loss": 0.048, "step": 11745 }, { "epoch": 8.254392129304287, "grad_norm": 0.13746513426303864, "learning_rate": 2.783274771609276e-05, "loss": 0.0174, "step": 11746 }, { "epoch": 8.255094869992973, "grad_norm": 0.19526022672653198, "learning_rate": 2.7832279222300304e-05, "loss": 0.0322, "step": 11747 }, { "epoch": 8.25579761068166, "grad_norm": 0.18969331681728363, "learning_rate": 2.7831810728507848e-05, "loss": 0.0159, "step": 11748 }, { "epoch": 8.256500351370345, "grad_norm": 0.16146984696388245, "learning_rate": 2.783134223471539e-05, "loss": 0.0288, "step": 11749 }, { "epoch": 8.25720309205903, "grad_norm": 0.4330422580242157, "learning_rate": 2.7830873740922932e-05, "loss": 0.0462, "step": 11750 }, { "epoch": 8.257905832747715, "grad_norm": 0.18083158135414124, "learning_rate": 2.7830405247130476e-05, "loss": 0.02, "step": 11751 }, { "epoch": 8.258608573436401, "grad_norm": 0.2048380821943283, "learning_rate": 2.782993675333802e-05, "loss": 0.04, "step": 11752 }, { "epoch": 8.259311314125087, "grad_norm": 0.3460836708545685, "learning_rate": 2.7829468259545563e-05, "loss": 0.0443, "step": 11753 }, { "epoch": 8.260014054813773, "grad_norm": 0.6711624264717102, "learning_rate": 2.7828999765753104e-05, "loss": 0.0655, "step": 11754 }, { "epoch": 8.260716795502459, "grad_norm": 0.7752638459205627, "learning_rate": 2.7828531271960647e-05, "loss": 0.0762, "step": 11755 }, { "epoch": 8.261419536191145, "grad_norm": 0.668938159942627, "learning_rate": 2.782806277816819e-05, "loss": 0.1053, "step": 11756 }, { "epoch": 8.26212227687983, "grad_norm": 0.9724336862564087, "learning_rate": 2.7827594284375735e-05, "loss": 0.1636, "step": 11757 }, { "epoch": 8.262825017568517, "grad_norm": 0.9019026756286621, "learning_rate": 2.7827125790583275e-05, "loss": 0.2088, "step": 11758 }, { "epoch": 8.263527758257203, "grad_norm": 1.6010985374450684, "learning_rate": 2.782665729679082e-05, "loss": 0.2359, "step": 11759 }, { "epoch": 8.264230498945889, "grad_norm": 0.237163707613945, "learning_rate": 2.7826188802998362e-05, "loss": 0.072, "step": 11760 }, { "epoch": 8.264933239634574, "grad_norm": 0.13932694494724274, "learning_rate": 2.7825720309205906e-05, "loss": 0.0243, "step": 11761 }, { "epoch": 8.26563598032326, "grad_norm": 0.20902307331562042, "learning_rate": 2.7825251815413447e-05, "loss": 0.0212, "step": 11762 }, { "epoch": 8.266338721011946, "grad_norm": 0.18120348453521729, "learning_rate": 2.7824783321620987e-05, "loss": 0.0274, "step": 11763 }, { "epoch": 8.267041461700632, "grad_norm": 0.2213527411222458, "learning_rate": 2.782431482782853e-05, "loss": 0.0269, "step": 11764 }, { "epoch": 8.267744202389318, "grad_norm": 0.23271101713180542, "learning_rate": 2.7823846334036074e-05, "loss": 0.0152, "step": 11765 }, { "epoch": 8.268446943078004, "grad_norm": 0.11808762699365616, "learning_rate": 2.7823377840243618e-05, "loss": 0.0072, "step": 11766 }, { "epoch": 8.26914968376669, "grad_norm": 0.43017125129699707, "learning_rate": 2.782290934645116e-05, "loss": 0.0371, "step": 11767 }, { "epoch": 8.269852424455376, "grad_norm": 0.14657893776893616, "learning_rate": 2.7822440852658702e-05, "loss": 0.0217, "step": 11768 }, { "epoch": 8.270555165144062, "grad_norm": 0.09080428630113602, "learning_rate": 2.7821972358866246e-05, "loss": 0.0092, "step": 11769 }, { "epoch": 8.271257905832748, "grad_norm": 0.22384604811668396, "learning_rate": 2.782150386507379e-05, "loss": 0.0286, "step": 11770 }, { "epoch": 8.271960646521434, "grad_norm": 0.21846109628677368, "learning_rate": 2.782103537128133e-05, "loss": 0.0155, "step": 11771 }, { "epoch": 8.27266338721012, "grad_norm": 0.33389991521835327, "learning_rate": 2.7820566877488874e-05, "loss": 0.0239, "step": 11772 }, { "epoch": 8.273366127898806, "grad_norm": 0.1389075666666031, "learning_rate": 2.7820098383696417e-05, "loss": 0.0157, "step": 11773 }, { "epoch": 8.274068868587491, "grad_norm": 0.5978154540061951, "learning_rate": 2.781962988990396e-05, "loss": 0.0504, "step": 11774 }, { "epoch": 8.274771609276177, "grad_norm": 0.2224135398864746, "learning_rate": 2.7819161396111505e-05, "loss": 0.0382, "step": 11775 }, { "epoch": 8.275474349964863, "grad_norm": 0.39628273248672485, "learning_rate": 2.7818692902319045e-05, "loss": 0.025, "step": 11776 }, { "epoch": 8.27617709065355, "grad_norm": 0.8777719736099243, "learning_rate": 2.781822440852659e-05, "loss": 0.0402, "step": 11777 }, { "epoch": 8.276879831342235, "grad_norm": 0.20812572538852692, "learning_rate": 2.7817755914734133e-05, "loss": 0.0374, "step": 11778 }, { "epoch": 8.277582572030921, "grad_norm": 1.9428492784500122, "learning_rate": 2.7817287420941673e-05, "loss": 0.0612, "step": 11779 }, { "epoch": 8.278285312719607, "grad_norm": 1.032944917678833, "learning_rate": 2.7816818927149213e-05, "loss": 0.0661, "step": 11780 }, { "epoch": 8.278988053408293, "grad_norm": 0.5966817140579224, "learning_rate": 2.7816350433356757e-05, "loss": 0.1115, "step": 11781 }, { "epoch": 8.279690794096979, "grad_norm": 1.0228956937789917, "learning_rate": 2.78158819395643e-05, "loss": 0.1841, "step": 11782 }, { "epoch": 8.280393534785665, "grad_norm": 2.0883548259735107, "learning_rate": 2.7815413445771844e-05, "loss": 0.2141, "step": 11783 }, { "epoch": 8.28109627547435, "grad_norm": 1.5813332796096802, "learning_rate": 2.7814944951979385e-05, "loss": 0.1976, "step": 11784 }, { "epoch": 8.281799016163037, "grad_norm": 0.28971049189567566, "learning_rate": 2.781447645818693e-05, "loss": 0.0831, "step": 11785 }, { "epoch": 8.282501756851723, "grad_norm": 0.15191179513931274, "learning_rate": 2.7814007964394472e-05, "loss": 0.0197, "step": 11786 }, { "epoch": 8.283204497540408, "grad_norm": 0.17613455653190613, "learning_rate": 2.7813539470602016e-05, "loss": 0.0279, "step": 11787 }, { "epoch": 8.283907238229094, "grad_norm": 0.15798017382621765, "learning_rate": 2.781307097680956e-05, "loss": 0.0253, "step": 11788 }, { "epoch": 8.284609978917779, "grad_norm": 0.1384691745042801, "learning_rate": 2.78126024830171e-05, "loss": 0.0222, "step": 11789 }, { "epoch": 8.285312719606464, "grad_norm": 0.13566572964191437, "learning_rate": 2.7812133989224644e-05, "loss": 0.0122, "step": 11790 }, { "epoch": 8.28601546029515, "grad_norm": 0.12655480206012726, "learning_rate": 2.7811665495432187e-05, "loss": 0.0129, "step": 11791 }, { "epoch": 8.286718200983836, "grad_norm": 0.3847452998161316, "learning_rate": 2.781119700163973e-05, "loss": 0.0263, "step": 11792 }, { "epoch": 8.287420941672522, "grad_norm": 0.48072072863578796, "learning_rate": 2.781072850784727e-05, "loss": 0.0397, "step": 11793 }, { "epoch": 8.288123682361208, "grad_norm": 0.14368189871311188, "learning_rate": 2.7810260014054815e-05, "loss": 0.0123, "step": 11794 }, { "epoch": 8.288826423049894, "grad_norm": 0.16527889668941498, "learning_rate": 2.780979152026236e-05, "loss": 0.0211, "step": 11795 }, { "epoch": 8.28952916373858, "grad_norm": 0.12418047338724136, "learning_rate": 2.78093230264699e-05, "loss": 0.0121, "step": 11796 }, { "epoch": 8.290231904427266, "grad_norm": 0.18298748135566711, "learning_rate": 2.780885453267744e-05, "loss": 0.0276, "step": 11797 }, { "epoch": 8.290934645115952, "grad_norm": 0.11820043623447418, "learning_rate": 2.7808386038884983e-05, "loss": 0.0222, "step": 11798 }, { "epoch": 8.291637385804638, "grad_norm": 0.1717178374528885, "learning_rate": 2.7807917545092527e-05, "loss": 0.0544, "step": 11799 }, { "epoch": 8.292340126493324, "grad_norm": 0.23957277834415436, "learning_rate": 2.780744905130007e-05, "loss": 0.0388, "step": 11800 }, { "epoch": 8.29304286718201, "grad_norm": 0.21701738238334656, "learning_rate": 2.7806980557507615e-05, "loss": 0.0328, "step": 11801 }, { "epoch": 8.293745607870695, "grad_norm": 0.2916702628135681, "learning_rate": 2.7806512063715155e-05, "loss": 0.0303, "step": 11802 }, { "epoch": 8.294448348559381, "grad_norm": 0.28879958391189575, "learning_rate": 2.78060435699227e-05, "loss": 0.0482, "step": 11803 }, { "epoch": 8.295151089248067, "grad_norm": 0.2813372313976288, "learning_rate": 2.7805575076130242e-05, "loss": 0.0516, "step": 11804 }, { "epoch": 8.295853829936753, "grad_norm": 0.5026065707206726, "learning_rate": 2.7805106582337786e-05, "loss": 0.0985, "step": 11805 }, { "epoch": 8.29655657062544, "grad_norm": 0.8564096689224243, "learning_rate": 2.7804638088545326e-05, "loss": 0.1128, "step": 11806 }, { "epoch": 8.297259311314125, "grad_norm": 1.1736992597579956, "learning_rate": 2.780416959475287e-05, "loss": 0.1571, "step": 11807 }, { "epoch": 8.297962052002811, "grad_norm": 0.9408254027366638, "learning_rate": 2.7803701100960414e-05, "loss": 0.2076, "step": 11808 }, { "epoch": 8.298664792691497, "grad_norm": 2.5541810989379883, "learning_rate": 2.7803232607167958e-05, "loss": 0.2105, "step": 11809 }, { "epoch": 8.299367533380183, "grad_norm": 0.23094753921031952, "learning_rate": 2.7802764113375498e-05, "loss": 0.0577, "step": 11810 }, { "epoch": 8.300070274068869, "grad_norm": 0.09288421273231506, "learning_rate": 2.780229561958304e-05, "loss": 0.0215, "step": 11811 }, { "epoch": 8.300773014757555, "grad_norm": 0.33129507303237915, "learning_rate": 2.7801827125790585e-05, "loss": 0.0414, "step": 11812 }, { "epoch": 8.30147575544624, "grad_norm": 0.2526299059391022, "learning_rate": 2.780135863199813e-05, "loss": 0.0203, "step": 11813 }, { "epoch": 8.302178496134927, "grad_norm": 0.10500980913639069, "learning_rate": 2.780089013820567e-05, "loss": 0.0138, "step": 11814 }, { "epoch": 8.302881236823612, "grad_norm": 0.1131206676363945, "learning_rate": 2.780042164441321e-05, "loss": 0.0197, "step": 11815 }, { "epoch": 8.303583977512298, "grad_norm": 0.1915738433599472, "learning_rate": 2.7799953150620753e-05, "loss": 0.0183, "step": 11816 }, { "epoch": 8.304286718200984, "grad_norm": 0.0825902447104454, "learning_rate": 2.7799484656828297e-05, "loss": 0.0108, "step": 11817 }, { "epoch": 8.30498945888967, "grad_norm": 0.2010875940322876, "learning_rate": 2.779901616303584e-05, "loss": 0.0271, "step": 11818 }, { "epoch": 8.305692199578356, "grad_norm": 0.15544933080673218, "learning_rate": 2.779854766924338e-05, "loss": 0.0116, "step": 11819 }, { "epoch": 8.306394940267042, "grad_norm": 0.12733705341815948, "learning_rate": 2.7798079175450925e-05, "loss": 0.0183, "step": 11820 }, { "epoch": 8.307097680955728, "grad_norm": 0.1851630061864853, "learning_rate": 2.779761068165847e-05, "loss": 0.0292, "step": 11821 }, { "epoch": 8.307800421644414, "grad_norm": 0.47059258818626404, "learning_rate": 2.7797142187866012e-05, "loss": 0.0184, "step": 11822 }, { "epoch": 8.3085031623331, "grad_norm": 0.14111977815628052, "learning_rate": 2.7796673694073553e-05, "loss": 0.0201, "step": 11823 }, { "epoch": 8.309205903021786, "grad_norm": 0.15489228069782257, "learning_rate": 2.7796205200281097e-05, "loss": 0.0268, "step": 11824 }, { "epoch": 8.309908643710472, "grad_norm": 1.1840304136276245, "learning_rate": 2.779573670648864e-05, "loss": 0.0314, "step": 11825 }, { "epoch": 8.310611384399156, "grad_norm": 0.1624889075756073, "learning_rate": 2.7795268212696184e-05, "loss": 0.0181, "step": 11826 }, { "epoch": 8.311314125087842, "grad_norm": 0.4506847560405731, "learning_rate": 2.7794799718903728e-05, "loss": 0.0337, "step": 11827 }, { "epoch": 8.312016865776528, "grad_norm": 0.17597419023513794, "learning_rate": 2.7794331225111268e-05, "loss": 0.0349, "step": 11828 }, { "epoch": 8.312719606465214, "grad_norm": 0.318925678730011, "learning_rate": 2.7793862731318812e-05, "loss": 0.0392, "step": 11829 }, { "epoch": 8.3134223471539, "grad_norm": 0.25078925490379333, "learning_rate": 2.7793394237526355e-05, "loss": 0.0887, "step": 11830 }, { "epoch": 8.314125087842585, "grad_norm": 1.6212193965911865, "learning_rate": 2.7792925743733896e-05, "loss": 0.1024, "step": 11831 }, { "epoch": 8.314827828531271, "grad_norm": 0.9162732362747192, "learning_rate": 2.7792457249941436e-05, "loss": 0.1549, "step": 11832 }, { "epoch": 8.315530569219957, "grad_norm": 0.7574057579040527, "learning_rate": 2.779198875614898e-05, "loss": 0.2198, "step": 11833 }, { "epoch": 8.316233309908643, "grad_norm": 1.0233944654464722, "learning_rate": 2.7791520262356524e-05, "loss": 0.1984, "step": 11834 }, { "epoch": 8.316936050597329, "grad_norm": 0.3634797930717468, "learning_rate": 2.7791051768564067e-05, "loss": 0.0662, "step": 11835 }, { "epoch": 8.317638791286015, "grad_norm": 0.16311529278755188, "learning_rate": 2.7790583274771608e-05, "loss": 0.0184, "step": 11836 }, { "epoch": 8.318341531974701, "grad_norm": 0.10119526088237762, "learning_rate": 2.779011478097915e-05, "loss": 0.0168, "step": 11837 }, { "epoch": 8.319044272663387, "grad_norm": 0.14634038507938385, "learning_rate": 2.7789646287186695e-05, "loss": 0.0196, "step": 11838 }, { "epoch": 8.319747013352073, "grad_norm": 0.11282512545585632, "learning_rate": 2.778917779339424e-05, "loss": 0.0274, "step": 11839 }, { "epoch": 8.320449754040759, "grad_norm": 0.1104278489947319, "learning_rate": 2.7788709299601783e-05, "loss": 0.0099, "step": 11840 }, { "epoch": 8.321152494729445, "grad_norm": 0.23915080726146698, "learning_rate": 2.7788240805809323e-05, "loss": 0.0299, "step": 11841 }, { "epoch": 8.32185523541813, "grad_norm": 0.1912708580493927, "learning_rate": 2.7787772312016867e-05, "loss": 0.0215, "step": 11842 }, { "epoch": 8.322557976106816, "grad_norm": 0.16596120595932007, "learning_rate": 2.778730381822441e-05, "loss": 0.0159, "step": 11843 }, { "epoch": 8.323260716795502, "grad_norm": 0.08517550677061081, "learning_rate": 2.7786835324431954e-05, "loss": 0.011, "step": 11844 }, { "epoch": 8.323963457484188, "grad_norm": 0.20877651870250702, "learning_rate": 2.7786366830639494e-05, "loss": 0.0344, "step": 11845 }, { "epoch": 8.324666198172874, "grad_norm": 0.16463950276374817, "learning_rate": 2.7785898336847038e-05, "loss": 0.0138, "step": 11846 }, { "epoch": 8.32536893886156, "grad_norm": 0.8538793325424194, "learning_rate": 2.7785429843054582e-05, "loss": 0.0345, "step": 11847 }, { "epoch": 8.326071679550246, "grad_norm": 0.3944620192050934, "learning_rate": 2.7784961349262126e-05, "loss": 0.0126, "step": 11848 }, { "epoch": 8.326774420238932, "grad_norm": 0.182118222117424, "learning_rate": 2.7784492855469663e-05, "loss": 0.0393, "step": 11849 }, { "epoch": 8.327477160927618, "grad_norm": 0.1800002008676529, "learning_rate": 2.7784024361677206e-05, "loss": 0.0276, "step": 11850 }, { "epoch": 8.328179901616304, "grad_norm": 0.22412854433059692, "learning_rate": 2.778355586788475e-05, "loss": 0.0446, "step": 11851 }, { "epoch": 8.32888264230499, "grad_norm": 0.30553314089775085, "learning_rate": 2.7783087374092294e-05, "loss": 0.0235, "step": 11852 }, { "epoch": 8.329585382993676, "grad_norm": 0.28008419275283813, "learning_rate": 2.7782618880299837e-05, "loss": 0.0631, "step": 11853 }, { "epoch": 8.330288123682362, "grad_norm": 0.5627357959747314, "learning_rate": 2.7782150386507378e-05, "loss": 0.0377, "step": 11854 }, { "epoch": 8.330990864371048, "grad_norm": 0.41943472623825073, "learning_rate": 2.778168189271492e-05, "loss": 0.0931, "step": 11855 }, { "epoch": 8.331693605059733, "grad_norm": 1.2347725629806519, "learning_rate": 2.7781213398922465e-05, "loss": 0.1331, "step": 11856 }, { "epoch": 8.33239634574842, "grad_norm": 0.4699656665325165, "learning_rate": 2.778074490513001e-05, "loss": 0.161, "step": 11857 }, { "epoch": 8.333099086437105, "grad_norm": 0.8820509910583496, "learning_rate": 2.778027641133755e-05, "loss": 0.2274, "step": 11858 }, { "epoch": 8.333801827125791, "grad_norm": 1.2085469961166382, "learning_rate": 2.7779807917545093e-05, "loss": 0.1973, "step": 11859 }, { "epoch": 8.334504567814477, "grad_norm": 0.27516186237335205, "learning_rate": 2.7779339423752637e-05, "loss": 0.0616, "step": 11860 }, { "epoch": 8.335207308503163, "grad_norm": 0.20331846177577972, "learning_rate": 2.777887092996018e-05, "loss": 0.0264, "step": 11861 }, { "epoch": 8.335910049191849, "grad_norm": 0.11170581728219986, "learning_rate": 2.777840243616772e-05, "loss": 0.0179, "step": 11862 }, { "epoch": 8.336612789880535, "grad_norm": 0.3498590886592865, "learning_rate": 2.7777933942375265e-05, "loss": 0.0213, "step": 11863 }, { "epoch": 8.33731553056922, "grad_norm": 0.2409301996231079, "learning_rate": 2.7777465448582808e-05, "loss": 0.0194, "step": 11864 }, { "epoch": 8.338018271257905, "grad_norm": 0.4114859998226166, "learning_rate": 2.7776996954790352e-05, "loss": 0.0154, "step": 11865 }, { "epoch": 8.33872101194659, "grad_norm": 0.24542628228664398, "learning_rate": 2.7776528460997892e-05, "loss": 0.013, "step": 11866 }, { "epoch": 8.339423752635277, "grad_norm": 0.1256725788116455, "learning_rate": 2.7776059967205433e-05, "loss": 0.0223, "step": 11867 }, { "epoch": 8.340126493323963, "grad_norm": 0.13764667510986328, "learning_rate": 2.7775591473412976e-05, "loss": 0.0214, "step": 11868 }, { "epoch": 8.340829234012649, "grad_norm": 0.21690957248210907, "learning_rate": 2.777512297962052e-05, "loss": 0.0105, "step": 11869 }, { "epoch": 8.341531974701335, "grad_norm": 0.3000432550907135, "learning_rate": 2.7774654485828064e-05, "loss": 0.0213, "step": 11870 }, { "epoch": 8.34223471539002, "grad_norm": 0.15527743101119995, "learning_rate": 2.7774185992035604e-05, "loss": 0.0158, "step": 11871 }, { "epoch": 8.342937456078706, "grad_norm": 0.17778296768665314, "learning_rate": 2.7773717498243148e-05, "loss": 0.0227, "step": 11872 }, { "epoch": 8.343640196767392, "grad_norm": 0.0718669593334198, "learning_rate": 2.777324900445069e-05, "loss": 0.0075, "step": 11873 }, { "epoch": 8.344342937456078, "grad_norm": 0.614504873752594, "learning_rate": 2.7772780510658235e-05, "loss": 0.0278, "step": 11874 }, { "epoch": 8.345045678144764, "grad_norm": 0.34009286761283875, "learning_rate": 2.7772312016865776e-05, "loss": 0.0513, "step": 11875 }, { "epoch": 8.34574841883345, "grad_norm": 0.21068695187568665, "learning_rate": 2.777184352307332e-05, "loss": 0.0186, "step": 11876 }, { "epoch": 8.346451159522136, "grad_norm": 0.28950509428977966, "learning_rate": 2.7771375029280863e-05, "loss": 0.0442, "step": 11877 }, { "epoch": 8.347153900210822, "grad_norm": 0.35407036542892456, "learning_rate": 2.7770906535488407e-05, "loss": 0.0422, "step": 11878 }, { "epoch": 8.347856640899508, "grad_norm": 0.692635715007782, "learning_rate": 2.777043804169595e-05, "loss": 0.0689, "step": 11879 }, { "epoch": 8.348559381588194, "grad_norm": 0.38849571347236633, "learning_rate": 2.776996954790349e-05, "loss": 0.0862, "step": 11880 }, { "epoch": 8.34926212227688, "grad_norm": 0.3460457921028137, "learning_rate": 2.7769501054111035e-05, "loss": 0.1089, "step": 11881 }, { "epoch": 8.349964862965566, "grad_norm": 0.591715395450592, "learning_rate": 2.776903256031858e-05, "loss": 0.1893, "step": 11882 }, { "epoch": 8.350667603654252, "grad_norm": 0.5627415180206299, "learning_rate": 2.7768564066526122e-05, "loss": 0.1719, "step": 11883 }, { "epoch": 8.351370344342937, "grad_norm": 1.3706443309783936, "learning_rate": 2.776809557273366e-05, "loss": 0.1931, "step": 11884 }, { "epoch": 8.352073085031623, "grad_norm": 0.24768656492233276, "learning_rate": 2.7767627078941203e-05, "loss": 0.0756, "step": 11885 }, { "epoch": 8.35277582572031, "grad_norm": 0.22620898485183716, "learning_rate": 2.7767158585148746e-05, "loss": 0.0356, "step": 11886 }, { "epoch": 8.353478566408995, "grad_norm": 0.128138929605484, "learning_rate": 2.776669009135629e-05, "loss": 0.0165, "step": 11887 }, { "epoch": 8.354181307097681, "grad_norm": 0.3248199224472046, "learning_rate": 2.7766221597563834e-05, "loss": 0.0209, "step": 11888 }, { "epoch": 8.354884047786367, "grad_norm": 0.08750776946544647, "learning_rate": 2.7765753103771374e-05, "loss": 0.0167, "step": 11889 }, { "epoch": 8.355586788475053, "grad_norm": 0.1494171917438507, "learning_rate": 2.7765284609978918e-05, "loss": 0.0186, "step": 11890 }, { "epoch": 8.356289529163739, "grad_norm": 0.1283876746892929, "learning_rate": 2.7764816116186462e-05, "loss": 0.0252, "step": 11891 }, { "epoch": 8.356992269852425, "grad_norm": 0.19678130745887756, "learning_rate": 2.7764347622394005e-05, "loss": 0.0241, "step": 11892 }, { "epoch": 8.35769501054111, "grad_norm": 0.1569080352783203, "learning_rate": 2.7763879128601546e-05, "loss": 0.0384, "step": 11893 }, { "epoch": 8.358397751229797, "grad_norm": 0.22623279690742493, "learning_rate": 2.776341063480909e-05, "loss": 0.0088, "step": 11894 }, { "epoch": 8.359100491918483, "grad_norm": 0.3481896221637726, "learning_rate": 2.7762942141016633e-05, "loss": 0.0351, "step": 11895 }, { "epoch": 8.359803232607169, "grad_norm": 0.1304425299167633, "learning_rate": 2.7762473647224177e-05, "loss": 0.0206, "step": 11896 }, { "epoch": 8.360505973295854, "grad_norm": 0.20512422919273376, "learning_rate": 2.7762005153431717e-05, "loss": 0.0395, "step": 11897 }, { "epoch": 8.36120871398454, "grad_norm": 0.1742391437292099, "learning_rate": 2.776153665963926e-05, "loss": 0.0122, "step": 11898 }, { "epoch": 8.361911454673226, "grad_norm": 0.5638222694396973, "learning_rate": 2.7761068165846805e-05, "loss": 0.0339, "step": 11899 }, { "epoch": 8.362614195361912, "grad_norm": 0.19798815250396729, "learning_rate": 2.776059967205435e-05, "loss": 0.0545, "step": 11900 }, { "epoch": 8.363316936050598, "grad_norm": 0.311311274766922, "learning_rate": 2.776013117826189e-05, "loss": 0.0187, "step": 11901 }, { "epoch": 8.364019676739284, "grad_norm": 0.3278113901615143, "learning_rate": 2.775966268446943e-05, "loss": 0.031, "step": 11902 }, { "epoch": 8.36472241742797, "grad_norm": 0.1937539130449295, "learning_rate": 2.7759194190676973e-05, "loss": 0.0362, "step": 11903 }, { "epoch": 8.365425158116654, "grad_norm": 0.24583058059215546, "learning_rate": 2.7758725696884517e-05, "loss": 0.0582, "step": 11904 }, { "epoch": 8.36612789880534, "grad_norm": 0.4072065055370331, "learning_rate": 2.775825720309206e-05, "loss": 0.0749, "step": 11905 }, { "epoch": 8.366830639494026, "grad_norm": 1.1325069665908813, "learning_rate": 2.77577887092996e-05, "loss": 0.1523, "step": 11906 }, { "epoch": 8.367533380182712, "grad_norm": 0.9147791862487793, "learning_rate": 2.7757320215507144e-05, "loss": 0.1693, "step": 11907 }, { "epoch": 8.368236120871398, "grad_norm": 0.8887282609939575, "learning_rate": 2.7756851721714688e-05, "loss": 0.1834, "step": 11908 }, { "epoch": 8.368938861560084, "grad_norm": 3.2395811080932617, "learning_rate": 2.7756383227922232e-05, "loss": 0.2994, "step": 11909 }, { "epoch": 8.36964160224877, "grad_norm": 0.32103732228279114, "learning_rate": 2.7755914734129772e-05, "loss": 0.0548, "step": 11910 }, { "epoch": 8.370344342937456, "grad_norm": 0.13445255160331726, "learning_rate": 2.7755446240337316e-05, "loss": 0.0225, "step": 11911 }, { "epoch": 8.371047083626141, "grad_norm": 0.17747433483600616, "learning_rate": 2.775497774654486e-05, "loss": 0.0201, "step": 11912 }, { "epoch": 8.371749824314827, "grad_norm": 0.30949893593788147, "learning_rate": 2.7754509252752403e-05, "loss": 0.0257, "step": 11913 }, { "epoch": 8.372452565003513, "grad_norm": 0.22146356105804443, "learning_rate": 2.7754040758959947e-05, "loss": 0.0245, "step": 11914 }, { "epoch": 8.3731553056922, "grad_norm": 0.08210308849811554, "learning_rate": 2.7753572265167487e-05, "loss": 0.0083, "step": 11915 }, { "epoch": 8.373858046380885, "grad_norm": 0.1354997307062149, "learning_rate": 2.775310377137503e-05, "loss": 0.019, "step": 11916 }, { "epoch": 8.374560787069571, "grad_norm": 0.11357199400663376, "learning_rate": 2.7752635277582575e-05, "loss": 0.0189, "step": 11917 }, { "epoch": 8.375263527758257, "grad_norm": 0.3386411964893341, "learning_rate": 2.7752166783790115e-05, "loss": 0.0193, "step": 11918 }, { "epoch": 8.375966268446943, "grad_norm": 0.12077043205499649, "learning_rate": 2.7751698289997656e-05, "loss": 0.0149, "step": 11919 }, { "epoch": 8.376669009135629, "grad_norm": 0.16153909265995026, "learning_rate": 2.77512297962052e-05, "loss": 0.0216, "step": 11920 }, { "epoch": 8.377371749824315, "grad_norm": 0.1999979466199875, "learning_rate": 2.7750761302412743e-05, "loss": 0.0205, "step": 11921 }, { "epoch": 8.378074490513, "grad_norm": 0.23790565133094788, "learning_rate": 2.7750292808620287e-05, "loss": 0.0398, "step": 11922 }, { "epoch": 8.378777231201687, "grad_norm": 0.2051401287317276, "learning_rate": 2.7749824314827827e-05, "loss": 0.0154, "step": 11923 }, { "epoch": 8.379479971890373, "grad_norm": 0.6281797885894775, "learning_rate": 2.774935582103537e-05, "loss": 0.0341, "step": 11924 }, { "epoch": 8.380182712579058, "grad_norm": 0.46448445320129395, "learning_rate": 2.7748887327242915e-05, "loss": 0.0285, "step": 11925 }, { "epoch": 8.380885453267744, "grad_norm": 0.2331634908914566, "learning_rate": 2.7748418833450458e-05, "loss": 0.0206, "step": 11926 }, { "epoch": 8.38158819395643, "grad_norm": 0.3309840261936188, "learning_rate": 2.7747950339658002e-05, "loss": 0.0452, "step": 11927 }, { "epoch": 8.382290934645116, "grad_norm": 0.15739500522613525, "learning_rate": 2.7747481845865542e-05, "loss": 0.038, "step": 11928 }, { "epoch": 8.382993675333802, "grad_norm": 0.7234190106391907, "learning_rate": 2.7747013352073086e-05, "loss": 0.0804, "step": 11929 }, { "epoch": 8.383696416022488, "grad_norm": 0.3603857159614563, "learning_rate": 2.774654485828063e-05, "loss": 0.0769, "step": 11930 }, { "epoch": 8.384399156711174, "grad_norm": 0.45271381735801697, "learning_rate": 2.7746076364488173e-05, "loss": 0.1107, "step": 11931 }, { "epoch": 8.38510189739986, "grad_norm": 0.49854013323783875, "learning_rate": 2.7745607870695714e-05, "loss": 0.1515, "step": 11932 }, { "epoch": 8.385804638088546, "grad_norm": 0.6437174677848816, "learning_rate": 2.7745139376903258e-05, "loss": 0.178, "step": 11933 }, { "epoch": 8.386507378777232, "grad_norm": 3.2207276821136475, "learning_rate": 2.77446708831108e-05, "loss": 0.1995, "step": 11934 }, { "epoch": 8.387210119465918, "grad_norm": 0.3873318135738373, "learning_rate": 2.7744202389318345e-05, "loss": 0.0657, "step": 11935 }, { "epoch": 8.387912860154604, "grad_norm": 0.10559376329183578, "learning_rate": 2.7743733895525882e-05, "loss": 0.0131, "step": 11936 }, { "epoch": 8.38861560084329, "grad_norm": 0.3151344954967499, "learning_rate": 2.7743265401733426e-05, "loss": 0.0382, "step": 11937 }, { "epoch": 8.389318341531975, "grad_norm": 0.14082932472229004, "learning_rate": 2.774279690794097e-05, "loss": 0.0256, "step": 11938 }, { "epoch": 8.390021082220661, "grad_norm": 0.18414713442325592, "learning_rate": 2.7742328414148513e-05, "loss": 0.038, "step": 11939 }, { "epoch": 8.390723822909347, "grad_norm": 0.24639852344989777, "learning_rate": 2.7741859920356057e-05, "loss": 0.0206, "step": 11940 }, { "epoch": 8.391426563598033, "grad_norm": 0.08897925913333893, "learning_rate": 2.7741391426563597e-05, "loss": 0.0066, "step": 11941 }, { "epoch": 8.392129304286719, "grad_norm": 0.2911665439605713, "learning_rate": 2.774092293277114e-05, "loss": 0.0366, "step": 11942 }, { "epoch": 8.392832044975403, "grad_norm": 0.10900534689426422, "learning_rate": 2.7740454438978685e-05, "loss": 0.0148, "step": 11943 }, { "epoch": 8.39353478566409, "grad_norm": 0.1261165589094162, "learning_rate": 2.773998594518623e-05, "loss": 0.0129, "step": 11944 }, { "epoch": 8.394237526352775, "grad_norm": 0.16094253957271576, "learning_rate": 2.773951745139377e-05, "loss": 0.0235, "step": 11945 }, { "epoch": 8.394940267041461, "grad_norm": 0.16200007498264313, "learning_rate": 2.7739048957601312e-05, "loss": 0.0077, "step": 11946 }, { "epoch": 8.395643007730147, "grad_norm": 0.15106631815433502, "learning_rate": 2.7738580463808856e-05, "loss": 0.0309, "step": 11947 }, { "epoch": 8.396345748418833, "grad_norm": 0.21165458858013153, "learning_rate": 2.77381119700164e-05, "loss": 0.0171, "step": 11948 }, { "epoch": 8.397048489107519, "grad_norm": 0.23918470740318298, "learning_rate": 2.773764347622394e-05, "loss": 0.0382, "step": 11949 }, { "epoch": 8.397751229796205, "grad_norm": 0.12110023200511932, "learning_rate": 2.7737174982431484e-05, "loss": 0.0244, "step": 11950 }, { "epoch": 8.39845397048489, "grad_norm": 0.13689590990543365, "learning_rate": 2.7736706488639028e-05, "loss": 0.0295, "step": 11951 }, { "epoch": 8.399156711173577, "grad_norm": 0.15360645949840546, "learning_rate": 2.773623799484657e-05, "loss": 0.0235, "step": 11952 }, { "epoch": 8.399859451862262, "grad_norm": 0.27071404457092285, "learning_rate": 2.7735769501054112e-05, "loss": 0.0549, "step": 11953 }, { "epoch": 8.400562192550948, "grad_norm": 0.4968218505382538, "learning_rate": 2.7735301007261652e-05, "loss": 0.0451, "step": 11954 }, { "epoch": 8.401264933239634, "grad_norm": 0.29594436287879944, "learning_rate": 2.7734832513469196e-05, "loss": 0.0658, "step": 11955 }, { "epoch": 8.40196767392832, "grad_norm": 0.5840374231338501, "learning_rate": 2.773436401967674e-05, "loss": 0.131, "step": 11956 }, { "epoch": 8.402670414617006, "grad_norm": 0.9585350751876831, "learning_rate": 2.7733895525884283e-05, "loss": 0.1852, "step": 11957 }, { "epoch": 8.403373155305692, "grad_norm": 0.8468101024627686, "learning_rate": 2.7733427032091824e-05, "loss": 0.1857, "step": 11958 }, { "epoch": 8.404075895994378, "grad_norm": 1.5275520086288452, "learning_rate": 2.7732958538299367e-05, "loss": 0.1808, "step": 11959 }, { "epoch": 8.404778636683064, "grad_norm": 0.2813529968261719, "learning_rate": 2.773249004450691e-05, "loss": 0.0721, "step": 11960 }, { "epoch": 8.40548137737175, "grad_norm": 0.8148516416549683, "learning_rate": 2.7732021550714455e-05, "loss": 0.0228, "step": 11961 }, { "epoch": 8.406184118060436, "grad_norm": 0.16737376153469086, "learning_rate": 2.7731553056921995e-05, "loss": 0.0228, "step": 11962 }, { "epoch": 8.406886858749122, "grad_norm": 0.1770060509443283, "learning_rate": 2.773108456312954e-05, "loss": 0.0229, "step": 11963 }, { "epoch": 8.407589599437808, "grad_norm": 0.11691310256719589, "learning_rate": 2.7730616069337083e-05, "loss": 0.0126, "step": 11964 }, { "epoch": 8.408292340126494, "grad_norm": 0.11853571981191635, "learning_rate": 2.7730147575544626e-05, "loss": 0.0127, "step": 11965 }, { "epoch": 8.40899508081518, "grad_norm": 0.15241307020187378, "learning_rate": 2.772967908175217e-05, "loss": 0.0145, "step": 11966 }, { "epoch": 8.409697821503865, "grad_norm": 0.1468041092157364, "learning_rate": 2.772921058795971e-05, "loss": 0.0216, "step": 11967 }, { "epoch": 8.410400562192551, "grad_norm": 0.13245391845703125, "learning_rate": 2.7728742094167254e-05, "loss": 0.0248, "step": 11968 }, { "epoch": 8.411103302881237, "grad_norm": 0.21064414083957672, "learning_rate": 2.7728273600374798e-05, "loss": 0.0162, "step": 11969 }, { "epoch": 8.411806043569923, "grad_norm": 0.173239067196846, "learning_rate": 2.772780510658234e-05, "loss": 0.0298, "step": 11970 }, { "epoch": 8.412508784258609, "grad_norm": 0.22724582254886627, "learning_rate": 2.772733661278988e-05, "loss": 0.024, "step": 11971 }, { "epoch": 8.413211524947295, "grad_norm": 0.225345179438591, "learning_rate": 2.7726868118997422e-05, "loss": 0.0194, "step": 11972 }, { "epoch": 8.41391426563598, "grad_norm": 0.2480628937482834, "learning_rate": 2.7726399625204966e-05, "loss": 0.0175, "step": 11973 }, { "epoch": 8.414617006324667, "grad_norm": 0.4334077835083008, "learning_rate": 2.772593113141251e-05, "loss": 0.0229, "step": 11974 }, { "epoch": 8.415319747013353, "grad_norm": 0.23708556592464447, "learning_rate": 2.772546263762005e-05, "loss": 0.0385, "step": 11975 }, { "epoch": 8.416022487702039, "grad_norm": 0.322773277759552, "learning_rate": 2.7724994143827594e-05, "loss": 0.0262, "step": 11976 }, { "epoch": 8.416725228390725, "grad_norm": 0.17750388383865356, "learning_rate": 2.7724525650035137e-05, "loss": 0.0419, "step": 11977 }, { "epoch": 8.41742796907941, "grad_norm": 0.33366069197654724, "learning_rate": 2.772405715624268e-05, "loss": 0.0425, "step": 11978 }, { "epoch": 8.418130709768096, "grad_norm": 0.25236737728118896, "learning_rate": 2.7723588662450225e-05, "loss": 0.0508, "step": 11979 }, { "epoch": 8.41883345045678, "grad_norm": 0.3226333260536194, "learning_rate": 2.7723120168657765e-05, "loss": 0.0651, "step": 11980 }, { "epoch": 8.419536191145466, "grad_norm": 0.7873384952545166, "learning_rate": 2.772265167486531e-05, "loss": 0.1093, "step": 11981 }, { "epoch": 8.420238931834152, "grad_norm": 0.717807948589325, "learning_rate": 2.7722183181072853e-05, "loss": 0.1364, "step": 11982 }, { "epoch": 8.420941672522838, "grad_norm": 0.7902241945266724, "learning_rate": 2.7721714687280396e-05, "loss": 0.1845, "step": 11983 }, { "epoch": 8.421644413211524, "grad_norm": 1.464255690574646, "learning_rate": 2.7721246193487937e-05, "loss": 0.1783, "step": 11984 }, { "epoch": 8.42234715390021, "grad_norm": 0.17089338600635529, "learning_rate": 2.772077769969548e-05, "loss": 0.053, "step": 11985 }, { "epoch": 8.423049894588896, "grad_norm": 0.19022999703884125, "learning_rate": 2.7720309205903024e-05, "loss": 0.0507, "step": 11986 }, { "epoch": 8.423752635277582, "grad_norm": 0.27563241124153137, "learning_rate": 2.7719840712110568e-05, "loss": 0.031, "step": 11987 }, { "epoch": 8.424455375966268, "grad_norm": 0.1479923278093338, "learning_rate": 2.7719372218318105e-05, "loss": 0.0238, "step": 11988 }, { "epoch": 8.425158116654954, "grad_norm": 0.1767912209033966, "learning_rate": 2.771890372452565e-05, "loss": 0.017, "step": 11989 }, { "epoch": 8.42586085734364, "grad_norm": 0.15985602140426636, "learning_rate": 2.7718435230733192e-05, "loss": 0.0139, "step": 11990 }, { "epoch": 8.426563598032326, "grad_norm": 0.18192218244075775, "learning_rate": 2.7717966736940736e-05, "loss": 0.0238, "step": 11991 }, { "epoch": 8.427266338721012, "grad_norm": 0.11830714344978333, "learning_rate": 2.771749824314828e-05, "loss": 0.0146, "step": 11992 }, { "epoch": 8.427969079409698, "grad_norm": 0.541102409362793, "learning_rate": 2.771702974935582e-05, "loss": 0.0299, "step": 11993 }, { "epoch": 8.428671820098383, "grad_norm": 0.1253473460674286, "learning_rate": 2.7716561255563364e-05, "loss": 0.0263, "step": 11994 }, { "epoch": 8.42937456078707, "grad_norm": 0.15757058560848236, "learning_rate": 2.7716092761770908e-05, "loss": 0.0291, "step": 11995 }, { "epoch": 8.430077301475755, "grad_norm": 0.19824768602848053, "learning_rate": 2.771562426797845e-05, "loss": 0.0223, "step": 11996 }, { "epoch": 8.430780042164441, "grad_norm": 0.39738595485687256, "learning_rate": 2.771515577418599e-05, "loss": 0.0384, "step": 11997 }, { "epoch": 8.431482782853127, "grad_norm": 0.12858939170837402, "learning_rate": 2.7714687280393535e-05, "loss": 0.0246, "step": 11998 }, { "epoch": 8.432185523541813, "grad_norm": 0.17191863059997559, "learning_rate": 2.771421878660108e-05, "loss": 0.0312, "step": 11999 }, { "epoch": 8.432888264230499, "grad_norm": 0.3138163089752197, "learning_rate": 2.7713750292808623e-05, "loss": 0.0342, "step": 12000 }, { "epoch": 8.432888264230499, "eval_cer": 0.1949417913216152, "eval_loss": 0.2715981602668762, "eval_runtime": 18.3957, "eval_samples_per_second": 246.688, "eval_steps_per_second": 0.815, "eval_wer": 0.3491522583168068, "step": 12000 }, { "epoch": 8.433591004919185, "grad_norm": 0.15184728801250458, "learning_rate": 2.7713281799016163e-05, "loss": 0.0213, "step": 12001 }, { "epoch": 8.43429374560787, "grad_norm": 0.20259851217269897, "learning_rate": 2.7712813305223707e-05, "loss": 0.0606, "step": 12002 }, { "epoch": 8.434996486296557, "grad_norm": 0.27380701899528503, "learning_rate": 2.771234481143125e-05, "loss": 0.0452, "step": 12003 }, { "epoch": 8.435699226985243, "grad_norm": 0.2944409251213074, "learning_rate": 2.7711876317638794e-05, "loss": 0.0756, "step": 12004 }, { "epoch": 8.436401967673929, "grad_norm": 0.5460500121116638, "learning_rate": 2.7711407823846335e-05, "loss": 0.0493, "step": 12005 }, { "epoch": 8.437104708362615, "grad_norm": 0.39569586515426636, "learning_rate": 2.7710939330053875e-05, "loss": 0.1092, "step": 12006 }, { "epoch": 8.4378074490513, "grad_norm": 0.4558204114437103, "learning_rate": 2.771047083626142e-05, "loss": 0.1501, "step": 12007 }, { "epoch": 8.438510189739986, "grad_norm": 0.9800627827644348, "learning_rate": 2.7710002342468962e-05, "loss": 0.1844, "step": 12008 }, { "epoch": 8.439212930428672, "grad_norm": 1.5343374013900757, "learning_rate": 2.7709533848676506e-05, "loss": 0.228, "step": 12009 }, { "epoch": 8.439915671117358, "grad_norm": 0.36223044991493225, "learning_rate": 2.7709065354884046e-05, "loss": 0.0835, "step": 12010 }, { "epoch": 8.440618411806044, "grad_norm": 0.13049820065498352, "learning_rate": 2.770859686109159e-05, "loss": 0.0229, "step": 12011 }, { "epoch": 8.44132115249473, "grad_norm": 0.15029054880142212, "learning_rate": 2.7708128367299134e-05, "loss": 0.028, "step": 12012 }, { "epoch": 8.442023893183416, "grad_norm": 0.10271131247282028, "learning_rate": 2.7707659873506678e-05, "loss": 0.0157, "step": 12013 }, { "epoch": 8.442726633872102, "grad_norm": 0.1061602532863617, "learning_rate": 2.7707191379714218e-05, "loss": 0.0162, "step": 12014 }, { "epoch": 8.443429374560788, "grad_norm": 0.33089959621429443, "learning_rate": 2.770672288592176e-05, "loss": 0.0134, "step": 12015 }, { "epoch": 8.444132115249474, "grad_norm": 0.12553055584430695, "learning_rate": 2.7706254392129305e-05, "loss": 0.0171, "step": 12016 }, { "epoch": 8.44483485593816, "grad_norm": 0.0769026055932045, "learning_rate": 2.770578589833685e-05, "loss": 0.0076, "step": 12017 }, { "epoch": 8.445537596626846, "grad_norm": 0.14254766702651978, "learning_rate": 2.7705317404544393e-05, "loss": 0.0194, "step": 12018 }, { "epoch": 8.44624033731553, "grad_norm": 0.2743963897228241, "learning_rate": 2.7704848910751933e-05, "loss": 0.0111, "step": 12019 }, { "epoch": 8.446943078004216, "grad_norm": 0.16809824109077454, "learning_rate": 2.7704380416959477e-05, "loss": 0.0228, "step": 12020 }, { "epoch": 8.447645818692902, "grad_norm": 0.12721599638462067, "learning_rate": 2.770391192316702e-05, "loss": 0.0087, "step": 12021 }, { "epoch": 8.448348559381587, "grad_norm": 0.148137629032135, "learning_rate": 2.7703443429374564e-05, "loss": 0.02, "step": 12022 }, { "epoch": 8.449051300070273, "grad_norm": 0.12077668309211731, "learning_rate": 2.77029749355821e-05, "loss": 0.0186, "step": 12023 }, { "epoch": 8.44975404075896, "grad_norm": 0.17777365446090698, "learning_rate": 2.7702506441789645e-05, "loss": 0.0268, "step": 12024 }, { "epoch": 8.450456781447645, "grad_norm": 0.46422868967056274, "learning_rate": 2.770203794799719e-05, "loss": 0.0427, "step": 12025 }, { "epoch": 8.451159522136331, "grad_norm": 0.20744846761226654, "learning_rate": 2.7701569454204732e-05, "loss": 0.0261, "step": 12026 }, { "epoch": 8.451862262825017, "grad_norm": 0.21962857246398926, "learning_rate": 2.7701100960412273e-05, "loss": 0.0329, "step": 12027 }, { "epoch": 8.452565003513703, "grad_norm": 0.3460487127304077, "learning_rate": 2.7700632466619817e-05, "loss": 0.0321, "step": 12028 }, { "epoch": 8.453267744202389, "grad_norm": 0.3054918944835663, "learning_rate": 2.770016397282736e-05, "loss": 0.0714, "step": 12029 }, { "epoch": 8.453970484891075, "grad_norm": 0.4148314595222473, "learning_rate": 2.7699695479034904e-05, "loss": 0.0724, "step": 12030 }, { "epoch": 8.45467322557976, "grad_norm": 1.3059452772140503, "learning_rate": 2.7699226985242448e-05, "loss": 0.1077, "step": 12031 }, { "epoch": 8.455375966268447, "grad_norm": 0.6753339171409607, "learning_rate": 2.7698758491449988e-05, "loss": 0.1661, "step": 12032 }, { "epoch": 8.456078706957133, "grad_norm": 0.9184413552284241, "learning_rate": 2.7698289997657532e-05, "loss": 0.1838, "step": 12033 }, { "epoch": 8.456781447645819, "grad_norm": 1.5503339767456055, "learning_rate": 2.7697821503865076e-05, "loss": 0.2549, "step": 12034 }, { "epoch": 8.457484188334504, "grad_norm": 0.20540285110473633, "learning_rate": 2.769735301007262e-05, "loss": 0.0778, "step": 12035 }, { "epoch": 8.45818692902319, "grad_norm": 0.20358355343341827, "learning_rate": 2.769688451628016e-05, "loss": 0.0255, "step": 12036 }, { "epoch": 8.458889669711876, "grad_norm": 0.13177995383739471, "learning_rate": 2.7696416022487703e-05, "loss": 0.0195, "step": 12037 }, { "epoch": 8.459592410400562, "grad_norm": 0.1836635321378708, "learning_rate": 2.7695947528695247e-05, "loss": 0.0301, "step": 12038 }, { "epoch": 8.460295151089248, "grad_norm": 0.3087073862552643, "learning_rate": 2.769547903490279e-05, "loss": 0.027, "step": 12039 }, { "epoch": 8.460997891777934, "grad_norm": 0.1036015972495079, "learning_rate": 2.7695010541110328e-05, "loss": 0.0105, "step": 12040 }, { "epoch": 8.46170063246662, "grad_norm": 0.08665789663791656, "learning_rate": 2.769454204731787e-05, "loss": 0.0125, "step": 12041 }, { "epoch": 8.462403373155306, "grad_norm": 0.2257915884256363, "learning_rate": 2.7694073553525415e-05, "loss": 0.0269, "step": 12042 }, { "epoch": 8.463106113843992, "grad_norm": 0.2862522304058075, "learning_rate": 2.769360505973296e-05, "loss": 0.0196, "step": 12043 }, { "epoch": 8.463808854532678, "grad_norm": 0.1786789745092392, "learning_rate": 2.7693136565940503e-05, "loss": 0.0121, "step": 12044 }, { "epoch": 8.464511595221364, "grad_norm": 0.1531863808631897, "learning_rate": 2.7692668072148043e-05, "loss": 0.0178, "step": 12045 }, { "epoch": 8.46521433591005, "grad_norm": 0.24344968795776367, "learning_rate": 2.7692199578355587e-05, "loss": 0.0188, "step": 12046 }, { "epoch": 8.465917076598735, "grad_norm": 0.24692100286483765, "learning_rate": 2.769173108456313e-05, "loss": 0.0216, "step": 12047 }, { "epoch": 8.466619817287421, "grad_norm": 0.13409438729286194, "learning_rate": 2.7691262590770674e-05, "loss": 0.0147, "step": 12048 }, { "epoch": 8.467322557976107, "grad_norm": 0.18596914410591125, "learning_rate": 2.7690794096978214e-05, "loss": 0.0256, "step": 12049 }, { "epoch": 8.468025298664793, "grad_norm": 0.2069394588470459, "learning_rate": 2.7690325603185758e-05, "loss": 0.0349, "step": 12050 }, { "epoch": 8.46872803935348, "grad_norm": 0.2689078748226166, "learning_rate": 2.7689857109393302e-05, "loss": 0.0192, "step": 12051 }, { "epoch": 8.469430780042165, "grad_norm": 0.2252141386270523, "learning_rate": 2.7689388615600846e-05, "loss": 0.0411, "step": 12052 }, { "epoch": 8.470133520730851, "grad_norm": 0.2874157130718231, "learning_rate": 2.7688920121808386e-05, "loss": 0.065, "step": 12053 }, { "epoch": 8.470836261419537, "grad_norm": 0.461755633354187, "learning_rate": 2.768845162801593e-05, "loss": 0.0576, "step": 12054 }, { "epoch": 8.471539002108223, "grad_norm": 2.655086040496826, "learning_rate": 2.7687983134223473e-05, "loss": 0.0905, "step": 12055 }, { "epoch": 8.472241742796909, "grad_norm": 0.40492090582847595, "learning_rate": 2.7687514640431017e-05, "loss": 0.1077, "step": 12056 }, { "epoch": 8.472944483485595, "grad_norm": 0.8352064490318298, "learning_rate": 2.768704614663856e-05, "loss": 0.151, "step": 12057 }, { "epoch": 8.473647224174279, "grad_norm": 2.4095218181610107, "learning_rate": 2.7686577652846098e-05, "loss": 0.2248, "step": 12058 }, { "epoch": 8.474349964862965, "grad_norm": 1.8038722276687622, "learning_rate": 2.768610915905364e-05, "loss": 0.2362, "step": 12059 }, { "epoch": 8.47505270555165, "grad_norm": 0.312947154045105, "learning_rate": 2.7685640665261185e-05, "loss": 0.0625, "step": 12060 }, { "epoch": 8.475755446240337, "grad_norm": 0.4765094220638275, "learning_rate": 2.768517217146873e-05, "loss": 0.028, "step": 12061 }, { "epoch": 8.476458186929023, "grad_norm": 0.10049708187580109, "learning_rate": 2.768470367767627e-05, "loss": 0.0183, "step": 12062 }, { "epoch": 8.477160927617708, "grad_norm": 0.06455709785223007, "learning_rate": 2.7684235183883813e-05, "loss": 0.0121, "step": 12063 }, { "epoch": 8.477863668306394, "grad_norm": 0.16169458627700806, "learning_rate": 2.7683766690091357e-05, "loss": 0.0162, "step": 12064 }, { "epoch": 8.47856640899508, "grad_norm": 0.21078982949256897, "learning_rate": 2.76832981962989e-05, "loss": 0.0152, "step": 12065 }, { "epoch": 8.479269149683766, "grad_norm": 0.08694775402545929, "learning_rate": 2.768282970250644e-05, "loss": 0.0127, "step": 12066 }, { "epoch": 8.479971890372452, "grad_norm": 0.2069719433784485, "learning_rate": 2.7682361208713985e-05, "loss": 0.0192, "step": 12067 }, { "epoch": 8.480674631061138, "grad_norm": 0.21945837140083313, "learning_rate": 2.7681892714921528e-05, "loss": 0.0288, "step": 12068 }, { "epoch": 8.481377371749824, "grad_norm": 0.1404285728931427, "learning_rate": 2.7681424221129072e-05, "loss": 0.0152, "step": 12069 }, { "epoch": 8.48208011243851, "grad_norm": 0.17607088387012482, "learning_rate": 2.7680955727336616e-05, "loss": 0.0369, "step": 12070 }, { "epoch": 8.482782853127196, "grad_norm": 0.13238617777824402, "learning_rate": 2.7680487233544156e-05, "loss": 0.0066, "step": 12071 }, { "epoch": 8.483485593815882, "grad_norm": 0.1817038655281067, "learning_rate": 2.76800187397517e-05, "loss": 0.0467, "step": 12072 }, { "epoch": 8.484188334504568, "grad_norm": 0.2008163332939148, "learning_rate": 2.7679550245959244e-05, "loss": 0.0205, "step": 12073 }, { "epoch": 8.484891075193254, "grad_norm": 0.5795685648918152, "learning_rate": 2.7679081752166787e-05, "loss": 0.0316, "step": 12074 }, { "epoch": 8.48559381588194, "grad_norm": 0.628176212310791, "learning_rate": 2.7678613258374324e-05, "loss": 0.0341, "step": 12075 }, { "epoch": 8.486296556570625, "grad_norm": 0.32784733176231384, "learning_rate": 2.7678144764581868e-05, "loss": 0.0494, "step": 12076 }, { "epoch": 8.486999297259311, "grad_norm": 0.13822315633296967, "learning_rate": 2.767767627078941e-05, "loss": 0.0267, "step": 12077 }, { "epoch": 8.487702037947997, "grad_norm": 0.30596357583999634, "learning_rate": 2.7677207776996955e-05, "loss": 0.0591, "step": 12078 }, { "epoch": 8.488404778636683, "grad_norm": 0.3333336412906647, "learning_rate": 2.76767392832045e-05, "loss": 0.0395, "step": 12079 }, { "epoch": 8.489107519325369, "grad_norm": 0.5451095104217529, "learning_rate": 2.767627078941204e-05, "loss": 0.0653, "step": 12080 }, { "epoch": 8.489810260014055, "grad_norm": 0.46780821681022644, "learning_rate": 2.7675802295619583e-05, "loss": 0.1372, "step": 12081 }, { "epoch": 8.490513000702741, "grad_norm": 0.5334280133247375, "learning_rate": 2.7675333801827127e-05, "loss": 0.1441, "step": 12082 }, { "epoch": 8.491215741391427, "grad_norm": 0.7373769879341125, "learning_rate": 2.767486530803467e-05, "loss": 0.2132, "step": 12083 }, { "epoch": 8.491918482080113, "grad_norm": 0.8194997310638428, "learning_rate": 2.767439681424221e-05, "loss": 0.2119, "step": 12084 }, { "epoch": 8.492621222768799, "grad_norm": 0.6172620058059692, "learning_rate": 2.7673928320449755e-05, "loss": 0.0591, "step": 12085 }, { "epoch": 8.493323963457485, "grad_norm": 0.16866609454154968, "learning_rate": 2.76734598266573e-05, "loss": 0.0255, "step": 12086 }, { "epoch": 8.49402670414617, "grad_norm": 0.1764494627714157, "learning_rate": 2.7672991332864842e-05, "loss": 0.0297, "step": 12087 }, { "epoch": 8.494729444834856, "grad_norm": 0.21529817581176758, "learning_rate": 2.7672522839072382e-05, "loss": 0.0204, "step": 12088 }, { "epoch": 8.495432185523542, "grad_norm": 0.3602442741394043, "learning_rate": 2.7672054345279926e-05, "loss": 0.0395, "step": 12089 }, { "epoch": 8.496134926212228, "grad_norm": 0.15378816425800323, "learning_rate": 2.767158585148747e-05, "loss": 0.0104, "step": 12090 }, { "epoch": 8.496837666900914, "grad_norm": 0.12956304848194122, "learning_rate": 2.7671117357695014e-05, "loss": 0.0224, "step": 12091 }, { "epoch": 8.4975404075896, "grad_norm": 0.24379609525203705, "learning_rate": 2.7670648863902554e-05, "loss": 0.0312, "step": 12092 }, { "epoch": 8.498243148278286, "grad_norm": 0.13984312117099762, "learning_rate": 2.7670180370110094e-05, "loss": 0.0265, "step": 12093 }, { "epoch": 8.498945888966972, "grad_norm": 0.08178309351205826, "learning_rate": 2.7669711876317638e-05, "loss": 0.0093, "step": 12094 }, { "epoch": 8.499648629655656, "grad_norm": 0.1473771631717682, "learning_rate": 2.7669243382525182e-05, "loss": 0.0232, "step": 12095 }, { "epoch": 8.500351370344344, "grad_norm": 0.08708195388317108, "learning_rate": 2.7668774888732725e-05, "loss": 0.0098, "step": 12096 }, { "epoch": 8.501054111033028, "grad_norm": 0.19632397592067719, "learning_rate": 2.7668306394940266e-05, "loss": 0.0143, "step": 12097 }, { "epoch": 8.501756851721714, "grad_norm": 0.18829776346683502, "learning_rate": 2.766783790114781e-05, "loss": 0.0201, "step": 12098 }, { "epoch": 8.5024595924104, "grad_norm": 0.33713963627815247, "learning_rate": 2.7667369407355353e-05, "loss": 0.0345, "step": 12099 }, { "epoch": 8.503162333099086, "grad_norm": 0.22668114304542542, "learning_rate": 2.7666900913562897e-05, "loss": 0.0268, "step": 12100 }, { "epoch": 8.503865073787772, "grad_norm": 0.12573020160198212, "learning_rate": 2.7666432419770437e-05, "loss": 0.0174, "step": 12101 }, { "epoch": 8.504567814476458, "grad_norm": 0.20716729760169983, "learning_rate": 2.766596392597798e-05, "loss": 0.0333, "step": 12102 }, { "epoch": 8.505270555165144, "grad_norm": 0.2633686363697052, "learning_rate": 2.7665495432185525e-05, "loss": 0.051, "step": 12103 }, { "epoch": 8.50597329585383, "grad_norm": 0.367801308631897, "learning_rate": 2.766502693839307e-05, "loss": 0.0702, "step": 12104 }, { "epoch": 8.506676036542515, "grad_norm": 0.5358773469924927, "learning_rate": 2.7664558444600612e-05, "loss": 0.0789, "step": 12105 }, { "epoch": 8.507378777231201, "grad_norm": 0.284435898065567, "learning_rate": 2.7664089950808153e-05, "loss": 0.0879, "step": 12106 }, { "epoch": 8.508081517919887, "grad_norm": 0.679416835308075, "learning_rate": 2.7663621457015696e-05, "loss": 0.159, "step": 12107 }, { "epoch": 8.508784258608573, "grad_norm": 1.3317177295684814, "learning_rate": 2.766315296322324e-05, "loss": 0.1928, "step": 12108 }, { "epoch": 8.509486999297259, "grad_norm": 1.179178237915039, "learning_rate": 2.7662684469430784e-05, "loss": 0.2489, "step": 12109 }, { "epoch": 8.510189739985945, "grad_norm": 0.18164744973182678, "learning_rate": 2.766221597563832e-05, "loss": 0.086, "step": 12110 }, { "epoch": 8.510892480674631, "grad_norm": 0.1778789758682251, "learning_rate": 2.7661747481845864e-05, "loss": 0.0317, "step": 12111 }, { "epoch": 8.511595221363317, "grad_norm": 0.24368999898433685, "learning_rate": 2.7661278988053408e-05, "loss": 0.036, "step": 12112 }, { "epoch": 8.512297962052003, "grad_norm": 0.24472053349018097, "learning_rate": 2.7660810494260952e-05, "loss": 0.0205, "step": 12113 }, { "epoch": 8.513000702740689, "grad_norm": 0.10945838689804077, "learning_rate": 2.7660342000468492e-05, "loss": 0.0161, "step": 12114 }, { "epoch": 8.513703443429375, "grad_norm": 0.18114562332630157, "learning_rate": 2.7659873506676036e-05, "loss": 0.0081, "step": 12115 }, { "epoch": 8.51440618411806, "grad_norm": 0.16865471005439758, "learning_rate": 2.765940501288358e-05, "loss": 0.0207, "step": 12116 }, { "epoch": 8.515108924806746, "grad_norm": 0.14654015004634857, "learning_rate": 2.7658936519091123e-05, "loss": 0.021, "step": 12117 }, { "epoch": 8.515811665495432, "grad_norm": 0.17792607843875885, "learning_rate": 2.7658468025298667e-05, "loss": 0.0227, "step": 12118 }, { "epoch": 8.516514406184118, "grad_norm": 0.16528119146823883, "learning_rate": 2.7657999531506207e-05, "loss": 0.011, "step": 12119 }, { "epoch": 8.517217146872804, "grad_norm": 0.1829753965139389, "learning_rate": 2.765753103771375e-05, "loss": 0.0238, "step": 12120 }, { "epoch": 8.51791988756149, "grad_norm": 0.11682511866092682, "learning_rate": 2.7657062543921295e-05, "loss": 0.0139, "step": 12121 }, { "epoch": 8.518622628250176, "grad_norm": 0.13333141803741455, "learning_rate": 2.765659405012884e-05, "loss": 0.027, "step": 12122 }, { "epoch": 8.519325368938862, "grad_norm": 0.12096332013607025, "learning_rate": 2.765612555633638e-05, "loss": 0.0201, "step": 12123 }, { "epoch": 8.520028109627548, "grad_norm": 0.1796162724494934, "learning_rate": 2.7655657062543923e-05, "loss": 0.0246, "step": 12124 }, { "epoch": 8.520730850316234, "grad_norm": 0.2021210491657257, "learning_rate": 2.7655188568751466e-05, "loss": 0.0379, "step": 12125 }, { "epoch": 8.52143359100492, "grad_norm": 0.23056919872760773, "learning_rate": 2.765472007495901e-05, "loss": 0.0294, "step": 12126 }, { "epoch": 8.522136331693606, "grad_norm": 0.22759833931922913, "learning_rate": 2.7654251581166547e-05, "loss": 0.0544, "step": 12127 }, { "epoch": 8.522839072382292, "grad_norm": 0.1815071851015091, "learning_rate": 2.765378308737409e-05, "loss": 0.0297, "step": 12128 }, { "epoch": 8.523541813070977, "grad_norm": 0.2512597441673279, "learning_rate": 2.7653314593581635e-05, "loss": 0.0596, "step": 12129 }, { "epoch": 8.524244553759663, "grad_norm": 0.4001903533935547, "learning_rate": 2.7652846099789178e-05, "loss": 0.0922, "step": 12130 }, { "epoch": 8.52494729444835, "grad_norm": 1.238315463066101, "learning_rate": 2.7652377605996722e-05, "loss": 0.1134, "step": 12131 }, { "epoch": 8.525650035137035, "grad_norm": 0.38359883427619934, "learning_rate": 2.7651909112204262e-05, "loss": 0.1353, "step": 12132 }, { "epoch": 8.526352775825721, "grad_norm": 0.6700802445411682, "learning_rate": 2.7651440618411806e-05, "loss": 0.1751, "step": 12133 }, { "epoch": 8.527055516514405, "grad_norm": 1.125074863433838, "learning_rate": 2.765097212461935e-05, "loss": 0.2329, "step": 12134 }, { "epoch": 8.527758257203093, "grad_norm": 0.240958109498024, "learning_rate": 2.7650503630826894e-05, "loss": 0.0656, "step": 12135 }, { "epoch": 8.528460997891777, "grad_norm": 0.1332854926586151, "learning_rate": 2.7650035137034434e-05, "loss": 0.0214, "step": 12136 }, { "epoch": 8.529163738580463, "grad_norm": 0.08363480865955353, "learning_rate": 2.7649566643241978e-05, "loss": 0.0168, "step": 12137 }, { "epoch": 8.529866479269149, "grad_norm": 0.23854337632656097, "learning_rate": 2.764909814944952e-05, "loss": 0.0162, "step": 12138 }, { "epoch": 8.530569219957835, "grad_norm": 0.09202978760004044, "learning_rate": 2.7648629655657065e-05, "loss": 0.0226, "step": 12139 }, { "epoch": 8.53127196064652, "grad_norm": 0.21427883207798004, "learning_rate": 2.7648161161864605e-05, "loss": 0.0108, "step": 12140 }, { "epoch": 8.531974701335207, "grad_norm": 0.2395777702331543, "learning_rate": 2.764769266807215e-05, "loss": 0.0229, "step": 12141 }, { "epoch": 8.532677442023893, "grad_norm": 0.18382439017295837, "learning_rate": 2.7647224174279693e-05, "loss": 0.0223, "step": 12142 }, { "epoch": 8.533380182712579, "grad_norm": 0.16709595918655396, "learning_rate": 2.7646755680487237e-05, "loss": 0.0253, "step": 12143 }, { "epoch": 8.534082923401265, "grad_norm": 0.13253024220466614, "learning_rate": 2.764628718669478e-05, "loss": 0.0166, "step": 12144 }, { "epoch": 8.53478566408995, "grad_norm": 0.25877252221107483, "learning_rate": 2.7645818692902317e-05, "loss": 0.0314, "step": 12145 }, { "epoch": 8.535488404778636, "grad_norm": 0.11904208362102509, "learning_rate": 2.764535019910986e-05, "loss": 0.0229, "step": 12146 }, { "epoch": 8.536191145467322, "grad_norm": 0.16900579631328583, "learning_rate": 2.7644881705317405e-05, "loss": 0.0299, "step": 12147 }, { "epoch": 8.536893886156008, "grad_norm": 0.11233015358448029, "learning_rate": 2.764441321152495e-05, "loss": 0.014, "step": 12148 }, { "epoch": 8.537596626844694, "grad_norm": 0.15326887369155884, "learning_rate": 2.764394471773249e-05, "loss": 0.0321, "step": 12149 }, { "epoch": 8.53829936753338, "grad_norm": 0.20938554406166077, "learning_rate": 2.7643476223940032e-05, "loss": 0.0273, "step": 12150 }, { "epoch": 8.539002108222066, "grad_norm": 0.1973017454147339, "learning_rate": 2.7643007730147576e-05, "loss": 0.0162, "step": 12151 }, { "epoch": 8.539704848910752, "grad_norm": 0.6672607064247131, "learning_rate": 2.764253923635512e-05, "loss": 0.0378, "step": 12152 }, { "epoch": 8.540407589599438, "grad_norm": 0.27796754240989685, "learning_rate": 2.764207074256266e-05, "loss": 0.0472, "step": 12153 }, { "epoch": 8.541110330288124, "grad_norm": 0.22288377583026886, "learning_rate": 2.7641602248770204e-05, "loss": 0.0556, "step": 12154 }, { "epoch": 8.54181307097681, "grad_norm": 0.3620879054069519, "learning_rate": 2.7641133754977748e-05, "loss": 0.0608, "step": 12155 }, { "epoch": 8.542515811665496, "grad_norm": 0.36451074481010437, "learning_rate": 2.764066526118529e-05, "loss": 0.094, "step": 12156 }, { "epoch": 8.543218552354181, "grad_norm": 0.6228469610214233, "learning_rate": 2.7640196767392835e-05, "loss": 0.1626, "step": 12157 }, { "epoch": 8.543921293042867, "grad_norm": 0.5669289827346802, "learning_rate": 2.7639728273600375e-05, "loss": 0.1982, "step": 12158 }, { "epoch": 8.544624033731553, "grad_norm": 1.3171260356903076, "learning_rate": 2.763925977980792e-05, "loss": 0.1934, "step": 12159 }, { "epoch": 8.54532677442024, "grad_norm": 0.22383835911750793, "learning_rate": 2.7638791286015463e-05, "loss": 0.0696, "step": 12160 }, { "epoch": 8.546029515108925, "grad_norm": 0.17577232420444489, "learning_rate": 2.7638322792223007e-05, "loss": 0.037, "step": 12161 }, { "epoch": 8.546732255797611, "grad_norm": 0.46276435256004333, "learning_rate": 2.7637854298430544e-05, "loss": 0.033, "step": 12162 }, { "epoch": 8.547434996486297, "grad_norm": 0.1820458173751831, "learning_rate": 2.7637385804638087e-05, "loss": 0.02, "step": 12163 }, { "epoch": 8.548137737174983, "grad_norm": 0.1685786098241806, "learning_rate": 2.763691731084563e-05, "loss": 0.0235, "step": 12164 }, { "epoch": 8.548840477863669, "grad_norm": 0.07685566693544388, "learning_rate": 2.7636448817053175e-05, "loss": 0.0152, "step": 12165 }, { "epoch": 8.549543218552355, "grad_norm": 0.2085922658443451, "learning_rate": 2.7635980323260715e-05, "loss": 0.011, "step": 12166 }, { "epoch": 8.55024595924104, "grad_norm": 0.1204366385936737, "learning_rate": 2.763551182946826e-05, "loss": 0.0165, "step": 12167 }, { "epoch": 8.550948699929727, "grad_norm": 0.1517690122127533, "learning_rate": 2.7635043335675803e-05, "loss": 0.0208, "step": 12168 }, { "epoch": 8.551651440618413, "grad_norm": 0.2739160656929016, "learning_rate": 2.7634574841883346e-05, "loss": 0.0254, "step": 12169 }, { "epoch": 8.552354181307098, "grad_norm": 0.22619302570819855, "learning_rate": 2.763410634809089e-05, "loss": 0.0266, "step": 12170 }, { "epoch": 8.553056921995784, "grad_norm": 0.10431744903326035, "learning_rate": 2.763363785429843e-05, "loss": 0.0169, "step": 12171 }, { "epoch": 8.55375966268447, "grad_norm": 0.15486831963062286, "learning_rate": 2.7633169360505974e-05, "loss": 0.0332, "step": 12172 }, { "epoch": 8.554462403373154, "grad_norm": 0.31771472096443176, "learning_rate": 2.7632700866713518e-05, "loss": 0.0247, "step": 12173 }, { "epoch": 8.55516514406184, "grad_norm": 0.12170208245515823, "learning_rate": 2.763223237292106e-05, "loss": 0.0235, "step": 12174 }, { "epoch": 8.555867884750526, "grad_norm": 0.13845491409301758, "learning_rate": 2.7631763879128602e-05, "loss": 0.0369, "step": 12175 }, { "epoch": 8.556570625439212, "grad_norm": 0.12468221038579941, "learning_rate": 2.7631295385336146e-05, "loss": 0.0136, "step": 12176 }, { "epoch": 8.557273366127898, "grad_norm": 0.18940787017345428, "learning_rate": 2.763082689154369e-05, "loss": 0.0353, "step": 12177 }, { "epoch": 8.557976106816584, "grad_norm": 0.22178685665130615, "learning_rate": 2.7630358397751233e-05, "loss": 0.0409, "step": 12178 }, { "epoch": 8.55867884750527, "grad_norm": 0.2590368688106537, "learning_rate": 2.7629889903958773e-05, "loss": 0.0403, "step": 12179 }, { "epoch": 8.559381588193956, "grad_norm": 0.28393128514289856, "learning_rate": 2.7629421410166314e-05, "loss": 0.0693, "step": 12180 }, { "epoch": 8.560084328882642, "grad_norm": 0.5126261115074158, "learning_rate": 2.7628952916373857e-05, "loss": 0.1154, "step": 12181 }, { "epoch": 8.560787069571328, "grad_norm": 0.5923712253570557, "learning_rate": 2.76284844225814e-05, "loss": 0.1307, "step": 12182 }, { "epoch": 8.561489810260014, "grad_norm": 1.2480783462524414, "learning_rate": 2.7628015928788945e-05, "loss": 0.1981, "step": 12183 }, { "epoch": 8.5621925509487, "grad_norm": 1.2287564277648926, "learning_rate": 2.7627547434996485e-05, "loss": 0.2278, "step": 12184 }, { "epoch": 8.562895291637385, "grad_norm": 0.2786920964717865, "learning_rate": 2.762707894120403e-05, "loss": 0.0582, "step": 12185 }, { "epoch": 8.563598032326071, "grad_norm": 0.11745749413967133, "learning_rate": 2.7626610447411573e-05, "loss": 0.0255, "step": 12186 }, { "epoch": 8.564300773014757, "grad_norm": 0.23517188429832458, "learning_rate": 2.7626141953619116e-05, "loss": 0.0374, "step": 12187 }, { "epoch": 8.565003513703443, "grad_norm": 0.17253944277763367, "learning_rate": 2.7625673459826657e-05, "loss": 0.022, "step": 12188 }, { "epoch": 8.56570625439213, "grad_norm": 0.12828339636325836, "learning_rate": 2.76252049660342e-05, "loss": 0.0259, "step": 12189 }, { "epoch": 8.566408995080815, "grad_norm": 0.19457672536373138, "learning_rate": 2.7624736472241744e-05, "loss": 0.0078, "step": 12190 }, { "epoch": 8.567111735769501, "grad_norm": 0.08910652995109558, "learning_rate": 2.7624267978449288e-05, "loss": 0.0133, "step": 12191 }, { "epoch": 8.567814476458187, "grad_norm": 0.08556453138589859, "learning_rate": 2.7623799484656828e-05, "loss": 0.0129, "step": 12192 }, { "epoch": 8.568517217146873, "grad_norm": 0.16101190447807312, "learning_rate": 2.7623330990864372e-05, "loss": 0.0226, "step": 12193 }, { "epoch": 8.569219957835559, "grad_norm": 0.1570153832435608, "learning_rate": 2.7622862497071916e-05, "loss": 0.0153, "step": 12194 }, { "epoch": 8.569922698524245, "grad_norm": 0.29275330901145935, "learning_rate": 2.762239400327946e-05, "loss": 0.037, "step": 12195 }, { "epoch": 8.57062543921293, "grad_norm": 0.16248168051242828, "learning_rate": 2.7621925509487003e-05, "loss": 0.0211, "step": 12196 }, { "epoch": 8.571328179901617, "grad_norm": 0.1429138034582138, "learning_rate": 2.762145701569454e-05, "loss": 0.0195, "step": 12197 }, { "epoch": 8.572030920590302, "grad_norm": 0.1253349930047989, "learning_rate": 2.7620988521902084e-05, "loss": 0.0235, "step": 12198 }, { "epoch": 8.572733661278988, "grad_norm": 0.31569063663482666, "learning_rate": 2.7620520028109628e-05, "loss": 0.0333, "step": 12199 }, { "epoch": 8.573436401967674, "grad_norm": 0.6286429166793823, "learning_rate": 2.762005153431717e-05, "loss": 0.0576, "step": 12200 }, { "epoch": 8.57413914265636, "grad_norm": 0.12674354016780853, "learning_rate": 2.761958304052471e-05, "loss": 0.0168, "step": 12201 }, { "epoch": 8.574841883345046, "grad_norm": 0.40259093046188354, "learning_rate": 2.7619114546732255e-05, "loss": 0.0378, "step": 12202 }, { "epoch": 8.575544624033732, "grad_norm": 0.26854148507118225, "learning_rate": 2.76186460529398e-05, "loss": 0.0454, "step": 12203 }, { "epoch": 8.576247364722418, "grad_norm": 0.24471822381019592, "learning_rate": 2.7618177559147343e-05, "loss": 0.0411, "step": 12204 }, { "epoch": 8.576950105411104, "grad_norm": 0.2615226209163666, "learning_rate": 2.7617709065354883e-05, "loss": 0.0784, "step": 12205 }, { "epoch": 8.57765284609979, "grad_norm": 0.3309192359447479, "learning_rate": 2.7617240571562427e-05, "loss": 0.0925, "step": 12206 }, { "epoch": 8.578355586788476, "grad_norm": 0.6184173822402954, "learning_rate": 2.761677207776997e-05, "loss": 0.1635, "step": 12207 }, { "epoch": 8.579058327477162, "grad_norm": 0.5308364629745483, "learning_rate": 2.7616303583977514e-05, "loss": 0.1615, "step": 12208 }, { "epoch": 8.579761068165848, "grad_norm": 1.1360262632369995, "learning_rate": 2.7615835090185058e-05, "loss": 0.2237, "step": 12209 }, { "epoch": 8.580463808854532, "grad_norm": 0.2012391835451126, "learning_rate": 2.76153665963926e-05, "loss": 0.0589, "step": 12210 }, { "epoch": 8.58116654954322, "grad_norm": 0.22234106063842773, "learning_rate": 2.7614898102600142e-05, "loss": 0.0333, "step": 12211 }, { "epoch": 8.581869290231904, "grad_norm": 0.09351925551891327, "learning_rate": 2.7614429608807686e-05, "loss": 0.0218, "step": 12212 }, { "epoch": 8.58257203092059, "grad_norm": 0.411399781703949, "learning_rate": 2.761396111501523e-05, "loss": 0.0142, "step": 12213 }, { "epoch": 8.583274771609275, "grad_norm": 0.33759573101997375, "learning_rate": 2.7613492621222766e-05, "loss": 0.0204, "step": 12214 }, { "epoch": 8.583977512297961, "grad_norm": 0.1374734491109848, "learning_rate": 2.761302412743031e-05, "loss": 0.0105, "step": 12215 }, { "epoch": 8.584680252986647, "grad_norm": 0.14858710765838623, "learning_rate": 2.7612555633637854e-05, "loss": 0.0297, "step": 12216 }, { "epoch": 8.585382993675333, "grad_norm": 0.29224225878715515, "learning_rate": 2.7612087139845398e-05, "loss": 0.0146, "step": 12217 }, { "epoch": 8.58608573436402, "grad_norm": 0.15229007601737976, "learning_rate": 2.7611618646052938e-05, "loss": 0.0342, "step": 12218 }, { "epoch": 8.586788475052705, "grad_norm": 0.23082488775253296, "learning_rate": 2.7611150152260482e-05, "loss": 0.0175, "step": 12219 }, { "epoch": 8.587491215741391, "grad_norm": 0.4677185118198395, "learning_rate": 2.7610681658468025e-05, "loss": 0.0202, "step": 12220 }, { "epoch": 8.588193956430077, "grad_norm": 0.15023337304592133, "learning_rate": 2.761021316467557e-05, "loss": 0.0201, "step": 12221 }, { "epoch": 8.588896697118763, "grad_norm": 0.24487991631031036, "learning_rate": 2.7609744670883113e-05, "loss": 0.0278, "step": 12222 }, { "epoch": 8.589599437807449, "grad_norm": 0.23263826966285706, "learning_rate": 2.7609276177090653e-05, "loss": 0.0185, "step": 12223 }, { "epoch": 8.590302178496135, "grad_norm": 0.14522366225719452, "learning_rate": 2.7608807683298197e-05, "loss": 0.0256, "step": 12224 }, { "epoch": 8.59100491918482, "grad_norm": 0.23653945326805115, "learning_rate": 2.760833918950574e-05, "loss": 0.0234, "step": 12225 }, { "epoch": 8.591707659873506, "grad_norm": 0.42643287777900696, "learning_rate": 2.7607870695713284e-05, "loss": 0.0196, "step": 12226 }, { "epoch": 8.592410400562192, "grad_norm": 0.3568241000175476, "learning_rate": 2.7607402201920825e-05, "loss": 0.0336, "step": 12227 }, { "epoch": 8.593113141250878, "grad_norm": 0.39621320366859436, "learning_rate": 2.760693370812837e-05, "loss": 0.0467, "step": 12228 }, { "epoch": 8.593815881939564, "grad_norm": 0.20470954477787018, "learning_rate": 2.7606465214335912e-05, "loss": 0.0389, "step": 12229 }, { "epoch": 8.59451862262825, "grad_norm": 0.5139558911323547, "learning_rate": 2.7605996720543456e-05, "loss": 0.0889, "step": 12230 }, { "epoch": 8.595221363316936, "grad_norm": 0.4067542254924774, "learning_rate": 2.7605528226750996e-05, "loss": 0.096, "step": 12231 }, { "epoch": 8.595924104005622, "grad_norm": 0.4159061312675476, "learning_rate": 2.7605059732958537e-05, "loss": 0.124, "step": 12232 }, { "epoch": 8.596626844694308, "grad_norm": 1.6380058526992798, "learning_rate": 2.760459123916608e-05, "loss": 0.2159, "step": 12233 }, { "epoch": 8.597329585382994, "grad_norm": 0.9857954978942871, "learning_rate": 2.7604122745373624e-05, "loss": 0.2325, "step": 12234 }, { "epoch": 8.59803232607168, "grad_norm": 0.28255054354667664, "learning_rate": 2.7603654251581168e-05, "loss": 0.045, "step": 12235 }, { "epoch": 8.598735066760366, "grad_norm": 0.2570160925388336, "learning_rate": 2.7603185757788708e-05, "loss": 0.0247, "step": 12236 }, { "epoch": 8.599437807449052, "grad_norm": 0.11501207202672958, "learning_rate": 2.7602717263996252e-05, "loss": 0.0247, "step": 12237 }, { "epoch": 8.600140548137738, "grad_norm": 0.0939122810959816, "learning_rate": 2.7602248770203796e-05, "loss": 0.0114, "step": 12238 }, { "epoch": 8.600843288826423, "grad_norm": 0.15253159403800964, "learning_rate": 2.760178027641134e-05, "loss": 0.0237, "step": 12239 }, { "epoch": 8.60154602951511, "grad_norm": 0.22150981426239014, "learning_rate": 2.760131178261888e-05, "loss": 0.0088, "step": 12240 }, { "epoch": 8.602248770203795, "grad_norm": 0.09828251600265503, "learning_rate": 2.7600843288826423e-05, "loss": 0.011, "step": 12241 }, { "epoch": 8.602951510892481, "grad_norm": 0.17368538677692413, "learning_rate": 2.7600374795033967e-05, "loss": 0.0376, "step": 12242 }, { "epoch": 8.603654251581167, "grad_norm": 0.10237167030572891, "learning_rate": 2.759990630124151e-05, "loss": 0.0252, "step": 12243 }, { "epoch": 8.604356992269853, "grad_norm": 0.09253427386283875, "learning_rate": 2.759943780744905e-05, "loss": 0.0107, "step": 12244 }, { "epoch": 8.605059732958539, "grad_norm": 0.21091540157794952, "learning_rate": 2.7598969313656595e-05, "loss": 0.0296, "step": 12245 }, { "epoch": 8.605762473647225, "grad_norm": 0.15243016183376312, "learning_rate": 2.759850081986414e-05, "loss": 0.0171, "step": 12246 }, { "epoch": 8.60646521433591, "grad_norm": 0.13307559490203857, "learning_rate": 2.7598032326071682e-05, "loss": 0.0203, "step": 12247 }, { "epoch": 8.607167955024597, "grad_norm": 0.20151999592781067, "learning_rate": 2.7597563832279226e-05, "loss": 0.0136, "step": 12248 }, { "epoch": 8.607870695713281, "grad_norm": 0.2084667980670929, "learning_rate": 2.7597095338486763e-05, "loss": 0.0243, "step": 12249 }, { "epoch": 8.608573436401969, "grad_norm": 0.5689617395401001, "learning_rate": 2.7596626844694307e-05, "loss": 0.0406, "step": 12250 }, { "epoch": 8.609276177090653, "grad_norm": 0.20988011360168457, "learning_rate": 2.759615835090185e-05, "loss": 0.0167, "step": 12251 }, { "epoch": 8.609978917779339, "grad_norm": 0.15508367121219635, "learning_rate": 2.7595689857109394e-05, "loss": 0.0266, "step": 12252 }, { "epoch": 8.610681658468025, "grad_norm": 0.2657913267612457, "learning_rate": 2.7595221363316934e-05, "loss": 0.0467, "step": 12253 }, { "epoch": 8.61138439915671, "grad_norm": 0.3397749364376068, "learning_rate": 2.7594752869524478e-05, "loss": 0.0458, "step": 12254 }, { "epoch": 8.612087139845396, "grad_norm": 0.292980432510376, "learning_rate": 2.7594284375732022e-05, "loss": 0.089, "step": 12255 }, { "epoch": 8.612789880534082, "grad_norm": 0.580999493598938, "learning_rate": 2.7593815881939566e-05, "loss": 0.1152, "step": 12256 }, { "epoch": 8.613492621222768, "grad_norm": 0.7126258015632629, "learning_rate": 2.7593347388147106e-05, "loss": 0.1522, "step": 12257 }, { "epoch": 8.614195361911454, "grad_norm": 0.9744446873664856, "learning_rate": 2.759287889435465e-05, "loss": 0.1678, "step": 12258 }, { "epoch": 8.61489810260014, "grad_norm": 1.0550354719161987, "learning_rate": 2.7592410400562193e-05, "loss": 0.19, "step": 12259 }, { "epoch": 8.615600843288826, "grad_norm": 0.3611619472503662, "learning_rate": 2.7591941906769737e-05, "loss": 0.0674, "step": 12260 }, { "epoch": 8.616303583977512, "grad_norm": 0.14432664215564728, "learning_rate": 2.759147341297728e-05, "loss": 0.0228, "step": 12261 }, { "epoch": 8.617006324666198, "grad_norm": 0.14357665181159973, "learning_rate": 2.759100491918482e-05, "loss": 0.019, "step": 12262 }, { "epoch": 8.617709065354884, "grad_norm": 0.1161438375711441, "learning_rate": 2.7590536425392365e-05, "loss": 0.013, "step": 12263 }, { "epoch": 8.61841180604357, "grad_norm": 0.8154304027557373, "learning_rate": 2.759006793159991e-05, "loss": 0.0262, "step": 12264 }, { "epoch": 8.619114546732256, "grad_norm": 0.18118232488632202, "learning_rate": 2.7589599437807452e-05, "loss": 0.0138, "step": 12265 }, { "epoch": 8.619817287420942, "grad_norm": 0.11664234846830368, "learning_rate": 2.7589130944014993e-05, "loss": 0.0191, "step": 12266 }, { "epoch": 8.620520028109627, "grad_norm": 0.39642009139060974, "learning_rate": 2.7588662450222533e-05, "loss": 0.0211, "step": 12267 }, { "epoch": 8.621222768798313, "grad_norm": 0.1259097307920456, "learning_rate": 2.7588193956430077e-05, "loss": 0.0221, "step": 12268 }, { "epoch": 8.621925509487, "grad_norm": 0.10396011173725128, "learning_rate": 2.758772546263762e-05, "loss": 0.0125, "step": 12269 }, { "epoch": 8.622628250175685, "grad_norm": 0.19456730782985687, "learning_rate": 2.758725696884516e-05, "loss": 0.0318, "step": 12270 }, { "epoch": 8.623330990864371, "grad_norm": 0.12339744716882706, "learning_rate": 2.7586788475052705e-05, "loss": 0.0173, "step": 12271 }, { "epoch": 8.624033731553057, "grad_norm": 0.19224469363689423, "learning_rate": 2.758631998126025e-05, "loss": 0.0428, "step": 12272 }, { "epoch": 8.624736472241743, "grad_norm": 0.1938212364912033, "learning_rate": 2.7585851487467792e-05, "loss": 0.0172, "step": 12273 }, { "epoch": 8.625439212930429, "grad_norm": 0.2069004327058792, "learning_rate": 2.7585382993675336e-05, "loss": 0.0361, "step": 12274 }, { "epoch": 8.626141953619115, "grad_norm": 1.8809173107147217, "learning_rate": 2.7584914499882876e-05, "loss": 0.0394, "step": 12275 }, { "epoch": 8.6268446943078, "grad_norm": 0.36113619804382324, "learning_rate": 2.758444600609042e-05, "loss": 0.0145, "step": 12276 }, { "epoch": 8.627547434996487, "grad_norm": 0.44439736008644104, "learning_rate": 2.7583977512297964e-05, "loss": 0.046, "step": 12277 }, { "epoch": 8.628250175685173, "grad_norm": 0.17116329073905945, "learning_rate": 2.7583509018505507e-05, "loss": 0.0432, "step": 12278 }, { "epoch": 8.628952916373859, "grad_norm": 0.3292076885700226, "learning_rate": 2.7583040524713048e-05, "loss": 0.0577, "step": 12279 }, { "epoch": 8.629655657062544, "grad_norm": 0.3082815706729889, "learning_rate": 2.758257203092059e-05, "loss": 0.0516, "step": 12280 }, { "epoch": 8.63035839775123, "grad_norm": 0.5566632151603699, "learning_rate": 2.7582103537128135e-05, "loss": 0.1108, "step": 12281 }, { "epoch": 8.631061138439916, "grad_norm": 1.305970549583435, "learning_rate": 2.758163504333568e-05, "loss": 0.1899, "step": 12282 }, { "epoch": 8.631763879128602, "grad_norm": 1.231845498085022, "learning_rate": 2.758116654954322e-05, "loss": 0.1865, "step": 12283 }, { "epoch": 8.632466619817288, "grad_norm": 2.647547960281372, "learning_rate": 2.758069805575076e-05, "loss": 0.2385, "step": 12284 }, { "epoch": 8.633169360505974, "grad_norm": 0.16217105090618134, "learning_rate": 2.7580229561958303e-05, "loss": 0.0752, "step": 12285 }, { "epoch": 8.63387210119466, "grad_norm": 1.1911078691482544, "learning_rate": 2.7579761068165847e-05, "loss": 0.0443, "step": 12286 }, { "epoch": 8.634574841883346, "grad_norm": 0.13588052988052368, "learning_rate": 2.757929257437339e-05, "loss": 0.0296, "step": 12287 }, { "epoch": 8.63527758257203, "grad_norm": 0.17129427194595337, "learning_rate": 2.757882408058093e-05, "loss": 0.0397, "step": 12288 }, { "epoch": 8.635980323260716, "grad_norm": 0.12813180685043335, "learning_rate": 2.7578355586788475e-05, "loss": 0.0168, "step": 12289 }, { "epoch": 8.636683063949402, "grad_norm": 0.1152791753411293, "learning_rate": 2.757788709299602e-05, "loss": 0.0222, "step": 12290 }, { "epoch": 8.637385804638088, "grad_norm": 0.11636149883270264, "learning_rate": 2.7577418599203562e-05, "loss": 0.0097, "step": 12291 }, { "epoch": 8.638088545326774, "grad_norm": 0.1562383472919464, "learning_rate": 2.7576950105411103e-05, "loss": 0.0278, "step": 12292 }, { "epoch": 8.63879128601546, "grad_norm": 0.10664365440607071, "learning_rate": 2.7576481611618646e-05, "loss": 0.0219, "step": 12293 }, { "epoch": 8.639494026704146, "grad_norm": 0.13709764182567596, "learning_rate": 2.757601311782619e-05, "loss": 0.0117, "step": 12294 }, { "epoch": 8.640196767392831, "grad_norm": 0.18981942534446716, "learning_rate": 2.7575544624033734e-05, "loss": 0.0264, "step": 12295 }, { "epoch": 8.640899508081517, "grad_norm": 0.3685530424118042, "learning_rate": 2.7575076130241274e-05, "loss": 0.0194, "step": 12296 }, { "epoch": 8.641602248770203, "grad_norm": 0.20521137118339539, "learning_rate": 2.7574607636448818e-05, "loss": 0.025, "step": 12297 }, { "epoch": 8.64230498945889, "grad_norm": 0.2992567718029022, "learning_rate": 2.757413914265636e-05, "loss": 0.035, "step": 12298 }, { "epoch": 8.643007730147575, "grad_norm": 0.3133951723575592, "learning_rate": 2.7573670648863905e-05, "loss": 0.0323, "step": 12299 }, { "epoch": 8.643710470836261, "grad_norm": 0.19806933403015137, "learning_rate": 2.757320215507145e-05, "loss": 0.0415, "step": 12300 }, { "epoch": 8.644413211524947, "grad_norm": 0.1687469184398651, "learning_rate": 2.7572733661278986e-05, "loss": 0.0247, "step": 12301 }, { "epoch": 8.645115952213633, "grad_norm": 0.8464400768280029, "learning_rate": 2.757226516748653e-05, "loss": 0.05, "step": 12302 }, { "epoch": 8.645818692902319, "grad_norm": 0.5194679498672485, "learning_rate": 2.7571796673694073e-05, "loss": 0.0559, "step": 12303 }, { "epoch": 8.646521433591005, "grad_norm": 0.2663823962211609, "learning_rate": 2.7571328179901617e-05, "loss": 0.0589, "step": 12304 }, { "epoch": 8.64722417427969, "grad_norm": 0.3378388583660126, "learning_rate": 2.7570859686109157e-05, "loss": 0.0963, "step": 12305 }, { "epoch": 8.647926914968377, "grad_norm": 0.6851024627685547, "learning_rate": 2.75703911923167e-05, "loss": 0.1488, "step": 12306 }, { "epoch": 8.648629655657063, "grad_norm": 0.7358816266059875, "learning_rate": 2.7569922698524245e-05, "loss": 0.1793, "step": 12307 }, { "epoch": 8.649332396345748, "grad_norm": 2.2824976444244385, "learning_rate": 2.756945420473179e-05, "loss": 0.198, "step": 12308 }, { "epoch": 8.650035137034434, "grad_norm": 0.9821956753730774, "learning_rate": 2.7568985710939332e-05, "loss": 0.2393, "step": 12309 }, { "epoch": 8.65073787772312, "grad_norm": 0.35449492931365967, "learning_rate": 2.7568517217146873e-05, "loss": 0.0713, "step": 12310 }, { "epoch": 8.651440618411806, "grad_norm": 0.15322276949882507, "learning_rate": 2.7568048723354416e-05, "loss": 0.0342, "step": 12311 }, { "epoch": 8.652143359100492, "grad_norm": 0.15960291028022766, "learning_rate": 2.756758022956196e-05, "loss": 0.0249, "step": 12312 }, { "epoch": 8.652846099789178, "grad_norm": 0.4089638888835907, "learning_rate": 2.7567111735769504e-05, "loss": 0.024, "step": 12313 }, { "epoch": 8.653548840477864, "grad_norm": 0.14178937673568726, "learning_rate": 2.7566643241977044e-05, "loss": 0.0188, "step": 12314 }, { "epoch": 8.65425158116655, "grad_norm": 0.09724703431129456, "learning_rate": 2.7566174748184588e-05, "loss": 0.0113, "step": 12315 }, { "epoch": 8.654954321855236, "grad_norm": 0.22900806367397308, "learning_rate": 2.756570625439213e-05, "loss": 0.0251, "step": 12316 }, { "epoch": 8.655657062543922, "grad_norm": 0.2489921599626541, "learning_rate": 2.7565237760599675e-05, "loss": 0.0199, "step": 12317 }, { "epoch": 8.656359803232608, "grad_norm": 0.12642033398151398, "learning_rate": 2.7564769266807216e-05, "loss": 0.0176, "step": 12318 }, { "epoch": 8.657062543921294, "grad_norm": 0.12563638389110565, "learning_rate": 2.7564300773014756e-05, "loss": 0.0113, "step": 12319 }, { "epoch": 8.65776528460998, "grad_norm": 0.1864279806613922, "learning_rate": 2.75638322792223e-05, "loss": 0.0184, "step": 12320 }, { "epoch": 8.658468025298665, "grad_norm": 0.20900927484035492, "learning_rate": 2.7563363785429843e-05, "loss": 0.0145, "step": 12321 }, { "epoch": 8.659170765987351, "grad_norm": 0.311470091342926, "learning_rate": 2.7562895291637387e-05, "loss": 0.0197, "step": 12322 }, { "epoch": 8.659873506676037, "grad_norm": 0.12923268973827362, "learning_rate": 2.7562426797844927e-05, "loss": 0.0178, "step": 12323 }, { "epoch": 8.660576247364723, "grad_norm": 0.34321969747543335, "learning_rate": 2.756195830405247e-05, "loss": 0.0417, "step": 12324 }, { "epoch": 8.66127898805341, "grad_norm": 0.25722646713256836, "learning_rate": 2.7561489810260015e-05, "loss": 0.0559, "step": 12325 }, { "epoch": 8.661981728742095, "grad_norm": 0.14761124551296234, "learning_rate": 2.756102131646756e-05, "loss": 0.0285, "step": 12326 }, { "epoch": 8.66268446943078, "grad_norm": 0.29616793990135193, "learning_rate": 2.75605528226751e-05, "loss": 0.0382, "step": 12327 }, { "epoch": 8.663387210119465, "grad_norm": 0.1768355518579483, "learning_rate": 2.7560084328882643e-05, "loss": 0.0285, "step": 12328 }, { "epoch": 8.664089950808151, "grad_norm": 0.43629714846611023, "learning_rate": 2.7559615835090186e-05, "loss": 0.0747, "step": 12329 }, { "epoch": 8.664792691496837, "grad_norm": 0.31522366404533386, "learning_rate": 2.755914734129773e-05, "loss": 0.0574, "step": 12330 }, { "epoch": 8.665495432185523, "grad_norm": 0.8513815402984619, "learning_rate": 2.755867884750527e-05, "loss": 0.0908, "step": 12331 }, { "epoch": 8.666198172874209, "grad_norm": 1.0407392978668213, "learning_rate": 2.7558210353712814e-05, "loss": 0.1423, "step": 12332 }, { "epoch": 8.666900913562895, "grad_norm": 1.1318259239196777, "learning_rate": 2.7557741859920358e-05, "loss": 0.1795, "step": 12333 }, { "epoch": 8.66760365425158, "grad_norm": 1.9428777694702148, "learning_rate": 2.7557273366127902e-05, "loss": 0.2441, "step": 12334 }, { "epoch": 8.668306394940267, "grad_norm": 0.3095157742500305, "learning_rate": 2.7556804872335445e-05, "loss": 0.0745, "step": 12335 }, { "epoch": 8.669009135628952, "grad_norm": 0.19540569186210632, "learning_rate": 2.7556336378542982e-05, "loss": 0.0291, "step": 12336 }, { "epoch": 8.669711876317638, "grad_norm": 0.1388939619064331, "learning_rate": 2.7555867884750526e-05, "loss": 0.0297, "step": 12337 }, { "epoch": 8.670414617006324, "grad_norm": 0.15679997205734253, "learning_rate": 2.755539939095807e-05, "loss": 0.013, "step": 12338 }, { "epoch": 8.67111735769501, "grad_norm": 0.22508284449577332, "learning_rate": 2.7554930897165614e-05, "loss": 0.0217, "step": 12339 }, { "epoch": 8.671820098383696, "grad_norm": 0.19190643727779388, "learning_rate": 2.7554462403373154e-05, "loss": 0.0127, "step": 12340 }, { "epoch": 8.672522839072382, "grad_norm": 0.20287060737609863, "learning_rate": 2.7553993909580698e-05, "loss": 0.0193, "step": 12341 }, { "epoch": 8.673225579761068, "grad_norm": 0.17378756403923035, "learning_rate": 2.755352541578824e-05, "loss": 0.0273, "step": 12342 }, { "epoch": 8.673928320449754, "grad_norm": 0.25243374705314636, "learning_rate": 2.7553056921995785e-05, "loss": 0.0177, "step": 12343 }, { "epoch": 8.67463106113844, "grad_norm": 0.09532121568918228, "learning_rate": 2.7552588428203325e-05, "loss": 0.0123, "step": 12344 }, { "epoch": 8.675333801827126, "grad_norm": 0.10300374031066895, "learning_rate": 2.755211993441087e-05, "loss": 0.0168, "step": 12345 }, { "epoch": 8.676036542515812, "grad_norm": 0.10346812754869461, "learning_rate": 2.7551651440618413e-05, "loss": 0.0122, "step": 12346 }, { "epoch": 8.676739283204498, "grad_norm": 0.14810866117477417, "learning_rate": 2.7551182946825957e-05, "loss": 0.0295, "step": 12347 }, { "epoch": 8.677442023893184, "grad_norm": 0.2876105010509491, "learning_rate": 2.75507144530335e-05, "loss": 0.016, "step": 12348 }, { "epoch": 8.67814476458187, "grad_norm": 0.21661534905433655, "learning_rate": 2.755024595924104e-05, "loss": 0.0381, "step": 12349 }, { "epoch": 8.678847505270555, "grad_norm": 0.17015771567821503, "learning_rate": 2.7549777465448584e-05, "loss": 0.0285, "step": 12350 }, { "epoch": 8.679550245959241, "grad_norm": 0.13350282609462738, "learning_rate": 2.7549308971656128e-05, "loss": 0.0137, "step": 12351 }, { "epoch": 8.680252986647927, "grad_norm": 0.20590327680110931, "learning_rate": 2.7548840477863672e-05, "loss": 0.0336, "step": 12352 }, { "epoch": 8.680955727336613, "grad_norm": 0.26187729835510254, "learning_rate": 2.7548371984071212e-05, "loss": 0.0368, "step": 12353 }, { "epoch": 8.681658468025299, "grad_norm": 0.23069792985916138, "learning_rate": 2.7547903490278752e-05, "loss": 0.0559, "step": 12354 }, { "epoch": 8.682361208713985, "grad_norm": 0.3729396164417267, "learning_rate": 2.7547434996486296e-05, "loss": 0.0618, "step": 12355 }, { "epoch": 8.683063949402671, "grad_norm": 1.5069198608398438, "learning_rate": 2.754696650269384e-05, "loss": 0.1288, "step": 12356 }, { "epoch": 8.683766690091357, "grad_norm": 0.6002113819122314, "learning_rate": 2.754649800890138e-05, "loss": 0.137, "step": 12357 }, { "epoch": 8.684469430780043, "grad_norm": 1.1125991344451904, "learning_rate": 2.7546029515108924e-05, "loss": 0.175, "step": 12358 }, { "epoch": 8.685172171468729, "grad_norm": 3.1495249271392822, "learning_rate": 2.7545561021316468e-05, "loss": 0.2221, "step": 12359 }, { "epoch": 8.685874912157415, "grad_norm": 0.32479938864707947, "learning_rate": 2.754509252752401e-05, "loss": 0.092, "step": 12360 }, { "epoch": 8.6865776528461, "grad_norm": 0.12533308565616608, "learning_rate": 2.7544624033731555e-05, "loss": 0.0236, "step": 12361 }, { "epoch": 8.687280393534786, "grad_norm": 0.20880499482154846, "learning_rate": 2.7544155539939096e-05, "loss": 0.0378, "step": 12362 }, { "epoch": 8.687983134223472, "grad_norm": 0.15016596019268036, "learning_rate": 2.754368704614664e-05, "loss": 0.0216, "step": 12363 }, { "epoch": 8.688685874912156, "grad_norm": 0.2059766948223114, "learning_rate": 2.7543218552354183e-05, "loss": 0.0086, "step": 12364 }, { "epoch": 8.689388615600844, "grad_norm": 0.12852942943572998, "learning_rate": 2.7542750058561727e-05, "loss": 0.014, "step": 12365 }, { "epoch": 8.690091356289528, "grad_norm": 0.20356762409210205, "learning_rate": 2.7542281564769267e-05, "loss": 0.0295, "step": 12366 }, { "epoch": 8.690794096978214, "grad_norm": 0.11632560193538666, "learning_rate": 2.754181307097681e-05, "loss": 0.019, "step": 12367 }, { "epoch": 8.6914968376669, "grad_norm": 0.1881098598241806, "learning_rate": 2.7541344577184354e-05, "loss": 0.0155, "step": 12368 }, { "epoch": 8.692199578355586, "grad_norm": 0.1810743510723114, "learning_rate": 2.7540876083391898e-05, "loss": 0.0123, "step": 12369 }, { "epoch": 8.692902319044272, "grad_norm": 0.18187077343463898, "learning_rate": 2.754040758959944e-05, "loss": 0.0243, "step": 12370 }, { "epoch": 8.693605059732958, "grad_norm": 0.11110500991344452, "learning_rate": 2.753993909580698e-05, "loss": 0.0178, "step": 12371 }, { "epoch": 8.694307800421644, "grad_norm": 0.13726267218589783, "learning_rate": 2.7539470602014523e-05, "loss": 0.0233, "step": 12372 }, { "epoch": 8.69501054111033, "grad_norm": 0.2054649144411087, "learning_rate": 2.7539002108222066e-05, "loss": 0.0161, "step": 12373 }, { "epoch": 8.695713281799016, "grad_norm": 0.15402758121490479, "learning_rate": 2.753853361442961e-05, "loss": 0.0374, "step": 12374 }, { "epoch": 8.696416022487702, "grad_norm": 0.3956260085105896, "learning_rate": 2.753806512063715e-05, "loss": 0.0299, "step": 12375 }, { "epoch": 8.697118763176388, "grad_norm": 0.15521332621574402, "learning_rate": 2.7537596626844694e-05, "loss": 0.0248, "step": 12376 }, { "epoch": 8.697821503865073, "grad_norm": 0.41614559292793274, "learning_rate": 2.7537128133052238e-05, "loss": 0.0311, "step": 12377 }, { "epoch": 8.69852424455376, "grad_norm": 0.3082752525806427, "learning_rate": 2.753665963925978e-05, "loss": 0.0564, "step": 12378 }, { "epoch": 8.699226985242445, "grad_norm": 0.2998494505882263, "learning_rate": 2.7536191145467322e-05, "loss": 0.0517, "step": 12379 }, { "epoch": 8.699929725931131, "grad_norm": 0.47263240814208984, "learning_rate": 2.7535722651674866e-05, "loss": 0.0986, "step": 12380 }, { "epoch": 8.700632466619817, "grad_norm": 0.5901572704315186, "learning_rate": 2.753525415788241e-05, "loss": 0.0921, "step": 12381 }, { "epoch": 8.701335207308503, "grad_norm": 0.6545228958129883, "learning_rate": 2.7534785664089953e-05, "loss": 0.1844, "step": 12382 }, { "epoch": 8.702037947997189, "grad_norm": 1.0119110345840454, "learning_rate": 2.7534317170297493e-05, "loss": 0.2041, "step": 12383 }, { "epoch": 8.702740688685875, "grad_norm": 3.184286117553711, "learning_rate": 2.7533848676505037e-05, "loss": 0.2019, "step": 12384 }, { "epoch": 8.70344342937456, "grad_norm": 0.3239738941192627, "learning_rate": 2.753338018271258e-05, "loss": 0.0654, "step": 12385 }, { "epoch": 8.704146170063247, "grad_norm": 0.15370330214500427, "learning_rate": 2.7532911688920125e-05, "loss": 0.0273, "step": 12386 }, { "epoch": 8.704848910751933, "grad_norm": 0.35706669092178345, "learning_rate": 2.753244319512767e-05, "loss": 0.0471, "step": 12387 }, { "epoch": 8.705551651440619, "grad_norm": 0.2608467936515808, "learning_rate": 2.753197470133521e-05, "loss": 0.0205, "step": 12388 }, { "epoch": 8.706254392129305, "grad_norm": 0.09409386664628983, "learning_rate": 2.753150620754275e-05, "loss": 0.0192, "step": 12389 }, { "epoch": 8.70695713281799, "grad_norm": 0.0857650563120842, "learning_rate": 2.7531037713750293e-05, "loss": 0.0086, "step": 12390 }, { "epoch": 8.707659873506676, "grad_norm": 0.11092487722635269, "learning_rate": 2.7530569219957836e-05, "loss": 0.0209, "step": 12391 }, { "epoch": 8.708362614195362, "grad_norm": 0.16782307624816895, "learning_rate": 2.7530100726165377e-05, "loss": 0.0182, "step": 12392 }, { "epoch": 8.709065354884048, "grad_norm": 0.151373028755188, "learning_rate": 2.752963223237292e-05, "loss": 0.0136, "step": 12393 }, { "epoch": 8.709768095572734, "grad_norm": 0.13994723558425903, "learning_rate": 2.7529163738580464e-05, "loss": 0.0258, "step": 12394 }, { "epoch": 8.71047083626142, "grad_norm": 0.20472100377082825, "learning_rate": 2.7528695244788008e-05, "loss": 0.0289, "step": 12395 }, { "epoch": 8.711173576950106, "grad_norm": 0.200114905834198, "learning_rate": 2.7528226750995548e-05, "loss": 0.0186, "step": 12396 }, { "epoch": 8.711876317638792, "grad_norm": 0.13575442135334015, "learning_rate": 2.7527758257203092e-05, "loss": 0.0374, "step": 12397 }, { "epoch": 8.712579058327478, "grad_norm": 0.17570289969444275, "learning_rate": 2.7527289763410636e-05, "loss": 0.0107, "step": 12398 }, { "epoch": 8.713281799016164, "grad_norm": 0.4341607689857483, "learning_rate": 2.752682126961818e-05, "loss": 0.0117, "step": 12399 }, { "epoch": 8.71398453970485, "grad_norm": 0.2523852288722992, "learning_rate": 2.7526352775825723e-05, "loss": 0.0491, "step": 12400 }, { "epoch": 8.714687280393536, "grad_norm": 0.20349706709384918, "learning_rate": 2.7525884282033264e-05, "loss": 0.0185, "step": 12401 }, { "epoch": 8.715390021082221, "grad_norm": 0.3129548132419586, "learning_rate": 2.7525415788240807e-05, "loss": 0.0341, "step": 12402 }, { "epoch": 8.716092761770906, "grad_norm": 0.42128270864486694, "learning_rate": 2.752494729444835e-05, "loss": 0.0356, "step": 12403 }, { "epoch": 8.716795502459593, "grad_norm": 0.8349596261978149, "learning_rate": 2.7524478800655895e-05, "loss": 0.0432, "step": 12404 }, { "epoch": 8.717498243148277, "grad_norm": 0.25002655386924744, "learning_rate": 2.7524010306863435e-05, "loss": 0.057, "step": 12405 }, { "epoch": 8.718200983836963, "grad_norm": 0.5572519898414612, "learning_rate": 2.7523541813070975e-05, "loss": 0.1197, "step": 12406 }, { "epoch": 8.71890372452565, "grad_norm": 0.7468071579933167, "learning_rate": 2.752307331927852e-05, "loss": 0.145, "step": 12407 }, { "epoch": 8.719606465214335, "grad_norm": 1.2133796215057373, "learning_rate": 2.7522604825486063e-05, "loss": 0.1852, "step": 12408 }, { "epoch": 8.720309205903021, "grad_norm": 1.370693325996399, "learning_rate": 2.7522136331693603e-05, "loss": 0.2558, "step": 12409 }, { "epoch": 8.721011946591707, "grad_norm": 0.2504313290119171, "learning_rate": 2.7521667837901147e-05, "loss": 0.0677, "step": 12410 }, { "epoch": 8.721714687280393, "grad_norm": 0.13798032701015472, "learning_rate": 2.752119934410869e-05, "loss": 0.0257, "step": 12411 }, { "epoch": 8.722417427969079, "grad_norm": 0.25991886854171753, "learning_rate": 2.7520730850316234e-05, "loss": 0.0226, "step": 12412 }, { "epoch": 8.723120168657765, "grad_norm": 0.10361813008785248, "learning_rate": 2.7520262356523778e-05, "loss": 0.0172, "step": 12413 }, { "epoch": 8.72382290934645, "grad_norm": 0.09746455401182175, "learning_rate": 2.751979386273132e-05, "loss": 0.0115, "step": 12414 }, { "epoch": 8.724525650035137, "grad_norm": 0.6052478551864624, "learning_rate": 2.7519325368938862e-05, "loss": 0.0236, "step": 12415 }, { "epoch": 8.725228390723823, "grad_norm": 0.10577581077814102, "learning_rate": 2.7518856875146406e-05, "loss": 0.0175, "step": 12416 }, { "epoch": 8.725931131412509, "grad_norm": 0.08262309432029724, "learning_rate": 2.751838838135395e-05, "loss": 0.0132, "step": 12417 }, { "epoch": 8.726633872101194, "grad_norm": 0.12877580523490906, "learning_rate": 2.751791988756149e-05, "loss": 0.0171, "step": 12418 }, { "epoch": 8.72733661278988, "grad_norm": 0.11956026405096054, "learning_rate": 2.7517451393769034e-05, "loss": 0.0141, "step": 12419 }, { "epoch": 8.728039353478566, "grad_norm": 0.12117037922143936, "learning_rate": 2.7516982899976577e-05, "loss": 0.0247, "step": 12420 }, { "epoch": 8.728742094167252, "grad_norm": 0.07408057153224945, "learning_rate": 2.751651440618412e-05, "loss": 0.0062, "step": 12421 }, { "epoch": 8.729444834855938, "grad_norm": 0.20324237644672394, "learning_rate": 2.751604591239166e-05, "loss": 0.043, "step": 12422 }, { "epoch": 8.730147575544624, "grad_norm": 0.16031938791275024, "learning_rate": 2.7515577418599202e-05, "loss": 0.0198, "step": 12423 }, { "epoch": 8.73085031623331, "grad_norm": 0.27298784255981445, "learning_rate": 2.7515108924806745e-05, "loss": 0.0496, "step": 12424 }, { "epoch": 8.731553056921996, "grad_norm": 0.17123201489448547, "learning_rate": 2.751464043101429e-05, "loss": 0.0402, "step": 12425 }, { "epoch": 8.732255797610682, "grad_norm": 0.2573363482952118, "learning_rate": 2.7514171937221833e-05, "loss": 0.0192, "step": 12426 }, { "epoch": 8.732958538299368, "grad_norm": 1.4157671928405762, "learning_rate": 2.7513703443429373e-05, "loss": 0.0708, "step": 12427 }, { "epoch": 8.733661278988054, "grad_norm": 0.3472272753715515, "learning_rate": 2.7513234949636917e-05, "loss": 0.0365, "step": 12428 }, { "epoch": 8.73436401967674, "grad_norm": 1.1759791374206543, "learning_rate": 2.751276645584446e-05, "loss": 0.0536, "step": 12429 }, { "epoch": 8.735066760365426, "grad_norm": 0.5206289887428284, "learning_rate": 2.7512297962052004e-05, "loss": 0.0692, "step": 12430 }, { "epoch": 8.735769501054111, "grad_norm": 0.39451417326927185, "learning_rate": 2.7511829468259545e-05, "loss": 0.123, "step": 12431 }, { "epoch": 8.736472241742797, "grad_norm": 0.6263315081596375, "learning_rate": 2.751136097446709e-05, "loss": 0.1445, "step": 12432 }, { "epoch": 8.737174982431483, "grad_norm": 0.9226484298706055, "learning_rate": 2.7510892480674632e-05, "loss": 0.2001, "step": 12433 }, { "epoch": 8.73787772312017, "grad_norm": 1.1523348093032837, "learning_rate": 2.7510423986882176e-05, "loss": 0.2569, "step": 12434 }, { "epoch": 8.738580463808855, "grad_norm": 0.21804942190647125, "learning_rate": 2.7509955493089716e-05, "loss": 0.0627, "step": 12435 }, { "epoch": 8.739283204497541, "grad_norm": 0.21655943989753723, "learning_rate": 2.750948699929726e-05, "loss": 0.0169, "step": 12436 }, { "epoch": 8.739985945186227, "grad_norm": 0.11956629902124405, "learning_rate": 2.7509018505504804e-05, "loss": 0.0155, "step": 12437 }, { "epoch": 8.740688685874913, "grad_norm": 0.22159573435783386, "learning_rate": 2.7508550011712347e-05, "loss": 0.045, "step": 12438 }, { "epoch": 8.741391426563599, "grad_norm": 0.17670148611068726, "learning_rate": 2.750808151791989e-05, "loss": 0.031, "step": 12439 }, { "epoch": 8.742094167252285, "grad_norm": 0.17302463948726654, "learning_rate": 2.750761302412743e-05, "loss": 0.0141, "step": 12440 }, { "epoch": 8.74279690794097, "grad_norm": 0.09529739618301392, "learning_rate": 2.7507144530334972e-05, "loss": 0.0121, "step": 12441 }, { "epoch": 8.743499648629655, "grad_norm": 0.29024478793144226, "learning_rate": 2.7506676036542516e-05, "loss": 0.0207, "step": 12442 }, { "epoch": 8.74420238931834, "grad_norm": 0.1434393674135208, "learning_rate": 2.750620754275006e-05, "loss": 0.0259, "step": 12443 }, { "epoch": 8.744905130007027, "grad_norm": 0.3044688105583191, "learning_rate": 2.75057390489576e-05, "loss": 0.0095, "step": 12444 }, { "epoch": 8.745607870695713, "grad_norm": 0.14303024113178253, "learning_rate": 2.7505270555165143e-05, "loss": 0.0191, "step": 12445 }, { "epoch": 8.746310611384398, "grad_norm": 0.12713287770748138, "learning_rate": 2.7504802061372687e-05, "loss": 0.0242, "step": 12446 }, { "epoch": 8.747013352073084, "grad_norm": 0.23629525303840637, "learning_rate": 2.750433356758023e-05, "loss": 0.0255, "step": 12447 }, { "epoch": 8.74771609276177, "grad_norm": 0.1716039776802063, "learning_rate": 2.750386507378777e-05, "loss": 0.0237, "step": 12448 }, { "epoch": 8.748418833450456, "grad_norm": 0.1642375886440277, "learning_rate": 2.7503396579995315e-05, "loss": 0.027, "step": 12449 }, { "epoch": 8.749121574139142, "grad_norm": 0.553109347820282, "learning_rate": 2.750292808620286e-05, "loss": 0.0372, "step": 12450 }, { "epoch": 8.749824314827828, "grad_norm": 0.2639562785625458, "learning_rate": 2.7502459592410402e-05, "loss": 0.0315, "step": 12451 }, { "epoch": 8.750527055516514, "grad_norm": 0.5867969989776611, "learning_rate": 2.7501991098617946e-05, "loss": 0.0396, "step": 12452 }, { "epoch": 8.7512297962052, "grad_norm": 0.434048056602478, "learning_rate": 2.7501522604825486e-05, "loss": 0.0331, "step": 12453 }, { "epoch": 8.751932536893886, "grad_norm": 0.33756640553474426, "learning_rate": 2.750105411103303e-05, "loss": 0.07, "step": 12454 }, { "epoch": 8.752635277582572, "grad_norm": 0.3812093734741211, "learning_rate": 2.7500585617240574e-05, "loss": 0.0987, "step": 12455 }, { "epoch": 8.753338018271258, "grad_norm": 0.5838929414749146, "learning_rate": 2.7500117123448118e-05, "loss": 0.1285, "step": 12456 }, { "epoch": 8.754040758959944, "grad_norm": 0.6237568855285645, "learning_rate": 2.7499648629655658e-05, "loss": 0.1523, "step": 12457 }, { "epoch": 8.75474349964863, "grad_norm": 0.9647257328033447, "learning_rate": 2.7499180135863198e-05, "loss": 0.1814, "step": 12458 }, { "epoch": 8.755446240337315, "grad_norm": 1.3462785482406616, "learning_rate": 2.7498711642070742e-05, "loss": 0.2442, "step": 12459 }, { "epoch": 8.756148981026001, "grad_norm": 0.3131149709224701, "learning_rate": 2.7498243148278286e-05, "loss": 0.0724, "step": 12460 }, { "epoch": 8.756851721714687, "grad_norm": 0.17670288681983948, "learning_rate": 2.7497774654485826e-05, "loss": 0.0275, "step": 12461 }, { "epoch": 8.757554462403373, "grad_norm": 0.09541445970535278, "learning_rate": 2.749730616069337e-05, "loss": 0.0203, "step": 12462 }, { "epoch": 8.75825720309206, "grad_norm": 0.11248476058244705, "learning_rate": 2.7496837666900914e-05, "loss": 0.0253, "step": 12463 }, { "epoch": 8.758959943780745, "grad_norm": 0.4889770448207855, "learning_rate": 2.7496369173108457e-05, "loss": 0.0136, "step": 12464 }, { "epoch": 8.759662684469431, "grad_norm": 0.24333599209785461, "learning_rate": 2.7495900679316e-05, "loss": 0.0174, "step": 12465 }, { "epoch": 8.760365425158117, "grad_norm": 0.1682395339012146, "learning_rate": 2.749543218552354e-05, "loss": 0.0234, "step": 12466 }, { "epoch": 8.761068165846803, "grad_norm": 0.24569633603096008, "learning_rate": 2.7494963691731085e-05, "loss": 0.0321, "step": 12467 }, { "epoch": 8.761770906535489, "grad_norm": 0.24879151582717896, "learning_rate": 2.749449519793863e-05, "loss": 0.0437, "step": 12468 }, { "epoch": 8.762473647224175, "grad_norm": 0.3482265770435333, "learning_rate": 2.7494026704146172e-05, "loss": 0.027, "step": 12469 }, { "epoch": 8.76317638791286, "grad_norm": 0.19875386357307434, "learning_rate": 2.7493558210353713e-05, "loss": 0.0453, "step": 12470 }, { "epoch": 8.763879128601546, "grad_norm": 0.10842438042163849, "learning_rate": 2.7493089716561257e-05, "loss": 0.0085, "step": 12471 }, { "epoch": 8.764581869290232, "grad_norm": 0.2369982898235321, "learning_rate": 2.74926212227688e-05, "loss": 0.0274, "step": 12472 }, { "epoch": 8.765284609978918, "grad_norm": 0.157306969165802, "learning_rate": 2.7492152728976344e-05, "loss": 0.025, "step": 12473 }, { "epoch": 8.765987350667604, "grad_norm": 0.1046488881111145, "learning_rate": 2.7491684235183884e-05, "loss": 0.017, "step": 12474 }, { "epoch": 8.76669009135629, "grad_norm": 0.17285466194152832, "learning_rate": 2.7491215741391428e-05, "loss": 0.0467, "step": 12475 }, { "epoch": 8.767392832044976, "grad_norm": 0.1925489902496338, "learning_rate": 2.749074724759897e-05, "loss": 0.0384, "step": 12476 }, { "epoch": 8.768095572733662, "grad_norm": 0.14753100275993347, "learning_rate": 2.7490278753806512e-05, "loss": 0.0215, "step": 12477 }, { "epoch": 8.768798313422348, "grad_norm": 0.3735833168029785, "learning_rate": 2.7489810260014056e-05, "loss": 0.0446, "step": 12478 }, { "epoch": 8.769501054111032, "grad_norm": 0.42818763852119446, "learning_rate": 2.7489341766221596e-05, "loss": 0.0568, "step": 12479 }, { "epoch": 8.77020379479972, "grad_norm": 0.48462116718292236, "learning_rate": 2.748887327242914e-05, "loss": 0.0856, "step": 12480 }, { "epoch": 8.770906535488404, "grad_norm": 1.5143052339553833, "learning_rate": 2.7488404778636684e-05, "loss": 0.1177, "step": 12481 }, { "epoch": 8.77160927617709, "grad_norm": 0.4800800681114197, "learning_rate": 2.7487936284844227e-05, "loss": 0.1519, "step": 12482 }, { "epoch": 8.772312016865776, "grad_norm": 0.6984863877296448, "learning_rate": 2.7487467791051768e-05, "loss": 0.1986, "step": 12483 }, { "epoch": 8.773014757554462, "grad_norm": 0.852312445640564, "learning_rate": 2.748699929725931e-05, "loss": 0.2363, "step": 12484 }, { "epoch": 8.773717498243148, "grad_norm": 0.2184598445892334, "learning_rate": 2.7486530803466855e-05, "loss": 0.0668, "step": 12485 }, { "epoch": 8.774420238931834, "grad_norm": 0.14527827501296997, "learning_rate": 2.74860623096744e-05, "loss": 0.0316, "step": 12486 }, { "epoch": 8.77512297962052, "grad_norm": 0.12226589769124985, "learning_rate": 2.748559381588194e-05, "loss": 0.0277, "step": 12487 }, { "epoch": 8.775825720309205, "grad_norm": 0.13916826248168945, "learning_rate": 2.7485125322089483e-05, "loss": 0.0264, "step": 12488 }, { "epoch": 8.776528460997891, "grad_norm": 0.13940097391605377, "learning_rate": 2.7484656828297027e-05, "loss": 0.0207, "step": 12489 }, { "epoch": 8.777231201686577, "grad_norm": 0.5438774824142456, "learning_rate": 2.748418833450457e-05, "loss": 0.0148, "step": 12490 }, { "epoch": 8.777933942375263, "grad_norm": 0.2205674648284912, "learning_rate": 2.7483719840712114e-05, "loss": 0.0331, "step": 12491 }, { "epoch": 8.778636683063949, "grad_norm": 0.14065757393836975, "learning_rate": 2.7483251346919654e-05, "loss": 0.02, "step": 12492 }, { "epoch": 8.779339423752635, "grad_norm": 0.15920676290988922, "learning_rate": 2.7482782853127195e-05, "loss": 0.0337, "step": 12493 }, { "epoch": 8.780042164441321, "grad_norm": 0.20460206270217896, "learning_rate": 2.748231435933474e-05, "loss": 0.0162, "step": 12494 }, { "epoch": 8.780744905130007, "grad_norm": 0.35324612259864807, "learning_rate": 2.7481845865542282e-05, "loss": 0.0186, "step": 12495 }, { "epoch": 8.781447645818693, "grad_norm": 0.13043563067913055, "learning_rate": 2.7481377371749823e-05, "loss": 0.0143, "step": 12496 }, { "epoch": 8.782150386507379, "grad_norm": 0.2042231559753418, "learning_rate": 2.7480908877957366e-05, "loss": 0.0409, "step": 12497 }, { "epoch": 8.782853127196065, "grad_norm": 0.13037188351154327, "learning_rate": 2.748044038416491e-05, "loss": 0.0112, "step": 12498 }, { "epoch": 8.78355586788475, "grad_norm": 0.2142050713300705, "learning_rate": 2.7479971890372454e-05, "loss": 0.0374, "step": 12499 }, { "epoch": 8.784258608573436, "grad_norm": 0.19052232801914215, "learning_rate": 2.7479503396579997e-05, "loss": 0.0353, "step": 12500 }, { "epoch": 8.784961349262122, "grad_norm": 0.1313854157924652, "learning_rate": 2.7479034902787538e-05, "loss": 0.0206, "step": 12501 }, { "epoch": 8.785664089950808, "grad_norm": 0.2810842990875244, "learning_rate": 2.747856640899508e-05, "loss": 0.032, "step": 12502 }, { "epoch": 8.786366830639494, "grad_norm": 1.5574325323104858, "learning_rate": 2.7478097915202625e-05, "loss": 0.055, "step": 12503 }, { "epoch": 8.78706957132818, "grad_norm": 0.2772204577922821, "learning_rate": 2.747762942141017e-05, "loss": 0.0708, "step": 12504 }, { "epoch": 8.787772312016866, "grad_norm": 0.9129993319511414, "learning_rate": 2.747716092761771e-05, "loss": 0.1108, "step": 12505 }, { "epoch": 8.788475052705552, "grad_norm": 0.7017931938171387, "learning_rate": 2.7476692433825253e-05, "loss": 0.1334, "step": 12506 }, { "epoch": 8.789177793394238, "grad_norm": 0.7105374932289124, "learning_rate": 2.7476223940032797e-05, "loss": 0.1736, "step": 12507 }, { "epoch": 8.789880534082924, "grad_norm": 0.8491537570953369, "learning_rate": 2.747575544624034e-05, "loss": 0.1636, "step": 12508 }, { "epoch": 8.79058327477161, "grad_norm": 4.667258262634277, "learning_rate": 2.747528695244788e-05, "loss": 0.294, "step": 12509 }, { "epoch": 8.791286015460296, "grad_norm": 0.35995498299598694, "learning_rate": 2.747481845865542e-05, "loss": 0.0827, "step": 12510 }, { "epoch": 8.791988756148982, "grad_norm": 0.2131284773349762, "learning_rate": 2.7474349964862965e-05, "loss": 0.0198, "step": 12511 }, { "epoch": 8.792691496837667, "grad_norm": 0.28457164764404297, "learning_rate": 2.747388147107051e-05, "loss": 0.0281, "step": 12512 }, { "epoch": 8.793394237526353, "grad_norm": 0.10931903123855591, "learning_rate": 2.7473412977278052e-05, "loss": 0.0149, "step": 12513 }, { "epoch": 8.79409697821504, "grad_norm": 0.10416794568300247, "learning_rate": 2.7472944483485593e-05, "loss": 0.0217, "step": 12514 }, { "epoch": 8.794799718903725, "grad_norm": 0.09286019951105118, "learning_rate": 2.7472475989693136e-05, "loss": 0.017, "step": 12515 }, { "epoch": 8.795502459592411, "grad_norm": 0.21081800758838654, "learning_rate": 2.747200749590068e-05, "loss": 0.0175, "step": 12516 }, { "epoch": 8.796205200281097, "grad_norm": 0.24780337512493134, "learning_rate": 2.7471539002108224e-05, "loss": 0.0278, "step": 12517 }, { "epoch": 8.796907940969781, "grad_norm": 0.23416149616241455, "learning_rate": 2.7471070508315764e-05, "loss": 0.0201, "step": 12518 }, { "epoch": 8.797610681658469, "grad_norm": 0.1607120782136917, "learning_rate": 2.7470602014523308e-05, "loss": 0.0193, "step": 12519 }, { "epoch": 8.798313422347153, "grad_norm": 0.17114916443824768, "learning_rate": 2.747013352073085e-05, "loss": 0.0332, "step": 12520 }, { "epoch": 8.799016163035839, "grad_norm": 0.11444017291069031, "learning_rate": 2.7469665026938395e-05, "loss": 0.0117, "step": 12521 }, { "epoch": 8.799718903724525, "grad_norm": 0.14436255395412445, "learning_rate": 2.7469196533145936e-05, "loss": 0.0269, "step": 12522 }, { "epoch": 8.80042164441321, "grad_norm": 0.3339534401893616, "learning_rate": 2.746872803935348e-05, "loss": 0.0169, "step": 12523 }, { "epoch": 8.801124385101897, "grad_norm": 0.19775186479091644, "learning_rate": 2.7468259545561023e-05, "loss": 0.0241, "step": 12524 }, { "epoch": 8.801827125790583, "grad_norm": 0.3395117223262787, "learning_rate": 2.7467791051768567e-05, "loss": 0.0361, "step": 12525 }, { "epoch": 8.802529866479269, "grad_norm": 0.21828609704971313, "learning_rate": 2.746732255797611e-05, "loss": 0.0311, "step": 12526 }, { "epoch": 8.803232607167955, "grad_norm": 0.1540652960538864, "learning_rate": 2.746685406418365e-05, "loss": 0.0298, "step": 12527 }, { "epoch": 8.80393534785664, "grad_norm": 0.315844863653183, "learning_rate": 2.746638557039119e-05, "loss": 0.0534, "step": 12528 }, { "epoch": 8.804638088545326, "grad_norm": 0.3411601483821869, "learning_rate": 2.7465917076598735e-05, "loss": 0.0688, "step": 12529 }, { "epoch": 8.805340829234012, "grad_norm": 0.5750501155853271, "learning_rate": 2.746544858280628e-05, "loss": 0.0744, "step": 12530 }, { "epoch": 8.806043569922698, "grad_norm": 0.788047730922699, "learning_rate": 2.746498008901382e-05, "loss": 0.1063, "step": 12531 }, { "epoch": 8.806746310611384, "grad_norm": 0.41057974100112915, "learning_rate": 2.7464511595221363e-05, "loss": 0.1267, "step": 12532 }, { "epoch": 8.80744905130007, "grad_norm": 1.2273151874542236, "learning_rate": 2.7464043101428907e-05, "loss": 0.1766, "step": 12533 }, { "epoch": 8.808151791988756, "grad_norm": 1.6213879585266113, "learning_rate": 2.746357460763645e-05, "loss": 0.1978, "step": 12534 }, { "epoch": 8.808854532677442, "grad_norm": 0.28099846839904785, "learning_rate": 2.746310611384399e-05, "loss": 0.059, "step": 12535 }, { "epoch": 8.809557273366128, "grad_norm": 0.10802201181650162, "learning_rate": 2.7462637620051534e-05, "loss": 0.0184, "step": 12536 }, { "epoch": 8.810260014054814, "grad_norm": 0.34101366996765137, "learning_rate": 2.7462169126259078e-05, "loss": 0.0193, "step": 12537 }, { "epoch": 8.8109627547435, "grad_norm": 0.14734159409999847, "learning_rate": 2.7461700632466622e-05, "loss": 0.0175, "step": 12538 }, { "epoch": 8.811665495432186, "grad_norm": 0.11064433306455612, "learning_rate": 2.7461232138674165e-05, "loss": 0.0173, "step": 12539 }, { "epoch": 8.812368236120872, "grad_norm": 0.13026084005832672, "learning_rate": 2.7460763644881706e-05, "loss": 0.0246, "step": 12540 }, { "epoch": 8.813070976809557, "grad_norm": 0.23699802160263062, "learning_rate": 2.746029515108925e-05, "loss": 0.0167, "step": 12541 }, { "epoch": 8.813773717498243, "grad_norm": 0.15763412415981293, "learning_rate": 2.7459826657296793e-05, "loss": 0.0209, "step": 12542 }, { "epoch": 8.81447645818693, "grad_norm": 0.17871400713920593, "learning_rate": 2.7459358163504337e-05, "loss": 0.0366, "step": 12543 }, { "epoch": 8.815179198875615, "grad_norm": 0.1796312928199768, "learning_rate": 2.7458889669711877e-05, "loss": 0.0112, "step": 12544 }, { "epoch": 8.815881939564301, "grad_norm": 0.17319229245185852, "learning_rate": 2.7458421175919418e-05, "loss": 0.0271, "step": 12545 }, { "epoch": 8.816584680252987, "grad_norm": 0.23329386115074158, "learning_rate": 2.745795268212696e-05, "loss": 0.0206, "step": 12546 }, { "epoch": 8.817287420941673, "grad_norm": 0.24403730034828186, "learning_rate": 2.7457484188334505e-05, "loss": 0.0458, "step": 12547 }, { "epoch": 8.817990161630359, "grad_norm": 0.5388445258140564, "learning_rate": 2.7457015694542045e-05, "loss": 0.0175, "step": 12548 }, { "epoch": 8.818692902319045, "grad_norm": 0.2643747627735138, "learning_rate": 2.745654720074959e-05, "loss": 0.0668, "step": 12549 }, { "epoch": 8.81939564300773, "grad_norm": 0.4540032744407654, "learning_rate": 2.7456078706957133e-05, "loss": 0.0264, "step": 12550 }, { "epoch": 8.820098383696417, "grad_norm": 0.16747312247753143, "learning_rate": 2.7455610213164677e-05, "loss": 0.0319, "step": 12551 }, { "epoch": 8.820801124385103, "grad_norm": 0.2679080069065094, "learning_rate": 2.745514171937222e-05, "loss": 0.0266, "step": 12552 }, { "epoch": 8.821503865073788, "grad_norm": 0.28711163997650146, "learning_rate": 2.745467322557976e-05, "loss": 0.0369, "step": 12553 }, { "epoch": 8.822206605762474, "grad_norm": 0.31655386090278625, "learning_rate": 2.7454204731787304e-05, "loss": 0.0705, "step": 12554 }, { "epoch": 8.82290934645116, "grad_norm": 0.24350154399871826, "learning_rate": 2.7453736237994848e-05, "loss": 0.0526, "step": 12555 }, { "epoch": 8.823612087139846, "grad_norm": 0.4438616931438446, "learning_rate": 2.7453267744202392e-05, "loss": 0.1036, "step": 12556 }, { "epoch": 8.82431482782853, "grad_norm": 0.603867769241333, "learning_rate": 2.7452799250409932e-05, "loss": 0.1416, "step": 12557 }, { "epoch": 8.825017568517218, "grad_norm": 1.6143710613250732, "learning_rate": 2.7452330756617476e-05, "loss": 0.2106, "step": 12558 }, { "epoch": 8.825720309205902, "grad_norm": 3.519824504852295, "learning_rate": 2.745186226282502e-05, "loss": 0.2382, "step": 12559 }, { "epoch": 8.826423049894588, "grad_norm": 0.6089478731155396, "learning_rate": 2.7451393769032563e-05, "loss": 0.0768, "step": 12560 }, { "epoch": 8.827125790583274, "grad_norm": 0.20589236915111542, "learning_rate": 2.7450925275240104e-05, "loss": 0.0289, "step": 12561 }, { "epoch": 8.82782853127196, "grad_norm": 0.15306326746940613, "learning_rate": 2.7450456781447647e-05, "loss": 0.0233, "step": 12562 }, { "epoch": 8.828531271960646, "grad_norm": 0.12241779267787933, "learning_rate": 2.7449988287655188e-05, "loss": 0.017, "step": 12563 }, { "epoch": 8.829234012649332, "grad_norm": 0.18874599039554596, "learning_rate": 2.744951979386273e-05, "loss": 0.0189, "step": 12564 }, { "epoch": 8.829936753338018, "grad_norm": 0.09207647293806076, "learning_rate": 2.7449051300070275e-05, "loss": 0.0086, "step": 12565 }, { "epoch": 8.830639494026704, "grad_norm": 0.14305619895458221, "learning_rate": 2.7448582806277816e-05, "loss": 0.0212, "step": 12566 }, { "epoch": 8.83134223471539, "grad_norm": 0.17714130878448486, "learning_rate": 2.744811431248536e-05, "loss": 0.0168, "step": 12567 }, { "epoch": 8.832044975404076, "grad_norm": 0.1393379271030426, "learning_rate": 2.7447645818692903e-05, "loss": 0.0202, "step": 12568 }, { "epoch": 8.832747716092761, "grad_norm": 0.1649627983570099, "learning_rate": 2.7447177324900447e-05, "loss": 0.0185, "step": 12569 }, { "epoch": 8.833450456781447, "grad_norm": 0.27823662757873535, "learning_rate": 2.7446708831107987e-05, "loss": 0.0191, "step": 12570 }, { "epoch": 8.834153197470133, "grad_norm": 0.15949074923992157, "learning_rate": 2.744624033731553e-05, "loss": 0.0161, "step": 12571 }, { "epoch": 8.83485593815882, "grad_norm": 0.1650865375995636, "learning_rate": 2.7445771843523075e-05, "loss": 0.0361, "step": 12572 }, { "epoch": 8.835558678847505, "grad_norm": 0.21278785169124603, "learning_rate": 2.7445303349730618e-05, "loss": 0.0163, "step": 12573 }, { "epoch": 8.836261419536191, "grad_norm": 0.3261761963367462, "learning_rate": 2.744483485593816e-05, "loss": 0.0162, "step": 12574 }, { "epoch": 8.836964160224877, "grad_norm": 0.136781707406044, "learning_rate": 2.7444366362145702e-05, "loss": 0.0305, "step": 12575 }, { "epoch": 8.837666900913563, "grad_norm": 0.18105080723762512, "learning_rate": 2.7443897868353246e-05, "loss": 0.0266, "step": 12576 }, { "epoch": 8.838369641602249, "grad_norm": 0.43890008330345154, "learning_rate": 2.744342937456079e-05, "loss": 0.0657, "step": 12577 }, { "epoch": 8.839072382290935, "grad_norm": 0.32597076892852783, "learning_rate": 2.7442960880768333e-05, "loss": 0.0602, "step": 12578 }, { "epoch": 8.83977512297962, "grad_norm": 0.40447813272476196, "learning_rate": 2.7442492386975874e-05, "loss": 0.068, "step": 12579 }, { "epoch": 8.840477863668307, "grad_norm": 0.3817969858646393, "learning_rate": 2.7442023893183414e-05, "loss": 0.0727, "step": 12580 }, { "epoch": 8.841180604356992, "grad_norm": 1.6256985664367676, "learning_rate": 2.7441555399390958e-05, "loss": 0.1238, "step": 12581 }, { "epoch": 8.841883345045678, "grad_norm": 0.8933383226394653, "learning_rate": 2.74410869055985e-05, "loss": 0.1692, "step": 12582 }, { "epoch": 8.842586085734364, "grad_norm": 1.5493587255477905, "learning_rate": 2.7440618411806042e-05, "loss": 0.1725, "step": 12583 }, { "epoch": 8.84328882642305, "grad_norm": 1.8523544073104858, "learning_rate": 2.7440149918013586e-05, "loss": 0.2214, "step": 12584 }, { "epoch": 8.843991567111736, "grad_norm": 0.7241846323013306, "learning_rate": 2.743968142422113e-05, "loss": 0.0829, "step": 12585 }, { "epoch": 8.844694307800422, "grad_norm": 0.4658578038215637, "learning_rate": 2.7439212930428673e-05, "loss": 0.0443, "step": 12586 }, { "epoch": 8.845397048489108, "grad_norm": 0.17019091546535492, "learning_rate": 2.7438744436636213e-05, "loss": 0.0216, "step": 12587 }, { "epoch": 8.846099789177794, "grad_norm": 0.1102680116891861, "learning_rate": 2.7438275942843757e-05, "loss": 0.0192, "step": 12588 }, { "epoch": 8.84680252986648, "grad_norm": 0.11233805119991302, "learning_rate": 2.74378074490513e-05, "loss": 0.0158, "step": 12589 }, { "epoch": 8.847505270555166, "grad_norm": 0.11775931715965271, "learning_rate": 2.7437338955258845e-05, "loss": 0.0129, "step": 12590 }, { "epoch": 8.848208011243852, "grad_norm": 0.23313309252262115, "learning_rate": 2.743687046146639e-05, "loss": 0.0195, "step": 12591 }, { "epoch": 8.848910751932538, "grad_norm": 0.21492737531661987, "learning_rate": 2.743640196767393e-05, "loss": 0.0193, "step": 12592 }, { "epoch": 8.849613492621224, "grad_norm": 0.2414754033088684, "learning_rate": 2.7435933473881472e-05, "loss": 0.0302, "step": 12593 }, { "epoch": 8.85031623330991, "grad_norm": 0.10225499421358109, "learning_rate": 2.7435464980089016e-05, "loss": 0.0177, "step": 12594 }, { "epoch": 8.851018973998595, "grad_norm": 0.30506259202957153, "learning_rate": 2.743499648629656e-05, "loss": 0.0245, "step": 12595 }, { "epoch": 8.85172171468728, "grad_norm": 0.15317444503307343, "learning_rate": 2.74345279925041e-05, "loss": 0.0129, "step": 12596 }, { "epoch": 8.852424455375965, "grad_norm": 0.1994018703699112, "learning_rate": 2.7434059498711644e-05, "loss": 0.0241, "step": 12597 }, { "epoch": 8.853127196064651, "grad_norm": 0.21941502392292023, "learning_rate": 2.7433591004919184e-05, "loss": 0.0207, "step": 12598 }, { "epoch": 8.853829936753337, "grad_norm": 0.1259562224149704, "learning_rate": 2.7433122511126728e-05, "loss": 0.0284, "step": 12599 }, { "epoch": 8.854532677442023, "grad_norm": 0.31069105863571167, "learning_rate": 2.743265401733427e-05, "loss": 0.0446, "step": 12600 }, { "epoch": 8.85523541813071, "grad_norm": 0.173010915517807, "learning_rate": 2.7432185523541812e-05, "loss": 0.02, "step": 12601 }, { "epoch": 8.855938158819395, "grad_norm": 0.5023786425590515, "learning_rate": 2.7431717029749356e-05, "loss": 0.0518, "step": 12602 }, { "epoch": 8.856640899508081, "grad_norm": 0.26091131567955017, "learning_rate": 2.74312485359569e-05, "loss": 0.0312, "step": 12603 }, { "epoch": 8.857343640196767, "grad_norm": 0.2475881725549698, "learning_rate": 2.7430780042164443e-05, "loss": 0.0407, "step": 12604 }, { "epoch": 8.858046380885453, "grad_norm": 0.5446662306785583, "learning_rate": 2.7430311548371984e-05, "loss": 0.0938, "step": 12605 }, { "epoch": 8.858749121574139, "grad_norm": 0.6567056179046631, "learning_rate": 2.7429843054579527e-05, "loss": 0.1225, "step": 12606 }, { "epoch": 8.859451862262825, "grad_norm": 1.2903167009353638, "learning_rate": 2.742937456078707e-05, "loss": 0.1568, "step": 12607 }, { "epoch": 8.86015460295151, "grad_norm": 1.4219887256622314, "learning_rate": 2.7428906066994615e-05, "loss": 0.1967, "step": 12608 }, { "epoch": 8.860857343640197, "grad_norm": 1.4764055013656616, "learning_rate": 2.7428437573202155e-05, "loss": 0.2366, "step": 12609 }, { "epoch": 8.861560084328882, "grad_norm": 0.2913193106651306, "learning_rate": 2.74279690794097e-05, "loss": 0.0788, "step": 12610 }, { "epoch": 8.862262825017568, "grad_norm": 0.15142560005187988, "learning_rate": 2.7427500585617243e-05, "loss": 0.0319, "step": 12611 }, { "epoch": 8.862965565706254, "grad_norm": 0.16971486806869507, "learning_rate": 2.7427032091824786e-05, "loss": 0.022, "step": 12612 }, { "epoch": 8.86366830639494, "grad_norm": 0.15438656508922577, "learning_rate": 2.7426563598032327e-05, "loss": 0.0194, "step": 12613 }, { "epoch": 8.864371047083626, "grad_norm": 0.2020561248064041, "learning_rate": 2.742609510423987e-05, "loss": 0.0266, "step": 12614 }, { "epoch": 8.865073787772312, "grad_norm": 0.11121494323015213, "learning_rate": 2.742562661044741e-05, "loss": 0.0117, "step": 12615 }, { "epoch": 8.865776528460998, "grad_norm": 0.17935305833816528, "learning_rate": 2.7425158116654954e-05, "loss": 0.034, "step": 12616 }, { "epoch": 8.866479269149684, "grad_norm": 0.1470237374305725, "learning_rate": 2.7424689622862498e-05, "loss": 0.0329, "step": 12617 }, { "epoch": 8.86718200983837, "grad_norm": 0.08415807783603668, "learning_rate": 2.742422112907004e-05, "loss": 0.0204, "step": 12618 }, { "epoch": 8.867884750527056, "grad_norm": 0.1588509976863861, "learning_rate": 2.7423752635277582e-05, "loss": 0.017, "step": 12619 }, { "epoch": 8.868587491215742, "grad_norm": 0.09045334160327911, "learning_rate": 2.7423284141485126e-05, "loss": 0.0178, "step": 12620 }, { "epoch": 8.869290231904428, "grad_norm": 0.11430448293685913, "learning_rate": 2.742281564769267e-05, "loss": 0.0156, "step": 12621 }, { "epoch": 8.869992972593113, "grad_norm": 0.1596895456314087, "learning_rate": 2.742234715390021e-05, "loss": 0.0195, "step": 12622 }, { "epoch": 8.8706957132818, "grad_norm": 0.18237482011318207, "learning_rate": 2.7421878660107754e-05, "loss": 0.0137, "step": 12623 }, { "epoch": 8.871398453970485, "grad_norm": 0.33554521203041077, "learning_rate": 2.7421410166315297e-05, "loss": 0.0405, "step": 12624 }, { "epoch": 8.872101194659171, "grad_norm": 0.19173681735992432, "learning_rate": 2.742094167252284e-05, "loss": 0.0248, "step": 12625 }, { "epoch": 8.872803935347857, "grad_norm": 0.12810735404491425, "learning_rate": 2.742047317873038e-05, "loss": 0.0181, "step": 12626 }, { "epoch": 8.873506676036543, "grad_norm": 0.2872281074523926, "learning_rate": 2.7420004684937925e-05, "loss": 0.044, "step": 12627 }, { "epoch": 8.874209416725229, "grad_norm": 0.21259136497974396, "learning_rate": 2.741953619114547e-05, "loss": 0.0562, "step": 12628 }, { "epoch": 8.874912157413915, "grad_norm": 0.5627405047416687, "learning_rate": 2.7419067697353013e-05, "loss": 0.0475, "step": 12629 }, { "epoch": 8.8756148981026, "grad_norm": 0.5504276156425476, "learning_rate": 2.7418599203560556e-05, "loss": 0.0947, "step": 12630 }, { "epoch": 8.876317638791287, "grad_norm": 0.3930051326751709, "learning_rate": 2.7418130709768097e-05, "loss": 0.1122, "step": 12631 }, { "epoch": 8.877020379479973, "grad_norm": 0.8438937664031982, "learning_rate": 2.7417662215975637e-05, "loss": 0.1653, "step": 12632 }, { "epoch": 8.877723120168657, "grad_norm": 1.103551983833313, "learning_rate": 2.741719372218318e-05, "loss": 0.1967, "step": 12633 }, { "epoch": 8.878425860857345, "grad_norm": 1.6252843141555786, "learning_rate": 2.7416725228390724e-05, "loss": 0.202, "step": 12634 }, { "epoch": 8.879128601546029, "grad_norm": 0.3033871054649353, "learning_rate": 2.7416256734598265e-05, "loss": 0.0625, "step": 12635 }, { "epoch": 8.879831342234715, "grad_norm": 0.10365862399339676, "learning_rate": 2.741578824080581e-05, "loss": 0.0191, "step": 12636 }, { "epoch": 8.8805340829234, "grad_norm": 0.14372876286506653, "learning_rate": 2.7415319747013352e-05, "loss": 0.0192, "step": 12637 }, { "epoch": 8.881236823612086, "grad_norm": 1.2244912385940552, "learning_rate": 2.7414851253220896e-05, "loss": 0.0162, "step": 12638 }, { "epoch": 8.881939564300772, "grad_norm": 0.09436022490262985, "learning_rate": 2.7414382759428436e-05, "loss": 0.016, "step": 12639 }, { "epoch": 8.882642304989458, "grad_norm": 0.12610238790512085, "learning_rate": 2.741391426563598e-05, "loss": 0.0089, "step": 12640 }, { "epoch": 8.883345045678144, "grad_norm": 0.23600785434246063, "learning_rate": 2.7413445771843524e-05, "loss": 0.0247, "step": 12641 }, { "epoch": 8.88404778636683, "grad_norm": 0.1978476643562317, "learning_rate": 2.7412977278051068e-05, "loss": 0.0202, "step": 12642 }, { "epoch": 8.884750527055516, "grad_norm": 0.12559938430786133, "learning_rate": 2.741250878425861e-05, "loss": 0.0196, "step": 12643 }, { "epoch": 8.885453267744202, "grad_norm": 0.11917315423488617, "learning_rate": 2.741204029046615e-05, "loss": 0.0096, "step": 12644 }, { "epoch": 8.886156008432888, "grad_norm": 0.18919232487678528, "learning_rate": 2.7411571796673695e-05, "loss": 0.0325, "step": 12645 }, { "epoch": 8.886858749121574, "grad_norm": 0.09387481212615967, "learning_rate": 2.741110330288124e-05, "loss": 0.0147, "step": 12646 }, { "epoch": 8.88756148981026, "grad_norm": 0.13896475732326508, "learning_rate": 2.7410634809088783e-05, "loss": 0.0364, "step": 12647 }, { "epoch": 8.888264230498946, "grad_norm": 0.23239675164222717, "learning_rate": 2.7410166315296323e-05, "loss": 0.0193, "step": 12648 }, { "epoch": 8.888966971187632, "grad_norm": 0.13684844970703125, "learning_rate": 2.7409697821503867e-05, "loss": 0.0304, "step": 12649 }, { "epoch": 8.889669711876317, "grad_norm": 0.2515307068824768, "learning_rate": 2.7409229327711407e-05, "loss": 0.0286, "step": 12650 }, { "epoch": 8.890372452565003, "grad_norm": 0.32987430691719055, "learning_rate": 2.740876083391895e-05, "loss": 0.036, "step": 12651 }, { "epoch": 8.89107519325369, "grad_norm": 0.193109393119812, "learning_rate": 2.740829234012649e-05, "loss": 0.0182, "step": 12652 }, { "epoch": 8.891777933942375, "grad_norm": 0.4778037667274475, "learning_rate": 2.7407823846334035e-05, "loss": 0.0555, "step": 12653 }, { "epoch": 8.892480674631061, "grad_norm": 0.298798143863678, "learning_rate": 2.740735535254158e-05, "loss": 0.0451, "step": 12654 }, { "epoch": 8.893183415319747, "grad_norm": 0.5740682482719421, "learning_rate": 2.7406886858749122e-05, "loss": 0.0989, "step": 12655 }, { "epoch": 8.893886156008433, "grad_norm": 0.5441758036613464, "learning_rate": 2.7406418364956666e-05, "loss": 0.1071, "step": 12656 }, { "epoch": 8.894588896697119, "grad_norm": 0.8830281496047974, "learning_rate": 2.7405949871164206e-05, "loss": 0.1299, "step": 12657 }, { "epoch": 8.895291637385805, "grad_norm": 2.2011473178863525, "learning_rate": 2.740548137737175e-05, "loss": 0.1697, "step": 12658 }, { "epoch": 8.89599437807449, "grad_norm": 1.2400370836257935, "learning_rate": 2.7405012883579294e-05, "loss": 0.2246, "step": 12659 }, { "epoch": 8.896697118763177, "grad_norm": 0.30994656682014465, "learning_rate": 2.7404544389786838e-05, "loss": 0.1043, "step": 12660 }, { "epoch": 8.897399859451863, "grad_norm": 0.1034136414527893, "learning_rate": 2.7404075895994378e-05, "loss": 0.0201, "step": 12661 }, { "epoch": 8.898102600140549, "grad_norm": 0.12227854132652283, "learning_rate": 2.740360740220192e-05, "loss": 0.0249, "step": 12662 }, { "epoch": 8.898805340829234, "grad_norm": 0.20844680070877075, "learning_rate": 2.7403138908409465e-05, "loss": 0.0208, "step": 12663 }, { "epoch": 8.89950808151792, "grad_norm": 0.15616865456104279, "learning_rate": 2.740267041461701e-05, "loss": 0.0166, "step": 12664 }, { "epoch": 8.900210822206606, "grad_norm": 0.11997151374816895, "learning_rate": 2.740220192082455e-05, "loss": 0.0154, "step": 12665 }, { "epoch": 8.900913562895292, "grad_norm": 0.10452290624380112, "learning_rate": 2.7401733427032093e-05, "loss": 0.014, "step": 12666 }, { "epoch": 8.901616303583978, "grad_norm": 0.17096756398677826, "learning_rate": 2.7401264933239634e-05, "loss": 0.027, "step": 12667 }, { "epoch": 8.902319044272664, "grad_norm": 0.12906059622764587, "learning_rate": 2.7400796439447177e-05, "loss": 0.0178, "step": 12668 }, { "epoch": 8.90302178496135, "grad_norm": 0.1387551873922348, "learning_rate": 2.740032794565472e-05, "loss": 0.0108, "step": 12669 }, { "epoch": 8.903724525650036, "grad_norm": 0.13271725177764893, "learning_rate": 2.739985945186226e-05, "loss": 0.0213, "step": 12670 }, { "epoch": 8.904427266338722, "grad_norm": 0.17809619009494781, "learning_rate": 2.7399390958069805e-05, "loss": 0.0167, "step": 12671 }, { "epoch": 8.905130007027406, "grad_norm": 0.15043148398399353, "learning_rate": 2.739892246427735e-05, "loss": 0.0368, "step": 12672 }, { "epoch": 8.905832747716094, "grad_norm": 0.18826378881931305, "learning_rate": 2.7398453970484893e-05, "loss": 0.0284, "step": 12673 }, { "epoch": 8.906535488404778, "grad_norm": 0.23133185505867004, "learning_rate": 2.7397985476692433e-05, "loss": 0.0311, "step": 12674 }, { "epoch": 8.907238229093464, "grad_norm": 0.2518756687641144, "learning_rate": 2.7397516982899977e-05, "loss": 0.0393, "step": 12675 }, { "epoch": 8.90794096978215, "grad_norm": 0.23224838078022003, "learning_rate": 2.739704848910752e-05, "loss": 0.0222, "step": 12676 }, { "epoch": 8.908643710470836, "grad_norm": 0.18238988518714905, "learning_rate": 2.7396579995315064e-05, "loss": 0.0276, "step": 12677 }, { "epoch": 8.909346451159522, "grad_norm": 0.38124969601631165, "learning_rate": 2.7396111501522604e-05, "loss": 0.0384, "step": 12678 }, { "epoch": 8.910049191848207, "grad_norm": 0.4465111196041107, "learning_rate": 2.7395643007730148e-05, "loss": 0.0601, "step": 12679 }, { "epoch": 8.910751932536893, "grad_norm": 0.3176014721393585, "learning_rate": 2.7395174513937692e-05, "loss": 0.0594, "step": 12680 }, { "epoch": 8.91145467322558, "grad_norm": 0.3563460409641266, "learning_rate": 2.7394706020145236e-05, "loss": 0.1055, "step": 12681 }, { "epoch": 8.912157413914265, "grad_norm": 0.7238901853561401, "learning_rate": 2.739423752635278e-05, "loss": 0.1722, "step": 12682 }, { "epoch": 8.912860154602951, "grad_norm": 1.3958481550216675, "learning_rate": 2.739376903256032e-05, "loss": 0.2198, "step": 12683 }, { "epoch": 8.913562895291637, "grad_norm": 1.3424581289291382, "learning_rate": 2.7393300538767863e-05, "loss": 0.2327, "step": 12684 }, { "epoch": 8.914265635980323, "grad_norm": 0.16511507332324982, "learning_rate": 2.7392832044975404e-05, "loss": 0.0674, "step": 12685 }, { "epoch": 8.914968376669009, "grad_norm": 0.10472095012664795, "learning_rate": 2.7392363551182947e-05, "loss": 0.0314, "step": 12686 }, { "epoch": 8.915671117357695, "grad_norm": 0.1851353496313095, "learning_rate": 2.7391895057390488e-05, "loss": 0.0313, "step": 12687 }, { "epoch": 8.91637385804638, "grad_norm": 0.16999100148677826, "learning_rate": 2.739142656359803e-05, "loss": 0.0355, "step": 12688 }, { "epoch": 8.917076598735067, "grad_norm": 0.09262800216674805, "learning_rate": 2.7390958069805575e-05, "loss": 0.0129, "step": 12689 }, { "epoch": 8.917779339423753, "grad_norm": 0.10859167575836182, "learning_rate": 2.739048957601312e-05, "loss": 0.0203, "step": 12690 }, { "epoch": 8.918482080112438, "grad_norm": 0.38243821263313293, "learning_rate": 2.7390021082220663e-05, "loss": 0.0257, "step": 12691 }, { "epoch": 8.919184820801124, "grad_norm": 0.21549241244792938, "learning_rate": 2.7389552588428203e-05, "loss": 0.0179, "step": 12692 }, { "epoch": 8.91988756148981, "grad_norm": 0.1339564472436905, "learning_rate": 2.7389084094635747e-05, "loss": 0.0165, "step": 12693 }, { "epoch": 8.920590302178496, "grad_norm": 0.13204841315746307, "learning_rate": 2.738861560084329e-05, "loss": 0.0303, "step": 12694 }, { "epoch": 8.921293042867182, "grad_norm": 0.1279565691947937, "learning_rate": 2.7388147107050834e-05, "loss": 0.0228, "step": 12695 }, { "epoch": 8.921995783555868, "grad_norm": 0.15118011832237244, "learning_rate": 2.7387678613258374e-05, "loss": 0.028, "step": 12696 }, { "epoch": 8.922698524244554, "grad_norm": 0.25506603717803955, "learning_rate": 2.7387210119465918e-05, "loss": 0.032, "step": 12697 }, { "epoch": 8.92340126493324, "grad_norm": 0.1921960860490799, "learning_rate": 2.7386741625673462e-05, "loss": 0.0229, "step": 12698 }, { "epoch": 8.924104005621926, "grad_norm": 0.27074331045150757, "learning_rate": 2.7386273131881006e-05, "loss": 0.0342, "step": 12699 }, { "epoch": 8.924806746310612, "grad_norm": 0.19626665115356445, "learning_rate": 2.7385804638088546e-05, "loss": 0.0422, "step": 12700 }, { "epoch": 8.925509486999298, "grad_norm": 0.114605613052845, "learning_rate": 2.738533614429609e-05, "loss": 0.0157, "step": 12701 }, { "epoch": 8.926212227687984, "grad_norm": 0.2018972784280777, "learning_rate": 2.738486765050363e-05, "loss": 0.0268, "step": 12702 }, { "epoch": 8.92691496837667, "grad_norm": 0.34632983803749084, "learning_rate": 2.7384399156711174e-05, "loss": 0.0397, "step": 12703 }, { "epoch": 8.927617709065355, "grad_norm": 1.824540615081787, "learning_rate": 2.7383930662918717e-05, "loss": 0.0328, "step": 12704 }, { "epoch": 8.928320449754041, "grad_norm": 1.2500441074371338, "learning_rate": 2.7383462169126258e-05, "loss": 0.0732, "step": 12705 }, { "epoch": 8.929023190442727, "grad_norm": 0.37622344493865967, "learning_rate": 2.73829936753338e-05, "loss": 0.1031, "step": 12706 }, { "epoch": 8.929725931131413, "grad_norm": 0.45953425765037537, "learning_rate": 2.7382525181541345e-05, "loss": 0.1405, "step": 12707 }, { "epoch": 8.9304286718201, "grad_norm": Infinity, "learning_rate": 2.7382525181541345e-05, "loss": 0.2021, "step": 12708 }, { "epoch": 8.931131412508785, "grad_norm": 1.2512747049331665, "learning_rate": 2.738205668774889e-05, "loss": 0.2066, "step": 12709 }, { "epoch": 8.931834153197471, "grad_norm": 0.24795007705688477, "learning_rate": 2.738158819395643e-05, "loss": 0.0681, "step": 12710 }, { "epoch": 8.932536893886155, "grad_norm": 0.171549454331398, "learning_rate": 2.7381119700163973e-05, "loss": 0.0374, "step": 12711 }, { "epoch": 8.933239634574843, "grad_norm": 0.11643499881029129, "learning_rate": 2.7380651206371517e-05, "loss": 0.0258, "step": 12712 }, { "epoch": 8.933942375263527, "grad_norm": 0.12169406563043594, "learning_rate": 2.738018271257906e-05, "loss": 0.0217, "step": 12713 }, { "epoch": 8.934645115952213, "grad_norm": 2.1249277591705322, "learning_rate": 2.73797142187866e-05, "loss": 0.0186, "step": 12714 }, { "epoch": 8.935347856640899, "grad_norm": 0.13331589102745056, "learning_rate": 2.7379245724994145e-05, "loss": 0.019, "step": 12715 }, { "epoch": 8.936050597329585, "grad_norm": 0.21438802778720856, "learning_rate": 2.7378777231201688e-05, "loss": 0.0295, "step": 12716 }, { "epoch": 8.93675333801827, "grad_norm": 0.19731445610523224, "learning_rate": 2.7378308737409232e-05, "loss": 0.0111, "step": 12717 }, { "epoch": 8.937456078706957, "grad_norm": 0.411281555891037, "learning_rate": 2.7377840243616776e-05, "loss": 0.0177, "step": 12718 }, { "epoch": 8.938158819395642, "grad_norm": 0.12120752781629562, "learning_rate": 2.7377371749824316e-05, "loss": 0.0159, "step": 12719 }, { "epoch": 8.938861560084328, "grad_norm": 0.49949589371681213, "learning_rate": 2.7376903256031856e-05, "loss": 0.0259, "step": 12720 }, { "epoch": 8.939564300773014, "grad_norm": 0.1689056009054184, "learning_rate": 2.73764347622394e-05, "loss": 0.0221, "step": 12721 }, { "epoch": 8.9402670414617, "grad_norm": 0.15228857100009918, "learning_rate": 2.7375966268446944e-05, "loss": 0.0214, "step": 12722 }, { "epoch": 8.940969782150386, "grad_norm": 0.28642094135284424, "learning_rate": 2.7375497774654484e-05, "loss": 0.0335, "step": 12723 }, { "epoch": 8.941672522839072, "grad_norm": 0.2048981934785843, "learning_rate": 2.7375029280862028e-05, "loss": 0.0282, "step": 12724 }, { "epoch": 8.942375263527758, "grad_norm": 0.20238660275936127, "learning_rate": 2.737456078706957e-05, "loss": 0.0288, "step": 12725 }, { "epoch": 8.943078004216444, "grad_norm": 0.15738502144813538, "learning_rate": 2.7374092293277115e-05, "loss": 0.0255, "step": 12726 }, { "epoch": 8.94378074490513, "grad_norm": 0.9004853367805481, "learning_rate": 2.7373623799484656e-05, "loss": 0.0568, "step": 12727 }, { "epoch": 8.944483485593816, "grad_norm": 0.1909022033214569, "learning_rate": 2.73731553056922e-05, "loss": 0.0422, "step": 12728 }, { "epoch": 8.945186226282502, "grad_norm": 0.26017647981643677, "learning_rate": 2.7372686811899743e-05, "loss": 0.0449, "step": 12729 }, { "epoch": 8.945888966971188, "grad_norm": 0.29344141483306885, "learning_rate": 2.7372218318107287e-05, "loss": 0.0945, "step": 12730 }, { "epoch": 8.946591707659874, "grad_norm": 0.4815972149372101, "learning_rate": 2.737174982431483e-05, "loss": 0.1122, "step": 12731 }, { "epoch": 8.94729444834856, "grad_norm": 0.4269634187221527, "learning_rate": 2.737128133052237e-05, "loss": 0.1547, "step": 12732 }, { "epoch": 8.947997189037245, "grad_norm": 0.7403770089149475, "learning_rate": 2.7370812836729915e-05, "loss": 0.1582, "step": 12733 }, { "epoch": 8.948699929725931, "grad_norm": 1.4417438507080078, "learning_rate": 2.737034434293746e-05, "loss": 0.2425, "step": 12734 }, { "epoch": 8.949402670414617, "grad_norm": 0.2468125820159912, "learning_rate": 2.7369875849145002e-05, "loss": 0.0555, "step": 12735 }, { "epoch": 8.950105411103303, "grad_norm": 0.14851994812488556, "learning_rate": 2.7369407355352542e-05, "loss": 0.0286, "step": 12736 }, { "epoch": 8.950808151791989, "grad_norm": 0.1307416409254074, "learning_rate": 2.7368938861560086e-05, "loss": 0.0288, "step": 12737 }, { "epoch": 8.951510892480675, "grad_norm": 0.19412505626678467, "learning_rate": 2.7368470367767627e-05, "loss": 0.0254, "step": 12738 }, { "epoch": 8.952213633169361, "grad_norm": 0.18011043965816498, "learning_rate": 2.736800187397517e-05, "loss": 0.0206, "step": 12739 }, { "epoch": 8.952916373858047, "grad_norm": 0.17643798887729645, "learning_rate": 2.736753338018271e-05, "loss": 0.0212, "step": 12740 }, { "epoch": 8.953619114546733, "grad_norm": 0.13676488399505615, "learning_rate": 2.7367064886390254e-05, "loss": 0.012, "step": 12741 }, { "epoch": 8.954321855235419, "grad_norm": 0.13024291396141052, "learning_rate": 2.7366596392597798e-05, "loss": 0.0139, "step": 12742 }, { "epoch": 8.955024595924105, "grad_norm": 0.39718690514564514, "learning_rate": 2.7366127898805342e-05, "loss": 0.0333, "step": 12743 }, { "epoch": 8.95572733661279, "grad_norm": 0.1989583820104599, "learning_rate": 2.7365659405012886e-05, "loss": 0.012, "step": 12744 }, { "epoch": 8.956430077301476, "grad_norm": 0.23188717663288116, "learning_rate": 2.7365190911220426e-05, "loss": 0.0511, "step": 12745 }, { "epoch": 8.957132817990162, "grad_norm": 0.5842920541763306, "learning_rate": 2.736472241742797e-05, "loss": 0.0182, "step": 12746 }, { "epoch": 8.957835558678848, "grad_norm": 0.15876784920692444, "learning_rate": 2.7364253923635513e-05, "loss": 0.0324, "step": 12747 }, { "epoch": 8.958538299367534, "grad_norm": 0.20273856818675995, "learning_rate": 2.7363785429843057e-05, "loss": 0.0182, "step": 12748 }, { "epoch": 8.95924104005622, "grad_norm": 0.24709747731685638, "learning_rate": 2.7363316936050597e-05, "loss": 0.0367, "step": 12749 }, { "epoch": 8.959943780744904, "grad_norm": 0.22766394913196564, "learning_rate": 2.736284844225814e-05, "loss": 0.044, "step": 12750 }, { "epoch": 8.96064652143359, "grad_norm": 0.15952430665493011, "learning_rate": 2.7362379948465685e-05, "loss": 0.0235, "step": 12751 }, { "epoch": 8.961349262122276, "grad_norm": 0.19551847875118256, "learning_rate": 2.736191145467323e-05, "loss": 0.0505, "step": 12752 }, { "epoch": 8.962052002810962, "grad_norm": 0.20500542223453522, "learning_rate": 2.736144296088077e-05, "loss": 0.03, "step": 12753 }, { "epoch": 8.962754743499648, "grad_norm": 0.5201916694641113, "learning_rate": 2.7360974467088313e-05, "loss": 0.0592, "step": 12754 }, { "epoch": 8.963457484188334, "grad_norm": 0.5746026039123535, "learning_rate": 2.7360505973295853e-05, "loss": 0.0742, "step": 12755 }, { "epoch": 8.96416022487702, "grad_norm": 0.4261625409126282, "learning_rate": 2.7360037479503397e-05, "loss": 0.1122, "step": 12756 }, { "epoch": 8.964862965565706, "grad_norm": 0.6606026291847229, "learning_rate": 2.735956898571094e-05, "loss": 0.1676, "step": 12757 }, { "epoch": 8.965565706254392, "grad_norm": 0.62077397108078, "learning_rate": 2.735910049191848e-05, "loss": 0.18, "step": 12758 }, { "epoch": 8.966268446943078, "grad_norm": 1.3403472900390625, "learning_rate": 2.7358631998126024e-05, "loss": 0.2213, "step": 12759 }, { "epoch": 8.966971187631763, "grad_norm": 0.35093483328819275, "learning_rate": 2.7358163504333568e-05, "loss": 0.0553, "step": 12760 }, { "epoch": 8.96767392832045, "grad_norm": 0.12357197701931, "learning_rate": 2.7357695010541112e-05, "loss": 0.0157, "step": 12761 }, { "epoch": 8.968376669009135, "grad_norm": 0.25933289527893066, "learning_rate": 2.7357226516748652e-05, "loss": 0.0324, "step": 12762 }, { "epoch": 8.969079409697821, "grad_norm": 0.27391839027404785, "learning_rate": 2.7356758022956196e-05, "loss": 0.0243, "step": 12763 }, { "epoch": 8.969782150386507, "grad_norm": 0.15987661480903625, "learning_rate": 2.735628952916374e-05, "loss": 0.0291, "step": 12764 }, { "epoch": 8.970484891075193, "grad_norm": 0.14412495493888855, "learning_rate": 2.7355821035371283e-05, "loss": 0.0174, "step": 12765 }, { "epoch": 8.971187631763879, "grad_norm": 0.1271897256374359, "learning_rate": 2.7355352541578824e-05, "loss": 0.0165, "step": 12766 }, { "epoch": 8.971890372452565, "grad_norm": 0.11830484122037888, "learning_rate": 2.7354884047786367e-05, "loss": 0.0209, "step": 12767 }, { "epoch": 8.97259311314125, "grad_norm": 0.12075887620449066, "learning_rate": 2.735441555399391e-05, "loss": 0.0257, "step": 12768 }, { "epoch": 8.973295853829937, "grad_norm": 0.1126939132809639, "learning_rate": 2.7353947060201455e-05, "loss": 0.0127, "step": 12769 }, { "epoch": 8.973998594518623, "grad_norm": 0.10556579381227493, "learning_rate": 2.7353478566409e-05, "loss": 0.023, "step": 12770 }, { "epoch": 8.974701335207309, "grad_norm": 0.1077767014503479, "learning_rate": 2.735301007261654e-05, "loss": 0.0148, "step": 12771 }, { "epoch": 8.975404075895995, "grad_norm": 0.15099962055683136, "learning_rate": 2.7352541578824083e-05, "loss": 0.0261, "step": 12772 }, { "epoch": 8.97610681658468, "grad_norm": 0.14649473130702972, "learning_rate": 2.7352073085031623e-05, "loss": 0.0154, "step": 12773 }, { "epoch": 8.976809557273366, "grad_norm": 0.23693251609802246, "learning_rate": 2.7351604591239167e-05, "loss": 0.044, "step": 12774 }, { "epoch": 8.977512297962052, "grad_norm": 0.1723589301109314, "learning_rate": 2.7351136097446707e-05, "loss": 0.0294, "step": 12775 }, { "epoch": 8.978215038650738, "grad_norm": 0.2180873453617096, "learning_rate": 2.735066760365425e-05, "loss": 0.0322, "step": 12776 }, { "epoch": 8.978917779339424, "grad_norm": 0.12086733430624008, "learning_rate": 2.7350199109861795e-05, "loss": 0.023, "step": 12777 }, { "epoch": 8.97962052002811, "grad_norm": 0.2611730396747589, "learning_rate": 2.7349730616069338e-05, "loss": 0.0543, "step": 12778 }, { "epoch": 8.980323260716796, "grad_norm": 0.33456820249557495, "learning_rate": 2.734926212227688e-05, "loss": 0.0495, "step": 12779 }, { "epoch": 8.981026001405482, "grad_norm": 0.4516826570034027, "learning_rate": 2.7348793628484422e-05, "loss": 0.098, "step": 12780 }, { "epoch": 8.981728742094168, "grad_norm": 0.497728168964386, "learning_rate": 2.7348325134691966e-05, "loss": 0.1333, "step": 12781 }, { "epoch": 8.982431482782854, "grad_norm": 0.8097407221794128, "learning_rate": 2.734785664089951e-05, "loss": 0.1672, "step": 12782 }, { "epoch": 8.98313422347154, "grad_norm": 0.9363358616828918, "learning_rate": 2.7347388147107054e-05, "loss": 0.1852, "step": 12783 }, { "epoch": 8.983836964160226, "grad_norm": 0.7366158962249756, "learning_rate": 2.7346919653314594e-05, "loss": 0.2102, "step": 12784 }, { "epoch": 8.984539704848912, "grad_norm": 0.2433173507452011, "learning_rate": 2.7346451159522138e-05, "loss": 0.081, "step": 12785 }, { "epoch": 8.985242445537597, "grad_norm": 0.11712923645973206, "learning_rate": 2.734598266572968e-05, "loss": 0.0342, "step": 12786 }, { "epoch": 8.985945186226282, "grad_norm": 0.11489985138177872, "learning_rate": 2.7345514171937225e-05, "loss": 0.017, "step": 12787 }, { "epoch": 8.98664792691497, "grad_norm": 0.22560839354991913, "learning_rate": 2.7345045678144765e-05, "loss": 0.0227, "step": 12788 }, { "epoch": 8.987350667603653, "grad_norm": 0.12121711671352386, "learning_rate": 2.734457718435231e-05, "loss": 0.0259, "step": 12789 }, { "epoch": 8.98805340829234, "grad_norm": 0.2410832643508911, "learning_rate": 2.734410869055985e-05, "loss": 0.0139, "step": 12790 }, { "epoch": 8.988756148981025, "grad_norm": 0.10561753064393997, "learning_rate": 2.7343640196767393e-05, "loss": 0.0206, "step": 12791 }, { "epoch": 8.989458889669711, "grad_norm": 0.2151097059249878, "learning_rate": 2.7343171702974933e-05, "loss": 0.0417, "step": 12792 }, { "epoch": 8.990161630358397, "grad_norm": 0.07560805231332779, "learning_rate": 2.7342703209182477e-05, "loss": 0.0062, "step": 12793 }, { "epoch": 8.990864371047083, "grad_norm": 0.15063004195690155, "learning_rate": 2.734223471539002e-05, "loss": 0.0393, "step": 12794 }, { "epoch": 8.991567111735769, "grad_norm": 0.11079605668783188, "learning_rate": 2.7341766221597565e-05, "loss": 0.0158, "step": 12795 }, { "epoch": 8.992269852424455, "grad_norm": 0.31361836194992065, "learning_rate": 2.734129772780511e-05, "loss": 0.0179, "step": 12796 }, { "epoch": 8.99297259311314, "grad_norm": 0.09268354624509811, "learning_rate": 2.734082923401265e-05, "loss": 0.0185, "step": 12797 }, { "epoch": 8.993675333801827, "grad_norm": 0.19803965091705322, "learning_rate": 2.7340360740220192e-05, "loss": 0.0375, "step": 12798 }, { "epoch": 8.994378074490513, "grad_norm": 0.16805525124073029, "learning_rate": 2.7339892246427736e-05, "loss": 0.0098, "step": 12799 }, { "epoch": 8.995080815179199, "grad_norm": 0.46779125928878784, "learning_rate": 2.733942375263528e-05, "loss": 0.0416, "step": 12800 }, { "epoch": 8.995783555867884, "grad_norm": 0.2655266225337982, "learning_rate": 2.733895525884282e-05, "loss": 0.0391, "step": 12801 }, { "epoch": 8.99648629655657, "grad_norm": 0.28489547967910767, "learning_rate": 2.7338486765050364e-05, "loss": 0.0685, "step": 12802 }, { "epoch": 8.997189037245256, "grad_norm": 0.26220667362213135, "learning_rate": 2.7338018271257908e-05, "loss": 0.0814, "step": 12803 }, { "epoch": 8.997891777933942, "grad_norm": 0.6928137540817261, "learning_rate": 2.733754977746545e-05, "loss": 0.0899, "step": 12804 }, { "epoch": 8.998594518622628, "grad_norm": 0.5705932378768921, "learning_rate": 2.7337081283672992e-05, "loss": 0.1692, "step": 12805 }, { "epoch": 8.999297259311314, "grad_norm": 1.05355966091156, "learning_rate": 2.7336612789880535e-05, "loss": 0.218, "step": 12806 }, { "epoch": 9.0, "grad_norm": 1.082077145576477, "learning_rate": 2.733614429608808e-05, "loss": 0.1342, "step": 12807 }, { "epoch": 9.000702740688686, "grad_norm": 0.19299714267253876, "learning_rate": 2.733567580229562e-05, "loss": 0.0652, "step": 12808 }, { "epoch": 9.001405481377372, "grad_norm": 0.12288883328437805, "learning_rate": 2.7335207308503163e-05, "loss": 0.0179, "step": 12809 }, { "epoch": 9.002108222066058, "grad_norm": 0.15008893609046936, "learning_rate": 2.7334738814710704e-05, "loss": 0.0252, "step": 12810 }, { "epoch": 9.002810962754744, "grad_norm": 0.1361800730228424, "learning_rate": 2.7334270320918247e-05, "loss": 0.0187, "step": 12811 }, { "epoch": 9.00351370344343, "grad_norm": 0.21415269374847412, "learning_rate": 2.733380182712579e-05, "loss": 0.0121, "step": 12812 }, { "epoch": 9.004216444132116, "grad_norm": 0.10595113039016724, "learning_rate": 2.7333333333333335e-05, "loss": 0.0111, "step": 12813 }, { "epoch": 9.004919184820801, "grad_norm": 0.10039471834897995, "learning_rate": 2.7332864839540875e-05, "loss": 0.0179, "step": 12814 }, { "epoch": 9.005621925509487, "grad_norm": 0.12806051969528198, "learning_rate": 2.733239634574842e-05, "loss": 0.0285, "step": 12815 }, { "epoch": 9.006324666198173, "grad_norm": 0.14525353908538818, "learning_rate": 2.7331927851955963e-05, "loss": 0.0186, "step": 12816 }, { "epoch": 9.00702740688686, "grad_norm": 0.14511241018772125, "learning_rate": 2.7331459358163506e-05, "loss": 0.0104, "step": 12817 }, { "epoch": 9.007730147575545, "grad_norm": 0.36715802550315857, "learning_rate": 2.7330990864371047e-05, "loss": 0.0364, "step": 12818 }, { "epoch": 9.008432888264231, "grad_norm": 0.1518687903881073, "learning_rate": 2.733052237057859e-05, "loss": 0.015, "step": 12819 }, { "epoch": 9.009135628952917, "grad_norm": 0.14931225776672363, "learning_rate": 2.7330053876786134e-05, "loss": 0.031, "step": 12820 }, { "epoch": 9.009838369641603, "grad_norm": 0.12338021397590637, "learning_rate": 2.7329585382993678e-05, "loss": 0.0164, "step": 12821 }, { "epoch": 9.010541110330289, "grad_norm": 0.8106324076652527, "learning_rate": 2.732911688920122e-05, "loss": 0.0256, "step": 12822 }, { "epoch": 9.011243851018975, "grad_norm": 0.42679712176322937, "learning_rate": 2.7328648395408762e-05, "loss": 0.0306, "step": 12823 }, { "epoch": 9.01194659170766, "grad_norm": 0.27150803804397583, "learning_rate": 2.7328179901616306e-05, "loss": 0.0187, "step": 12824 }, { "epoch": 9.012649332396347, "grad_norm": 0.18923316895961761, "learning_rate": 2.7327711407823846e-05, "loss": 0.0403, "step": 12825 }, { "epoch": 9.013352073085033, "grad_norm": 0.1525508016347885, "learning_rate": 2.732724291403139e-05, "loss": 0.0341, "step": 12826 }, { "epoch": 9.014054813773717, "grad_norm": 0.3018178343772888, "learning_rate": 2.732677442023893e-05, "loss": 0.0568, "step": 12827 }, { "epoch": 9.014757554462403, "grad_norm": 0.31109511852264404, "learning_rate": 2.7326305926446474e-05, "loss": 0.0764, "step": 12828 }, { "epoch": 9.015460295151088, "grad_norm": 0.8238580822944641, "learning_rate": 2.7325837432654017e-05, "loss": 0.1301, "step": 12829 }, { "epoch": 9.016163035839774, "grad_norm": 0.4211392104625702, "learning_rate": 2.732536893886156e-05, "loss": 0.1063, "step": 12830 }, { "epoch": 9.01686577652846, "grad_norm": 0.7824185490608215, "learning_rate": 2.73249004450691e-05, "loss": 0.1711, "step": 12831 }, { "epoch": 9.017568517217146, "grad_norm": 2.3132822513580322, "learning_rate": 2.7324431951276645e-05, "loss": 0.2083, "step": 12832 }, { "epoch": 9.018271257905832, "grad_norm": 0.19264930486679077, "learning_rate": 2.732396345748419e-05, "loss": 0.0734, "step": 12833 }, { "epoch": 9.018973998594518, "grad_norm": 0.11741270869970322, "learning_rate": 2.7323494963691733e-05, "loss": 0.0236, "step": 12834 }, { "epoch": 9.019676739283204, "grad_norm": 0.11699142307043076, "learning_rate": 2.7323026469899276e-05, "loss": 0.0225, "step": 12835 }, { "epoch": 9.02037947997189, "grad_norm": 0.0925755724310875, "learning_rate": 2.7322557976106817e-05, "loss": 0.015, "step": 12836 }, { "epoch": 9.021082220660576, "grad_norm": 0.11549033969640732, "learning_rate": 2.732208948231436e-05, "loss": 0.0214, "step": 12837 }, { "epoch": 9.021784961349262, "grad_norm": 0.109420046210289, "learning_rate": 2.7321620988521904e-05, "loss": 0.0175, "step": 12838 }, { "epoch": 9.022487702037948, "grad_norm": 0.19551822543144226, "learning_rate": 2.7321152494729448e-05, "loss": 0.0406, "step": 12839 }, { "epoch": 9.023190442726634, "grad_norm": 0.16299490630626678, "learning_rate": 2.7320684000936988e-05, "loss": 0.0206, "step": 12840 }, { "epoch": 9.02389318341532, "grad_norm": 0.2819867730140686, "learning_rate": 2.7320215507144532e-05, "loss": 0.018, "step": 12841 }, { "epoch": 9.024595924104005, "grad_norm": 0.1595449298620224, "learning_rate": 2.7319747013352072e-05, "loss": 0.0136, "step": 12842 }, { "epoch": 9.025298664792691, "grad_norm": 0.17825277149677277, "learning_rate": 2.7319278519559616e-05, "loss": 0.0325, "step": 12843 }, { "epoch": 9.026001405481377, "grad_norm": 0.12333620339632034, "learning_rate": 2.7318810025767156e-05, "loss": 0.0177, "step": 12844 }, { "epoch": 9.026704146170063, "grad_norm": 0.15850292146205902, "learning_rate": 2.73183415319747e-05, "loss": 0.0283, "step": 12845 }, { "epoch": 9.02740688685875, "grad_norm": 0.13697810471057892, "learning_rate": 2.7317873038182244e-05, "loss": 0.0135, "step": 12846 }, { "epoch": 9.028109627547435, "grad_norm": 0.10694018751382828, "learning_rate": 2.7317404544389788e-05, "loss": 0.0304, "step": 12847 }, { "epoch": 9.028812368236121, "grad_norm": 0.1596524864435196, "learning_rate": 2.731693605059733e-05, "loss": 0.0368, "step": 12848 }, { "epoch": 9.029515108924807, "grad_norm": 0.23198701441287994, "learning_rate": 2.731646755680487e-05, "loss": 0.0316, "step": 12849 }, { "epoch": 9.030217849613493, "grad_norm": 0.5128378868103027, "learning_rate": 2.7315999063012415e-05, "loss": 0.0263, "step": 12850 }, { "epoch": 9.030920590302179, "grad_norm": 0.26263025403022766, "learning_rate": 2.731553056921996e-05, "loss": 0.0457, "step": 12851 }, { "epoch": 9.031623330990865, "grad_norm": 0.28864017128944397, "learning_rate": 2.7315062075427503e-05, "loss": 0.0704, "step": 12852 }, { "epoch": 9.03232607167955, "grad_norm": 0.7050281763076782, "learning_rate": 2.7314593581635043e-05, "loss": 0.1289, "step": 12853 }, { "epoch": 9.033028812368237, "grad_norm": 0.7787280678749084, "learning_rate": 2.7314125087842587e-05, "loss": 0.1391, "step": 12854 }, { "epoch": 9.033731553056922, "grad_norm": 0.526435911655426, "learning_rate": 2.731365659405013e-05, "loss": 0.1793, "step": 12855 }, { "epoch": 9.034434293745608, "grad_norm": 0.5769402980804443, "learning_rate": 2.7313188100257674e-05, "loss": 0.1965, "step": 12856 }, { "epoch": 9.035137034434294, "grad_norm": 0.9585477113723755, "learning_rate": 2.7312719606465215e-05, "loss": 0.217, "step": 12857 }, { "epoch": 9.03583977512298, "grad_norm": 0.18301457166671753, "learning_rate": 2.731225111267276e-05, "loss": 0.0684, "step": 12858 }, { "epoch": 9.036542515811666, "grad_norm": 0.30878737568855286, "learning_rate": 2.7311782618880302e-05, "loss": 0.0397, "step": 12859 }, { "epoch": 9.037245256500352, "grad_norm": 0.11651365458965302, "learning_rate": 2.7311314125087842e-05, "loss": 0.0281, "step": 12860 }, { "epoch": 9.037947997189038, "grad_norm": 0.12920509278774261, "learning_rate": 2.7310845631295386e-05, "loss": 0.0151, "step": 12861 }, { "epoch": 9.038650737877724, "grad_norm": 0.1742883175611496, "learning_rate": 2.7310377137502926e-05, "loss": 0.0169, "step": 12862 }, { "epoch": 9.03935347856641, "grad_norm": 0.11637543886899948, "learning_rate": 2.730990864371047e-05, "loss": 0.0124, "step": 12863 }, { "epoch": 9.040056219255096, "grad_norm": 0.12026554346084595, "learning_rate": 2.7309440149918014e-05, "loss": 0.0096, "step": 12864 }, { "epoch": 9.04075895994378, "grad_norm": 0.15380439162254333, "learning_rate": 2.7308971656125558e-05, "loss": 0.0239, "step": 12865 }, { "epoch": 9.041461700632466, "grad_norm": 0.13819023966789246, "learning_rate": 2.7308503162333098e-05, "loss": 0.0129, "step": 12866 }, { "epoch": 9.042164441321152, "grad_norm": 0.24322019517421722, "learning_rate": 2.7308034668540642e-05, "loss": 0.0172, "step": 12867 }, { "epoch": 9.042867182009838, "grad_norm": 0.23348185420036316, "learning_rate": 2.7307566174748185e-05, "loss": 0.0274, "step": 12868 }, { "epoch": 9.043569922698524, "grad_norm": 0.1368301957845688, "learning_rate": 2.730709768095573e-05, "loss": 0.0091, "step": 12869 }, { "epoch": 9.04427266338721, "grad_norm": 0.10607058554887772, "learning_rate": 2.730662918716327e-05, "loss": 0.0239, "step": 12870 }, { "epoch": 9.044975404075895, "grad_norm": 0.11401457339525223, "learning_rate": 2.7306160693370813e-05, "loss": 0.0072, "step": 12871 }, { "epoch": 9.045678144764581, "grad_norm": 0.2540489137172699, "learning_rate": 2.7305692199578357e-05, "loss": 0.0406, "step": 12872 }, { "epoch": 9.046380885453267, "grad_norm": 0.607736349105835, "learning_rate": 2.73052237057859e-05, "loss": 0.0614, "step": 12873 }, { "epoch": 9.047083626141953, "grad_norm": 0.1414267122745514, "learning_rate": 2.7304755211993444e-05, "loss": 0.0223, "step": 12874 }, { "epoch": 9.047786366830639, "grad_norm": 0.3122410178184509, "learning_rate": 2.7304286718200985e-05, "loss": 0.0435, "step": 12875 }, { "epoch": 9.048489107519325, "grad_norm": 0.24194855988025665, "learning_rate": 2.730381822440853e-05, "loss": 0.0369, "step": 12876 }, { "epoch": 9.049191848208011, "grad_norm": 0.5871606469154358, "learning_rate": 2.730334973061607e-05, "loss": 0.0408, "step": 12877 }, { "epoch": 9.049894588896697, "grad_norm": 0.2876923084259033, "learning_rate": 2.7302881236823613e-05, "loss": 0.0578, "step": 12878 }, { "epoch": 9.050597329585383, "grad_norm": 0.44717416167259216, "learning_rate": 2.7302412743031153e-05, "loss": 0.104, "step": 12879 }, { "epoch": 9.051300070274069, "grad_norm": 0.7036941051483154, "learning_rate": 2.7301944249238697e-05, "loss": 0.1375, "step": 12880 }, { "epoch": 9.052002810962755, "grad_norm": 0.7206048369407654, "learning_rate": 2.730147575544624e-05, "loss": 0.1448, "step": 12881 }, { "epoch": 9.05270555165144, "grad_norm": 0.8887888193130493, "learning_rate": 2.7301007261653784e-05, "loss": 0.1729, "step": 12882 }, { "epoch": 9.053408292340126, "grad_norm": 0.3733268082141876, "learning_rate": 2.7300538767861324e-05, "loss": 0.0865, "step": 12883 }, { "epoch": 9.054111033028812, "grad_norm": 0.13413771986961365, "learning_rate": 2.7300070274068868e-05, "loss": 0.0306, "step": 12884 }, { "epoch": 9.054813773717498, "grad_norm": 0.1522665023803711, "learning_rate": 2.7299601780276412e-05, "loss": 0.031, "step": 12885 }, { "epoch": 9.055516514406184, "grad_norm": 0.12191415578126907, "learning_rate": 2.7299133286483956e-05, "loss": 0.0197, "step": 12886 }, { "epoch": 9.05621925509487, "grad_norm": 0.30815938115119934, "learning_rate": 2.72986647926915e-05, "loss": 0.0167, "step": 12887 }, { "epoch": 9.056921995783556, "grad_norm": 0.06934140622615814, "learning_rate": 2.729819629889904e-05, "loss": 0.007, "step": 12888 }, { "epoch": 9.057624736472242, "grad_norm": 0.137517049908638, "learning_rate": 2.7297727805106583e-05, "loss": 0.0161, "step": 12889 }, { "epoch": 9.058327477160928, "grad_norm": 0.10059939324855804, "learning_rate": 2.7297259311314127e-05, "loss": 0.0077, "step": 12890 }, { "epoch": 9.059030217849614, "grad_norm": 0.11032011359930038, "learning_rate": 2.729679081752167e-05, "loss": 0.0171, "step": 12891 }, { "epoch": 9.0597329585383, "grad_norm": 0.1534455269575119, "learning_rate": 2.729632232372921e-05, "loss": 0.0106, "step": 12892 }, { "epoch": 9.060435699226986, "grad_norm": 0.17678222060203552, "learning_rate": 2.7295853829936755e-05, "loss": 0.0247, "step": 12893 }, { "epoch": 9.061138439915672, "grad_norm": 0.12766091525554657, "learning_rate": 2.72953853361443e-05, "loss": 0.0137, "step": 12894 }, { "epoch": 9.061841180604358, "grad_norm": 0.42426785826683044, "learning_rate": 2.729491684235184e-05, "loss": 0.0358, "step": 12895 }, { "epoch": 9.062543921293043, "grad_norm": 0.2325870543718338, "learning_rate": 2.7294448348559383e-05, "loss": 0.0304, "step": 12896 }, { "epoch": 9.06324666198173, "grad_norm": 0.1967407763004303, "learning_rate": 2.7293979854766923e-05, "loss": 0.0276, "step": 12897 }, { "epoch": 9.063949402670415, "grad_norm": 0.3913496434688568, "learning_rate": 2.7293511360974467e-05, "loss": 0.0264, "step": 12898 }, { "epoch": 9.064652143359101, "grad_norm": 0.13590006530284882, "learning_rate": 2.729304286718201e-05, "loss": 0.0256, "step": 12899 }, { "epoch": 9.065354884047787, "grad_norm": 0.29887655377388, "learning_rate": 2.7292574373389554e-05, "loss": 0.0376, "step": 12900 }, { "epoch": 9.066057624736473, "grad_norm": 0.1448042243719101, "learning_rate": 2.7292105879597095e-05, "loss": 0.0217, "step": 12901 }, { "epoch": 9.066760365425159, "grad_norm": 0.2813962697982788, "learning_rate": 2.7291637385804638e-05, "loss": 0.0324, "step": 12902 }, { "epoch": 9.067463106113845, "grad_norm": 0.41635072231292725, "learning_rate": 2.7291168892012182e-05, "loss": 0.0731, "step": 12903 }, { "epoch": 9.068165846802529, "grad_norm": 0.48729097843170166, "learning_rate": 2.7290700398219726e-05, "loss": 0.1137, "step": 12904 }, { "epoch": 9.068868587491215, "grad_norm": 0.5588338375091553, "learning_rate": 2.7290231904427266e-05, "loss": 0.1357, "step": 12905 }, { "epoch": 9.0695713281799, "grad_norm": 1.107039213180542, "learning_rate": 2.728976341063481e-05, "loss": 0.1598, "step": 12906 }, { "epoch": 9.070274068868587, "grad_norm": 2.3478405475616455, "learning_rate": 2.7289294916842353e-05, "loss": 0.2491, "step": 12907 }, { "epoch": 9.070976809557273, "grad_norm": 0.22473403811454773, "learning_rate": 2.7288826423049897e-05, "loss": 0.0657, "step": 12908 }, { "epoch": 9.071679550245959, "grad_norm": 0.15798059105873108, "learning_rate": 2.7288357929257438e-05, "loss": 0.0286, "step": 12909 }, { "epoch": 9.072382290934645, "grad_norm": 0.4035682678222656, "learning_rate": 2.728788943546498e-05, "loss": 0.0176, "step": 12910 }, { "epoch": 9.07308503162333, "grad_norm": 0.08013729751110077, "learning_rate": 2.7287420941672525e-05, "loss": 0.0147, "step": 12911 }, { "epoch": 9.073787772312016, "grad_norm": 0.10537933558225632, "learning_rate": 2.7286952447880065e-05, "loss": 0.0174, "step": 12912 }, { "epoch": 9.074490513000702, "grad_norm": 0.09912201762199402, "learning_rate": 2.728648395408761e-05, "loss": 0.0238, "step": 12913 }, { "epoch": 9.075193253689388, "grad_norm": 0.1606885939836502, "learning_rate": 2.728601546029515e-05, "loss": 0.0136, "step": 12914 }, { "epoch": 9.075895994378074, "grad_norm": 0.06962109357118607, "learning_rate": 2.7285546966502693e-05, "loss": 0.0132, "step": 12915 }, { "epoch": 9.07659873506676, "grad_norm": 0.16818636655807495, "learning_rate": 2.7285078472710237e-05, "loss": 0.0206, "step": 12916 }, { "epoch": 9.077301475755446, "grad_norm": 0.10170937329530716, "learning_rate": 2.728460997891778e-05, "loss": 0.0106, "step": 12917 }, { "epoch": 9.078004216444132, "grad_norm": 0.13254468142986298, "learning_rate": 2.728414148512532e-05, "loss": 0.0291, "step": 12918 }, { "epoch": 9.078706957132818, "grad_norm": 0.08435142785310745, "learning_rate": 2.7283672991332865e-05, "loss": 0.0117, "step": 12919 }, { "epoch": 9.079409697821504, "grad_norm": 0.33174464106559753, "learning_rate": 2.728320449754041e-05, "loss": 0.0257, "step": 12920 }, { "epoch": 9.08011243851019, "grad_norm": 0.22599615156650543, "learning_rate": 2.7282736003747952e-05, "loss": 0.0175, "step": 12921 }, { "epoch": 9.080815179198876, "grad_norm": 0.1866915374994278, "learning_rate": 2.7282267509955496e-05, "loss": 0.0294, "step": 12922 }, { "epoch": 9.081517919887562, "grad_norm": 0.17404574155807495, "learning_rate": 2.7281799016163036e-05, "loss": 0.0277, "step": 12923 }, { "epoch": 9.082220660576247, "grad_norm": 0.32658717036247253, "learning_rate": 2.728133052237058e-05, "loss": 0.0268, "step": 12924 }, { "epoch": 9.082923401264933, "grad_norm": 0.8308296799659729, "learning_rate": 2.7280862028578124e-05, "loss": 0.034, "step": 12925 }, { "epoch": 9.08362614195362, "grad_norm": 0.15394756197929382, "learning_rate": 2.7280393534785667e-05, "loss": 0.0298, "step": 12926 }, { "epoch": 9.084328882642305, "grad_norm": 0.21653559803962708, "learning_rate": 2.7279925040993208e-05, "loss": 0.0402, "step": 12927 }, { "epoch": 9.085031623330991, "grad_norm": 0.37357595562934875, "learning_rate": 2.727945654720075e-05, "loss": 0.0715, "step": 12928 }, { "epoch": 9.085734364019677, "grad_norm": 0.6061606407165527, "learning_rate": 2.7278988053408292e-05, "loss": 0.0972, "step": 12929 }, { "epoch": 9.086437104708363, "grad_norm": 0.4420895576477051, "learning_rate": 2.7278519559615835e-05, "loss": 0.1343, "step": 12930 }, { "epoch": 9.087139845397049, "grad_norm": 0.7760117053985596, "learning_rate": 2.7278051065823376e-05, "loss": 0.2024, "step": 12931 }, { "epoch": 9.087842586085735, "grad_norm": 0.9462016224861145, "learning_rate": 2.727758257203092e-05, "loss": 0.2018, "step": 12932 }, { "epoch": 9.08854532677442, "grad_norm": 0.24766044318675995, "learning_rate": 2.7277114078238463e-05, "loss": 0.0647, "step": 12933 }, { "epoch": 9.089248067463107, "grad_norm": 0.1204947680234909, "learning_rate": 2.7276645584446007e-05, "loss": 0.0323, "step": 12934 }, { "epoch": 9.089950808151793, "grad_norm": 0.08760435879230499, "learning_rate": 2.727617709065355e-05, "loss": 0.0166, "step": 12935 }, { "epoch": 9.090653548840478, "grad_norm": 0.14999797940254211, "learning_rate": 2.727570859686109e-05, "loss": 0.023, "step": 12936 }, { "epoch": 9.091356289529164, "grad_norm": 0.10203853249549866, "learning_rate": 2.7275240103068635e-05, "loss": 0.0141, "step": 12937 }, { "epoch": 9.09205903021785, "grad_norm": 0.13204029202461243, "learning_rate": 2.727477160927618e-05, "loss": 0.0113, "step": 12938 }, { "epoch": 9.092761770906536, "grad_norm": 0.09109573811292648, "learning_rate": 2.7274303115483722e-05, "loss": 0.0107, "step": 12939 }, { "epoch": 9.093464511595222, "grad_norm": 0.1763102263212204, "learning_rate": 2.7273834621691263e-05, "loss": 0.0231, "step": 12940 }, { "epoch": 9.094167252283908, "grad_norm": 0.14603547751903534, "learning_rate": 2.7273366127898806e-05, "loss": 0.021, "step": 12941 }, { "epoch": 9.094869992972592, "grad_norm": 0.17626096308231354, "learning_rate": 2.727289763410635e-05, "loss": 0.0164, "step": 12942 }, { "epoch": 9.095572733661278, "grad_norm": 0.18303468823432922, "learning_rate": 2.7272429140313894e-05, "loss": 0.0122, "step": 12943 }, { "epoch": 9.096275474349964, "grad_norm": 0.11374927312135696, "learning_rate": 2.7271960646521434e-05, "loss": 0.0255, "step": 12944 }, { "epoch": 9.09697821503865, "grad_norm": 0.2419251799583435, "learning_rate": 2.7271492152728978e-05, "loss": 0.0342, "step": 12945 }, { "epoch": 9.097680955727336, "grad_norm": 0.10445449501276016, "learning_rate": 2.727102365893652e-05, "loss": 0.0116, "step": 12946 }, { "epoch": 9.098383696416022, "grad_norm": 0.13217277824878693, "learning_rate": 2.7270555165144062e-05, "loss": 0.0235, "step": 12947 }, { "epoch": 9.099086437104708, "grad_norm": 0.21477755904197693, "learning_rate": 2.7270086671351606e-05, "loss": 0.0339, "step": 12948 }, { "epoch": 9.099789177793394, "grad_norm": 0.11004794389009476, "learning_rate": 2.7269618177559146e-05, "loss": 0.0136, "step": 12949 }, { "epoch": 9.10049191848208, "grad_norm": 0.2795473039150238, "learning_rate": 2.726914968376669e-05, "loss": 0.0351, "step": 12950 }, { "epoch": 9.101194659170766, "grad_norm": 0.4008581340312958, "learning_rate": 2.7268681189974233e-05, "loss": 0.0287, "step": 12951 }, { "epoch": 9.101897399859451, "grad_norm": 0.3895147442817688, "learning_rate": 2.7268212696181777e-05, "loss": 0.077, "step": 12952 }, { "epoch": 9.102600140548137, "grad_norm": 1.138615369796753, "learning_rate": 2.7267744202389317e-05, "loss": 0.0553, "step": 12953 }, { "epoch": 9.103302881236823, "grad_norm": 0.4294387996196747, "learning_rate": 2.726727570859686e-05, "loss": 0.0922, "step": 12954 }, { "epoch": 9.10400562192551, "grad_norm": 1.1570212841033936, "learning_rate": 2.7266807214804405e-05, "loss": 0.166, "step": 12955 }, { "epoch": 9.104708362614195, "grad_norm": 0.9423289895057678, "learning_rate": 2.726633872101195e-05, "loss": 0.1966, "step": 12956 }, { "epoch": 9.105411103302881, "grad_norm": 1.7000714540481567, "learning_rate": 2.726587022721949e-05, "loss": 0.228, "step": 12957 }, { "epoch": 9.106113843991567, "grad_norm": 0.24032582342624664, "learning_rate": 2.7265401733427033e-05, "loss": 0.0623, "step": 12958 }, { "epoch": 9.106816584680253, "grad_norm": 0.17702616751194, "learning_rate": 2.7264933239634576e-05, "loss": 0.0282, "step": 12959 }, { "epoch": 9.107519325368939, "grad_norm": 0.1280171126127243, "learning_rate": 2.726446474584212e-05, "loss": 0.02, "step": 12960 }, { "epoch": 9.108222066057625, "grad_norm": 0.20484641194343567, "learning_rate": 2.7263996252049664e-05, "loss": 0.0145, "step": 12961 }, { "epoch": 9.10892480674631, "grad_norm": 0.19198325276374817, "learning_rate": 2.7263527758257204e-05, "loss": 0.0226, "step": 12962 }, { "epoch": 9.109627547434997, "grad_norm": 0.4439423978328705, "learning_rate": 2.7263059264464748e-05, "loss": 0.0158, "step": 12963 }, { "epoch": 9.110330288123683, "grad_norm": 0.3030052185058594, "learning_rate": 2.7262590770672288e-05, "loss": 0.0191, "step": 12964 }, { "epoch": 9.111033028812368, "grad_norm": 0.31835442781448364, "learning_rate": 2.7262122276879832e-05, "loss": 0.0179, "step": 12965 }, { "epoch": 9.111735769501054, "grad_norm": 0.17715506255626678, "learning_rate": 2.7261653783087372e-05, "loss": 0.0164, "step": 12966 }, { "epoch": 9.11243851018974, "grad_norm": 0.19657349586486816, "learning_rate": 2.7261185289294916e-05, "loss": 0.0322, "step": 12967 }, { "epoch": 9.113141250878426, "grad_norm": 0.16028042137622833, "learning_rate": 2.726071679550246e-05, "loss": 0.0209, "step": 12968 }, { "epoch": 9.113843991567112, "grad_norm": 0.21298205852508545, "learning_rate": 2.7260248301710003e-05, "loss": 0.0195, "step": 12969 }, { "epoch": 9.114546732255798, "grad_norm": 0.1439351737499237, "learning_rate": 2.7259779807917544e-05, "loss": 0.0277, "step": 12970 }, { "epoch": 9.115249472944484, "grad_norm": 0.13534395396709442, "learning_rate": 2.7259311314125088e-05, "loss": 0.0125, "step": 12971 }, { "epoch": 9.11595221363317, "grad_norm": 0.21232284605503082, "learning_rate": 2.725884282033263e-05, "loss": 0.0314, "step": 12972 }, { "epoch": 9.116654954321856, "grad_norm": 0.199678435921669, "learning_rate": 2.7258374326540175e-05, "loss": 0.036, "step": 12973 }, { "epoch": 9.117357695010542, "grad_norm": 0.17447379231452942, "learning_rate": 2.725790583274772e-05, "loss": 0.0148, "step": 12974 }, { "epoch": 9.118060435699228, "grad_norm": 0.29254111647605896, "learning_rate": 2.725743733895526e-05, "loss": 0.0286, "step": 12975 }, { "epoch": 9.118763176387914, "grad_norm": 0.2407137155532837, "learning_rate": 2.7256968845162803e-05, "loss": 0.043, "step": 12976 }, { "epoch": 9.1194659170766, "grad_norm": 0.2808404564857483, "learning_rate": 2.7256500351370346e-05, "loss": 0.0392, "step": 12977 }, { "epoch": 9.120168657765285, "grad_norm": 0.3137761652469635, "learning_rate": 2.725603185757789e-05, "loss": 0.0698, "step": 12978 }, { "epoch": 9.120871398453971, "grad_norm": 1.0149924755096436, "learning_rate": 2.725556336378543e-05, "loss": 0.0947, "step": 12979 }, { "epoch": 9.121574139142655, "grad_norm": 0.724829375743866, "learning_rate": 2.7255094869992974e-05, "loss": 0.1983, "step": 12980 }, { "epoch": 9.122276879831341, "grad_norm": 0.8689862489700317, "learning_rate": 2.7254626376200518e-05, "loss": 0.1692, "step": 12981 }, { "epoch": 9.122979620520027, "grad_norm": 2.017590045928955, "learning_rate": 2.725415788240806e-05, "loss": 0.2084, "step": 12982 }, { "epoch": 9.123682361208713, "grad_norm": 0.2428252249956131, "learning_rate": 2.72536893886156e-05, "loss": 0.0702, "step": 12983 }, { "epoch": 9.1243851018974, "grad_norm": 0.2251884490251541, "learning_rate": 2.7253220894823142e-05, "loss": 0.0275, "step": 12984 }, { "epoch": 9.125087842586085, "grad_norm": 0.12301775813102722, "learning_rate": 2.7252752401030686e-05, "loss": 0.0416, "step": 12985 }, { "epoch": 9.125790583274771, "grad_norm": 0.10959857702255249, "learning_rate": 2.725228390723823e-05, "loss": 0.0165, "step": 12986 }, { "epoch": 9.126493323963457, "grad_norm": 0.12084221839904785, "learning_rate": 2.7251815413445774e-05, "loss": 0.0248, "step": 12987 }, { "epoch": 9.127196064652143, "grad_norm": 0.13787105679512024, "learning_rate": 2.7251346919653314e-05, "loss": 0.0186, "step": 12988 }, { "epoch": 9.127898805340829, "grad_norm": 0.2356913983821869, "learning_rate": 2.7250878425860858e-05, "loss": 0.0233, "step": 12989 }, { "epoch": 9.128601546029515, "grad_norm": 0.18176871538162231, "learning_rate": 2.72504099320684e-05, "loss": 0.0269, "step": 12990 }, { "epoch": 9.1293042867182, "grad_norm": 0.15500816702842712, "learning_rate": 2.7249941438275945e-05, "loss": 0.0216, "step": 12991 }, { "epoch": 9.130007027406887, "grad_norm": 0.11067783832550049, "learning_rate": 2.7249472944483485e-05, "loss": 0.015, "step": 12992 }, { "epoch": 9.130709768095572, "grad_norm": 0.17753714323043823, "learning_rate": 2.724900445069103e-05, "loss": 0.0221, "step": 12993 }, { "epoch": 9.131412508784258, "grad_norm": 0.08381900936365128, "learning_rate": 2.7248535956898573e-05, "loss": 0.0124, "step": 12994 }, { "epoch": 9.132115249472944, "grad_norm": 0.21421298384666443, "learning_rate": 2.7248067463106117e-05, "loss": 0.0398, "step": 12995 }, { "epoch": 9.13281799016163, "grad_norm": 0.10112890601158142, "learning_rate": 2.7247598969313657e-05, "loss": 0.0207, "step": 12996 }, { "epoch": 9.133520730850316, "grad_norm": 0.19518673419952393, "learning_rate": 2.72471304755212e-05, "loss": 0.0276, "step": 12997 }, { "epoch": 9.134223471539002, "grad_norm": 0.2521696388721466, "learning_rate": 2.7246661981728744e-05, "loss": 0.0205, "step": 12998 }, { "epoch": 9.134926212227688, "grad_norm": 0.226626455783844, "learning_rate": 2.7246193487936285e-05, "loss": 0.0243, "step": 12999 }, { "epoch": 9.135628952916374, "grad_norm": 0.16753043234348297, "learning_rate": 2.724572499414383e-05, "loss": 0.0307, "step": 13000 }, { "epoch": 9.135628952916374, "eval_cer": 0.1952200146203616, "eval_loss": 0.2777820825576782, "eval_runtime": 18.2055, "eval_samples_per_second": 249.265, "eval_steps_per_second": 0.824, "eval_wer": 0.34892880087148404, "step": 13000 }, { "epoch": 9.13633169360506, "grad_norm": 0.4279053509235382, "learning_rate": 2.724525650035137e-05, "loss": 0.0631, "step": 13001 }, { "epoch": 9.137034434293746, "grad_norm": 0.2676025927066803, "learning_rate": 2.7244788006558912e-05, "loss": 0.0465, "step": 13002 }, { "epoch": 9.137737174982432, "grad_norm": 0.49121877551078796, "learning_rate": 2.7244319512766456e-05, "loss": 0.085, "step": 13003 }, { "epoch": 9.138439915671118, "grad_norm": 0.36176809668540955, "learning_rate": 2.7243851018974e-05, "loss": 0.0896, "step": 13004 }, { "epoch": 9.139142656359803, "grad_norm": 0.618577241897583, "learning_rate": 2.724338252518154e-05, "loss": 0.1472, "step": 13005 }, { "epoch": 9.13984539704849, "grad_norm": 0.517368495464325, "learning_rate": 2.7242914031389084e-05, "loss": 0.1522, "step": 13006 }, { "epoch": 9.140548137737175, "grad_norm": 1.0423191785812378, "learning_rate": 2.7242445537596628e-05, "loss": 0.219, "step": 13007 }, { "epoch": 9.141250878425861, "grad_norm": 0.33486872911453247, "learning_rate": 2.724197704380417e-05, "loss": 0.0931, "step": 13008 }, { "epoch": 9.141953619114547, "grad_norm": 0.1393963247537613, "learning_rate": 2.7241508550011712e-05, "loss": 0.0353, "step": 13009 }, { "epoch": 9.142656359803233, "grad_norm": 0.09380281716585159, "learning_rate": 2.7241040056219256e-05, "loss": 0.0174, "step": 13010 }, { "epoch": 9.143359100491919, "grad_norm": 0.14013421535491943, "learning_rate": 2.72405715624268e-05, "loss": 0.015, "step": 13011 }, { "epoch": 9.144061841180605, "grad_norm": 0.15173594653606415, "learning_rate": 2.7240103068634343e-05, "loss": 0.0305, "step": 13012 }, { "epoch": 9.14476458186929, "grad_norm": 0.16501480340957642, "learning_rate": 2.7239634574841887e-05, "loss": 0.02, "step": 13013 }, { "epoch": 9.145467322557977, "grad_norm": 0.1222338080406189, "learning_rate": 2.7239166081049427e-05, "loss": 0.0151, "step": 13014 }, { "epoch": 9.146170063246663, "grad_norm": 0.21920160949230194, "learning_rate": 2.723869758725697e-05, "loss": 0.0324, "step": 13015 }, { "epoch": 9.146872803935349, "grad_norm": 0.10101243108510971, "learning_rate": 2.723822909346451e-05, "loss": 0.0208, "step": 13016 }, { "epoch": 9.147575544624035, "grad_norm": 0.08070849627256393, "learning_rate": 2.7237760599672055e-05, "loss": 0.0125, "step": 13017 }, { "epoch": 9.14827828531272, "grad_norm": 0.13332702219486237, "learning_rate": 2.7237292105879595e-05, "loss": 0.0248, "step": 13018 }, { "epoch": 9.148981026001405, "grad_norm": 0.07842998206615448, "learning_rate": 2.723682361208714e-05, "loss": 0.0109, "step": 13019 }, { "epoch": 9.14968376669009, "grad_norm": 0.28387248516082764, "learning_rate": 2.7236355118294683e-05, "loss": 0.0306, "step": 13020 }, { "epoch": 9.150386507378776, "grad_norm": 0.34059470891952515, "learning_rate": 2.7235886624502226e-05, "loss": 0.0128, "step": 13021 }, { "epoch": 9.151089248067462, "grad_norm": 0.11616265028715134, "learning_rate": 2.7235418130709767e-05, "loss": 0.0229, "step": 13022 }, { "epoch": 9.151791988756148, "grad_norm": 0.1912158876657486, "learning_rate": 2.723494963691731e-05, "loss": 0.0358, "step": 13023 }, { "epoch": 9.152494729444834, "grad_norm": 0.10425782948732376, "learning_rate": 2.7234481143124854e-05, "loss": 0.0266, "step": 13024 }, { "epoch": 9.15319747013352, "grad_norm": 0.61726313829422, "learning_rate": 2.7234012649332398e-05, "loss": 0.0268, "step": 13025 }, { "epoch": 9.153900210822206, "grad_norm": 0.19713307917118073, "learning_rate": 2.723354415553994e-05, "loss": 0.0365, "step": 13026 }, { "epoch": 9.154602951510892, "grad_norm": 0.545785129070282, "learning_rate": 2.7233075661747482e-05, "loss": 0.1042, "step": 13027 }, { "epoch": 9.155305692199578, "grad_norm": 0.3039553761482239, "learning_rate": 2.7232607167955026e-05, "loss": 0.0825, "step": 13028 }, { "epoch": 9.156008432888264, "grad_norm": 0.41144055128097534, "learning_rate": 2.723213867416257e-05, "loss": 0.1082, "step": 13029 }, { "epoch": 9.15671117357695, "grad_norm": 0.7662582397460938, "learning_rate": 2.7231670180370113e-05, "loss": 0.1898, "step": 13030 }, { "epoch": 9.157413914265636, "grad_norm": 1.8921095132827759, "learning_rate": 2.7231201686577653e-05, "loss": 0.1762, "step": 13031 }, { "epoch": 9.158116654954322, "grad_norm": 1.9215940237045288, "learning_rate": 2.7230733192785197e-05, "loss": 0.2066, "step": 13032 }, { "epoch": 9.158819395643008, "grad_norm": 0.1538333296775818, "learning_rate": 2.723026469899274e-05, "loss": 0.0592, "step": 13033 }, { "epoch": 9.159522136331693, "grad_norm": 0.10101956129074097, "learning_rate": 2.722979620520028e-05, "loss": 0.021, "step": 13034 }, { "epoch": 9.16022487702038, "grad_norm": 0.1785973310470581, "learning_rate": 2.722932771140782e-05, "loss": 0.0248, "step": 13035 }, { "epoch": 9.160927617709065, "grad_norm": 0.1593962162733078, "learning_rate": 2.7228859217615365e-05, "loss": 0.0281, "step": 13036 }, { "epoch": 9.161630358397751, "grad_norm": 0.10819797217845917, "learning_rate": 2.722839072382291e-05, "loss": 0.0162, "step": 13037 }, { "epoch": 9.162333099086437, "grad_norm": 0.23831336200237274, "learning_rate": 2.7227922230030453e-05, "loss": 0.0231, "step": 13038 }, { "epoch": 9.163035839775123, "grad_norm": 0.16666947305202484, "learning_rate": 2.7227453736237996e-05, "loss": 0.007, "step": 13039 }, { "epoch": 9.163738580463809, "grad_norm": 0.12681429088115692, "learning_rate": 2.7226985242445537e-05, "loss": 0.0219, "step": 13040 }, { "epoch": 9.164441321152495, "grad_norm": 0.1639745831489563, "learning_rate": 2.722651674865308e-05, "loss": 0.0213, "step": 13041 }, { "epoch": 9.16514406184118, "grad_norm": 0.14989344775676727, "learning_rate": 2.7226048254860624e-05, "loss": 0.0096, "step": 13042 }, { "epoch": 9.165846802529867, "grad_norm": 0.10401967167854309, "learning_rate": 2.7225579761068168e-05, "loss": 0.023, "step": 13043 }, { "epoch": 9.166549543218553, "grad_norm": 0.642135739326477, "learning_rate": 2.7225111267275708e-05, "loss": 0.0153, "step": 13044 }, { "epoch": 9.167252283907239, "grad_norm": 0.10673955827951431, "learning_rate": 2.7224642773483252e-05, "loss": 0.0139, "step": 13045 }, { "epoch": 9.167955024595924, "grad_norm": 0.08933260291814804, "learning_rate": 2.7224174279690796e-05, "loss": 0.0116, "step": 13046 }, { "epoch": 9.16865776528461, "grad_norm": 0.15614299476146698, "learning_rate": 2.722370578589834e-05, "loss": 0.0241, "step": 13047 }, { "epoch": 9.169360505973296, "grad_norm": 0.20405620336532593, "learning_rate": 2.722323729210588e-05, "loss": 0.0223, "step": 13048 }, { "epoch": 9.170063246661982, "grad_norm": 0.21924611926078796, "learning_rate": 2.7222768798313424e-05, "loss": 0.0199, "step": 13049 }, { "epoch": 9.170765987350668, "grad_norm": 0.7423213124275208, "learning_rate": 2.7222300304520967e-05, "loss": 0.0311, "step": 13050 }, { "epoch": 9.171468728039354, "grad_norm": 0.4998873174190521, "learning_rate": 2.7221831810728508e-05, "loss": 0.0418, "step": 13051 }, { "epoch": 9.17217146872804, "grad_norm": 0.7593291401863098, "learning_rate": 2.722136331693605e-05, "loss": 0.0564, "step": 13052 }, { "epoch": 9.172874209416726, "grad_norm": 0.40276187658309937, "learning_rate": 2.722089482314359e-05, "loss": 0.0838, "step": 13053 }, { "epoch": 9.173576950105412, "grad_norm": 0.32577502727508545, "learning_rate": 2.7220426329351135e-05, "loss": 0.1107, "step": 13054 }, { "epoch": 9.174279690794098, "grad_norm": 0.5941523313522339, "learning_rate": 2.721995783555868e-05, "loss": 0.1507, "step": 13055 }, { "epoch": 9.174982431482784, "grad_norm": 0.9542060494422913, "learning_rate": 2.7219489341766223e-05, "loss": 0.166, "step": 13056 }, { "epoch": 9.17568517217147, "grad_norm": 0.8695201277732849, "learning_rate": 2.7219020847973763e-05, "loss": 0.2221, "step": 13057 }, { "epoch": 9.176387912860154, "grad_norm": 0.19273242354393005, "learning_rate": 2.7218552354181307e-05, "loss": 0.0564, "step": 13058 }, { "epoch": 9.17709065354884, "grad_norm": 0.11324156075716019, "learning_rate": 2.721808386038885e-05, "loss": 0.0211, "step": 13059 }, { "epoch": 9.177793394237526, "grad_norm": 0.13270814716815948, "learning_rate": 2.7217615366596394e-05, "loss": 0.0411, "step": 13060 }, { "epoch": 9.178496134926212, "grad_norm": 0.08612427115440369, "learning_rate": 2.7217146872803935e-05, "loss": 0.0107, "step": 13061 }, { "epoch": 9.179198875614897, "grad_norm": 0.15102215111255646, "learning_rate": 2.721667837901148e-05, "loss": 0.0257, "step": 13062 }, { "epoch": 9.179901616303583, "grad_norm": 0.07691062986850739, "learning_rate": 2.7216209885219022e-05, "loss": 0.0091, "step": 13063 }, { "epoch": 9.18060435699227, "grad_norm": 0.1667085438966751, "learning_rate": 2.7215741391426566e-05, "loss": 0.0243, "step": 13064 }, { "epoch": 9.181307097680955, "grad_norm": 0.14565354585647583, "learning_rate": 2.721527289763411e-05, "loss": 0.02, "step": 13065 }, { "epoch": 9.182009838369641, "grad_norm": 0.10258514434099197, "learning_rate": 2.721480440384165e-05, "loss": 0.017, "step": 13066 }, { "epoch": 9.182712579058327, "grad_norm": 0.12227876484394073, "learning_rate": 2.7214335910049194e-05, "loss": 0.0202, "step": 13067 }, { "epoch": 9.183415319747013, "grad_norm": 0.24392223358154297, "learning_rate": 2.7213867416256737e-05, "loss": 0.0276, "step": 13068 }, { "epoch": 9.184118060435699, "grad_norm": 0.12179838120937347, "learning_rate": 2.7213398922464278e-05, "loss": 0.0186, "step": 13069 }, { "epoch": 9.184820801124385, "grad_norm": 0.2510210871696472, "learning_rate": 2.7212930428671818e-05, "loss": 0.0242, "step": 13070 }, { "epoch": 9.18552354181307, "grad_norm": 0.17818300426006317, "learning_rate": 2.7212461934879362e-05, "loss": 0.0166, "step": 13071 }, { "epoch": 9.186226282501757, "grad_norm": 0.22416982054710388, "learning_rate": 2.7211993441086905e-05, "loss": 0.0329, "step": 13072 }, { "epoch": 9.186929023190443, "grad_norm": 0.250863254070282, "learning_rate": 2.721152494729445e-05, "loss": 0.0448, "step": 13073 }, { "epoch": 9.187631763879128, "grad_norm": 0.18765588104724884, "learning_rate": 2.721105645350199e-05, "loss": 0.0331, "step": 13074 }, { "epoch": 9.188334504567814, "grad_norm": 0.1500394195318222, "learning_rate": 2.7210587959709533e-05, "loss": 0.023, "step": 13075 }, { "epoch": 9.1890372452565, "grad_norm": 0.23529304563999176, "learning_rate": 2.7210119465917077e-05, "loss": 0.0458, "step": 13076 }, { "epoch": 9.189739985945186, "grad_norm": 0.2567824721336365, "learning_rate": 2.720965097212462e-05, "loss": 0.0496, "step": 13077 }, { "epoch": 9.190442726633872, "grad_norm": 0.38332104682922363, "learning_rate": 2.7209182478332164e-05, "loss": 0.1021, "step": 13078 }, { "epoch": 9.191145467322558, "grad_norm": 0.4113841652870178, "learning_rate": 2.7208713984539705e-05, "loss": 0.1291, "step": 13079 }, { "epoch": 9.191848208011244, "grad_norm": 1.709681749343872, "learning_rate": 2.720824549074725e-05, "loss": 0.1832, "step": 13080 }, { "epoch": 9.19255094869993, "grad_norm": 0.6453352570533752, "learning_rate": 2.7207776996954792e-05, "loss": 0.1974, "step": 13081 }, { "epoch": 9.193253689388616, "grad_norm": 1.298415184020996, "learning_rate": 2.7207308503162336e-05, "loss": 0.2894, "step": 13082 }, { "epoch": 9.193956430077302, "grad_norm": 0.27368393540382385, "learning_rate": 2.7206840009369876e-05, "loss": 0.0647, "step": 13083 }, { "epoch": 9.194659170765988, "grad_norm": 0.2255186140537262, "learning_rate": 2.720637151557742e-05, "loss": 0.0245, "step": 13084 }, { "epoch": 9.195361911454674, "grad_norm": 0.14411243796348572, "learning_rate": 2.7205903021784964e-05, "loss": 0.0302, "step": 13085 }, { "epoch": 9.19606465214336, "grad_norm": 0.12258706986904144, "learning_rate": 2.7205434527992504e-05, "loss": 0.0241, "step": 13086 }, { "epoch": 9.196767392832045, "grad_norm": 0.4040243923664093, "learning_rate": 2.7204966034200048e-05, "loss": 0.0168, "step": 13087 }, { "epoch": 9.197470133520731, "grad_norm": 0.23536019027233124, "learning_rate": 2.7204497540407588e-05, "loss": 0.0185, "step": 13088 }, { "epoch": 9.198172874209417, "grad_norm": 0.19198203086853027, "learning_rate": 2.7204029046615132e-05, "loss": 0.0201, "step": 13089 }, { "epoch": 9.198875614898103, "grad_norm": 0.23209823668003082, "learning_rate": 2.7203560552822676e-05, "loss": 0.022, "step": 13090 }, { "epoch": 9.19957835558679, "grad_norm": 0.1695926934480667, "learning_rate": 2.720309205903022e-05, "loss": 0.0225, "step": 13091 }, { "epoch": 9.200281096275475, "grad_norm": 0.10357952117919922, "learning_rate": 2.720262356523776e-05, "loss": 0.0104, "step": 13092 }, { "epoch": 9.200983836964161, "grad_norm": 0.1421007364988327, "learning_rate": 2.7202155071445303e-05, "loss": 0.0237, "step": 13093 }, { "epoch": 9.201686577652847, "grad_norm": 0.13963761925697327, "learning_rate": 2.7201686577652847e-05, "loss": 0.0099, "step": 13094 }, { "epoch": 9.202389318341533, "grad_norm": 0.27907896041870117, "learning_rate": 2.720121808386039e-05, "loss": 0.0347, "step": 13095 }, { "epoch": 9.203092059030217, "grad_norm": 0.15308472514152527, "learning_rate": 2.720074959006793e-05, "loss": 0.0168, "step": 13096 }, { "epoch": 9.203794799718903, "grad_norm": 0.16730564832687378, "learning_rate": 2.7200281096275475e-05, "loss": 0.0292, "step": 13097 }, { "epoch": 9.204497540407589, "grad_norm": 0.17739750444889069, "learning_rate": 2.719981260248302e-05, "loss": 0.0276, "step": 13098 }, { "epoch": 9.205200281096275, "grad_norm": 0.4873010218143463, "learning_rate": 2.7199344108690562e-05, "loss": 0.025, "step": 13099 }, { "epoch": 9.20590302178496, "grad_norm": 0.843565821647644, "learning_rate": 2.7198875614898103e-05, "loss": 0.0319, "step": 13100 }, { "epoch": 9.206605762473647, "grad_norm": 0.20945188403129578, "learning_rate": 2.7198407121105646e-05, "loss": 0.0413, "step": 13101 }, { "epoch": 9.207308503162333, "grad_norm": 0.29834428429603577, "learning_rate": 2.719793862731319e-05, "loss": 0.0766, "step": 13102 }, { "epoch": 9.208011243851018, "grad_norm": 0.9851646423339844, "learning_rate": 2.7197470133520734e-05, "loss": 0.081, "step": 13103 }, { "epoch": 9.208713984539704, "grad_norm": 0.3495718240737915, "learning_rate": 2.7197001639728274e-05, "loss": 0.1033, "step": 13104 }, { "epoch": 9.20941672522839, "grad_norm": 0.6805979013442993, "learning_rate": 2.7196533145935815e-05, "loss": 0.1399, "step": 13105 }, { "epoch": 9.210119465917076, "grad_norm": 0.5047782063484192, "learning_rate": 2.7196064652143358e-05, "loss": 0.148, "step": 13106 }, { "epoch": 9.210822206605762, "grad_norm": 0.910194993019104, "learning_rate": 2.7195596158350902e-05, "loss": 0.189, "step": 13107 }, { "epoch": 9.211524947294448, "grad_norm": 0.25113120675086975, "learning_rate": 2.7195127664558446e-05, "loss": 0.0736, "step": 13108 }, { "epoch": 9.212227687983134, "grad_norm": 0.11401958018541336, "learning_rate": 2.7194659170765986e-05, "loss": 0.0197, "step": 13109 }, { "epoch": 9.21293042867182, "grad_norm": 0.17599858343601227, "learning_rate": 2.719419067697353e-05, "loss": 0.0226, "step": 13110 }, { "epoch": 9.213633169360506, "grad_norm": 0.1440432220697403, "learning_rate": 2.7193722183181074e-05, "loss": 0.0184, "step": 13111 }, { "epoch": 9.214335910049192, "grad_norm": 0.1753114014863968, "learning_rate": 2.7193253689388617e-05, "loss": 0.0118, "step": 13112 }, { "epoch": 9.215038650737878, "grad_norm": 0.10241333395242691, "learning_rate": 2.719278519559616e-05, "loss": 0.0144, "step": 13113 }, { "epoch": 9.215741391426564, "grad_norm": 0.1280709207057953, "learning_rate": 2.71923167018037e-05, "loss": 0.0097, "step": 13114 }, { "epoch": 9.21644413211525, "grad_norm": 0.29373934864997864, "learning_rate": 2.7191848208011245e-05, "loss": 0.0305, "step": 13115 }, { "epoch": 9.217146872803935, "grad_norm": 0.24430683255195618, "learning_rate": 2.719137971421879e-05, "loss": 0.027, "step": 13116 }, { "epoch": 9.217849613492621, "grad_norm": 0.06289765983819962, "learning_rate": 2.7190911220426332e-05, "loss": 0.0082, "step": 13117 }, { "epoch": 9.218552354181307, "grad_norm": 0.1515532284975052, "learning_rate": 2.7190442726633873e-05, "loss": 0.025, "step": 13118 }, { "epoch": 9.219255094869993, "grad_norm": 0.10506784915924072, "learning_rate": 2.7189974232841417e-05, "loss": 0.0141, "step": 13119 }, { "epoch": 9.219957835558679, "grad_norm": 0.6957598328590393, "learning_rate": 2.718950573904896e-05, "loss": 0.0195, "step": 13120 }, { "epoch": 9.220660576247365, "grad_norm": 0.13200855255126953, "learning_rate": 2.71890372452565e-05, "loss": 0.014, "step": 13121 }, { "epoch": 9.221363316936051, "grad_norm": 0.3750894367694855, "learning_rate": 2.718856875146404e-05, "loss": 0.0244, "step": 13122 }, { "epoch": 9.222066057624737, "grad_norm": 0.13779497146606445, "learning_rate": 2.7188100257671585e-05, "loss": 0.0331, "step": 13123 }, { "epoch": 9.222768798313423, "grad_norm": 0.23624688386917114, "learning_rate": 2.718763176387913e-05, "loss": 0.015, "step": 13124 }, { "epoch": 9.223471539002109, "grad_norm": 0.29211822152137756, "learning_rate": 2.7187163270086672e-05, "loss": 0.0445, "step": 13125 }, { "epoch": 9.224174279690795, "grad_norm": 0.8966199159622192, "learning_rate": 2.7186694776294216e-05, "loss": 0.0423, "step": 13126 }, { "epoch": 9.22487702037948, "grad_norm": 0.5637885928153992, "learning_rate": 2.7186226282501756e-05, "loss": 0.0671, "step": 13127 }, { "epoch": 9.225579761068166, "grad_norm": 0.692601203918457, "learning_rate": 2.71857577887093e-05, "loss": 0.0827, "step": 13128 }, { "epoch": 9.226282501756852, "grad_norm": 0.5178441405296326, "learning_rate": 2.7185289294916844e-05, "loss": 0.1299, "step": 13129 }, { "epoch": 9.226985242445538, "grad_norm": 0.696983814239502, "learning_rate": 2.7184820801124387e-05, "loss": 0.1517, "step": 13130 }, { "epoch": 9.227687983134224, "grad_norm": 1.359904170036316, "learning_rate": 2.7184352307331928e-05, "loss": 0.1985, "step": 13131 }, { "epoch": 9.22839072382291, "grad_norm": 1.5829503536224365, "learning_rate": 2.718388381353947e-05, "loss": 0.1875, "step": 13132 }, { "epoch": 9.229093464511596, "grad_norm": 0.20084743201732635, "learning_rate": 2.7183415319747015e-05, "loss": 0.0821, "step": 13133 }, { "epoch": 9.22979620520028, "grad_norm": 0.13170892000198364, "learning_rate": 2.718294682595456e-05, "loss": 0.0318, "step": 13134 }, { "epoch": 9.230498945888966, "grad_norm": 0.15522971749305725, "learning_rate": 2.71824783321621e-05, "loss": 0.0261, "step": 13135 }, { "epoch": 9.231201686577652, "grad_norm": 0.1001509577035904, "learning_rate": 2.7182009838369643e-05, "loss": 0.0149, "step": 13136 }, { "epoch": 9.231904427266338, "grad_norm": 0.1458946019411087, "learning_rate": 2.7181541344577187e-05, "loss": 0.031, "step": 13137 }, { "epoch": 9.232607167955024, "grad_norm": 0.422880619764328, "learning_rate": 2.7181072850784727e-05, "loss": 0.0129, "step": 13138 }, { "epoch": 9.23330990864371, "grad_norm": 0.08533342182636261, "learning_rate": 2.718060435699227e-05, "loss": 0.0161, "step": 13139 }, { "epoch": 9.234012649332396, "grad_norm": 0.3167298436164856, "learning_rate": 2.718013586319981e-05, "loss": 0.0443, "step": 13140 }, { "epoch": 9.234715390021082, "grad_norm": 0.10260944813489914, "learning_rate": 2.7179667369407355e-05, "loss": 0.0214, "step": 13141 }, { "epoch": 9.235418130709768, "grad_norm": 0.10182947665452957, "learning_rate": 2.71791988756149e-05, "loss": 0.0294, "step": 13142 }, { "epoch": 9.236120871398454, "grad_norm": 0.14717967808246613, "learning_rate": 2.7178730381822442e-05, "loss": 0.0398, "step": 13143 }, { "epoch": 9.23682361208714, "grad_norm": 0.10538748651742935, "learning_rate": 2.7178261888029983e-05, "loss": 0.009, "step": 13144 }, { "epoch": 9.237526352775825, "grad_norm": 0.14574624598026276, "learning_rate": 2.7177793394237526e-05, "loss": 0.0265, "step": 13145 }, { "epoch": 9.238229093464511, "grad_norm": 0.115060955286026, "learning_rate": 2.717732490044507e-05, "loss": 0.0262, "step": 13146 }, { "epoch": 9.238931834153197, "grad_norm": 0.33223816752433777, "learning_rate": 2.7176856406652614e-05, "loss": 0.0212, "step": 13147 }, { "epoch": 9.239634574841883, "grad_norm": 0.256510853767395, "learning_rate": 2.7176387912860154e-05, "loss": 0.0384, "step": 13148 }, { "epoch": 9.240337315530569, "grad_norm": 0.16693544387817383, "learning_rate": 2.7175919419067698e-05, "loss": 0.02, "step": 13149 }, { "epoch": 9.241040056219255, "grad_norm": 0.2644510269165039, "learning_rate": 2.717545092527524e-05, "loss": 0.0508, "step": 13150 }, { "epoch": 9.24174279690794, "grad_norm": 0.23723439872264862, "learning_rate": 2.7174982431482785e-05, "loss": 0.0522, "step": 13151 }, { "epoch": 9.242445537596627, "grad_norm": 0.23051075637340546, "learning_rate": 2.717451393769033e-05, "loss": 0.0498, "step": 13152 }, { "epoch": 9.243148278285313, "grad_norm": 0.7886036038398743, "learning_rate": 2.717404544389787e-05, "loss": 0.0821, "step": 13153 }, { "epoch": 9.243851018973999, "grad_norm": 0.4512055218219757, "learning_rate": 2.7173576950105413e-05, "loss": 0.1045, "step": 13154 }, { "epoch": 9.244553759662685, "grad_norm": 0.4475705921649933, "learning_rate": 2.7173108456312957e-05, "loss": 0.1221, "step": 13155 }, { "epoch": 9.24525650035137, "grad_norm": 1.447519302368164, "learning_rate": 2.7172639962520497e-05, "loss": 0.155, "step": 13156 }, { "epoch": 9.245959241040056, "grad_norm": 3.8504886627197266, "learning_rate": 2.7172171468728037e-05, "loss": 0.1806, "step": 13157 }, { "epoch": 9.246661981728742, "grad_norm": 0.3302026689052582, "learning_rate": 2.717170297493558e-05, "loss": 0.1028, "step": 13158 }, { "epoch": 9.247364722417428, "grad_norm": 0.28976428508758545, "learning_rate": 2.7171234481143125e-05, "loss": 0.0436, "step": 13159 }, { "epoch": 9.248067463106114, "grad_norm": 0.19783838093280792, "learning_rate": 2.717076598735067e-05, "loss": 0.0291, "step": 13160 }, { "epoch": 9.2487702037948, "grad_norm": 0.15384547412395477, "learning_rate": 2.717029749355821e-05, "loss": 0.0227, "step": 13161 }, { "epoch": 9.249472944483486, "grad_norm": 0.14729420840740204, "learning_rate": 2.7169828999765753e-05, "loss": 0.0164, "step": 13162 }, { "epoch": 9.250175685172172, "grad_norm": 0.08648239076137543, "learning_rate": 2.7169360505973296e-05, "loss": 0.0136, "step": 13163 }, { "epoch": 9.250878425860858, "grad_norm": 0.18624791502952576, "learning_rate": 2.716889201218084e-05, "loss": 0.0205, "step": 13164 }, { "epoch": 9.251581166549544, "grad_norm": 0.08321633189916611, "learning_rate": 2.7168423518388384e-05, "loss": 0.0075, "step": 13165 }, { "epoch": 9.25228390723823, "grad_norm": 0.15336297452449799, "learning_rate": 2.7167955024595924e-05, "loss": 0.0336, "step": 13166 }, { "epoch": 9.252986647926916, "grad_norm": 0.16561304032802582, "learning_rate": 2.7167486530803468e-05, "loss": 0.0146, "step": 13167 }, { "epoch": 9.253689388615602, "grad_norm": 0.11436895281076431, "learning_rate": 2.716701803701101e-05, "loss": 0.0229, "step": 13168 }, { "epoch": 9.254392129304287, "grad_norm": 0.15885309875011444, "learning_rate": 2.7166549543218555e-05, "loss": 0.0204, "step": 13169 }, { "epoch": 9.255094869992973, "grad_norm": 0.1488550454378128, "learning_rate": 2.7166081049426096e-05, "loss": 0.026, "step": 13170 }, { "epoch": 9.25579761068166, "grad_norm": 0.19611656665802002, "learning_rate": 2.716561255563364e-05, "loss": 0.0249, "step": 13171 }, { "epoch": 9.256500351370345, "grad_norm": 0.1402842253446579, "learning_rate": 2.7165144061841183e-05, "loss": 0.0187, "step": 13172 }, { "epoch": 9.25720309205903, "grad_norm": 0.22420921921730042, "learning_rate": 2.7164675568048723e-05, "loss": 0.0447, "step": 13173 }, { "epoch": 9.257905832747715, "grad_norm": 0.10937285423278809, "learning_rate": 2.7164207074256264e-05, "loss": 0.018, "step": 13174 }, { "epoch": 9.258608573436401, "grad_norm": 0.30984169244766235, "learning_rate": 2.7163738580463808e-05, "loss": 0.0358, "step": 13175 }, { "epoch": 9.259311314125087, "grad_norm": 0.19639025628566742, "learning_rate": 2.716327008667135e-05, "loss": 0.0445, "step": 13176 }, { "epoch": 9.260014054813773, "grad_norm": 0.4606807231903076, "learning_rate": 2.7162801592878895e-05, "loss": 0.0516, "step": 13177 }, { "epoch": 9.260716795502459, "grad_norm": 0.2883003354072571, "learning_rate": 2.716233309908644e-05, "loss": 0.0593, "step": 13178 }, { "epoch": 9.261419536191145, "grad_norm": 0.6469007134437561, "learning_rate": 2.716186460529398e-05, "loss": 0.1031, "step": 13179 }, { "epoch": 9.26212227687983, "grad_norm": 0.7436271905899048, "learning_rate": 2.7161396111501523e-05, "loss": 0.1299, "step": 13180 }, { "epoch": 9.262825017568517, "grad_norm": 0.6507363319396973, "learning_rate": 2.7160927617709067e-05, "loss": 0.1804, "step": 13181 }, { "epoch": 9.263527758257203, "grad_norm": 1.111067295074463, "learning_rate": 2.716045912391661e-05, "loss": 0.2107, "step": 13182 }, { "epoch": 9.264230498945889, "grad_norm": 0.39526498317718506, "learning_rate": 2.715999063012415e-05, "loss": 0.0765, "step": 13183 }, { "epoch": 9.264933239634574, "grad_norm": 0.13898001611232758, "learning_rate": 2.7159522136331694e-05, "loss": 0.0231, "step": 13184 }, { "epoch": 9.26563598032326, "grad_norm": 0.9766921997070312, "learning_rate": 2.7159053642539238e-05, "loss": 0.0311, "step": 13185 }, { "epoch": 9.266338721011946, "grad_norm": 0.0998847708106041, "learning_rate": 2.7158585148746782e-05, "loss": 0.0227, "step": 13186 }, { "epoch": 9.267041461700632, "grad_norm": 0.11488979309797287, "learning_rate": 2.7158116654954322e-05, "loss": 0.0157, "step": 13187 }, { "epoch": 9.267744202389318, "grad_norm": 0.11229850351810455, "learning_rate": 2.7157648161161866e-05, "loss": 0.0115, "step": 13188 }, { "epoch": 9.268446943078004, "grad_norm": 0.14179007709026337, "learning_rate": 2.715717966736941e-05, "loss": 0.0151, "step": 13189 }, { "epoch": 9.26914968376669, "grad_norm": 0.1479020118713379, "learning_rate": 2.7156711173576953e-05, "loss": 0.0132, "step": 13190 }, { "epoch": 9.269852424455376, "grad_norm": 0.14122043550014496, "learning_rate": 2.7156242679784494e-05, "loss": 0.0293, "step": 13191 }, { "epoch": 9.270555165144062, "grad_norm": 0.1152305081486702, "learning_rate": 2.7155774185992034e-05, "loss": 0.0117, "step": 13192 }, { "epoch": 9.271257905832748, "grad_norm": 0.1306096911430359, "learning_rate": 2.7155305692199578e-05, "loss": 0.0158, "step": 13193 }, { "epoch": 9.271960646521434, "grad_norm": 0.16025133430957794, "learning_rate": 2.715483719840712e-05, "loss": 0.0198, "step": 13194 }, { "epoch": 9.27266338721012, "grad_norm": 0.14020799100399017, "learning_rate": 2.7154368704614665e-05, "loss": 0.0285, "step": 13195 }, { "epoch": 9.273366127898806, "grad_norm": 0.08824501186609268, "learning_rate": 2.7153900210822205e-05, "loss": 0.0101, "step": 13196 }, { "epoch": 9.274068868587491, "grad_norm": 0.1211482584476471, "learning_rate": 2.715343171702975e-05, "loss": 0.0327, "step": 13197 }, { "epoch": 9.274771609276177, "grad_norm": 0.11816717684268951, "learning_rate": 2.7152963223237293e-05, "loss": 0.031, "step": 13198 }, { "epoch": 9.275474349964863, "grad_norm": 0.10605504363775253, "learning_rate": 2.7152494729444837e-05, "loss": 0.0159, "step": 13199 }, { "epoch": 9.27617709065355, "grad_norm": 0.40519315004348755, "learning_rate": 2.7152026235652377e-05, "loss": 0.049, "step": 13200 }, { "epoch": 9.276879831342235, "grad_norm": 0.35434967279434204, "learning_rate": 2.715155774185992e-05, "loss": 0.0546, "step": 13201 }, { "epoch": 9.277582572030921, "grad_norm": 0.269113153219223, "learning_rate": 2.7151089248067464e-05, "loss": 0.0619, "step": 13202 }, { "epoch": 9.278285312719607, "grad_norm": 0.7769583463668823, "learning_rate": 2.7150620754275008e-05, "loss": 0.0753, "step": 13203 }, { "epoch": 9.278988053408293, "grad_norm": 0.5676798224449158, "learning_rate": 2.7150152260482552e-05, "loss": 0.1022, "step": 13204 }, { "epoch": 9.279690794096979, "grad_norm": 0.5099859237670898, "learning_rate": 2.7149683766690092e-05, "loss": 0.178, "step": 13205 }, { "epoch": 9.280393534785665, "grad_norm": 0.6036282181739807, "learning_rate": 2.7149215272897636e-05, "loss": 0.2049, "step": 13206 }, { "epoch": 9.28109627547435, "grad_norm": 2.0578019618988037, "learning_rate": 2.714874677910518e-05, "loss": 0.2276, "step": 13207 }, { "epoch": 9.281799016163037, "grad_norm": 0.16516442596912384, "learning_rate": 2.714827828531272e-05, "loss": 0.0727, "step": 13208 }, { "epoch": 9.282501756851723, "grad_norm": 0.15478138625621796, "learning_rate": 2.714780979152026e-05, "loss": 0.0379, "step": 13209 }, { "epoch": 9.283204497540408, "grad_norm": 0.19669930636882782, "learning_rate": 2.7147341297727804e-05, "loss": 0.034, "step": 13210 }, { "epoch": 9.283907238229094, "grad_norm": 0.12833856046199799, "learning_rate": 2.7146872803935348e-05, "loss": 0.0214, "step": 13211 }, { "epoch": 9.284609978917779, "grad_norm": 0.2661011219024658, "learning_rate": 2.714640431014289e-05, "loss": 0.0245, "step": 13212 }, { "epoch": 9.285312719606464, "grad_norm": 0.08757489919662476, "learning_rate": 2.7145935816350432e-05, "loss": 0.0162, "step": 13213 }, { "epoch": 9.28601546029515, "grad_norm": 0.12733106315135956, "learning_rate": 2.7145467322557976e-05, "loss": 0.0198, "step": 13214 }, { "epoch": 9.286718200983836, "grad_norm": 0.17594470083713531, "learning_rate": 2.714499882876552e-05, "loss": 0.0256, "step": 13215 }, { "epoch": 9.287420941672522, "grad_norm": 0.10333044826984406, "learning_rate": 2.7144530334973063e-05, "loss": 0.0193, "step": 13216 }, { "epoch": 9.288123682361208, "grad_norm": 0.12040833383798599, "learning_rate": 2.7144061841180607e-05, "loss": 0.0161, "step": 13217 }, { "epoch": 9.288826423049894, "grad_norm": 0.11626776307821274, "learning_rate": 2.7143593347388147e-05, "loss": 0.026, "step": 13218 }, { "epoch": 9.28952916373858, "grad_norm": 0.10034378618001938, "learning_rate": 2.714312485359569e-05, "loss": 0.0115, "step": 13219 }, { "epoch": 9.290231904427266, "grad_norm": 0.27745968103408813, "learning_rate": 2.7142656359803235e-05, "loss": 0.0265, "step": 13220 }, { "epoch": 9.290934645115952, "grad_norm": 0.12890160083770752, "learning_rate": 2.7142187866010778e-05, "loss": 0.0267, "step": 13221 }, { "epoch": 9.291637385804638, "grad_norm": 0.16351889073848724, "learning_rate": 2.714171937221832e-05, "loss": 0.0236, "step": 13222 }, { "epoch": 9.292340126493324, "grad_norm": 0.09807422757148743, "learning_rate": 2.7141250878425862e-05, "loss": 0.0217, "step": 13223 }, { "epoch": 9.29304286718201, "grad_norm": 0.09774955362081528, "learning_rate": 2.7140782384633406e-05, "loss": 0.0137, "step": 13224 }, { "epoch": 9.293745607870695, "grad_norm": 0.12635330855846405, "learning_rate": 2.7140313890840946e-05, "loss": 0.0261, "step": 13225 }, { "epoch": 9.294448348559381, "grad_norm": 0.48405569791793823, "learning_rate": 2.7139845397048487e-05, "loss": 0.0394, "step": 13226 }, { "epoch": 9.295151089248067, "grad_norm": 0.7052330374717712, "learning_rate": 2.713937690325603e-05, "loss": 0.072, "step": 13227 }, { "epoch": 9.295853829936753, "grad_norm": 0.30920183658599854, "learning_rate": 2.7138908409463574e-05, "loss": 0.0759, "step": 13228 }, { "epoch": 9.29655657062544, "grad_norm": 0.30822503566741943, "learning_rate": 2.7138439915671118e-05, "loss": 0.1039, "step": 13229 }, { "epoch": 9.297259311314125, "grad_norm": 0.39593416452407837, "learning_rate": 2.713797142187866e-05, "loss": 0.1303, "step": 13230 }, { "epoch": 9.297962052002811, "grad_norm": 0.5390492081642151, "learning_rate": 2.7137502928086202e-05, "loss": 0.173, "step": 13231 }, { "epoch": 9.298664792691497, "grad_norm": 0.8164365291595459, "learning_rate": 2.7137034434293746e-05, "loss": 0.193, "step": 13232 }, { "epoch": 9.299367533380183, "grad_norm": 0.24673300981521606, "learning_rate": 2.713656594050129e-05, "loss": 0.0527, "step": 13233 }, { "epoch": 9.300070274068869, "grad_norm": 0.07600860297679901, "learning_rate": 2.7136097446708833e-05, "loss": 0.0176, "step": 13234 }, { "epoch": 9.300773014757555, "grad_norm": 0.12573382258415222, "learning_rate": 2.7135628952916373e-05, "loss": 0.022, "step": 13235 }, { "epoch": 9.30147575544624, "grad_norm": 0.08795009553432465, "learning_rate": 2.7135160459123917e-05, "loss": 0.0177, "step": 13236 }, { "epoch": 9.302178496134927, "grad_norm": 0.12361452728509903, "learning_rate": 2.713469196533146e-05, "loss": 0.0214, "step": 13237 }, { "epoch": 9.302881236823612, "grad_norm": 0.08061929047107697, "learning_rate": 2.7134223471539005e-05, "loss": 0.0089, "step": 13238 }, { "epoch": 9.303583977512298, "grad_norm": 0.11547470092773438, "learning_rate": 2.7133754977746545e-05, "loss": 0.0162, "step": 13239 }, { "epoch": 9.304286718200984, "grad_norm": 0.2731703817844391, "learning_rate": 2.713328648395409e-05, "loss": 0.0077, "step": 13240 }, { "epoch": 9.30498945888967, "grad_norm": 0.08551862835884094, "learning_rate": 2.7132817990161632e-05, "loss": 0.0133, "step": 13241 }, { "epoch": 9.305692199578356, "grad_norm": 0.11328302323818207, "learning_rate": 2.7132349496369176e-05, "loss": 0.0213, "step": 13242 }, { "epoch": 9.306394940267042, "grad_norm": 0.17870034277439117, "learning_rate": 2.7131881002576716e-05, "loss": 0.0314, "step": 13243 }, { "epoch": 9.307097680955728, "grad_norm": 0.0722295418381691, "learning_rate": 2.7131412508784257e-05, "loss": 0.0139, "step": 13244 }, { "epoch": 9.307800421644414, "grad_norm": 0.17990374565124512, "learning_rate": 2.71309440149918e-05, "loss": 0.0214, "step": 13245 }, { "epoch": 9.3085031623331, "grad_norm": 0.1469632387161255, "learning_rate": 2.7130475521199344e-05, "loss": 0.0243, "step": 13246 }, { "epoch": 9.309205903021786, "grad_norm": 0.1607045829296112, "learning_rate": 2.7130007027406888e-05, "loss": 0.023, "step": 13247 }, { "epoch": 9.309908643710472, "grad_norm": 0.13322651386260986, "learning_rate": 2.712953853361443e-05, "loss": 0.0196, "step": 13248 }, { "epoch": 9.310611384399156, "grad_norm": 0.12810520827770233, "learning_rate": 2.7129070039821972e-05, "loss": 0.0203, "step": 13249 }, { "epoch": 9.311314125087842, "grad_norm": 0.28595319390296936, "learning_rate": 2.7128601546029516e-05, "loss": 0.0213, "step": 13250 }, { "epoch": 9.312016865776528, "grad_norm": 0.4085759222507477, "learning_rate": 2.712813305223706e-05, "loss": 0.0568, "step": 13251 }, { "epoch": 9.312719606465214, "grad_norm": 0.4011700749397278, "learning_rate": 2.71276645584446e-05, "loss": 0.0436, "step": 13252 }, { "epoch": 9.3134223471539, "grad_norm": 0.5620911717414856, "learning_rate": 2.7127196064652144e-05, "loss": 0.0783, "step": 13253 }, { "epoch": 9.314125087842585, "grad_norm": 0.43851250410079956, "learning_rate": 2.7126727570859687e-05, "loss": 0.1145, "step": 13254 }, { "epoch": 9.314827828531271, "grad_norm": 1.0984365940093994, "learning_rate": 2.712625907706723e-05, "loss": 0.1431, "step": 13255 }, { "epoch": 9.315530569219957, "grad_norm": 1.0631062984466553, "learning_rate": 2.7125790583274775e-05, "loss": 0.2204, "step": 13256 }, { "epoch": 9.316233309908643, "grad_norm": 0.8560409545898438, "learning_rate": 2.7125322089482315e-05, "loss": 0.245, "step": 13257 }, { "epoch": 9.316936050597329, "grad_norm": 0.266989141702652, "learning_rate": 2.712485359568986e-05, "loss": 0.0549, "step": 13258 }, { "epoch": 9.317638791286015, "grad_norm": 0.13226304948329926, "learning_rate": 2.7124385101897403e-05, "loss": 0.0181, "step": 13259 }, { "epoch": 9.318341531974701, "grad_norm": 0.11279649287462234, "learning_rate": 2.7123916608104943e-05, "loss": 0.0272, "step": 13260 }, { "epoch": 9.319044272663387, "grad_norm": 0.10703255236148834, "learning_rate": 2.7123448114312483e-05, "loss": 0.0275, "step": 13261 }, { "epoch": 9.319747013352073, "grad_norm": 0.10770543664693832, "learning_rate": 2.7122979620520027e-05, "loss": 0.0308, "step": 13262 }, { "epoch": 9.320449754040759, "grad_norm": 0.18434028327465057, "learning_rate": 2.712251112672757e-05, "loss": 0.0074, "step": 13263 }, { "epoch": 9.321152494729445, "grad_norm": 0.1677500307559967, "learning_rate": 2.7122042632935114e-05, "loss": 0.0207, "step": 13264 }, { "epoch": 9.32185523541813, "grad_norm": 0.11501289159059525, "learning_rate": 2.7121574139142655e-05, "loss": 0.0193, "step": 13265 }, { "epoch": 9.322557976106816, "grad_norm": 0.11527309566736221, "learning_rate": 2.71211056453502e-05, "loss": 0.0253, "step": 13266 }, { "epoch": 9.323260716795502, "grad_norm": 0.20419180393218994, "learning_rate": 2.7120637151557742e-05, "loss": 0.0113, "step": 13267 }, { "epoch": 9.323963457484188, "grad_norm": 0.20163166522979736, "learning_rate": 2.7120168657765286e-05, "loss": 0.0292, "step": 13268 }, { "epoch": 9.324666198172874, "grad_norm": 0.12008988112211227, "learning_rate": 2.711970016397283e-05, "loss": 0.0099, "step": 13269 }, { "epoch": 9.32536893886156, "grad_norm": 0.458075612783432, "learning_rate": 2.711923167018037e-05, "loss": 0.0349, "step": 13270 }, { "epoch": 9.326071679550246, "grad_norm": 0.6157525777816772, "learning_rate": 2.7118763176387914e-05, "loss": 0.024, "step": 13271 }, { "epoch": 9.326774420238932, "grad_norm": 0.14415615797042847, "learning_rate": 2.7118294682595457e-05, "loss": 0.026, "step": 13272 }, { "epoch": 9.327477160927618, "grad_norm": 0.16482964158058167, "learning_rate": 2.7117826188803e-05, "loss": 0.0283, "step": 13273 }, { "epoch": 9.328179901616304, "grad_norm": 0.15338176488876343, "learning_rate": 2.711735769501054e-05, "loss": 0.0286, "step": 13274 }, { "epoch": 9.32888264230499, "grad_norm": 0.2780884802341461, "learning_rate": 2.7116889201218085e-05, "loss": 0.0446, "step": 13275 }, { "epoch": 9.329585382993676, "grad_norm": 0.15924528241157532, "learning_rate": 2.711642070742563e-05, "loss": 0.0436, "step": 13276 }, { "epoch": 9.330288123682362, "grad_norm": 0.5362390875816345, "learning_rate": 2.7115952213633173e-05, "loss": 0.0645, "step": 13277 }, { "epoch": 9.330990864371048, "grad_norm": 0.4380698800086975, "learning_rate": 2.711548371984071e-05, "loss": 0.0649, "step": 13278 }, { "epoch": 9.331693605059733, "grad_norm": 0.8107923269271851, "learning_rate": 2.7115015226048253e-05, "loss": 0.1117, "step": 13279 }, { "epoch": 9.33239634574842, "grad_norm": 0.5587785840034485, "learning_rate": 2.7114546732255797e-05, "loss": 0.1689, "step": 13280 }, { "epoch": 9.333099086437105, "grad_norm": 0.5326409935951233, "learning_rate": 2.711407823846334e-05, "loss": 0.169, "step": 13281 }, { "epoch": 9.333801827125791, "grad_norm": 1.8829140663146973, "learning_rate": 2.7113609744670884e-05, "loss": 0.1721, "step": 13282 }, { "epoch": 9.334504567814477, "grad_norm": 0.1306750774383545, "learning_rate": 2.7113141250878425e-05, "loss": 0.0548, "step": 13283 }, { "epoch": 9.335207308503163, "grad_norm": 0.14397768676280975, "learning_rate": 2.711267275708597e-05, "loss": 0.0289, "step": 13284 }, { "epoch": 9.335910049191849, "grad_norm": 0.46454674005508423, "learning_rate": 2.7112204263293512e-05, "loss": 0.0234, "step": 13285 }, { "epoch": 9.336612789880535, "grad_norm": 0.13308727741241455, "learning_rate": 2.7111735769501056e-05, "loss": 0.0161, "step": 13286 }, { "epoch": 9.33731553056922, "grad_norm": 0.10967095196247101, "learning_rate": 2.7111267275708596e-05, "loss": 0.0173, "step": 13287 }, { "epoch": 9.338018271257905, "grad_norm": 0.10700114816427231, "learning_rate": 2.711079878191614e-05, "loss": 0.0105, "step": 13288 }, { "epoch": 9.33872101194659, "grad_norm": 0.22428864240646362, "learning_rate": 2.7110330288123684e-05, "loss": 0.0188, "step": 13289 }, { "epoch": 9.339423752635277, "grad_norm": 0.2600392997264862, "learning_rate": 2.7109861794331228e-05, "loss": 0.0332, "step": 13290 }, { "epoch": 9.340126493323963, "grad_norm": 0.7166242599487305, "learning_rate": 2.7109393300538768e-05, "loss": 0.0261, "step": 13291 }, { "epoch": 9.340829234012649, "grad_norm": 0.09280503541231155, "learning_rate": 2.710892480674631e-05, "loss": 0.0099, "step": 13292 }, { "epoch": 9.341531974701335, "grad_norm": 0.10532442480325699, "learning_rate": 2.7108456312953855e-05, "loss": 0.0229, "step": 13293 }, { "epoch": 9.34223471539002, "grad_norm": 0.22781682014465332, "learning_rate": 2.71079878191614e-05, "loss": 0.0222, "step": 13294 }, { "epoch": 9.342937456078706, "grad_norm": 0.15132030844688416, "learning_rate": 2.710751932536894e-05, "loss": 0.0444, "step": 13295 }, { "epoch": 9.343640196767392, "grad_norm": 0.13634169101715088, "learning_rate": 2.710705083157648e-05, "loss": 0.0211, "step": 13296 }, { "epoch": 9.344342937456078, "grad_norm": 0.3786498010158539, "learning_rate": 2.7106582337784023e-05, "loss": 0.0255, "step": 13297 }, { "epoch": 9.345045678144764, "grad_norm": 0.14651450514793396, "learning_rate": 2.7106113843991567e-05, "loss": 0.0195, "step": 13298 }, { "epoch": 9.34574841883345, "grad_norm": 0.16899502277374268, "learning_rate": 2.710564535019911e-05, "loss": 0.0221, "step": 13299 }, { "epoch": 9.346451159522136, "grad_norm": 0.3818177282810211, "learning_rate": 2.710517685640665e-05, "loss": 0.0379, "step": 13300 }, { "epoch": 9.347153900210822, "grad_norm": 0.15840613842010498, "learning_rate": 2.7104708362614195e-05, "loss": 0.0276, "step": 13301 }, { "epoch": 9.347856640899508, "grad_norm": 0.4753897786140442, "learning_rate": 2.710423986882174e-05, "loss": 0.0393, "step": 13302 }, { "epoch": 9.348559381588194, "grad_norm": 0.5158341526985168, "learning_rate": 2.7103771375029282e-05, "loss": 0.0908, "step": 13303 }, { "epoch": 9.34926212227688, "grad_norm": 0.5549956560134888, "learning_rate": 2.7103302881236826e-05, "loss": 0.1095, "step": 13304 }, { "epoch": 9.349964862965566, "grad_norm": 0.9438602328300476, "learning_rate": 2.7102834387444366e-05, "loss": 0.1461, "step": 13305 }, { "epoch": 9.350667603654252, "grad_norm": 0.8102531433105469, "learning_rate": 2.710236589365191e-05, "loss": 0.1892, "step": 13306 }, { "epoch": 9.351370344342937, "grad_norm": 0.9518353939056396, "learning_rate": 2.7101897399859454e-05, "loss": 0.2662, "step": 13307 }, { "epoch": 9.352073085031623, "grad_norm": 0.4714028537273407, "learning_rate": 2.7101428906066998e-05, "loss": 0.0937, "step": 13308 }, { "epoch": 9.35277582572031, "grad_norm": 0.15607155859470367, "learning_rate": 2.7100960412274538e-05, "loss": 0.027, "step": 13309 }, { "epoch": 9.353478566408995, "grad_norm": 0.16176484525203705, "learning_rate": 2.7100491918482082e-05, "loss": 0.0175, "step": 13310 }, { "epoch": 9.354181307097681, "grad_norm": 0.122939333319664, "learning_rate": 2.7100023424689625e-05, "loss": 0.0139, "step": 13311 }, { "epoch": 9.354884047786367, "grad_norm": 0.09837902337312698, "learning_rate": 2.709955493089717e-05, "loss": 0.0142, "step": 13312 }, { "epoch": 9.355586788475053, "grad_norm": 0.17523421347141266, "learning_rate": 2.7099086437104706e-05, "loss": 0.0145, "step": 13313 }, { "epoch": 9.356289529163739, "grad_norm": 0.13230571150779724, "learning_rate": 2.709861794331225e-05, "loss": 0.0166, "step": 13314 }, { "epoch": 9.356992269852425, "grad_norm": 0.19608251750469208, "learning_rate": 2.7098149449519794e-05, "loss": 0.0246, "step": 13315 }, { "epoch": 9.35769501054111, "grad_norm": 0.17602773010730743, "learning_rate": 2.7097680955727337e-05, "loss": 0.0251, "step": 13316 }, { "epoch": 9.358397751229797, "grad_norm": 0.14643433690071106, "learning_rate": 2.709721246193488e-05, "loss": 0.0084, "step": 13317 }, { "epoch": 9.359100491918483, "grad_norm": 0.13099315762519836, "learning_rate": 2.709674396814242e-05, "loss": 0.019, "step": 13318 }, { "epoch": 9.359803232607169, "grad_norm": 0.1601012647151947, "learning_rate": 2.7096275474349965e-05, "loss": 0.0161, "step": 13319 }, { "epoch": 9.360505973295854, "grad_norm": 0.3616722822189331, "learning_rate": 2.709580698055751e-05, "loss": 0.029, "step": 13320 }, { "epoch": 9.36120871398454, "grad_norm": 0.16576451063156128, "learning_rate": 2.7095338486765053e-05, "loss": 0.0223, "step": 13321 }, { "epoch": 9.361911454673226, "grad_norm": 0.14943461120128632, "learning_rate": 2.7094869992972593e-05, "loss": 0.0236, "step": 13322 }, { "epoch": 9.362614195361912, "grad_norm": 0.23438702523708344, "learning_rate": 2.7094401499180137e-05, "loss": 0.0402, "step": 13323 }, { "epoch": 9.363316936050598, "grad_norm": 0.15902592241764069, "learning_rate": 2.709393300538768e-05, "loss": 0.0154, "step": 13324 }, { "epoch": 9.364019676739284, "grad_norm": 0.2560403347015381, "learning_rate": 2.7093464511595224e-05, "loss": 0.0407, "step": 13325 }, { "epoch": 9.36472241742797, "grad_norm": 0.23332390189170837, "learning_rate": 2.7092996017802764e-05, "loss": 0.0331, "step": 13326 }, { "epoch": 9.365425158116654, "grad_norm": 0.33809396624565125, "learning_rate": 2.7092527524010308e-05, "loss": 0.0524, "step": 13327 }, { "epoch": 9.36612789880534, "grad_norm": 0.5468038320541382, "learning_rate": 2.7092059030217852e-05, "loss": 0.0864, "step": 13328 }, { "epoch": 9.366830639494026, "grad_norm": 0.4763562083244324, "learning_rate": 2.7091590536425396e-05, "loss": 0.0947, "step": 13329 }, { "epoch": 9.367533380182712, "grad_norm": 0.5899361968040466, "learning_rate": 2.7091122042632936e-05, "loss": 0.1351, "step": 13330 }, { "epoch": 9.368236120871398, "grad_norm": 1.1535265445709229, "learning_rate": 2.7090653548840476e-05, "loss": 0.1784, "step": 13331 }, { "epoch": 9.368938861560084, "grad_norm": 3.044285774230957, "learning_rate": 2.709018505504802e-05, "loss": 0.1941, "step": 13332 }, { "epoch": 9.36964160224877, "grad_norm": 0.3137022852897644, "learning_rate": 2.7089716561255564e-05, "loss": 0.0882, "step": 13333 }, { "epoch": 9.370344342937456, "grad_norm": 0.1258481740951538, "learning_rate": 2.7089248067463107e-05, "loss": 0.0214, "step": 13334 }, { "epoch": 9.371047083626141, "grad_norm": 0.12182947993278503, "learning_rate": 2.7088779573670648e-05, "loss": 0.0248, "step": 13335 }, { "epoch": 9.371749824314827, "grad_norm": 0.08964136242866516, "learning_rate": 2.708831107987819e-05, "loss": 0.0196, "step": 13336 }, { "epoch": 9.372452565003513, "grad_norm": 0.11942704021930695, "learning_rate": 2.7087842586085735e-05, "loss": 0.0163, "step": 13337 }, { "epoch": 9.3731553056922, "grad_norm": 0.12668731808662415, "learning_rate": 2.708737409229328e-05, "loss": 0.0121, "step": 13338 }, { "epoch": 9.373858046380885, "grad_norm": 0.15275385975837708, "learning_rate": 2.708690559850082e-05, "loss": 0.015, "step": 13339 }, { "epoch": 9.374560787069571, "grad_norm": 0.2394934743642807, "learning_rate": 2.7086437104708363e-05, "loss": 0.0277, "step": 13340 }, { "epoch": 9.375263527758257, "grad_norm": 0.13770367205142975, "learning_rate": 2.7085968610915907e-05, "loss": 0.0214, "step": 13341 }, { "epoch": 9.375966268446943, "grad_norm": 0.1798195093870163, "learning_rate": 2.708550011712345e-05, "loss": 0.0281, "step": 13342 }, { "epoch": 9.376669009135629, "grad_norm": 0.22907087206840515, "learning_rate": 2.7085031623330994e-05, "loss": 0.0356, "step": 13343 }, { "epoch": 9.377371749824315, "grad_norm": 0.1291390210390091, "learning_rate": 2.7084563129538534e-05, "loss": 0.0154, "step": 13344 }, { "epoch": 9.378074490513, "grad_norm": 0.14684294164180756, "learning_rate": 2.7084094635746078e-05, "loss": 0.0261, "step": 13345 }, { "epoch": 9.378777231201687, "grad_norm": 0.2441496104001999, "learning_rate": 2.7083626141953622e-05, "loss": 0.0256, "step": 13346 }, { "epoch": 9.379479971890373, "grad_norm": 0.25828495621681213, "learning_rate": 2.7083157648161162e-05, "loss": 0.0629, "step": 13347 }, { "epoch": 9.380182712579058, "grad_norm": 0.19542239606380463, "learning_rate": 2.7082689154368703e-05, "loss": 0.0375, "step": 13348 }, { "epoch": 9.380885453267744, "grad_norm": 0.2349252551794052, "learning_rate": 2.7082220660576246e-05, "loss": 0.0273, "step": 13349 }, { "epoch": 9.38158819395643, "grad_norm": 0.16287267208099365, "learning_rate": 2.708175216678379e-05, "loss": 0.023, "step": 13350 }, { "epoch": 9.382290934645116, "grad_norm": 0.4630149304866791, "learning_rate": 2.7081283672991334e-05, "loss": 0.0645, "step": 13351 }, { "epoch": 9.382993675333802, "grad_norm": 1.6094028949737549, "learning_rate": 2.7080815179198874e-05, "loss": 0.045, "step": 13352 }, { "epoch": 9.383696416022488, "grad_norm": 0.3848614990711212, "learning_rate": 2.7080346685406418e-05, "loss": 0.1022, "step": 13353 }, { "epoch": 9.384399156711174, "grad_norm": 0.40528181195259094, "learning_rate": 2.707987819161396e-05, "loss": 0.1243, "step": 13354 }, { "epoch": 9.38510189739986, "grad_norm": 1.1787440776824951, "learning_rate": 2.7079409697821505e-05, "loss": 0.1545, "step": 13355 }, { "epoch": 9.385804638088546, "grad_norm": 0.9275158643722534, "learning_rate": 2.707894120402905e-05, "loss": 0.2076, "step": 13356 }, { "epoch": 9.386507378777232, "grad_norm": 1.3269755840301514, "learning_rate": 2.707847271023659e-05, "loss": 0.2026, "step": 13357 }, { "epoch": 9.387210119465918, "grad_norm": 0.2527470588684082, "learning_rate": 2.7078004216444133e-05, "loss": 0.0613, "step": 13358 }, { "epoch": 9.387912860154604, "grad_norm": 0.10875622183084488, "learning_rate": 2.7077535722651677e-05, "loss": 0.0253, "step": 13359 }, { "epoch": 9.38861560084329, "grad_norm": 0.08536022156476974, "learning_rate": 2.707706722885922e-05, "loss": 0.0151, "step": 13360 }, { "epoch": 9.389318341531975, "grad_norm": 0.14134350419044495, "learning_rate": 2.707659873506676e-05, "loss": 0.014, "step": 13361 }, { "epoch": 9.390021082220661, "grad_norm": 0.11462143063545227, "learning_rate": 2.7076130241274305e-05, "loss": 0.028, "step": 13362 }, { "epoch": 9.390723822909347, "grad_norm": 0.2828492820262909, "learning_rate": 2.707566174748185e-05, "loss": 0.0088, "step": 13363 }, { "epoch": 9.391426563598033, "grad_norm": 0.07522402703762054, "learning_rate": 2.7075193253689392e-05, "loss": 0.012, "step": 13364 }, { "epoch": 9.392129304286719, "grad_norm": 0.131561279296875, "learning_rate": 2.707472475989693e-05, "loss": 0.0183, "step": 13365 }, { "epoch": 9.392832044975403, "grad_norm": 0.20065560936927795, "learning_rate": 2.7074256266104473e-05, "loss": 0.035, "step": 13366 }, { "epoch": 9.39353478566409, "grad_norm": 0.14440402388572693, "learning_rate": 2.7073787772312016e-05, "loss": 0.0254, "step": 13367 }, { "epoch": 9.394237526352775, "grad_norm": 0.24574097990989685, "learning_rate": 2.707331927851956e-05, "loss": 0.0349, "step": 13368 }, { "epoch": 9.394940267041461, "grad_norm": 0.11408702284097672, "learning_rate": 2.7072850784727104e-05, "loss": 0.0192, "step": 13369 }, { "epoch": 9.395643007730147, "grad_norm": 0.21457688510417938, "learning_rate": 2.7072382290934644e-05, "loss": 0.0185, "step": 13370 }, { "epoch": 9.396345748418833, "grad_norm": 0.26757749915122986, "learning_rate": 2.7071913797142188e-05, "loss": 0.0094, "step": 13371 }, { "epoch": 9.397048489107519, "grad_norm": 0.11220264434814453, "learning_rate": 2.707144530334973e-05, "loss": 0.0283, "step": 13372 }, { "epoch": 9.397751229796205, "grad_norm": 0.3443185091018677, "learning_rate": 2.7070976809557275e-05, "loss": 0.0345, "step": 13373 }, { "epoch": 9.39845397048489, "grad_norm": 0.12695416808128357, "learning_rate": 2.7070508315764816e-05, "loss": 0.0215, "step": 13374 }, { "epoch": 9.399156711173577, "grad_norm": 0.16733913123607635, "learning_rate": 2.707003982197236e-05, "loss": 0.0298, "step": 13375 }, { "epoch": 9.399859451862262, "grad_norm": 0.9174492955207825, "learning_rate": 2.7069571328179903e-05, "loss": 0.0651, "step": 13376 }, { "epoch": 9.400562192550948, "grad_norm": 0.25601714849472046, "learning_rate": 2.7069102834387447e-05, "loss": 0.0512, "step": 13377 }, { "epoch": 9.401264933239634, "grad_norm": 0.21177732944488525, "learning_rate": 2.7068634340594987e-05, "loss": 0.0623, "step": 13378 }, { "epoch": 9.40196767392832, "grad_norm": 0.6944045424461365, "learning_rate": 2.706816584680253e-05, "loss": 0.1049, "step": 13379 }, { "epoch": 9.402670414617006, "grad_norm": 0.47040730714797974, "learning_rate": 2.7067697353010075e-05, "loss": 0.1498, "step": 13380 }, { "epoch": 9.403373155305692, "grad_norm": 0.6466898322105408, "learning_rate": 2.706722885921762e-05, "loss": 0.1736, "step": 13381 }, { "epoch": 9.404075895994378, "grad_norm": 1.0280946493148804, "learning_rate": 2.706676036542516e-05, "loss": 0.2086, "step": 13382 }, { "epoch": 9.404778636683064, "grad_norm": 0.2744412422180176, "learning_rate": 2.70662918716327e-05, "loss": 0.0694, "step": 13383 }, { "epoch": 9.40548137737175, "grad_norm": 0.2683347463607788, "learning_rate": 2.7065823377840243e-05, "loss": 0.0243, "step": 13384 }, { "epoch": 9.406184118060436, "grad_norm": 0.41010984778404236, "learning_rate": 2.7065354884047787e-05, "loss": 0.0145, "step": 13385 }, { "epoch": 9.406886858749122, "grad_norm": 0.1016215831041336, "learning_rate": 2.706488639025533e-05, "loss": 0.0163, "step": 13386 }, { "epoch": 9.407589599437808, "grad_norm": 0.08377879112958908, "learning_rate": 2.706441789646287e-05, "loss": 0.0176, "step": 13387 }, { "epoch": 9.408292340126494, "grad_norm": 0.09975653141736984, "learning_rate": 2.7063949402670414e-05, "loss": 0.009, "step": 13388 }, { "epoch": 9.40899508081518, "grad_norm": 0.5952807068824768, "learning_rate": 2.7063480908877958e-05, "loss": 0.0201, "step": 13389 }, { "epoch": 9.409697821503865, "grad_norm": 0.10290145128965378, "learning_rate": 2.7063012415085502e-05, "loss": 0.0218, "step": 13390 }, { "epoch": 9.410400562192551, "grad_norm": 0.13523054122924805, "learning_rate": 2.7062543921293042e-05, "loss": 0.0212, "step": 13391 }, { "epoch": 9.411103302881237, "grad_norm": 0.15629303455352783, "learning_rate": 2.7062075427500586e-05, "loss": 0.0265, "step": 13392 }, { "epoch": 9.411806043569923, "grad_norm": 0.12118319422006607, "learning_rate": 2.706160693370813e-05, "loss": 0.0201, "step": 13393 }, { "epoch": 9.412508784258609, "grad_norm": 0.5429221391677856, "learning_rate": 2.7061138439915673e-05, "loss": 0.0191, "step": 13394 }, { "epoch": 9.413211524947295, "grad_norm": 0.14337779581546783, "learning_rate": 2.7060669946123217e-05, "loss": 0.0384, "step": 13395 }, { "epoch": 9.41391426563598, "grad_norm": 0.11463840305805206, "learning_rate": 2.7060201452330757e-05, "loss": 0.0222, "step": 13396 }, { "epoch": 9.414617006324667, "grad_norm": 0.2225663810968399, "learning_rate": 2.70597329585383e-05, "loss": 0.0362, "step": 13397 }, { "epoch": 9.415319747013353, "grad_norm": 0.24631883203983307, "learning_rate": 2.7059264464745845e-05, "loss": 0.0474, "step": 13398 }, { "epoch": 9.416022487702039, "grad_norm": 0.21581417322158813, "learning_rate": 2.705879597095339e-05, "loss": 0.0445, "step": 13399 }, { "epoch": 9.416725228390725, "grad_norm": 0.1602419912815094, "learning_rate": 2.7058327477160925e-05, "loss": 0.0232, "step": 13400 }, { "epoch": 9.41742796907941, "grad_norm": 0.3121863603591919, "learning_rate": 2.705785898336847e-05, "loss": 0.062, "step": 13401 }, { "epoch": 9.418130709768096, "grad_norm": 0.5590468049049377, "learning_rate": 2.7057390489576013e-05, "loss": 0.0486, "step": 13402 }, { "epoch": 9.41883345045678, "grad_norm": 0.3262491226196289, "learning_rate": 2.7056921995783557e-05, "loss": 0.0713, "step": 13403 }, { "epoch": 9.419536191145466, "grad_norm": 0.33343297243118286, "learning_rate": 2.7056453501991097e-05, "loss": 0.1107, "step": 13404 }, { "epoch": 9.420238931834152, "grad_norm": 0.5526849627494812, "learning_rate": 2.705598500819864e-05, "loss": 0.1506, "step": 13405 }, { "epoch": 9.420941672522838, "grad_norm": 0.5640955567359924, "learning_rate": 2.7055516514406184e-05, "loss": 0.181, "step": 13406 }, { "epoch": 9.421644413211524, "grad_norm": 0.8942168951034546, "learning_rate": 2.7055048020613728e-05, "loss": 0.2081, "step": 13407 }, { "epoch": 9.42234715390021, "grad_norm": 0.22654099762439728, "learning_rate": 2.7054579526821272e-05, "loss": 0.0654, "step": 13408 }, { "epoch": 9.423049894588896, "grad_norm": 0.10444122552871704, "learning_rate": 2.7054111033028812e-05, "loss": 0.0263, "step": 13409 }, { "epoch": 9.423752635277582, "grad_norm": 0.13296334445476532, "learning_rate": 2.7053642539236356e-05, "loss": 0.0228, "step": 13410 }, { "epoch": 9.424455375966268, "grad_norm": 0.19870536029338837, "learning_rate": 2.70531740454439e-05, "loss": 0.0189, "step": 13411 }, { "epoch": 9.425158116654954, "grad_norm": 0.08979009836912155, "learning_rate": 2.7052705551651443e-05, "loss": 0.0081, "step": 13412 }, { "epoch": 9.42586085734364, "grad_norm": 0.22777687013149261, "learning_rate": 2.7052237057858984e-05, "loss": 0.0103, "step": 13413 }, { "epoch": 9.426563598032326, "grad_norm": 0.12172619253396988, "learning_rate": 2.7051768564066527e-05, "loss": 0.0117, "step": 13414 }, { "epoch": 9.427266338721012, "grad_norm": 0.20474134385585785, "learning_rate": 2.705130007027407e-05, "loss": 0.0254, "step": 13415 }, { "epoch": 9.427969079409698, "grad_norm": 0.1972273588180542, "learning_rate": 2.7050831576481615e-05, "loss": 0.0316, "step": 13416 }, { "epoch": 9.428671820098383, "grad_norm": 0.14884279668331146, "learning_rate": 2.7050363082689152e-05, "loss": 0.0113, "step": 13417 }, { "epoch": 9.42937456078707, "grad_norm": 0.13522309064865112, "learning_rate": 2.7049894588896696e-05, "loss": 0.021, "step": 13418 }, { "epoch": 9.430077301475755, "grad_norm": 0.08102010935544968, "learning_rate": 2.704942609510424e-05, "loss": 0.0133, "step": 13419 }, { "epoch": 9.430780042164441, "grad_norm": 0.23816512525081635, "learning_rate": 2.7048957601311783e-05, "loss": 0.0275, "step": 13420 }, { "epoch": 9.431482782853127, "grad_norm": 0.19976744055747986, "learning_rate": 2.7048489107519327e-05, "loss": 0.026, "step": 13421 }, { "epoch": 9.432185523541813, "grad_norm": 0.11225400865077972, "learning_rate": 2.7048020613726867e-05, "loss": 0.0192, "step": 13422 }, { "epoch": 9.432888264230499, "grad_norm": 0.18046130239963531, "learning_rate": 2.704755211993441e-05, "loss": 0.0262, "step": 13423 }, { "epoch": 9.433591004919185, "grad_norm": 0.12070966511964798, "learning_rate": 2.7047083626141955e-05, "loss": 0.0286, "step": 13424 }, { "epoch": 9.43429374560787, "grad_norm": 0.17591845989227295, "learning_rate": 2.7046615132349498e-05, "loss": 0.041, "step": 13425 }, { "epoch": 9.434996486296557, "grad_norm": 0.23177003860473633, "learning_rate": 2.704614663855704e-05, "loss": 0.0444, "step": 13426 }, { "epoch": 9.435699226985243, "grad_norm": 0.21474871039390564, "learning_rate": 2.7045678144764582e-05, "loss": 0.0619, "step": 13427 }, { "epoch": 9.436401967673929, "grad_norm": 0.2795363664627075, "learning_rate": 2.7045209650972126e-05, "loss": 0.066, "step": 13428 }, { "epoch": 9.437104708362615, "grad_norm": 0.41990843415260315, "learning_rate": 2.704474115717967e-05, "loss": 0.1116, "step": 13429 }, { "epoch": 9.4378074490513, "grad_norm": 0.5254747271537781, "learning_rate": 2.704427266338721e-05, "loss": 0.1413, "step": 13430 }, { "epoch": 9.438510189739986, "grad_norm": 1.2695039510726929, "learning_rate": 2.7043804169594754e-05, "loss": 0.1528, "step": 13431 }, { "epoch": 9.439212930428672, "grad_norm": 1.5733953714370728, "learning_rate": 2.7043335675802298e-05, "loss": 0.218, "step": 13432 }, { "epoch": 9.439915671117358, "grad_norm": 0.29932788014411926, "learning_rate": 2.704286718200984e-05, "loss": 0.0869, "step": 13433 }, { "epoch": 9.440618411806044, "grad_norm": 0.10664419829845428, "learning_rate": 2.704239868821738e-05, "loss": 0.0268, "step": 13434 }, { "epoch": 9.44132115249473, "grad_norm": 0.18047425150871277, "learning_rate": 2.7041930194424922e-05, "loss": 0.026, "step": 13435 }, { "epoch": 9.442023893183416, "grad_norm": 0.16163986921310425, "learning_rate": 2.7041461700632466e-05, "loss": 0.0153, "step": 13436 }, { "epoch": 9.442726633872102, "grad_norm": 0.13702863454818726, "learning_rate": 2.704099320684001e-05, "loss": 0.0204, "step": 13437 }, { "epoch": 9.443429374560788, "grad_norm": 0.10175186395645142, "learning_rate": 2.7040524713047553e-05, "loss": 0.0168, "step": 13438 }, { "epoch": 9.444132115249474, "grad_norm": 0.20390811562538147, "learning_rate": 2.7040056219255093e-05, "loss": 0.0302, "step": 13439 }, { "epoch": 9.44483485593816, "grad_norm": 0.13965526223182678, "learning_rate": 2.7039587725462637e-05, "loss": 0.025, "step": 13440 }, { "epoch": 9.445537596626846, "grad_norm": 0.18503738939762115, "learning_rate": 2.703911923167018e-05, "loss": 0.0388, "step": 13441 }, { "epoch": 9.44624033731553, "grad_norm": 0.11404722929000854, "learning_rate": 2.7038650737877725e-05, "loss": 0.0112, "step": 13442 }, { "epoch": 9.446943078004216, "grad_norm": 0.16363386809825897, "learning_rate": 2.7038182244085265e-05, "loss": 0.0248, "step": 13443 }, { "epoch": 9.447645818692902, "grad_norm": 0.09826099127531052, "learning_rate": 2.703771375029281e-05, "loss": 0.0184, "step": 13444 }, { "epoch": 9.448348559381587, "grad_norm": 0.19516970217227936, "learning_rate": 2.7037245256500352e-05, "loss": 0.0324, "step": 13445 }, { "epoch": 9.449051300070273, "grad_norm": 0.22947773337364197, "learning_rate": 2.7036776762707896e-05, "loss": 0.0177, "step": 13446 }, { "epoch": 9.44975404075896, "grad_norm": 0.4166252315044403, "learning_rate": 2.703630826891544e-05, "loss": 0.0281, "step": 13447 }, { "epoch": 9.450456781447645, "grad_norm": 0.2450287640094757, "learning_rate": 2.703583977512298e-05, "loss": 0.0208, "step": 13448 }, { "epoch": 9.451159522136331, "grad_norm": 0.2527465224266052, "learning_rate": 2.7035371281330524e-05, "loss": 0.0258, "step": 13449 }, { "epoch": 9.451862262825017, "grad_norm": 0.5485187768936157, "learning_rate": 2.7034902787538068e-05, "loss": 0.036, "step": 13450 }, { "epoch": 9.452565003513703, "grad_norm": 0.33181944489479065, "learning_rate": 2.703443429374561e-05, "loss": 0.0632, "step": 13451 }, { "epoch": 9.453267744202389, "grad_norm": 0.28072434663772583, "learning_rate": 2.703396579995315e-05, "loss": 0.0548, "step": 13452 }, { "epoch": 9.453970484891075, "grad_norm": 0.23615890741348267, "learning_rate": 2.7033497306160692e-05, "loss": 0.0692, "step": 13453 }, { "epoch": 9.45467322557976, "grad_norm": 1.1959246397018433, "learning_rate": 2.7033028812368236e-05, "loss": 0.0891, "step": 13454 }, { "epoch": 9.455375966268447, "grad_norm": 0.7068175673484802, "learning_rate": 2.703256031857578e-05, "loss": 0.2091, "step": 13455 }, { "epoch": 9.456078706957133, "grad_norm": 0.7321760654449463, "learning_rate": 2.703209182478332e-05, "loss": 0.1722, "step": 13456 }, { "epoch": 9.456781447645819, "grad_norm": 0.8048975467681885, "learning_rate": 2.7031623330990864e-05, "loss": 0.2434, "step": 13457 }, { "epoch": 9.457484188334504, "grad_norm": 0.3004826009273529, "learning_rate": 2.7031154837198407e-05, "loss": 0.0896, "step": 13458 }, { "epoch": 9.45818692902319, "grad_norm": 0.5519185066223145, "learning_rate": 2.703068634340595e-05, "loss": 0.0197, "step": 13459 }, { "epoch": 9.458889669711876, "grad_norm": 0.16647350788116455, "learning_rate": 2.7030217849613495e-05, "loss": 0.0235, "step": 13460 }, { "epoch": 9.459592410400562, "grad_norm": 0.14349804818630219, "learning_rate": 2.7029749355821035e-05, "loss": 0.021, "step": 13461 }, { "epoch": 9.460295151089248, "grad_norm": 0.16011156141757965, "learning_rate": 2.702928086202858e-05, "loss": 0.0277, "step": 13462 }, { "epoch": 9.460997891777934, "grad_norm": 0.1128024011850357, "learning_rate": 2.7028812368236123e-05, "loss": 0.0167, "step": 13463 }, { "epoch": 9.46170063246662, "grad_norm": 0.21117930114269257, "learning_rate": 2.7028343874443666e-05, "loss": 0.0191, "step": 13464 }, { "epoch": 9.462403373155306, "grad_norm": 0.0866582915186882, "learning_rate": 2.7027875380651207e-05, "loss": 0.017, "step": 13465 }, { "epoch": 9.463106113843992, "grad_norm": 0.10688292980194092, "learning_rate": 2.702740688685875e-05, "loss": 0.0223, "step": 13466 }, { "epoch": 9.463808854532678, "grad_norm": 0.13603495061397552, "learning_rate": 2.7026938393066294e-05, "loss": 0.0162, "step": 13467 }, { "epoch": 9.464511595221364, "grad_norm": 0.2087731957435608, "learning_rate": 2.7026469899273838e-05, "loss": 0.0205, "step": 13468 }, { "epoch": 9.46521433591005, "grad_norm": 0.10883773118257523, "learning_rate": 2.7026001405481375e-05, "loss": 0.0107, "step": 13469 }, { "epoch": 9.465917076598735, "grad_norm": 0.16061347723007202, "learning_rate": 2.702553291168892e-05, "loss": 0.0247, "step": 13470 }, { "epoch": 9.466619817287421, "grad_norm": 0.1351398080587387, "learning_rate": 2.7025064417896462e-05, "loss": 0.0117, "step": 13471 }, { "epoch": 9.467322557976107, "grad_norm": 0.25054433941841125, "learning_rate": 2.7024595924104006e-05, "loss": 0.027, "step": 13472 }, { "epoch": 9.468025298664793, "grad_norm": 0.28266140818595886, "learning_rate": 2.702412743031155e-05, "loss": 0.0406, "step": 13473 }, { "epoch": 9.46872803935348, "grad_norm": 0.11714685708284378, "learning_rate": 2.702365893651909e-05, "loss": 0.012, "step": 13474 }, { "epoch": 9.469430780042165, "grad_norm": 0.18088576197624207, "learning_rate": 2.7023190442726634e-05, "loss": 0.0372, "step": 13475 }, { "epoch": 9.470133520730851, "grad_norm": 0.5362526178359985, "learning_rate": 2.7022721948934177e-05, "loss": 0.0394, "step": 13476 }, { "epoch": 9.470836261419537, "grad_norm": 0.32041555643081665, "learning_rate": 2.702225345514172e-05, "loss": 0.0688, "step": 13477 }, { "epoch": 9.471539002108223, "grad_norm": 0.47572338581085205, "learning_rate": 2.702178496134926e-05, "loss": 0.0719, "step": 13478 }, { "epoch": 9.472241742796909, "grad_norm": 0.438571959733963, "learning_rate": 2.7021316467556805e-05, "loss": 0.1245, "step": 13479 }, { "epoch": 9.472944483485595, "grad_norm": 0.5091730952262878, "learning_rate": 2.702084797376435e-05, "loss": 0.1335, "step": 13480 }, { "epoch": 9.473647224174279, "grad_norm": 0.6090230941772461, "learning_rate": 2.7020379479971893e-05, "loss": 0.2041, "step": 13481 }, { "epoch": 9.474349964862965, "grad_norm": 1.4807947874069214, "learning_rate": 2.7019910986179433e-05, "loss": 0.2024, "step": 13482 }, { "epoch": 9.47505270555165, "grad_norm": 0.1912887990474701, "learning_rate": 2.7019442492386977e-05, "loss": 0.0716, "step": 13483 }, { "epoch": 9.475755446240337, "grad_norm": 0.10287711024284363, "learning_rate": 2.701897399859452e-05, "loss": 0.0231, "step": 13484 }, { "epoch": 9.476458186929023, "grad_norm": 0.17165516316890717, "learning_rate": 2.7018505504802064e-05, "loss": 0.0172, "step": 13485 }, { "epoch": 9.477160927617708, "grad_norm": 0.11480113863945007, "learning_rate": 2.7018037011009608e-05, "loss": 0.0255, "step": 13486 }, { "epoch": 9.477863668306394, "grad_norm": 0.09137042611837387, "learning_rate": 2.7017568517217145e-05, "loss": 0.0171, "step": 13487 }, { "epoch": 9.47856640899508, "grad_norm": 0.0891290232539177, "learning_rate": 2.701710002342469e-05, "loss": 0.0145, "step": 13488 }, { "epoch": 9.479269149683766, "grad_norm": 0.09270914644002914, "learning_rate": 2.7016631529632232e-05, "loss": 0.0093, "step": 13489 }, { "epoch": 9.479971890372452, "grad_norm": 0.1431179940700531, "learning_rate": 2.7016163035839776e-05, "loss": 0.0269, "step": 13490 }, { "epoch": 9.480674631061138, "grad_norm": 0.12308775633573532, "learning_rate": 2.7015694542047316e-05, "loss": 0.0258, "step": 13491 }, { "epoch": 9.481377371749824, "grad_norm": 0.08428964018821716, "learning_rate": 2.701522604825486e-05, "loss": 0.0118, "step": 13492 }, { "epoch": 9.48208011243851, "grad_norm": 0.1551184058189392, "learning_rate": 2.7014757554462404e-05, "loss": 0.0345, "step": 13493 }, { "epoch": 9.482782853127196, "grad_norm": 0.16185486316680908, "learning_rate": 2.7014289060669948e-05, "loss": 0.0179, "step": 13494 }, { "epoch": 9.483485593815882, "grad_norm": 0.35772934556007385, "learning_rate": 2.7013820566877488e-05, "loss": 0.0291, "step": 13495 }, { "epoch": 9.484188334504568, "grad_norm": 0.1527344137430191, "learning_rate": 2.701335207308503e-05, "loss": 0.014, "step": 13496 }, { "epoch": 9.484891075193254, "grad_norm": 0.23435388505458832, "learning_rate": 2.7012883579292575e-05, "loss": 0.041, "step": 13497 }, { "epoch": 9.48559381588194, "grad_norm": 0.14410945773124695, "learning_rate": 2.701241508550012e-05, "loss": 0.0279, "step": 13498 }, { "epoch": 9.486296556570625, "grad_norm": 0.22248515486717224, "learning_rate": 2.7011946591707663e-05, "loss": 0.0222, "step": 13499 }, { "epoch": 9.486999297259311, "grad_norm": 0.13628222048282623, "learning_rate": 2.7011478097915203e-05, "loss": 0.0247, "step": 13500 }, { "epoch": 9.487702037947997, "grad_norm": 0.2736893594264984, "learning_rate": 2.7011009604122747e-05, "loss": 0.0436, "step": 13501 }, { "epoch": 9.488404778636683, "grad_norm": 0.4326474368572235, "learning_rate": 2.701054111033029e-05, "loss": 0.0505, "step": 13502 }, { "epoch": 9.489107519325369, "grad_norm": 0.3583493232727051, "learning_rate": 2.7010072616537834e-05, "loss": 0.0761, "step": 13503 }, { "epoch": 9.489810260014055, "grad_norm": 0.555221438407898, "learning_rate": 2.700960412274537e-05, "loss": 0.1075, "step": 13504 }, { "epoch": 9.490513000702741, "grad_norm": 0.8677630424499512, "learning_rate": 2.7009135628952915e-05, "loss": 0.1638, "step": 13505 }, { "epoch": 9.491215741391427, "grad_norm": 0.641148030757904, "learning_rate": 2.700866713516046e-05, "loss": 0.2219, "step": 13506 }, { "epoch": 9.491918482080113, "grad_norm": 1.9534411430358887, "learning_rate": 2.7008198641368002e-05, "loss": 0.2144, "step": 13507 }, { "epoch": 9.492621222768799, "grad_norm": 0.2699021100997925, "learning_rate": 2.7007730147575546e-05, "loss": 0.0698, "step": 13508 }, { "epoch": 9.493323963457485, "grad_norm": 0.15131765604019165, "learning_rate": 2.7007261653783086e-05, "loss": 0.0211, "step": 13509 }, { "epoch": 9.49402670414617, "grad_norm": 1.2443560361862183, "learning_rate": 2.700679315999063e-05, "loss": 0.0153, "step": 13510 }, { "epoch": 9.494729444834856, "grad_norm": 0.1806585043668747, "learning_rate": 2.7006324666198174e-05, "loss": 0.016, "step": 13511 }, { "epoch": 9.495432185523542, "grad_norm": 0.11451322585344315, "learning_rate": 2.7005856172405718e-05, "loss": 0.0132, "step": 13512 }, { "epoch": 9.496134926212228, "grad_norm": 0.11531934142112732, "learning_rate": 2.7005387678613258e-05, "loss": 0.0188, "step": 13513 }, { "epoch": 9.496837666900914, "grad_norm": 0.12042248994112015, "learning_rate": 2.7004919184820802e-05, "loss": 0.022, "step": 13514 }, { "epoch": 9.4975404075896, "grad_norm": 0.12956802546977997, "learning_rate": 2.7004450691028345e-05, "loss": 0.0212, "step": 13515 }, { "epoch": 9.498243148278286, "grad_norm": 0.19870510697364807, "learning_rate": 2.700398219723589e-05, "loss": 0.0259, "step": 13516 }, { "epoch": 9.498945888966972, "grad_norm": 0.12097281962633133, "learning_rate": 2.700351370344343e-05, "loss": 0.0149, "step": 13517 }, { "epoch": 9.499648629655656, "grad_norm": 0.17536213994026184, "learning_rate": 2.7003045209650973e-05, "loss": 0.0249, "step": 13518 }, { "epoch": 9.500351370344344, "grad_norm": 0.29005321860313416, "learning_rate": 2.7002576715858517e-05, "loss": 0.0194, "step": 13519 }, { "epoch": 9.501054111033028, "grad_norm": 0.21564482152462006, "learning_rate": 2.700210822206606e-05, "loss": 0.033, "step": 13520 }, { "epoch": 9.501756851721714, "grad_norm": 0.07200592011213303, "learning_rate": 2.70016397282736e-05, "loss": 0.011, "step": 13521 }, { "epoch": 9.5024595924104, "grad_norm": 0.30499473214149475, "learning_rate": 2.700117123448114e-05, "loss": 0.0527, "step": 13522 }, { "epoch": 9.503162333099086, "grad_norm": 0.2100924253463745, "learning_rate": 2.7000702740688685e-05, "loss": 0.0281, "step": 13523 }, { "epoch": 9.503865073787772, "grad_norm": 0.24659541249275208, "learning_rate": 2.700023424689623e-05, "loss": 0.0205, "step": 13524 }, { "epoch": 9.504567814476458, "grad_norm": 1.0090641975402832, "learning_rate": 2.6999765753103773e-05, "loss": 0.0382, "step": 13525 }, { "epoch": 9.505270555165144, "grad_norm": 0.184398353099823, "learning_rate": 2.6999297259311313e-05, "loss": 0.0375, "step": 13526 }, { "epoch": 9.50597329585383, "grad_norm": 0.37844881415367126, "learning_rate": 2.6998828765518857e-05, "loss": 0.0573, "step": 13527 }, { "epoch": 9.506676036542515, "grad_norm": 0.2275547981262207, "learning_rate": 2.69983602717264e-05, "loss": 0.0598, "step": 13528 }, { "epoch": 9.507378777231201, "grad_norm": 0.40277689695358276, "learning_rate": 2.6997891777933944e-05, "loss": 0.106, "step": 13529 }, { "epoch": 9.508081517919887, "grad_norm": 0.6277948021888733, "learning_rate": 2.6997423284141484e-05, "loss": 0.1404, "step": 13530 }, { "epoch": 9.508784258608573, "grad_norm": 1.31759512424469, "learning_rate": 2.6996954790349028e-05, "loss": 0.1752, "step": 13531 }, { "epoch": 9.509486999297259, "grad_norm": 1.3348089456558228, "learning_rate": 2.6996486296556572e-05, "loss": 0.2119, "step": 13532 }, { "epoch": 9.510189739985945, "grad_norm": 0.1700696349143982, "learning_rate": 2.6996017802764116e-05, "loss": 0.052, "step": 13533 }, { "epoch": 9.510892480674631, "grad_norm": 0.1519991159439087, "learning_rate": 2.699554930897166e-05, "loss": 0.029, "step": 13534 }, { "epoch": 9.511595221363317, "grad_norm": 0.24380037188529968, "learning_rate": 2.69950808151792e-05, "loss": 0.024, "step": 13535 }, { "epoch": 9.512297962052003, "grad_norm": 0.11335338652133942, "learning_rate": 2.6994612321386743e-05, "loss": 0.0181, "step": 13536 }, { "epoch": 9.513000702740689, "grad_norm": 0.26245537400245667, "learning_rate": 2.6994143827594287e-05, "loss": 0.0284, "step": 13537 }, { "epoch": 9.513703443429375, "grad_norm": 0.08240391314029694, "learning_rate": 2.699367533380183e-05, "loss": 0.0058, "step": 13538 }, { "epoch": 9.51440618411806, "grad_norm": 0.28649407625198364, "learning_rate": 2.6993206840009368e-05, "loss": 0.0163, "step": 13539 }, { "epoch": 9.515108924806746, "grad_norm": 0.2598539888858795, "learning_rate": 2.699273834621691e-05, "loss": 0.0212, "step": 13540 }, { "epoch": 9.515811665495432, "grad_norm": 0.14366796612739563, "learning_rate": 2.6992269852424455e-05, "loss": 0.012, "step": 13541 }, { "epoch": 9.516514406184118, "grad_norm": 0.5283547043800354, "learning_rate": 2.6991801358632e-05, "loss": 0.0233, "step": 13542 }, { "epoch": 9.517217146872804, "grad_norm": 0.27565959095954895, "learning_rate": 2.699133286483954e-05, "loss": 0.028, "step": 13543 }, { "epoch": 9.51791988756149, "grad_norm": 0.15411537885665894, "learning_rate": 2.6990864371047083e-05, "loss": 0.0229, "step": 13544 }, { "epoch": 9.518622628250176, "grad_norm": 0.18153175711631775, "learning_rate": 2.6990395877254627e-05, "loss": 0.0347, "step": 13545 }, { "epoch": 9.519325368938862, "grad_norm": 0.3097589910030365, "learning_rate": 2.698992738346217e-05, "loss": 0.0203, "step": 13546 }, { "epoch": 9.520028109627548, "grad_norm": 0.2343832552433014, "learning_rate": 2.6989458889669714e-05, "loss": 0.0386, "step": 13547 }, { "epoch": 9.520730850316234, "grad_norm": 0.28149527311325073, "learning_rate": 2.6988990395877255e-05, "loss": 0.0294, "step": 13548 }, { "epoch": 9.52143359100492, "grad_norm": 0.11686982214450836, "learning_rate": 2.6988521902084798e-05, "loss": 0.0258, "step": 13549 }, { "epoch": 9.522136331693606, "grad_norm": 0.5101634860038757, "learning_rate": 2.6988053408292342e-05, "loss": 0.04, "step": 13550 }, { "epoch": 9.522839072382292, "grad_norm": 0.2866484522819519, "learning_rate": 2.6987584914499886e-05, "loss": 0.0436, "step": 13551 }, { "epoch": 9.523541813070977, "grad_norm": 0.21191783249378204, "learning_rate": 2.6987116420707426e-05, "loss": 0.0406, "step": 13552 }, { "epoch": 9.524244553759663, "grad_norm": 0.2538187801837921, "learning_rate": 2.698664792691497e-05, "loss": 0.0829, "step": 13553 }, { "epoch": 9.52494729444835, "grad_norm": 0.9406041502952576, "learning_rate": 2.6986179433122513e-05, "loss": 0.164, "step": 13554 }, { "epoch": 9.525650035137035, "grad_norm": 1.4372663497924805, "learning_rate": 2.6985710939330057e-05, "loss": 0.1488, "step": 13555 }, { "epoch": 9.526352775825721, "grad_norm": 0.6043457984924316, "learning_rate": 2.6985242445537594e-05, "loss": 0.1854, "step": 13556 }, { "epoch": 9.527055516514405, "grad_norm": 1.015512228012085, "learning_rate": 2.6984773951745138e-05, "loss": 0.2186, "step": 13557 }, { "epoch": 9.527758257203093, "grad_norm": 0.3356172740459442, "learning_rate": 2.698430545795268e-05, "loss": 0.062, "step": 13558 }, { "epoch": 9.528460997891777, "grad_norm": 0.24624906480312347, "learning_rate": 2.6983836964160225e-05, "loss": 0.0239, "step": 13559 }, { "epoch": 9.529163738580463, "grad_norm": 0.1707640290260315, "learning_rate": 2.698336847036777e-05, "loss": 0.0235, "step": 13560 }, { "epoch": 9.529866479269149, "grad_norm": 0.2382035106420517, "learning_rate": 2.698289997657531e-05, "loss": 0.013, "step": 13561 }, { "epoch": 9.530569219957835, "grad_norm": 0.08168525993824005, "learning_rate": 2.6982431482782853e-05, "loss": 0.0136, "step": 13562 }, { "epoch": 9.53127196064652, "grad_norm": 0.10515305399894714, "learning_rate": 2.6981962988990397e-05, "loss": 0.0123, "step": 13563 }, { "epoch": 9.531974701335207, "grad_norm": 0.2508939504623413, "learning_rate": 2.698149449519794e-05, "loss": 0.0209, "step": 13564 }, { "epoch": 9.532677442023893, "grad_norm": 0.1136549785733223, "learning_rate": 2.698102600140548e-05, "loss": 0.0265, "step": 13565 }, { "epoch": 9.533380182712579, "grad_norm": 0.5373113751411438, "learning_rate": 2.6980557507613025e-05, "loss": 0.0139, "step": 13566 }, { "epoch": 9.534082923401265, "grad_norm": 0.14536887407302856, "learning_rate": 2.698008901382057e-05, "loss": 0.0128, "step": 13567 }, { "epoch": 9.53478566408995, "grad_norm": 0.16237206757068634, "learning_rate": 2.6979620520028112e-05, "loss": 0.023, "step": 13568 }, { "epoch": 9.535488404778636, "grad_norm": 0.14144350588321686, "learning_rate": 2.6979152026235652e-05, "loss": 0.0156, "step": 13569 }, { "epoch": 9.536191145467322, "grad_norm": 0.3926927447319031, "learning_rate": 2.6978683532443196e-05, "loss": 0.0282, "step": 13570 }, { "epoch": 9.536893886156008, "grad_norm": 0.1345929205417633, "learning_rate": 2.697821503865074e-05, "loss": 0.0108, "step": 13571 }, { "epoch": 9.537596626844694, "grad_norm": 0.4671023488044739, "learning_rate": 2.6977746544858284e-05, "loss": 0.0374, "step": 13572 }, { "epoch": 9.53829936753338, "grad_norm": 0.16980168223381042, "learning_rate": 2.6977278051065827e-05, "loss": 0.0284, "step": 13573 }, { "epoch": 9.539002108222066, "grad_norm": 0.3484533429145813, "learning_rate": 2.6976809557273364e-05, "loss": 0.0249, "step": 13574 }, { "epoch": 9.539704848910752, "grad_norm": 0.9367671012878418, "learning_rate": 2.6976341063480908e-05, "loss": 0.0351, "step": 13575 }, { "epoch": 9.540407589599438, "grad_norm": 0.5000987648963928, "learning_rate": 2.6975872569688452e-05, "loss": 0.0483, "step": 13576 }, { "epoch": 9.541110330288124, "grad_norm": 0.5422404408454895, "learning_rate": 2.6975404075895995e-05, "loss": 0.0441, "step": 13577 }, { "epoch": 9.54181307097681, "grad_norm": 0.3491804897785187, "learning_rate": 2.6974935582103536e-05, "loss": 0.0754, "step": 13578 }, { "epoch": 9.542515811665496, "grad_norm": 0.3262627124786377, "learning_rate": 2.697446708831108e-05, "loss": 0.0957, "step": 13579 }, { "epoch": 9.543218552354181, "grad_norm": 0.6137742400169373, "learning_rate": 2.6973998594518623e-05, "loss": 0.135, "step": 13580 }, { "epoch": 9.543921293042867, "grad_norm": 0.7580654621124268, "learning_rate": 2.6973530100726167e-05, "loss": 0.1914, "step": 13581 }, { "epoch": 9.544624033731553, "grad_norm": 1.6176341772079468, "learning_rate": 2.6973061606933707e-05, "loss": 0.2375, "step": 13582 }, { "epoch": 9.54532677442024, "grad_norm": 0.3003992736339569, "learning_rate": 2.697259311314125e-05, "loss": 0.0641, "step": 13583 }, { "epoch": 9.546029515108925, "grad_norm": 0.14366158843040466, "learning_rate": 2.6972124619348795e-05, "loss": 0.0388, "step": 13584 }, { "epoch": 9.546732255797611, "grad_norm": 0.11355667561292648, "learning_rate": 2.697165612555634e-05, "loss": 0.0166, "step": 13585 }, { "epoch": 9.547434996486297, "grad_norm": 0.1749121993780136, "learning_rate": 2.6971187631763882e-05, "loss": 0.0402, "step": 13586 }, { "epoch": 9.548137737174983, "grad_norm": 0.09419983625411987, "learning_rate": 2.6970719137971423e-05, "loss": 0.0131, "step": 13587 }, { "epoch": 9.548840477863669, "grad_norm": 0.08861439675092697, "learning_rate": 2.6970250644178966e-05, "loss": 0.0106, "step": 13588 }, { "epoch": 9.549543218552355, "grad_norm": 0.1957615315914154, "learning_rate": 2.696978215038651e-05, "loss": 0.0216, "step": 13589 }, { "epoch": 9.55024595924104, "grad_norm": 0.14310982823371887, "learning_rate": 2.6969313656594054e-05, "loss": 0.028, "step": 13590 }, { "epoch": 9.550948699929727, "grad_norm": 0.17669200897216797, "learning_rate": 2.696884516280159e-05, "loss": 0.022, "step": 13591 }, { "epoch": 9.551651440618413, "grad_norm": 0.06289119273424149, "learning_rate": 2.6968376669009134e-05, "loss": 0.0114, "step": 13592 }, { "epoch": 9.552354181307098, "grad_norm": 0.14072373509407043, "learning_rate": 2.6967908175216678e-05, "loss": 0.0192, "step": 13593 }, { "epoch": 9.553056921995784, "grad_norm": 0.148252472281456, "learning_rate": 2.6967439681424222e-05, "loss": 0.0107, "step": 13594 }, { "epoch": 9.55375966268447, "grad_norm": 0.16791868209838867, "learning_rate": 2.6966971187631762e-05, "loss": 0.0344, "step": 13595 }, { "epoch": 9.554462403373154, "grad_norm": 0.2460082620382309, "learning_rate": 2.6966502693839306e-05, "loss": 0.0128, "step": 13596 }, { "epoch": 9.55516514406184, "grad_norm": 0.2237064391374588, "learning_rate": 2.696603420004685e-05, "loss": 0.0308, "step": 13597 }, { "epoch": 9.555867884750526, "grad_norm": 0.1963912695646286, "learning_rate": 2.6965565706254393e-05, "loss": 0.0438, "step": 13598 }, { "epoch": 9.556570625439212, "grad_norm": 0.3048458397388458, "learning_rate": 2.6965097212461937e-05, "loss": 0.0183, "step": 13599 }, { "epoch": 9.557273366127898, "grad_norm": 0.1831764131784439, "learning_rate": 2.6964628718669477e-05, "loss": 0.0426, "step": 13600 }, { "epoch": 9.557976106816584, "grad_norm": 0.24799342453479767, "learning_rate": 2.696416022487702e-05, "loss": 0.0444, "step": 13601 }, { "epoch": 9.55867884750527, "grad_norm": 0.4653533697128296, "learning_rate": 2.6963691731084565e-05, "loss": 0.0632, "step": 13602 }, { "epoch": 9.559381588193956, "grad_norm": 0.26301470398902893, "learning_rate": 2.696322323729211e-05, "loss": 0.0641, "step": 13603 }, { "epoch": 9.560084328882642, "grad_norm": 0.9166905879974365, "learning_rate": 2.696275474349965e-05, "loss": 0.1382, "step": 13604 }, { "epoch": 9.560787069571328, "grad_norm": 0.618887186050415, "learning_rate": 2.6962286249707193e-05, "loss": 0.1514, "step": 13605 }, { "epoch": 9.561489810260014, "grad_norm": 0.5591427683830261, "learning_rate": 2.6961817755914736e-05, "loss": 0.1387, "step": 13606 }, { "epoch": 9.5621925509487, "grad_norm": 2.4697353839874268, "learning_rate": 2.696134926212228e-05, "loss": 0.2727, "step": 13607 }, { "epoch": 9.562895291637385, "grad_norm": 0.2187965214252472, "learning_rate": 2.696088076832982e-05, "loss": 0.0775, "step": 13608 }, { "epoch": 9.563598032326071, "grad_norm": 0.13020405173301697, "learning_rate": 2.696041227453736e-05, "loss": 0.0341, "step": 13609 }, { "epoch": 9.564300773014757, "grad_norm": 0.12212114036083221, "learning_rate": 2.6959943780744904e-05, "loss": 0.019, "step": 13610 }, { "epoch": 9.565003513703443, "grad_norm": 0.12944567203521729, "learning_rate": 2.6959475286952448e-05, "loss": 0.0219, "step": 13611 }, { "epoch": 9.56570625439213, "grad_norm": 0.1052851676940918, "learning_rate": 2.6959006793159992e-05, "loss": 0.0184, "step": 13612 }, { "epoch": 9.566408995080815, "grad_norm": 0.13244646787643433, "learning_rate": 2.6958538299367532e-05, "loss": 0.0106, "step": 13613 }, { "epoch": 9.567111735769501, "grad_norm": 0.09116899222135544, "learning_rate": 2.6958069805575076e-05, "loss": 0.0133, "step": 13614 }, { "epoch": 9.567814476458187, "grad_norm": 0.1954842060804367, "learning_rate": 2.695760131178262e-05, "loss": 0.0233, "step": 13615 }, { "epoch": 9.568517217146873, "grad_norm": 0.20722998678684235, "learning_rate": 2.6957132817990163e-05, "loss": 0.0327, "step": 13616 }, { "epoch": 9.569219957835559, "grad_norm": 0.1525775045156479, "learning_rate": 2.6956664324197704e-05, "loss": 0.0113, "step": 13617 }, { "epoch": 9.569922698524245, "grad_norm": 0.11752741783857346, "learning_rate": 2.6956195830405248e-05, "loss": 0.0154, "step": 13618 }, { "epoch": 9.57062543921293, "grad_norm": 0.12425391376018524, "learning_rate": 2.695572733661279e-05, "loss": 0.0144, "step": 13619 }, { "epoch": 9.571328179901617, "grad_norm": 0.15364022552967072, "learning_rate": 2.6955258842820335e-05, "loss": 0.0271, "step": 13620 }, { "epoch": 9.572030920590302, "grad_norm": 0.16950565576553345, "learning_rate": 2.6954790349027875e-05, "loss": 0.0121, "step": 13621 }, { "epoch": 9.572733661278988, "grad_norm": 0.21661771833896637, "learning_rate": 2.695432185523542e-05, "loss": 0.0319, "step": 13622 }, { "epoch": 9.573436401967674, "grad_norm": 0.3250299096107483, "learning_rate": 2.6953853361442963e-05, "loss": 0.0309, "step": 13623 }, { "epoch": 9.57413914265636, "grad_norm": 0.25550904870033264, "learning_rate": 2.6953384867650506e-05, "loss": 0.0232, "step": 13624 }, { "epoch": 9.574841883345046, "grad_norm": 2.3460235595703125, "learning_rate": 2.695291637385805e-05, "loss": 0.053, "step": 13625 }, { "epoch": 9.575544624033732, "grad_norm": 0.28896766901016235, "learning_rate": 2.6952447880065587e-05, "loss": 0.0462, "step": 13626 }, { "epoch": 9.576247364722418, "grad_norm": 0.290140300989151, "learning_rate": 2.695197938627313e-05, "loss": 0.0385, "step": 13627 }, { "epoch": 9.576950105411104, "grad_norm": 0.6176428198814392, "learning_rate": 2.6951510892480675e-05, "loss": 0.0625, "step": 13628 }, { "epoch": 9.57765284609979, "grad_norm": 0.5024946331977844, "learning_rate": 2.695104239868822e-05, "loss": 0.1046, "step": 13629 }, { "epoch": 9.578355586788476, "grad_norm": 0.6782647371292114, "learning_rate": 2.695057390489576e-05, "loss": 0.1449, "step": 13630 }, { "epoch": 9.579058327477162, "grad_norm": 0.6336104869842529, "learning_rate": 2.6950105411103302e-05, "loss": 0.1917, "step": 13631 }, { "epoch": 9.579761068165848, "grad_norm": 1.1977360248565674, "learning_rate": 2.6949636917310846e-05, "loss": 0.2076, "step": 13632 }, { "epoch": 9.580463808854532, "grad_norm": 0.16881874203681946, "learning_rate": 2.694916842351839e-05, "loss": 0.0572, "step": 13633 }, { "epoch": 9.58116654954322, "grad_norm": 0.09866936504840851, "learning_rate": 2.694869992972593e-05, "loss": 0.0207, "step": 13634 }, { "epoch": 9.581869290231904, "grad_norm": 0.18508437275886536, "learning_rate": 2.6948231435933474e-05, "loss": 0.026, "step": 13635 }, { "epoch": 9.58257203092059, "grad_norm": 0.1039511188864708, "learning_rate": 2.6947762942141018e-05, "loss": 0.0167, "step": 13636 }, { "epoch": 9.583274771609275, "grad_norm": 0.12364094704389572, "learning_rate": 2.694729444834856e-05, "loss": 0.0157, "step": 13637 }, { "epoch": 9.583977512297961, "grad_norm": 0.213545024394989, "learning_rate": 2.6946825954556105e-05, "loss": 0.02, "step": 13638 }, { "epoch": 9.584680252986647, "grad_norm": 0.08918914943933487, "learning_rate": 2.6946357460763645e-05, "loss": 0.0093, "step": 13639 }, { "epoch": 9.585382993675333, "grad_norm": 0.3471282720565796, "learning_rate": 2.694588896697119e-05, "loss": 0.0197, "step": 13640 }, { "epoch": 9.58608573436402, "grad_norm": 0.8474369049072266, "learning_rate": 2.6945420473178733e-05, "loss": 0.0326, "step": 13641 }, { "epoch": 9.586788475052705, "grad_norm": 0.09337438642978668, "learning_rate": 2.6944951979386277e-05, "loss": 0.0147, "step": 13642 }, { "epoch": 9.587491215741391, "grad_norm": 0.12261932343244553, "learning_rate": 2.6944483485593814e-05, "loss": 0.0196, "step": 13643 }, { "epoch": 9.588193956430077, "grad_norm": 0.32804179191589355, "learning_rate": 2.6944014991801357e-05, "loss": 0.0171, "step": 13644 }, { "epoch": 9.588896697118763, "grad_norm": 0.1571367233991623, "learning_rate": 2.69435464980089e-05, "loss": 0.0226, "step": 13645 }, { "epoch": 9.589599437807449, "grad_norm": 0.4036216735839844, "learning_rate": 2.6943078004216445e-05, "loss": 0.0172, "step": 13646 }, { "epoch": 9.590302178496135, "grad_norm": 0.20642100274562836, "learning_rate": 2.6942609510423985e-05, "loss": 0.0336, "step": 13647 }, { "epoch": 9.59100491918482, "grad_norm": 0.20400308072566986, "learning_rate": 2.694214101663153e-05, "loss": 0.033, "step": 13648 }, { "epoch": 9.591707659873506, "grad_norm": 0.11066102236509323, "learning_rate": 2.6941672522839072e-05, "loss": 0.018, "step": 13649 }, { "epoch": 9.592410400562192, "grad_norm": 0.7999298572540283, "learning_rate": 2.6941204029046616e-05, "loss": 0.0585, "step": 13650 }, { "epoch": 9.593113141250878, "grad_norm": 0.1530563235282898, "learning_rate": 2.694073553525416e-05, "loss": 0.0298, "step": 13651 }, { "epoch": 9.593815881939564, "grad_norm": 0.3120841681957245, "learning_rate": 2.69402670414617e-05, "loss": 0.0566, "step": 13652 }, { "epoch": 9.59451862262825, "grad_norm": 0.388899564743042, "learning_rate": 2.6939798547669244e-05, "loss": 0.0831, "step": 13653 }, { "epoch": 9.595221363316936, "grad_norm": 0.677011251449585, "learning_rate": 2.6939330053876788e-05, "loss": 0.131, "step": 13654 }, { "epoch": 9.595924104005622, "grad_norm": 0.6990036368370056, "learning_rate": 2.693886156008433e-05, "loss": 0.1862, "step": 13655 }, { "epoch": 9.596626844694308, "grad_norm": 1.2762534618377686, "learning_rate": 2.6938393066291872e-05, "loss": 0.1514, "step": 13656 }, { "epoch": 9.597329585382994, "grad_norm": 2.1411068439483643, "learning_rate": 2.6937924572499416e-05, "loss": 0.1966, "step": 13657 }, { "epoch": 9.59803232607168, "grad_norm": 0.25545257329940796, "learning_rate": 2.693745607870696e-05, "loss": 0.0906, "step": 13658 }, { "epoch": 9.598735066760366, "grad_norm": 0.180571049451828, "learning_rate": 2.6936987584914503e-05, "loss": 0.0243, "step": 13659 }, { "epoch": 9.599437807449052, "grad_norm": 0.18308411538600922, "learning_rate": 2.6936519091122043e-05, "loss": 0.022, "step": 13660 }, { "epoch": 9.600140548137738, "grad_norm": 0.20168565213680267, "learning_rate": 2.6936050597329584e-05, "loss": 0.0206, "step": 13661 }, { "epoch": 9.600843288826423, "grad_norm": 0.13465465605258942, "learning_rate": 2.6935582103537127e-05, "loss": 0.0187, "step": 13662 }, { "epoch": 9.60154602951511, "grad_norm": 0.07903899252414703, "learning_rate": 2.693511360974467e-05, "loss": 0.0127, "step": 13663 }, { "epoch": 9.602248770203795, "grad_norm": 0.1033717468380928, "learning_rate": 2.6934645115952215e-05, "loss": 0.0158, "step": 13664 }, { "epoch": 9.602951510892481, "grad_norm": 0.3933793902397156, "learning_rate": 2.6934176622159755e-05, "loss": 0.0134, "step": 13665 }, { "epoch": 9.603654251581167, "grad_norm": 0.18592733144760132, "learning_rate": 2.69337081283673e-05, "loss": 0.0204, "step": 13666 }, { "epoch": 9.604356992269853, "grad_norm": 0.17702428996562958, "learning_rate": 2.6933239634574843e-05, "loss": 0.0179, "step": 13667 }, { "epoch": 9.605059732958539, "grad_norm": 0.22466671466827393, "learning_rate": 2.6932771140782386e-05, "loss": 0.0302, "step": 13668 }, { "epoch": 9.605762473647225, "grad_norm": 0.15306848287582397, "learning_rate": 2.6932302646989927e-05, "loss": 0.0192, "step": 13669 }, { "epoch": 9.60646521433591, "grad_norm": 0.2556350827217102, "learning_rate": 2.693183415319747e-05, "loss": 0.0446, "step": 13670 }, { "epoch": 9.607167955024597, "grad_norm": 0.2815510928630829, "learning_rate": 2.6931365659405014e-05, "loss": 0.0328, "step": 13671 }, { "epoch": 9.607870695713281, "grad_norm": 0.1805151402950287, "learning_rate": 2.6930897165612558e-05, "loss": 0.033, "step": 13672 }, { "epoch": 9.608573436401969, "grad_norm": 0.23854240775108337, "learning_rate": 2.6930428671820098e-05, "loss": 0.0429, "step": 13673 }, { "epoch": 9.609276177090653, "grad_norm": 0.23210248351097107, "learning_rate": 2.6929960178027642e-05, "loss": 0.0251, "step": 13674 }, { "epoch": 9.609978917779339, "grad_norm": 0.8119791150093079, "learning_rate": 2.6929491684235186e-05, "loss": 0.0367, "step": 13675 }, { "epoch": 9.610681658468025, "grad_norm": 0.18759317696094513, "learning_rate": 2.692902319044273e-05, "loss": 0.0397, "step": 13676 }, { "epoch": 9.61138439915671, "grad_norm": 0.2378111034631729, "learning_rate": 2.6928554696650273e-05, "loss": 0.0487, "step": 13677 }, { "epoch": 9.612087139845396, "grad_norm": 0.2941174805164337, "learning_rate": 2.692808620285781e-05, "loss": 0.054, "step": 13678 }, { "epoch": 9.612789880534082, "grad_norm": 0.5079416036605835, "learning_rate": 2.6927617709065354e-05, "loss": 0.1008, "step": 13679 }, { "epoch": 9.613492621222768, "grad_norm": 0.5732165575027466, "learning_rate": 2.6927149215272897e-05, "loss": 0.121, "step": 13680 }, { "epoch": 9.614195361911454, "grad_norm": 0.9729206562042236, "learning_rate": 2.692668072148044e-05, "loss": 0.2219, "step": 13681 }, { "epoch": 9.61489810260014, "grad_norm": 1.2434693574905396, "learning_rate": 2.692621222768798e-05, "loss": 0.1953, "step": 13682 }, { "epoch": 9.615600843288826, "grad_norm": 0.4515845775604248, "learning_rate": 2.6925743733895525e-05, "loss": 0.0886, "step": 13683 }, { "epoch": 9.616303583977512, "grad_norm": 0.4437156021595001, "learning_rate": 2.692527524010307e-05, "loss": 0.0334, "step": 13684 }, { "epoch": 9.617006324666198, "grad_norm": 0.27841436862945557, "learning_rate": 2.6924806746310613e-05, "loss": 0.028, "step": 13685 }, { "epoch": 9.617709065354884, "grad_norm": 0.14723551273345947, "learning_rate": 2.6924338252518153e-05, "loss": 0.0177, "step": 13686 }, { "epoch": 9.61841180604357, "grad_norm": 0.2340822070837021, "learning_rate": 2.6923869758725697e-05, "loss": 0.022, "step": 13687 }, { "epoch": 9.619114546732256, "grad_norm": 0.20698049664497375, "learning_rate": 2.692340126493324e-05, "loss": 0.0141, "step": 13688 }, { "epoch": 9.619817287420942, "grad_norm": 1.2111190557479858, "learning_rate": 2.6922932771140784e-05, "loss": 0.0384, "step": 13689 }, { "epoch": 9.620520028109627, "grad_norm": 0.5209121108055115, "learning_rate": 2.6922464277348328e-05, "loss": 0.0189, "step": 13690 }, { "epoch": 9.621222768798313, "grad_norm": 0.18921086192131042, "learning_rate": 2.6921995783555868e-05, "loss": 0.0252, "step": 13691 }, { "epoch": 9.621925509487, "grad_norm": 0.16195428371429443, "learning_rate": 2.6921527289763412e-05, "loss": 0.0178, "step": 13692 }, { "epoch": 9.622628250175685, "grad_norm": 0.31136831641197205, "learning_rate": 2.6921058795970956e-05, "loss": 0.0276, "step": 13693 }, { "epoch": 9.623330990864371, "grad_norm": 0.533248782157898, "learning_rate": 2.69205903021785e-05, "loss": 0.0177, "step": 13694 }, { "epoch": 9.624033731553057, "grad_norm": 0.16241510212421417, "learning_rate": 2.692012180838604e-05, "loss": 0.0252, "step": 13695 }, { "epoch": 9.624736472241743, "grad_norm": 0.2403000444173813, "learning_rate": 2.691965331459358e-05, "loss": 0.0113, "step": 13696 }, { "epoch": 9.625439212930429, "grad_norm": 0.34137237071990967, "learning_rate": 2.6919184820801124e-05, "loss": 0.0295, "step": 13697 }, { "epoch": 9.626141953619115, "grad_norm": 0.35529637336730957, "learning_rate": 2.6918716327008668e-05, "loss": 0.0313, "step": 13698 }, { "epoch": 9.6268446943078, "grad_norm": 0.16664162278175354, "learning_rate": 2.691824783321621e-05, "loss": 0.0216, "step": 13699 }, { "epoch": 9.627547434996487, "grad_norm": 0.18587520718574524, "learning_rate": 2.691777933942375e-05, "loss": 0.0362, "step": 13700 }, { "epoch": 9.628250175685173, "grad_norm": 0.2534102499485016, "learning_rate": 2.6917310845631295e-05, "loss": 0.0421, "step": 13701 }, { "epoch": 9.628952916373859, "grad_norm": 0.8056074380874634, "learning_rate": 2.691684235183884e-05, "loss": 0.0644, "step": 13702 }, { "epoch": 9.629655657062544, "grad_norm": 0.4125673472881317, "learning_rate": 2.6916373858046383e-05, "loss": 0.0561, "step": 13703 }, { "epoch": 9.63035839775123, "grad_norm": 0.49812376499176025, "learning_rate": 2.6915905364253923e-05, "loss": 0.1076, "step": 13704 }, { "epoch": 9.631061138439916, "grad_norm": 1.0507837533950806, "learning_rate": 2.6915436870461467e-05, "loss": 0.1469, "step": 13705 }, { "epoch": 9.631763879128602, "grad_norm": 0.7198270559310913, "learning_rate": 2.691496837666901e-05, "loss": 0.2189, "step": 13706 }, { "epoch": 9.632466619817288, "grad_norm": 1.3918955326080322, "learning_rate": 2.6914499882876554e-05, "loss": 0.227, "step": 13707 }, { "epoch": 9.633169360505974, "grad_norm": 0.20899948477745056, "learning_rate": 2.6914031389084095e-05, "loss": 0.0633, "step": 13708 }, { "epoch": 9.63387210119466, "grad_norm": 0.11917359381914139, "learning_rate": 2.691356289529164e-05, "loss": 0.021, "step": 13709 }, { "epoch": 9.634574841883346, "grad_norm": 0.1700243055820465, "learning_rate": 2.6913094401499182e-05, "loss": 0.0271, "step": 13710 }, { "epoch": 9.63527758257203, "grad_norm": 0.13880032300949097, "learning_rate": 2.6912625907706726e-05, "loss": 0.0162, "step": 13711 }, { "epoch": 9.635980323260716, "grad_norm": 0.14652246236801147, "learning_rate": 2.6912157413914266e-05, "loss": 0.0216, "step": 13712 }, { "epoch": 9.636683063949402, "grad_norm": 0.21795707941055298, "learning_rate": 2.6911688920121807e-05, "loss": 0.0264, "step": 13713 }, { "epoch": 9.637385804638088, "grad_norm": 0.11350642144680023, "learning_rate": 2.691122042632935e-05, "loss": 0.0158, "step": 13714 }, { "epoch": 9.638088545326774, "grad_norm": 0.08910644054412842, "learning_rate": 2.6910751932536894e-05, "loss": 0.02, "step": 13715 }, { "epoch": 9.63879128601546, "grad_norm": 0.2984362244606018, "learning_rate": 2.6910283438744438e-05, "loss": 0.0284, "step": 13716 }, { "epoch": 9.639494026704146, "grad_norm": 0.2439652383327484, "learning_rate": 2.6909814944951978e-05, "loss": 0.0144, "step": 13717 }, { "epoch": 9.640196767392831, "grad_norm": 0.20376023650169373, "learning_rate": 2.6909346451159522e-05, "loss": 0.0377, "step": 13718 }, { "epoch": 9.640899508081517, "grad_norm": 0.12985140085220337, "learning_rate": 2.6908877957367065e-05, "loss": 0.0253, "step": 13719 }, { "epoch": 9.641602248770203, "grad_norm": 0.3278812766075134, "learning_rate": 2.690840946357461e-05, "loss": 0.0156, "step": 13720 }, { "epoch": 9.64230498945889, "grad_norm": 0.1442316472530365, "learning_rate": 2.690794096978215e-05, "loss": 0.0241, "step": 13721 }, { "epoch": 9.643007730147575, "grad_norm": 0.19913826882839203, "learning_rate": 2.6907472475989693e-05, "loss": 0.0256, "step": 13722 }, { "epoch": 9.643710470836261, "grad_norm": 0.15879641473293304, "learning_rate": 2.6907003982197237e-05, "loss": 0.0259, "step": 13723 }, { "epoch": 9.644413211524947, "grad_norm": 0.20356880128383636, "learning_rate": 2.690653548840478e-05, "loss": 0.0157, "step": 13724 }, { "epoch": 9.645115952213633, "grad_norm": 0.14970612525939941, "learning_rate": 2.6906066994612324e-05, "loss": 0.0339, "step": 13725 }, { "epoch": 9.645818692902319, "grad_norm": 0.27361539006233215, "learning_rate": 2.6905598500819865e-05, "loss": 0.0245, "step": 13726 }, { "epoch": 9.646521433591005, "grad_norm": 0.3095136880874634, "learning_rate": 2.690513000702741e-05, "loss": 0.0618, "step": 13727 }, { "epoch": 9.64722417427969, "grad_norm": 0.3487769067287445, "learning_rate": 2.6904661513234952e-05, "loss": 0.0844, "step": 13728 }, { "epoch": 9.647926914968377, "grad_norm": 0.35219064354896545, "learning_rate": 2.6904193019442496e-05, "loss": 0.097, "step": 13729 }, { "epoch": 9.648629655657063, "grad_norm": 1.6073094606399536, "learning_rate": 2.6903724525650033e-05, "loss": 0.1244, "step": 13730 }, { "epoch": 9.649332396345748, "grad_norm": 1.0954722166061401, "learning_rate": 2.6903256031857577e-05, "loss": 0.2062, "step": 13731 }, { "epoch": 9.650035137034434, "grad_norm": 1.0629849433898926, "learning_rate": 2.690278753806512e-05, "loss": 0.2701, "step": 13732 }, { "epoch": 9.65073787772312, "grad_norm": 0.22303426265716553, "learning_rate": 2.6902319044272664e-05, "loss": 0.0701, "step": 13733 }, { "epoch": 9.651440618411806, "grad_norm": 0.11955109983682632, "learning_rate": 2.6901850550480204e-05, "loss": 0.0286, "step": 13734 }, { "epoch": 9.652143359100492, "grad_norm": 0.25147056579589844, "learning_rate": 2.6901382056687748e-05, "loss": 0.0278, "step": 13735 }, { "epoch": 9.652846099789178, "grad_norm": 0.10595709830522537, "learning_rate": 2.6900913562895292e-05, "loss": 0.02, "step": 13736 }, { "epoch": 9.653548840477864, "grad_norm": 0.16739404201507568, "learning_rate": 2.6900445069102836e-05, "loss": 0.0249, "step": 13737 }, { "epoch": 9.65425158116655, "grad_norm": 0.08958331495523453, "learning_rate": 2.689997657531038e-05, "loss": 0.0183, "step": 13738 }, { "epoch": 9.654954321855236, "grad_norm": 0.10084649175405502, "learning_rate": 2.689950808151792e-05, "loss": 0.015, "step": 13739 }, { "epoch": 9.655657062543922, "grad_norm": 0.21670612692832947, "learning_rate": 2.6899039587725463e-05, "loss": 0.0248, "step": 13740 }, { "epoch": 9.656359803232608, "grad_norm": 0.11979050189256668, "learning_rate": 2.6898571093933007e-05, "loss": 0.0242, "step": 13741 }, { "epoch": 9.657062543921294, "grad_norm": 0.13688012957572937, "learning_rate": 2.689810260014055e-05, "loss": 0.0176, "step": 13742 }, { "epoch": 9.65776528460998, "grad_norm": 0.11447598785161972, "learning_rate": 2.689763410634809e-05, "loss": 0.02, "step": 13743 }, { "epoch": 9.658468025298665, "grad_norm": 0.0933012068271637, "learning_rate": 2.6897165612555635e-05, "loss": 0.0093, "step": 13744 }, { "epoch": 9.659170765987351, "grad_norm": 0.14139869809150696, "learning_rate": 2.689669711876318e-05, "loss": 0.0265, "step": 13745 }, { "epoch": 9.659873506676037, "grad_norm": 0.185139000415802, "learning_rate": 2.6896228624970722e-05, "loss": 0.0154, "step": 13746 }, { "epoch": 9.660576247364723, "grad_norm": 0.4641578495502472, "learning_rate": 2.6895760131178263e-05, "loss": 0.0283, "step": 13747 }, { "epoch": 9.66127898805341, "grad_norm": 0.16934798657894135, "learning_rate": 2.6895291637385803e-05, "loss": 0.0306, "step": 13748 }, { "epoch": 9.661981728742095, "grad_norm": 0.14480186998844147, "learning_rate": 2.6894823143593347e-05, "loss": 0.0321, "step": 13749 }, { "epoch": 9.66268446943078, "grad_norm": 0.10475269705057144, "learning_rate": 2.689435464980089e-05, "loss": 0.0151, "step": 13750 }, { "epoch": 9.663387210119465, "grad_norm": 0.3108024001121521, "learning_rate": 2.6893886156008434e-05, "loss": 0.1019, "step": 13751 }, { "epoch": 9.664089950808151, "grad_norm": 0.20583266019821167, "learning_rate": 2.6893417662215975e-05, "loss": 0.0629, "step": 13752 }, { "epoch": 9.664792691496837, "grad_norm": 0.311859130859375, "learning_rate": 2.6892949168423518e-05, "loss": 0.0807, "step": 13753 }, { "epoch": 9.665495432185523, "grad_norm": 0.3390776515007019, "learning_rate": 2.6892480674631062e-05, "loss": 0.0981, "step": 13754 }, { "epoch": 9.666198172874209, "grad_norm": 1.1236377954483032, "learning_rate": 2.6892012180838606e-05, "loss": 0.1707, "step": 13755 }, { "epoch": 9.666900913562895, "grad_norm": 1.07374906539917, "learning_rate": 2.6891543687046146e-05, "loss": 0.1776, "step": 13756 }, { "epoch": 9.66760365425158, "grad_norm": 0.9972128868103027, "learning_rate": 2.689107519325369e-05, "loss": 0.2488, "step": 13757 }, { "epoch": 9.668306394940267, "grad_norm": 0.22415930032730103, "learning_rate": 2.6890606699461234e-05, "loss": 0.0678, "step": 13758 }, { "epoch": 9.669009135628952, "grad_norm": 0.22591958940029144, "learning_rate": 2.6890138205668777e-05, "loss": 0.0317, "step": 13759 }, { "epoch": 9.669711876317638, "grad_norm": 0.1177646815776825, "learning_rate": 2.6889669711876318e-05, "loss": 0.0104, "step": 13760 }, { "epoch": 9.670414617006324, "grad_norm": 0.185777485370636, "learning_rate": 2.688920121808386e-05, "loss": 0.0202, "step": 13761 }, { "epoch": 9.67111735769501, "grad_norm": 0.13231952488422394, "learning_rate": 2.6888732724291405e-05, "loss": 0.014, "step": 13762 }, { "epoch": 9.671820098383696, "grad_norm": 0.06228543817996979, "learning_rate": 2.688826423049895e-05, "loss": 0.0047, "step": 13763 }, { "epoch": 9.672522839072382, "grad_norm": 0.09162794798612595, "learning_rate": 2.6887795736706492e-05, "loss": 0.0106, "step": 13764 }, { "epoch": 9.673225579761068, "grad_norm": 0.1867845505475998, "learning_rate": 2.688732724291403e-05, "loss": 0.0263, "step": 13765 }, { "epoch": 9.673928320449754, "grad_norm": 0.13783493638038635, "learning_rate": 2.6886858749121573e-05, "loss": 0.0327, "step": 13766 }, { "epoch": 9.67463106113844, "grad_norm": 0.11752216517925262, "learning_rate": 2.6886390255329117e-05, "loss": 0.0124, "step": 13767 }, { "epoch": 9.675333801827126, "grad_norm": 0.1626995950937271, "learning_rate": 2.688592176153666e-05, "loss": 0.0263, "step": 13768 }, { "epoch": 9.676036542515812, "grad_norm": 0.09122905135154724, "learning_rate": 2.68854532677442e-05, "loss": 0.0083, "step": 13769 }, { "epoch": 9.676739283204498, "grad_norm": 0.14248725771903992, "learning_rate": 2.6884984773951745e-05, "loss": 0.0414, "step": 13770 }, { "epoch": 9.677442023893184, "grad_norm": 0.20103175938129425, "learning_rate": 2.688451628015929e-05, "loss": 0.0159, "step": 13771 }, { "epoch": 9.67814476458187, "grad_norm": 0.1338907927274704, "learning_rate": 2.6884047786366832e-05, "loss": 0.0136, "step": 13772 }, { "epoch": 9.678847505270555, "grad_norm": 0.2691953182220459, "learning_rate": 2.6883579292574372e-05, "loss": 0.0394, "step": 13773 }, { "epoch": 9.679550245959241, "grad_norm": 0.23729577660560608, "learning_rate": 2.6883110798781916e-05, "loss": 0.0311, "step": 13774 }, { "epoch": 9.680252986647927, "grad_norm": 0.33223018050193787, "learning_rate": 2.688264230498946e-05, "loss": 0.049, "step": 13775 }, { "epoch": 9.680955727336613, "grad_norm": 0.14334769546985626, "learning_rate": 2.6882173811197004e-05, "loss": 0.048, "step": 13776 }, { "epoch": 9.681658468025299, "grad_norm": 0.205271378159523, "learning_rate": 2.6881705317404547e-05, "loss": 0.0661, "step": 13777 }, { "epoch": 9.682361208713985, "grad_norm": 0.19289007782936096, "learning_rate": 2.6881236823612088e-05, "loss": 0.0737, "step": 13778 }, { "epoch": 9.683063949402671, "grad_norm": 0.5127153992652893, "learning_rate": 2.688076832981963e-05, "loss": 0.0937, "step": 13779 }, { "epoch": 9.683766690091357, "grad_norm": 0.5585649609565735, "learning_rate": 2.6880299836027175e-05, "loss": 0.1773, "step": 13780 }, { "epoch": 9.684469430780043, "grad_norm": 0.9101743102073669, "learning_rate": 2.687983134223472e-05, "loss": 0.1851, "step": 13781 }, { "epoch": 9.685172171468729, "grad_norm": 0.8014460802078247, "learning_rate": 2.687936284844226e-05, "loss": 0.2375, "step": 13782 }, { "epoch": 9.685874912157415, "grad_norm": 0.24745431542396545, "learning_rate": 2.68788943546498e-05, "loss": 0.0745, "step": 13783 }, { "epoch": 9.6865776528461, "grad_norm": 0.14537641406059265, "learning_rate": 2.6878425860857343e-05, "loss": 0.0292, "step": 13784 }, { "epoch": 9.687280393534786, "grad_norm": 0.12099022418260574, "learning_rate": 2.6877957367064887e-05, "loss": 0.026, "step": 13785 }, { "epoch": 9.687983134223472, "grad_norm": 0.0920349732041359, "learning_rate": 2.6877488873272427e-05, "loss": 0.0164, "step": 13786 }, { "epoch": 9.688685874912156, "grad_norm": 0.10721894353628159, "learning_rate": 2.687702037947997e-05, "loss": 0.0173, "step": 13787 }, { "epoch": 9.689388615600844, "grad_norm": 0.09108640998601913, "learning_rate": 2.6876551885687515e-05, "loss": 0.0096, "step": 13788 }, { "epoch": 9.690091356289528, "grad_norm": 0.21799542009830475, "learning_rate": 2.687608339189506e-05, "loss": 0.0167, "step": 13789 }, { "epoch": 9.690794096978214, "grad_norm": 0.1402466744184494, "learning_rate": 2.6875614898102602e-05, "loss": 0.0165, "step": 13790 }, { "epoch": 9.6914968376669, "grad_norm": 0.08644713461399078, "learning_rate": 2.6875146404310143e-05, "loss": 0.0136, "step": 13791 }, { "epoch": 9.692199578355586, "grad_norm": 0.13255298137664795, "learning_rate": 2.6874677910517686e-05, "loss": 0.0113, "step": 13792 }, { "epoch": 9.692902319044272, "grad_norm": 0.5902727842330933, "learning_rate": 2.687420941672523e-05, "loss": 0.02, "step": 13793 }, { "epoch": 9.693605059732958, "grad_norm": 0.11609893292188644, "learning_rate": 2.6873740922932774e-05, "loss": 0.0155, "step": 13794 }, { "epoch": 9.694307800421644, "grad_norm": 0.31573572754859924, "learning_rate": 2.6873272429140314e-05, "loss": 0.0217, "step": 13795 }, { "epoch": 9.69501054111033, "grad_norm": 0.12432003766298294, "learning_rate": 2.6872803935347858e-05, "loss": 0.0183, "step": 13796 }, { "epoch": 9.695713281799016, "grad_norm": 0.19360366463661194, "learning_rate": 2.68723354415554e-05, "loss": 0.0345, "step": 13797 }, { "epoch": 9.696416022487702, "grad_norm": 0.2809596657752991, "learning_rate": 2.6871866947762945e-05, "loss": 0.0599, "step": 13798 }, { "epoch": 9.697118763176388, "grad_norm": 0.18327321112155914, "learning_rate": 2.6871398453970486e-05, "loss": 0.0385, "step": 13799 }, { "epoch": 9.697821503865073, "grad_norm": 0.11986713111400604, "learning_rate": 2.6870929960178026e-05, "loss": 0.0289, "step": 13800 }, { "epoch": 9.69852424455376, "grad_norm": 0.13765409588813782, "learning_rate": 2.687046146638557e-05, "loss": 0.0296, "step": 13801 }, { "epoch": 9.699226985242445, "grad_norm": 0.2710495889186859, "learning_rate": 2.6869992972593113e-05, "loss": 0.0544, "step": 13802 }, { "epoch": 9.699929725931131, "grad_norm": 0.44405397772789, "learning_rate": 2.6869524478800657e-05, "loss": 0.0669, "step": 13803 }, { "epoch": 9.700632466619817, "grad_norm": 0.3366107940673828, "learning_rate": 2.6869055985008197e-05, "loss": 0.1039, "step": 13804 }, { "epoch": 9.701335207308503, "grad_norm": 0.5274109840393066, "learning_rate": 2.686858749121574e-05, "loss": 0.1512, "step": 13805 }, { "epoch": 9.702037947997189, "grad_norm": 0.8794665932655334, "learning_rate": 2.6868118997423285e-05, "loss": 0.1933, "step": 13806 }, { "epoch": 9.702740688685875, "grad_norm": 0.6512903571128845, "learning_rate": 2.686765050363083e-05, "loss": 0.2144, "step": 13807 }, { "epoch": 9.70344342937456, "grad_norm": 0.14574213325977325, "learning_rate": 2.686718200983837e-05, "loss": 0.0518, "step": 13808 }, { "epoch": 9.704146170063247, "grad_norm": 0.14299540221691132, "learning_rate": 2.6866713516045913e-05, "loss": 0.0294, "step": 13809 }, { "epoch": 9.704848910751933, "grad_norm": 0.1866086721420288, "learning_rate": 2.6866245022253456e-05, "loss": 0.0303, "step": 13810 }, { "epoch": 9.705551651440619, "grad_norm": 0.09795096516609192, "learning_rate": 2.6865776528461e-05, "loss": 0.0191, "step": 13811 }, { "epoch": 9.706254392129305, "grad_norm": 0.28943678736686707, "learning_rate": 2.686530803466854e-05, "loss": 0.0305, "step": 13812 }, { "epoch": 9.70695713281799, "grad_norm": 0.09990016371011734, "learning_rate": 2.6864839540876084e-05, "loss": 0.015, "step": 13813 }, { "epoch": 9.707659873506676, "grad_norm": 0.10488173365592957, "learning_rate": 2.6864371047083628e-05, "loss": 0.0062, "step": 13814 }, { "epoch": 9.708362614195362, "grad_norm": 0.24743331968784332, "learning_rate": 2.686390255329117e-05, "loss": 0.0266, "step": 13815 }, { "epoch": 9.709065354884048, "grad_norm": 0.1531960517168045, "learning_rate": 2.6863434059498715e-05, "loss": 0.0218, "step": 13816 }, { "epoch": 9.709768095572734, "grad_norm": 0.11382516473531723, "learning_rate": 2.6862965565706256e-05, "loss": 0.0115, "step": 13817 }, { "epoch": 9.71047083626142, "grad_norm": 0.16805803775787354, "learning_rate": 2.6862497071913796e-05, "loss": 0.0411, "step": 13818 }, { "epoch": 9.711173576950106, "grad_norm": 0.42241641879081726, "learning_rate": 2.686202857812134e-05, "loss": 0.0134, "step": 13819 }, { "epoch": 9.711876317638792, "grad_norm": 0.12941035628318787, "learning_rate": 2.6861560084328883e-05, "loss": 0.023, "step": 13820 }, { "epoch": 9.712579058327478, "grad_norm": 0.13466984033584595, "learning_rate": 2.6861091590536424e-05, "loss": 0.0161, "step": 13821 }, { "epoch": 9.713281799016164, "grad_norm": 0.2758996784687042, "learning_rate": 2.6860623096743968e-05, "loss": 0.0197, "step": 13822 }, { "epoch": 9.71398453970485, "grad_norm": 0.2034228891134262, "learning_rate": 2.686015460295151e-05, "loss": 0.0325, "step": 13823 }, { "epoch": 9.714687280393536, "grad_norm": 0.19663909077644348, "learning_rate": 2.6859686109159055e-05, "loss": 0.0328, "step": 13824 }, { "epoch": 9.715390021082221, "grad_norm": 0.34691953659057617, "learning_rate": 2.6859217615366595e-05, "loss": 0.0529, "step": 13825 }, { "epoch": 9.716092761770906, "grad_norm": 0.1247362270951271, "learning_rate": 2.685874912157414e-05, "loss": 0.0276, "step": 13826 }, { "epoch": 9.716795502459593, "grad_norm": 0.26710546016693115, "learning_rate": 2.6858280627781683e-05, "loss": 0.0356, "step": 13827 }, { "epoch": 9.717498243148277, "grad_norm": 0.3210235834121704, "learning_rate": 2.6857812133989227e-05, "loss": 0.0993, "step": 13828 }, { "epoch": 9.718200983836963, "grad_norm": 0.4847860634326935, "learning_rate": 2.685734364019677e-05, "loss": 0.0998, "step": 13829 }, { "epoch": 9.71890372452565, "grad_norm": 0.6417699456214905, "learning_rate": 2.685687514640431e-05, "loss": 0.1749, "step": 13830 }, { "epoch": 9.719606465214335, "grad_norm": 0.5764870643615723, "learning_rate": 2.6856406652611854e-05, "loss": 0.1725, "step": 13831 }, { "epoch": 9.720309205903021, "grad_norm": 1.0610089302062988, "learning_rate": 2.6855938158819398e-05, "loss": 0.2096, "step": 13832 }, { "epoch": 9.721011946591707, "grad_norm": 0.40457504987716675, "learning_rate": 2.6855469665026942e-05, "loss": 0.0733, "step": 13833 }, { "epoch": 9.721714687280393, "grad_norm": 0.15762929618358612, "learning_rate": 2.6855001171234482e-05, "loss": 0.035, "step": 13834 }, { "epoch": 9.722417427969079, "grad_norm": 0.17798110842704773, "learning_rate": 2.6854532677442022e-05, "loss": 0.0257, "step": 13835 }, { "epoch": 9.723120168657765, "grad_norm": 0.19891797006130219, "learning_rate": 2.6854064183649566e-05, "loss": 0.0141, "step": 13836 }, { "epoch": 9.72382290934645, "grad_norm": 0.2453276813030243, "learning_rate": 2.685359568985711e-05, "loss": 0.0201, "step": 13837 }, { "epoch": 9.724525650035137, "grad_norm": 0.0515238381922245, "learning_rate": 2.685312719606465e-05, "loss": 0.0063, "step": 13838 }, { "epoch": 9.725228390723823, "grad_norm": 0.08769747614860535, "learning_rate": 2.6852658702272194e-05, "loss": 0.009, "step": 13839 }, { "epoch": 9.725931131412509, "grad_norm": 0.1177215650677681, "learning_rate": 2.6852190208479738e-05, "loss": 0.0205, "step": 13840 }, { "epoch": 9.726633872101194, "grad_norm": 0.1879081279039383, "learning_rate": 2.685172171468728e-05, "loss": 0.02, "step": 13841 }, { "epoch": 9.72733661278988, "grad_norm": 0.1082492545247078, "learning_rate": 2.6851253220894825e-05, "loss": 0.0152, "step": 13842 }, { "epoch": 9.728039353478566, "grad_norm": 0.1346958577632904, "learning_rate": 2.6850784727102365e-05, "loss": 0.029, "step": 13843 }, { "epoch": 9.728742094167252, "grad_norm": 0.18531660735607147, "learning_rate": 2.685031623330991e-05, "loss": 0.0263, "step": 13844 }, { "epoch": 9.729444834855938, "grad_norm": 0.11528562009334564, "learning_rate": 2.6849847739517453e-05, "loss": 0.022, "step": 13845 }, { "epoch": 9.730147575544624, "grad_norm": 0.09448254853487015, "learning_rate": 2.6849379245724997e-05, "loss": 0.0135, "step": 13846 }, { "epoch": 9.73085031623331, "grad_norm": 0.4145374298095703, "learning_rate": 2.6848910751932537e-05, "loss": 0.0483, "step": 13847 }, { "epoch": 9.731553056921996, "grad_norm": 0.18022866547107697, "learning_rate": 2.684844225814008e-05, "loss": 0.0331, "step": 13848 }, { "epoch": 9.732255797610682, "grad_norm": 0.1299799233675003, "learning_rate": 2.6847973764347624e-05, "loss": 0.0142, "step": 13849 }, { "epoch": 9.732958538299368, "grad_norm": 0.18034565448760986, "learning_rate": 2.6847505270555168e-05, "loss": 0.0381, "step": 13850 }, { "epoch": 9.733661278988054, "grad_norm": 0.27811071276664734, "learning_rate": 2.684703677676271e-05, "loss": 0.0439, "step": 13851 }, { "epoch": 9.73436401967674, "grad_norm": 0.21772442758083344, "learning_rate": 2.684656828297025e-05, "loss": 0.0474, "step": 13852 }, { "epoch": 9.735066760365426, "grad_norm": 0.20898275077342987, "learning_rate": 2.6846099789177793e-05, "loss": 0.0658, "step": 13853 }, { "epoch": 9.735769501054111, "grad_norm": 1.031551480293274, "learning_rate": 2.6845631295385336e-05, "loss": 0.1296, "step": 13854 }, { "epoch": 9.736472241742797, "grad_norm": 0.6488060355186462, "learning_rate": 2.684516280159288e-05, "loss": 0.139, "step": 13855 }, { "epoch": 9.737174982431483, "grad_norm": 0.681562602519989, "learning_rate": 2.684469430780042e-05, "loss": 0.1886, "step": 13856 }, { "epoch": 9.73787772312017, "grad_norm": 1.3813962936401367, "learning_rate": 2.6844225814007964e-05, "loss": 0.2294, "step": 13857 }, { "epoch": 9.738580463808855, "grad_norm": 0.2126973420381546, "learning_rate": 2.6843757320215508e-05, "loss": 0.0652, "step": 13858 }, { "epoch": 9.739283204497541, "grad_norm": 0.25063255429267883, "learning_rate": 2.684328882642305e-05, "loss": 0.0258, "step": 13859 }, { "epoch": 9.739985945186227, "grad_norm": 0.17549021542072296, "learning_rate": 2.6842820332630592e-05, "loss": 0.0424, "step": 13860 }, { "epoch": 9.740688685874913, "grad_norm": 0.38386020064353943, "learning_rate": 2.6842351838838136e-05, "loss": 0.0302, "step": 13861 }, { "epoch": 9.741391426563599, "grad_norm": 0.06596552580595016, "learning_rate": 2.684188334504568e-05, "loss": 0.0124, "step": 13862 }, { "epoch": 9.742094167252285, "grad_norm": 0.1639784872531891, "learning_rate": 2.6841414851253223e-05, "loss": 0.0139, "step": 13863 }, { "epoch": 9.74279690794097, "grad_norm": 0.15337568521499634, "learning_rate": 2.6840946357460763e-05, "loss": 0.0159, "step": 13864 }, { "epoch": 9.743499648629655, "grad_norm": 0.18047122657299042, "learning_rate": 2.6840477863668307e-05, "loss": 0.0251, "step": 13865 }, { "epoch": 9.74420238931834, "grad_norm": 0.115516796708107, "learning_rate": 2.684000936987585e-05, "loss": 0.0241, "step": 13866 }, { "epoch": 9.744905130007027, "grad_norm": 0.14420248568058014, "learning_rate": 2.6839540876083395e-05, "loss": 0.0148, "step": 13867 }, { "epoch": 9.745607870695713, "grad_norm": 0.17739103734493256, "learning_rate": 2.6839072382290938e-05, "loss": 0.0302, "step": 13868 }, { "epoch": 9.746310611384398, "grad_norm": 0.0966721847653389, "learning_rate": 2.683860388849848e-05, "loss": 0.0115, "step": 13869 }, { "epoch": 9.747013352073084, "grad_norm": 0.4162541925907135, "learning_rate": 2.683813539470602e-05, "loss": 0.0258, "step": 13870 }, { "epoch": 9.74771609276177, "grad_norm": 0.17135363817214966, "learning_rate": 2.6837666900913563e-05, "loss": 0.0137, "step": 13871 }, { "epoch": 9.748418833450456, "grad_norm": 0.21664045751094818, "learning_rate": 2.6837198407121106e-05, "loss": 0.0359, "step": 13872 }, { "epoch": 9.749121574139142, "grad_norm": 0.14775079488754272, "learning_rate": 2.6836729913328647e-05, "loss": 0.0248, "step": 13873 }, { "epoch": 9.749824314827828, "grad_norm": 0.2091469168663025, "learning_rate": 2.683626141953619e-05, "loss": 0.0351, "step": 13874 }, { "epoch": 9.750527055516514, "grad_norm": 1.1287909746170044, "learning_rate": 2.6835792925743734e-05, "loss": 0.0254, "step": 13875 }, { "epoch": 9.7512297962052, "grad_norm": 0.3054180443286896, "learning_rate": 2.6835324431951278e-05, "loss": 0.0639, "step": 13876 }, { "epoch": 9.751932536893886, "grad_norm": 1.057991862297058, "learning_rate": 2.6834855938158818e-05, "loss": 0.0788, "step": 13877 }, { "epoch": 9.752635277582572, "grad_norm": 0.2553311288356781, "learning_rate": 2.6834387444366362e-05, "loss": 0.0719, "step": 13878 }, { "epoch": 9.753338018271258, "grad_norm": 0.4930953085422516, "learning_rate": 2.6833918950573906e-05, "loss": 0.1165, "step": 13879 }, { "epoch": 9.754040758959944, "grad_norm": 0.6167026162147522, "learning_rate": 2.683345045678145e-05, "loss": 0.14, "step": 13880 }, { "epoch": 9.75474349964863, "grad_norm": 0.5285847187042236, "learning_rate": 2.6832981962988993e-05, "loss": 0.1963, "step": 13881 }, { "epoch": 9.755446240337315, "grad_norm": 0.8595433235168457, "learning_rate": 2.6832513469196533e-05, "loss": 0.2156, "step": 13882 }, { "epoch": 9.756148981026001, "grad_norm": 0.7800493240356445, "learning_rate": 2.6832044975404077e-05, "loss": 0.0725, "step": 13883 }, { "epoch": 9.756851721714687, "grad_norm": 0.14968489110469818, "learning_rate": 2.683157648161162e-05, "loss": 0.0304, "step": 13884 }, { "epoch": 9.757554462403373, "grad_norm": 0.2939518690109253, "learning_rate": 2.6831107987819165e-05, "loss": 0.0278, "step": 13885 }, { "epoch": 9.75825720309206, "grad_norm": 0.20777671039104462, "learning_rate": 2.6830639494026705e-05, "loss": 0.0233, "step": 13886 }, { "epoch": 9.758959943780745, "grad_norm": 0.19491875171661377, "learning_rate": 2.6830171000234245e-05, "loss": 0.0233, "step": 13887 }, { "epoch": 9.759662684469431, "grad_norm": 0.16966161131858826, "learning_rate": 2.682970250644179e-05, "loss": 0.0153, "step": 13888 }, { "epoch": 9.760365425158117, "grad_norm": 0.1829618662595749, "learning_rate": 2.6829234012649333e-05, "loss": 0.0262, "step": 13889 }, { "epoch": 9.761068165846803, "grad_norm": 0.31916552782058716, "learning_rate": 2.6828765518856873e-05, "loss": 0.0295, "step": 13890 }, { "epoch": 9.761770906535489, "grad_norm": 0.13588005304336548, "learning_rate": 2.6828297025064417e-05, "loss": 0.026, "step": 13891 }, { "epoch": 9.762473647224175, "grad_norm": 0.14075617492198944, "learning_rate": 2.682782853127196e-05, "loss": 0.0201, "step": 13892 }, { "epoch": 9.76317638791286, "grad_norm": 0.11827301979064941, "learning_rate": 2.6827360037479504e-05, "loss": 0.0249, "step": 13893 }, { "epoch": 9.763879128601546, "grad_norm": 0.08866655081510544, "learning_rate": 2.6826891543687048e-05, "loss": 0.0109, "step": 13894 }, { "epoch": 9.764581869290232, "grad_norm": 0.17927666008472443, "learning_rate": 2.682642304989459e-05, "loss": 0.0218, "step": 13895 }, { "epoch": 9.765284609978918, "grad_norm": 0.3413528501987457, "learning_rate": 2.6825954556102132e-05, "loss": 0.0235, "step": 13896 }, { "epoch": 9.765987350667604, "grad_norm": 0.14427365362644196, "learning_rate": 2.6825486062309676e-05, "loss": 0.0247, "step": 13897 }, { "epoch": 9.76669009135629, "grad_norm": 0.2871835231781006, "learning_rate": 2.682501756851722e-05, "loss": 0.0293, "step": 13898 }, { "epoch": 9.767392832044976, "grad_norm": 0.21288305521011353, "learning_rate": 2.682454907472476e-05, "loss": 0.0282, "step": 13899 }, { "epoch": 9.768095572733662, "grad_norm": 0.1483369916677475, "learning_rate": 2.6824080580932304e-05, "loss": 0.0283, "step": 13900 }, { "epoch": 9.768798313422348, "grad_norm": 0.222802996635437, "learning_rate": 2.6823612087139847e-05, "loss": 0.0561, "step": 13901 }, { "epoch": 9.769501054111032, "grad_norm": 0.26229265332221985, "learning_rate": 2.682314359334739e-05, "loss": 0.0527, "step": 13902 }, { "epoch": 9.77020379479972, "grad_norm": 0.2572087049484253, "learning_rate": 2.682267509955493e-05, "loss": 0.0773, "step": 13903 }, { "epoch": 9.770906535488404, "grad_norm": 0.33752647042274475, "learning_rate": 2.6822206605762475e-05, "loss": 0.1062, "step": 13904 }, { "epoch": 9.77160927617709, "grad_norm": 0.5149563550949097, "learning_rate": 2.6821738111970015e-05, "loss": 0.1466, "step": 13905 }, { "epoch": 9.772312016865776, "grad_norm": 0.6068510413169861, "learning_rate": 2.682126961817756e-05, "loss": 0.2033, "step": 13906 }, { "epoch": 9.773014757554462, "grad_norm": 1.0341254472732544, "learning_rate": 2.6820801124385103e-05, "loss": 0.216, "step": 13907 }, { "epoch": 9.773717498243148, "grad_norm": 0.2600971758365631, "learning_rate": 2.6820332630592643e-05, "loss": 0.0622, "step": 13908 }, { "epoch": 9.774420238931834, "grad_norm": 0.21626664698123932, "learning_rate": 2.6819864136800187e-05, "loss": 0.051, "step": 13909 }, { "epoch": 9.77512297962052, "grad_norm": 0.17272856831550598, "learning_rate": 2.681939564300773e-05, "loss": 0.0258, "step": 13910 }, { "epoch": 9.775825720309205, "grad_norm": 0.260164350271225, "learning_rate": 2.6818927149215274e-05, "loss": 0.0211, "step": 13911 }, { "epoch": 9.776528460997891, "grad_norm": 0.29225778579711914, "learning_rate": 2.6818458655422815e-05, "loss": 0.0172, "step": 13912 }, { "epoch": 9.777231201686577, "grad_norm": 0.20841245353221893, "learning_rate": 2.681799016163036e-05, "loss": 0.0116, "step": 13913 }, { "epoch": 9.777933942375263, "grad_norm": 0.11615928262472153, "learning_rate": 2.6817521667837902e-05, "loss": 0.0191, "step": 13914 }, { "epoch": 9.778636683063949, "grad_norm": 0.18603619933128357, "learning_rate": 2.6817053174045446e-05, "loss": 0.0307, "step": 13915 }, { "epoch": 9.779339423752635, "grad_norm": 0.18123631179332733, "learning_rate": 2.6816584680252986e-05, "loss": 0.0201, "step": 13916 }, { "epoch": 9.780042164441321, "grad_norm": 0.13861292600631714, "learning_rate": 2.681611618646053e-05, "loss": 0.0107, "step": 13917 }, { "epoch": 9.780744905130007, "grad_norm": 0.1723635494709015, "learning_rate": 2.6815647692668074e-05, "loss": 0.0353, "step": 13918 }, { "epoch": 9.781447645818693, "grad_norm": 0.16918352246284485, "learning_rate": 2.6815179198875617e-05, "loss": 0.0137, "step": 13919 }, { "epoch": 9.782150386507379, "grad_norm": 0.15976238250732422, "learning_rate": 2.681471070508316e-05, "loss": 0.0262, "step": 13920 }, { "epoch": 9.782853127196065, "grad_norm": 0.26201725006103516, "learning_rate": 2.68142422112907e-05, "loss": 0.0195, "step": 13921 }, { "epoch": 9.78355586788475, "grad_norm": 0.17386369407176971, "learning_rate": 2.6813773717498242e-05, "loss": 0.0237, "step": 13922 }, { "epoch": 9.784258608573436, "grad_norm": 0.22163648903369904, "learning_rate": 2.6813305223705786e-05, "loss": 0.034, "step": 13923 }, { "epoch": 9.784961349262122, "grad_norm": 0.14974914491176605, "learning_rate": 2.681283672991333e-05, "loss": 0.0172, "step": 13924 }, { "epoch": 9.785664089950808, "grad_norm": 0.22135066986083984, "learning_rate": 2.681236823612087e-05, "loss": 0.0356, "step": 13925 }, { "epoch": 9.786366830639494, "grad_norm": 0.3939768671989441, "learning_rate": 2.6811899742328413e-05, "loss": 0.0433, "step": 13926 }, { "epoch": 9.78706957132818, "grad_norm": 0.1895390897989273, "learning_rate": 2.6811431248535957e-05, "loss": 0.0433, "step": 13927 }, { "epoch": 9.787772312016866, "grad_norm": 0.2409852296113968, "learning_rate": 2.68109627547435e-05, "loss": 0.0517, "step": 13928 }, { "epoch": 9.788475052705552, "grad_norm": 0.37898439168930054, "learning_rate": 2.6810494260951044e-05, "loss": 0.1167, "step": 13929 }, { "epoch": 9.789177793394238, "grad_norm": 0.5259461402893066, "learning_rate": 2.6810025767158585e-05, "loss": 0.1088, "step": 13930 }, { "epoch": 9.789880534082924, "grad_norm": 0.6779031753540039, "learning_rate": 2.680955727336613e-05, "loss": 0.1754, "step": 13931 }, { "epoch": 9.79058327477161, "grad_norm": 1.1374691724777222, "learning_rate": 2.6809088779573672e-05, "loss": 0.2068, "step": 13932 }, { "epoch": 9.791286015460296, "grad_norm": 0.20445062220096588, "learning_rate": 2.6808620285781216e-05, "loss": 0.0602, "step": 13933 }, { "epoch": 9.791988756148982, "grad_norm": 0.4760993421077728, "learning_rate": 2.6808151791988756e-05, "loss": 0.023, "step": 13934 }, { "epoch": 9.792691496837667, "grad_norm": 0.13261117041110992, "learning_rate": 2.68076832981963e-05, "loss": 0.0348, "step": 13935 }, { "epoch": 9.793394237526353, "grad_norm": 0.13797156512737274, "learning_rate": 2.6807214804403844e-05, "loss": 0.0313, "step": 13936 }, { "epoch": 9.79409697821504, "grad_norm": 0.12165108323097229, "learning_rate": 2.6806746310611388e-05, "loss": 0.0176, "step": 13937 }, { "epoch": 9.794799718903725, "grad_norm": 0.09827686846256256, "learning_rate": 2.6806277816818928e-05, "loss": 0.0193, "step": 13938 }, { "epoch": 9.795502459592411, "grad_norm": 0.08453802019357681, "learning_rate": 2.6805809323026468e-05, "loss": 0.0172, "step": 13939 }, { "epoch": 9.796205200281097, "grad_norm": 0.21670565009117126, "learning_rate": 2.6805340829234012e-05, "loss": 0.0192, "step": 13940 }, { "epoch": 9.796907940969781, "grad_norm": 0.30234357714653015, "learning_rate": 2.6804872335441556e-05, "loss": 0.0258, "step": 13941 }, { "epoch": 9.797610681658469, "grad_norm": 0.07578236609697342, "learning_rate": 2.68044038416491e-05, "loss": 0.0098, "step": 13942 }, { "epoch": 9.798313422347153, "grad_norm": 0.15458492934703827, "learning_rate": 2.680393534785664e-05, "loss": 0.0257, "step": 13943 }, { "epoch": 9.799016163035839, "grad_norm": 0.13734053075313568, "learning_rate": 2.6803466854064183e-05, "loss": 0.0204, "step": 13944 }, { "epoch": 9.799718903724525, "grad_norm": 0.25346213579177856, "learning_rate": 2.6802998360271727e-05, "loss": 0.031, "step": 13945 }, { "epoch": 9.80042164441321, "grad_norm": 0.15452025830745697, "learning_rate": 2.680252986647927e-05, "loss": 0.0153, "step": 13946 }, { "epoch": 9.801124385101897, "grad_norm": 0.18934482336044312, "learning_rate": 2.680206137268681e-05, "loss": 0.03, "step": 13947 }, { "epoch": 9.801827125790583, "grad_norm": 0.19839949905872345, "learning_rate": 2.6801592878894355e-05, "loss": 0.0382, "step": 13948 }, { "epoch": 9.802529866479269, "grad_norm": 0.1377178281545639, "learning_rate": 2.68011243851019e-05, "loss": 0.0229, "step": 13949 }, { "epoch": 9.803232607167955, "grad_norm": 0.14050088822841644, "learning_rate": 2.6800655891309442e-05, "loss": 0.0309, "step": 13950 }, { "epoch": 9.80393534785664, "grad_norm": 0.2575764060020447, "learning_rate": 2.6800187397516983e-05, "loss": 0.0522, "step": 13951 }, { "epoch": 9.804638088545326, "grad_norm": 0.19040155410766602, "learning_rate": 2.6799718903724526e-05, "loss": 0.0368, "step": 13952 }, { "epoch": 9.805340829234012, "grad_norm": 0.3325352966785431, "learning_rate": 2.679925040993207e-05, "loss": 0.0817, "step": 13953 }, { "epoch": 9.806043569922698, "grad_norm": 0.322055459022522, "learning_rate": 2.6798781916139614e-05, "loss": 0.1232, "step": 13954 }, { "epoch": 9.806746310611384, "grad_norm": 0.5815073251724243, "learning_rate": 2.6798313422347158e-05, "loss": 0.1472, "step": 13955 }, { "epoch": 9.80744905130007, "grad_norm": 0.7571860551834106, "learning_rate": 2.6797844928554698e-05, "loss": 0.1638, "step": 13956 }, { "epoch": 9.808151791988756, "grad_norm": 0.9339292049407959, "learning_rate": 2.679737643476224e-05, "loss": 0.193, "step": 13957 }, { "epoch": 9.808854532677442, "grad_norm": 0.2745918929576874, "learning_rate": 2.6796907940969782e-05, "loss": 0.0539, "step": 13958 }, { "epoch": 9.809557273366128, "grad_norm": 0.20507334172725677, "learning_rate": 2.6796439447177326e-05, "loss": 0.02, "step": 13959 }, { "epoch": 9.810260014054814, "grad_norm": 1.082309603691101, "learning_rate": 2.6795970953384866e-05, "loss": 0.0279, "step": 13960 }, { "epoch": 9.8109627547435, "grad_norm": 0.1371675431728363, "learning_rate": 2.679550245959241e-05, "loss": 0.0214, "step": 13961 }, { "epoch": 9.811665495432186, "grad_norm": 0.25188642740249634, "learning_rate": 2.6795033965799954e-05, "loss": 0.0202, "step": 13962 }, { "epoch": 9.812368236120872, "grad_norm": 0.09737219661474228, "learning_rate": 2.6794565472007497e-05, "loss": 0.0173, "step": 13963 }, { "epoch": 9.813070976809557, "grad_norm": 0.23522993922233582, "learning_rate": 2.6794096978215038e-05, "loss": 0.0225, "step": 13964 }, { "epoch": 9.813773717498243, "grad_norm": 0.3133547306060791, "learning_rate": 2.679362848442258e-05, "loss": 0.0307, "step": 13965 }, { "epoch": 9.81447645818693, "grad_norm": 0.4897475838661194, "learning_rate": 2.6793159990630125e-05, "loss": 0.022, "step": 13966 }, { "epoch": 9.815179198875615, "grad_norm": 0.09732068330049515, "learning_rate": 2.679269149683767e-05, "loss": 0.013, "step": 13967 }, { "epoch": 9.815881939564301, "grad_norm": 0.2529418468475342, "learning_rate": 2.6792223003045213e-05, "loss": 0.0346, "step": 13968 }, { "epoch": 9.816584680252987, "grad_norm": 0.13572928309440613, "learning_rate": 2.6791754509252753e-05, "loss": 0.0187, "step": 13969 }, { "epoch": 9.817287420941673, "grad_norm": 0.11636634171009064, "learning_rate": 2.6791286015460297e-05, "loss": 0.0253, "step": 13970 }, { "epoch": 9.817990161630359, "grad_norm": 0.08349436521530151, "learning_rate": 2.679081752166784e-05, "loss": 0.0133, "step": 13971 }, { "epoch": 9.818692902319045, "grad_norm": 0.1510942280292511, "learning_rate": 2.6790349027875384e-05, "loss": 0.0293, "step": 13972 }, { "epoch": 9.81939564300773, "grad_norm": 0.32199734449386597, "learning_rate": 2.6789880534082924e-05, "loss": 0.0483, "step": 13973 }, { "epoch": 9.820098383696417, "grad_norm": 0.11091427505016327, "learning_rate": 2.6789412040290465e-05, "loss": 0.0245, "step": 13974 }, { "epoch": 9.820801124385103, "grad_norm": 0.2635304033756256, "learning_rate": 2.678894354649801e-05, "loss": 0.0502, "step": 13975 }, { "epoch": 9.821503865073788, "grad_norm": 0.18787962198257446, "learning_rate": 2.6788475052705552e-05, "loss": 0.0303, "step": 13976 }, { "epoch": 9.822206605762474, "grad_norm": 0.20545673370361328, "learning_rate": 2.6788006558913092e-05, "loss": 0.0378, "step": 13977 }, { "epoch": 9.82290934645116, "grad_norm": 0.4234279990196228, "learning_rate": 2.6787538065120636e-05, "loss": 0.0715, "step": 13978 }, { "epoch": 9.823612087139846, "grad_norm": 0.7854675054550171, "learning_rate": 2.678706957132818e-05, "loss": 0.0934, "step": 13979 }, { "epoch": 9.82431482782853, "grad_norm": 0.6100966930389404, "learning_rate": 2.6786601077535724e-05, "loss": 0.1365, "step": 13980 }, { "epoch": 9.825017568517218, "grad_norm": 0.8587914109230042, "learning_rate": 2.6786132583743267e-05, "loss": 0.1846, "step": 13981 }, { "epoch": 9.825720309205902, "grad_norm": 0.9438179135322571, "learning_rate": 2.6785664089950808e-05, "loss": 0.1933, "step": 13982 }, { "epoch": 9.826423049894588, "grad_norm": 0.2772950232028961, "learning_rate": 2.678519559615835e-05, "loss": 0.0657, "step": 13983 }, { "epoch": 9.827125790583274, "grad_norm": 0.19456657767295837, "learning_rate": 2.6784727102365895e-05, "loss": 0.0205, "step": 13984 }, { "epoch": 9.82782853127196, "grad_norm": 0.17489223182201385, "learning_rate": 2.678425860857344e-05, "loss": 0.0288, "step": 13985 }, { "epoch": 9.828531271960646, "grad_norm": 0.14270950853824615, "learning_rate": 2.678379011478098e-05, "loss": 0.0228, "step": 13986 }, { "epoch": 9.829234012649332, "grad_norm": 0.2301284670829773, "learning_rate": 2.6783321620988523e-05, "loss": 0.0344, "step": 13987 }, { "epoch": 9.829936753338018, "grad_norm": 0.15725968778133392, "learning_rate": 2.6782853127196067e-05, "loss": 0.0149, "step": 13988 }, { "epoch": 9.830639494026704, "grad_norm": 0.07753939181566238, "learning_rate": 2.678238463340361e-05, "loss": 0.0119, "step": 13989 }, { "epoch": 9.83134223471539, "grad_norm": 0.12486239522695541, "learning_rate": 2.678191613961115e-05, "loss": 0.0208, "step": 13990 }, { "epoch": 9.832044975404076, "grad_norm": 0.1375054270029068, "learning_rate": 2.6781447645818694e-05, "loss": 0.0296, "step": 13991 }, { "epoch": 9.832747716092761, "grad_norm": 0.1745447814464569, "learning_rate": 2.6780979152026235e-05, "loss": 0.0191, "step": 13992 }, { "epoch": 9.833450456781447, "grad_norm": 0.12245136499404907, "learning_rate": 2.678051065823378e-05, "loss": 0.024, "step": 13993 }, { "epoch": 9.834153197470133, "grad_norm": 0.14567624032497406, "learning_rate": 2.6780042164441322e-05, "loss": 0.0168, "step": 13994 }, { "epoch": 9.83485593815882, "grad_norm": 0.08722609281539917, "learning_rate": 2.6779573670648863e-05, "loss": 0.0222, "step": 13995 }, { "epoch": 9.835558678847505, "grad_norm": 0.17487472295761108, "learning_rate": 2.6779105176856406e-05, "loss": 0.0284, "step": 13996 }, { "epoch": 9.836261419536191, "grad_norm": 0.19707976281642914, "learning_rate": 2.677863668306395e-05, "loss": 0.0221, "step": 13997 }, { "epoch": 9.836964160224877, "grad_norm": 0.12536923587322235, "learning_rate": 2.6778168189271494e-05, "loss": 0.0452, "step": 13998 }, { "epoch": 9.837666900913563, "grad_norm": 0.23294533789157867, "learning_rate": 2.6777699695479034e-05, "loss": 0.0172, "step": 13999 }, { "epoch": 9.838369641602249, "grad_norm": 0.21970771253108978, "learning_rate": 2.6777231201686578e-05, "loss": 0.0339, "step": 14000 }, { "epoch": 9.838369641602249, "eval_cer": 0.19327790688793603, "eval_loss": 0.2675686478614807, "eval_runtime": 18.3174, "eval_samples_per_second": 247.743, "eval_steps_per_second": 0.819, "eval_wer": 0.34342616128041115, "step": 14000 }, { "epoch": 9.839072382290935, "grad_norm": 0.20513921976089478, "learning_rate": 2.677676270789412e-05, "loss": 0.0572, "step": 14001 }, { "epoch": 9.83977512297962, "grad_norm": 0.7433884143829346, "learning_rate": 2.6776294214101665e-05, "loss": 0.0507, "step": 14002 }, { "epoch": 9.840477863668307, "grad_norm": 0.4333968162536621, "learning_rate": 2.6775825720309206e-05, "loss": 0.1334, "step": 14003 }, { "epoch": 9.841180604356992, "grad_norm": 0.3303965628147125, "learning_rate": 2.677535722651675e-05, "loss": 0.0965, "step": 14004 }, { "epoch": 9.841883345045678, "grad_norm": 0.44892874360084534, "learning_rate": 2.6774888732724293e-05, "loss": 0.1317, "step": 14005 }, { "epoch": 9.842586085734364, "grad_norm": 0.8225806951522827, "learning_rate": 2.6774420238931837e-05, "loss": 0.2072, "step": 14006 }, { "epoch": 9.84328882642305, "grad_norm": 1.1820791959762573, "learning_rate": 2.677395174513938e-05, "loss": 0.1995, "step": 14007 }, { "epoch": 9.843991567111736, "grad_norm": 0.1767168492078781, "learning_rate": 2.677348325134692e-05, "loss": 0.0643, "step": 14008 }, { "epoch": 9.844694307800422, "grad_norm": 0.15425801277160645, "learning_rate": 2.677301475755446e-05, "loss": 0.0275, "step": 14009 }, { "epoch": 9.845397048489108, "grad_norm": 0.11537112295627594, "learning_rate": 2.6772546263762005e-05, "loss": 0.0244, "step": 14010 }, { "epoch": 9.846099789177794, "grad_norm": 0.24645791947841644, "learning_rate": 2.677207776996955e-05, "loss": 0.0172, "step": 14011 }, { "epoch": 9.84680252986648, "grad_norm": 0.4021894335746765, "learning_rate": 2.677160927617709e-05, "loss": 0.0157, "step": 14012 }, { "epoch": 9.847505270555166, "grad_norm": 0.24212653934955597, "learning_rate": 2.6771140782384633e-05, "loss": 0.0395, "step": 14013 }, { "epoch": 9.848208011243852, "grad_norm": 0.143766388297081, "learning_rate": 2.6770672288592176e-05, "loss": 0.0116, "step": 14014 }, { "epoch": 9.848910751932538, "grad_norm": 0.5487292408943176, "learning_rate": 2.677020379479972e-05, "loss": 0.0208, "step": 14015 }, { "epoch": 9.849613492621224, "grad_norm": 0.15160436928272247, "learning_rate": 2.676973530100726e-05, "loss": 0.0226, "step": 14016 }, { "epoch": 9.85031623330991, "grad_norm": 0.11606188118457794, "learning_rate": 2.6769266807214804e-05, "loss": 0.0099, "step": 14017 }, { "epoch": 9.851018973998595, "grad_norm": 0.37812283635139465, "learning_rate": 2.6768798313422348e-05, "loss": 0.0266, "step": 14018 }, { "epoch": 9.85172171468728, "grad_norm": 0.14236722886562347, "learning_rate": 2.676832981962989e-05, "loss": 0.0143, "step": 14019 }, { "epoch": 9.852424455375965, "grad_norm": 0.14201204478740692, "learning_rate": 2.6767861325837435e-05, "loss": 0.0285, "step": 14020 }, { "epoch": 9.853127196064651, "grad_norm": 0.10434990376234055, "learning_rate": 2.6767392832044976e-05, "loss": 0.0112, "step": 14021 }, { "epoch": 9.853829936753337, "grad_norm": 0.1985095590353012, "learning_rate": 2.676692433825252e-05, "loss": 0.0266, "step": 14022 }, { "epoch": 9.854532677442023, "grad_norm": 0.14770662784576416, "learning_rate": 2.6766455844460063e-05, "loss": 0.0373, "step": 14023 }, { "epoch": 9.85523541813071, "grad_norm": 0.12218047678470612, "learning_rate": 2.6765987350667607e-05, "loss": 0.0143, "step": 14024 }, { "epoch": 9.855938158819395, "grad_norm": 0.26821327209472656, "learning_rate": 2.6765518856875147e-05, "loss": 0.0686, "step": 14025 }, { "epoch": 9.856640899508081, "grad_norm": 0.2860829830169678, "learning_rate": 2.676505036308269e-05, "loss": 0.0352, "step": 14026 }, { "epoch": 9.857343640196767, "grad_norm": 0.17115534842014313, "learning_rate": 2.676458186929023e-05, "loss": 0.0365, "step": 14027 }, { "epoch": 9.858046380885453, "grad_norm": 0.3385591208934784, "learning_rate": 2.6764113375497775e-05, "loss": 0.0871, "step": 14028 }, { "epoch": 9.858749121574139, "grad_norm": 0.4670141637325287, "learning_rate": 2.6763644881705315e-05, "loss": 0.125, "step": 14029 }, { "epoch": 9.859451862262825, "grad_norm": 0.4526801109313965, "learning_rate": 2.676317638791286e-05, "loss": 0.1484, "step": 14030 }, { "epoch": 9.86015460295151, "grad_norm": 0.5480131506919861, "learning_rate": 2.6762707894120403e-05, "loss": 0.1504, "step": 14031 }, { "epoch": 9.860857343640197, "grad_norm": 1.2524213790893555, "learning_rate": 2.6762239400327947e-05, "loss": 0.2067, "step": 14032 }, { "epoch": 9.861560084328882, "grad_norm": 0.18565143644809723, "learning_rate": 2.676177090653549e-05, "loss": 0.0577, "step": 14033 }, { "epoch": 9.862262825017568, "grad_norm": 0.11731710284948349, "learning_rate": 2.676130241274303e-05, "loss": 0.0239, "step": 14034 }, { "epoch": 9.862965565706254, "grad_norm": 0.12682434916496277, "learning_rate": 2.6760833918950574e-05, "loss": 0.0175, "step": 14035 }, { "epoch": 9.86366830639494, "grad_norm": 0.19355663657188416, "learning_rate": 2.6760365425158118e-05, "loss": 0.0157, "step": 14036 }, { "epoch": 9.864371047083626, "grad_norm": 0.10608836263418198, "learning_rate": 2.6759896931365662e-05, "loss": 0.0263, "step": 14037 }, { "epoch": 9.865073787772312, "grad_norm": 0.12839090824127197, "learning_rate": 2.6759428437573202e-05, "loss": 0.0146, "step": 14038 }, { "epoch": 9.865776528460998, "grad_norm": 0.12325411289930344, "learning_rate": 2.6758959943780746e-05, "loss": 0.0198, "step": 14039 }, { "epoch": 9.866479269149684, "grad_norm": 0.1184464618563652, "learning_rate": 2.675849144998829e-05, "loss": 0.0175, "step": 14040 }, { "epoch": 9.86718200983837, "grad_norm": 0.09176228940486908, "learning_rate": 2.6758022956195833e-05, "loss": 0.0139, "step": 14041 }, { "epoch": 9.867884750527056, "grad_norm": 0.12022705376148224, "learning_rate": 2.6757554462403374e-05, "loss": 0.0144, "step": 14042 }, { "epoch": 9.868587491215742, "grad_norm": 0.3319837749004364, "learning_rate": 2.6757085968610917e-05, "loss": 0.0159, "step": 14043 }, { "epoch": 9.869290231904428, "grad_norm": 0.11036588996648788, "learning_rate": 2.6756617474818458e-05, "loss": 0.0167, "step": 14044 }, { "epoch": 9.869992972593113, "grad_norm": 0.3814385235309601, "learning_rate": 2.6756148981026e-05, "loss": 0.0329, "step": 14045 }, { "epoch": 9.8706957132818, "grad_norm": 0.15224723517894745, "learning_rate": 2.6755680487233545e-05, "loss": 0.0147, "step": 14046 }, { "epoch": 9.871398453970485, "grad_norm": 0.12340209633111954, "learning_rate": 2.6755211993441085e-05, "loss": 0.0213, "step": 14047 }, { "epoch": 9.872101194659171, "grad_norm": 0.28701940178871155, "learning_rate": 2.675474349964863e-05, "loss": 0.0359, "step": 14048 }, { "epoch": 9.872803935347857, "grad_norm": 0.11421424150466919, "learning_rate": 2.6754275005856173e-05, "loss": 0.0161, "step": 14049 }, { "epoch": 9.873506676036543, "grad_norm": 0.18761301040649414, "learning_rate": 2.6753806512063717e-05, "loss": 0.0321, "step": 14050 }, { "epoch": 9.874209416725229, "grad_norm": 0.17143285274505615, "learning_rate": 2.6753338018271257e-05, "loss": 0.0343, "step": 14051 }, { "epoch": 9.874912157413915, "grad_norm": 0.23424780368804932, "learning_rate": 2.67528695244788e-05, "loss": 0.0611, "step": 14052 }, { "epoch": 9.8756148981026, "grad_norm": 0.28808069229125977, "learning_rate": 2.6752401030686344e-05, "loss": 0.0723, "step": 14053 }, { "epoch": 9.876317638791287, "grad_norm": 0.388190358877182, "learning_rate": 2.6751932536893888e-05, "loss": 0.1203, "step": 14054 }, { "epoch": 9.877020379479973, "grad_norm": 0.5243847370147705, "learning_rate": 2.675146404310143e-05, "loss": 0.127, "step": 14055 }, { "epoch": 9.877723120168657, "grad_norm": 0.5319430828094482, "learning_rate": 2.6750995549308972e-05, "loss": 0.1454, "step": 14056 }, { "epoch": 9.878425860857345, "grad_norm": 0.9023975729942322, "learning_rate": 2.6750527055516516e-05, "loss": 0.2104, "step": 14057 }, { "epoch": 9.879128601546029, "grad_norm": 0.1594804972410202, "learning_rate": 2.675005856172406e-05, "loss": 0.056, "step": 14058 }, { "epoch": 9.879831342234715, "grad_norm": 0.11623550951480865, "learning_rate": 2.6749590067931603e-05, "loss": 0.0158, "step": 14059 }, { "epoch": 9.8805340829234, "grad_norm": 0.20299604535102844, "learning_rate": 2.6749121574139144e-05, "loss": 0.0356, "step": 14060 }, { "epoch": 9.881236823612086, "grad_norm": 0.11238342523574829, "learning_rate": 2.6748653080346684e-05, "loss": 0.0132, "step": 14061 }, { "epoch": 9.881939564300772, "grad_norm": 0.13901278376579285, "learning_rate": 2.6748184586554228e-05, "loss": 0.0324, "step": 14062 }, { "epoch": 9.882642304989458, "grad_norm": 0.08591455221176147, "learning_rate": 2.674771609276177e-05, "loss": 0.0088, "step": 14063 }, { "epoch": 9.883345045678144, "grad_norm": 0.0864042192697525, "learning_rate": 2.6747247598969312e-05, "loss": 0.0165, "step": 14064 }, { "epoch": 9.88404778636683, "grad_norm": 0.11716686189174652, "learning_rate": 2.6746779105176856e-05, "loss": 0.0217, "step": 14065 }, { "epoch": 9.884750527055516, "grad_norm": 0.0728164091706276, "learning_rate": 2.67463106113844e-05, "loss": 0.0203, "step": 14066 }, { "epoch": 9.885453267744202, "grad_norm": 0.1331486999988556, "learning_rate": 2.6745842117591943e-05, "loss": 0.0304, "step": 14067 }, { "epoch": 9.886156008432888, "grad_norm": 0.13438287377357483, "learning_rate": 2.6745373623799483e-05, "loss": 0.0213, "step": 14068 }, { "epoch": 9.886858749121574, "grad_norm": 0.18967784941196442, "learning_rate": 2.6744905130007027e-05, "loss": 0.0088, "step": 14069 }, { "epoch": 9.88756148981026, "grad_norm": 0.2201228141784668, "learning_rate": 2.674443663621457e-05, "loss": 0.0218, "step": 14070 }, { "epoch": 9.888264230498946, "grad_norm": 0.10946758091449738, "learning_rate": 2.6743968142422115e-05, "loss": 0.0197, "step": 14071 }, { "epoch": 9.888966971187632, "grad_norm": 0.24493812024593353, "learning_rate": 2.6743499648629658e-05, "loss": 0.0353, "step": 14072 }, { "epoch": 9.889669711876317, "grad_norm": 0.13437721133232117, "learning_rate": 2.67430311548372e-05, "loss": 0.0314, "step": 14073 }, { "epoch": 9.890372452565003, "grad_norm": 0.10806465148925781, "learning_rate": 2.6742562661044742e-05, "loss": 0.0267, "step": 14074 }, { "epoch": 9.89107519325369, "grad_norm": 0.18492716550827026, "learning_rate": 2.6742094167252286e-05, "loss": 0.0287, "step": 14075 }, { "epoch": 9.891777933942375, "grad_norm": 0.31060680747032166, "learning_rate": 2.674162567345983e-05, "loss": 0.0479, "step": 14076 }, { "epoch": 9.892480674631061, "grad_norm": 0.39876359701156616, "learning_rate": 2.674115717966737e-05, "loss": 0.0822, "step": 14077 }, { "epoch": 9.893183415319747, "grad_norm": 0.259841650724411, "learning_rate": 2.6740688685874914e-05, "loss": 0.0587, "step": 14078 }, { "epoch": 9.893886156008433, "grad_norm": 0.4316580593585968, "learning_rate": 2.6740220192082454e-05, "loss": 0.104, "step": 14079 }, { "epoch": 9.894588896697119, "grad_norm": 0.6349793076515198, "learning_rate": 2.6739751698289998e-05, "loss": 0.144, "step": 14080 }, { "epoch": 9.895291637385805, "grad_norm": 1.2283817529678345, "learning_rate": 2.6739283204497538e-05, "loss": 0.2147, "step": 14081 }, { "epoch": 9.89599437807449, "grad_norm": 1.1433286666870117, "learning_rate": 2.6738814710705082e-05, "loss": 0.1968, "step": 14082 }, { "epoch": 9.896697118763177, "grad_norm": 0.33472007513046265, "learning_rate": 2.6738346216912626e-05, "loss": 0.0721, "step": 14083 }, { "epoch": 9.897399859451863, "grad_norm": 0.26545414328575134, "learning_rate": 2.673787772312017e-05, "loss": 0.027, "step": 14084 }, { "epoch": 9.898102600140549, "grad_norm": 0.08118689060211182, "learning_rate": 2.6737409229327713e-05, "loss": 0.0177, "step": 14085 }, { "epoch": 9.898805340829234, "grad_norm": 0.0954945906996727, "learning_rate": 2.6736940735535254e-05, "loss": 0.0171, "step": 14086 }, { "epoch": 9.89950808151792, "grad_norm": 0.14744803309440613, "learning_rate": 2.6736472241742797e-05, "loss": 0.0129, "step": 14087 }, { "epoch": 9.900210822206606, "grad_norm": 0.12302379310131073, "learning_rate": 2.673600374795034e-05, "loss": 0.0181, "step": 14088 }, { "epoch": 9.900913562895292, "grad_norm": 0.23294614255428314, "learning_rate": 2.6735535254157885e-05, "loss": 0.0164, "step": 14089 }, { "epoch": 9.901616303583978, "grad_norm": 0.11888010054826736, "learning_rate": 2.6735066760365425e-05, "loss": 0.0124, "step": 14090 }, { "epoch": 9.902319044272664, "grad_norm": 0.12936186790466309, "learning_rate": 2.673459826657297e-05, "loss": 0.0182, "step": 14091 }, { "epoch": 9.90302178496135, "grad_norm": 0.11790244281291962, "learning_rate": 2.6734129772780512e-05, "loss": 0.0135, "step": 14092 }, { "epoch": 9.903724525650036, "grad_norm": 0.14910128712654114, "learning_rate": 2.6733661278988056e-05, "loss": 0.0177, "step": 14093 }, { "epoch": 9.904427266338722, "grad_norm": 0.11550956219434738, "learning_rate": 2.6733192785195597e-05, "loss": 0.0112, "step": 14094 }, { "epoch": 9.905130007027406, "grad_norm": 0.14756788313388824, "learning_rate": 2.673272429140314e-05, "loss": 0.0373, "step": 14095 }, { "epoch": 9.905832747716094, "grad_norm": 0.12916556000709534, "learning_rate": 2.673225579761068e-05, "loss": 0.0213, "step": 14096 }, { "epoch": 9.906535488404778, "grad_norm": 0.12100288271903992, "learning_rate": 2.6731787303818224e-05, "loss": 0.0245, "step": 14097 }, { "epoch": 9.907238229093464, "grad_norm": 0.1185087189078331, "learning_rate": 2.6731318810025768e-05, "loss": 0.0292, "step": 14098 }, { "epoch": 9.90794096978215, "grad_norm": 0.16902682185173035, "learning_rate": 2.673085031623331e-05, "loss": 0.0168, "step": 14099 }, { "epoch": 9.908643710470836, "grad_norm": 0.22745448350906372, "learning_rate": 2.6730381822440852e-05, "loss": 0.0415, "step": 14100 }, { "epoch": 9.909346451159522, "grad_norm": 0.1650567650794983, "learning_rate": 2.6729913328648396e-05, "loss": 0.0352, "step": 14101 }, { "epoch": 9.910049191848207, "grad_norm": 0.14981739223003387, "learning_rate": 2.672944483485594e-05, "loss": 0.0428, "step": 14102 }, { "epoch": 9.910751932536893, "grad_norm": 0.2977166771888733, "learning_rate": 2.672897634106348e-05, "loss": 0.0612, "step": 14103 }, { "epoch": 9.91145467322558, "grad_norm": 0.39500120282173157, "learning_rate": 2.6728507847271024e-05, "loss": 0.115, "step": 14104 }, { "epoch": 9.912157413914265, "grad_norm": 0.4133642315864563, "learning_rate": 2.6728039353478567e-05, "loss": 0.1687, "step": 14105 }, { "epoch": 9.912860154602951, "grad_norm": 0.46248674392700195, "learning_rate": 2.672757085968611e-05, "loss": 0.1639, "step": 14106 }, { "epoch": 9.913562895291637, "grad_norm": 0.8132138252258301, "learning_rate": 2.672710236589365e-05, "loss": 0.1986, "step": 14107 }, { "epoch": 9.914265635980323, "grad_norm": 0.18330402672290802, "learning_rate": 2.6726633872101195e-05, "loss": 0.0592, "step": 14108 }, { "epoch": 9.914968376669009, "grad_norm": 0.2080662101507187, "learning_rate": 2.672616537830874e-05, "loss": 0.0341, "step": 14109 }, { "epoch": 9.915671117357695, "grad_norm": 0.10876841098070145, "learning_rate": 2.6725696884516283e-05, "loss": 0.0226, "step": 14110 }, { "epoch": 9.91637385804638, "grad_norm": 0.22397036850452423, "learning_rate": 2.6725228390723826e-05, "loss": 0.0237, "step": 14111 }, { "epoch": 9.917076598735067, "grad_norm": 0.1432296484708786, "learning_rate": 2.6724759896931367e-05, "loss": 0.0247, "step": 14112 }, { "epoch": 9.917779339423753, "grad_norm": 0.11190640926361084, "learning_rate": 2.672429140313891e-05, "loss": 0.0146, "step": 14113 }, { "epoch": 9.918482080112438, "grad_norm": 0.10246257483959198, "learning_rate": 2.672382290934645e-05, "loss": 0.014, "step": 14114 }, { "epoch": 9.919184820801124, "grad_norm": 0.15349550545215607, "learning_rate": 2.6723354415553994e-05, "loss": 0.0114, "step": 14115 }, { "epoch": 9.91988756148981, "grad_norm": 0.18937399983406067, "learning_rate": 2.6722885921761535e-05, "loss": 0.0176, "step": 14116 }, { "epoch": 9.920590302178496, "grad_norm": 0.2147834151983261, "learning_rate": 2.672241742796908e-05, "loss": 0.0265, "step": 14117 }, { "epoch": 9.921293042867182, "grad_norm": 0.27704885601997375, "learning_rate": 2.6721948934176622e-05, "loss": 0.0403, "step": 14118 }, { "epoch": 9.921995783555868, "grad_norm": 0.17403829097747803, "learning_rate": 2.6721480440384166e-05, "loss": 0.0123, "step": 14119 }, { "epoch": 9.922698524244554, "grad_norm": 0.3168219327926636, "learning_rate": 2.672101194659171e-05, "loss": 0.0352, "step": 14120 }, { "epoch": 9.92340126493324, "grad_norm": 0.19542677700519562, "learning_rate": 2.672054345279925e-05, "loss": 0.0148, "step": 14121 }, { "epoch": 9.924104005621926, "grad_norm": 0.19752222299575806, "learning_rate": 2.6720074959006794e-05, "loss": 0.035, "step": 14122 }, { "epoch": 9.924806746310612, "grad_norm": 0.19518426060676575, "learning_rate": 2.6719606465214337e-05, "loss": 0.0321, "step": 14123 }, { "epoch": 9.925509486999298, "grad_norm": 0.14716939628124237, "learning_rate": 2.671913797142188e-05, "loss": 0.0227, "step": 14124 }, { "epoch": 9.926212227687984, "grad_norm": 0.19195839762687683, "learning_rate": 2.671866947762942e-05, "loss": 0.0307, "step": 14125 }, { "epoch": 9.92691496837667, "grad_norm": 0.16411729156970978, "learning_rate": 2.6718200983836965e-05, "loss": 0.044, "step": 14126 }, { "epoch": 9.927617709065355, "grad_norm": 0.30847442150115967, "learning_rate": 2.671773249004451e-05, "loss": 0.0615, "step": 14127 }, { "epoch": 9.928320449754041, "grad_norm": 0.28283384442329407, "learning_rate": 2.6717263996252053e-05, "loss": 0.0681, "step": 14128 }, { "epoch": 9.929023190442727, "grad_norm": 0.300422340631485, "learning_rate": 2.6716795502459593e-05, "loss": 0.0852, "step": 14129 }, { "epoch": 9.929725931131413, "grad_norm": 0.37851086258888245, "learning_rate": 2.6716327008667137e-05, "loss": 0.1157, "step": 14130 }, { "epoch": 9.9304286718201, "grad_norm": 0.6249216198921204, "learning_rate": 2.6715858514874677e-05, "loss": 0.1679, "step": 14131 }, { "epoch": 9.931131412508785, "grad_norm": 1.2789313793182373, "learning_rate": 2.671539002108222e-05, "loss": 0.2235, "step": 14132 }, { "epoch": 9.931834153197471, "grad_norm": 0.1954975426197052, "learning_rate": 2.6714921527289765e-05, "loss": 0.0662, "step": 14133 }, { "epoch": 9.932536893886155, "grad_norm": 0.1869889348745346, "learning_rate": 2.6714453033497305e-05, "loss": 0.0285, "step": 14134 }, { "epoch": 9.933239634574843, "grad_norm": 0.2506960928440094, "learning_rate": 2.671398453970485e-05, "loss": 0.0151, "step": 14135 }, { "epoch": 9.933942375263527, "grad_norm": 0.41634324193000793, "learning_rate": 2.6713516045912392e-05, "loss": 0.0174, "step": 14136 }, { "epoch": 9.934645115952213, "grad_norm": 0.12835636734962463, "learning_rate": 2.6713047552119936e-05, "loss": 0.0118, "step": 14137 }, { "epoch": 9.935347856640899, "grad_norm": 0.12613840401172638, "learning_rate": 2.6712579058327476e-05, "loss": 0.0137, "step": 14138 }, { "epoch": 9.936050597329585, "grad_norm": 0.16552886366844177, "learning_rate": 2.671211056453502e-05, "loss": 0.0284, "step": 14139 }, { "epoch": 9.93675333801827, "grad_norm": 0.08504678308963776, "learning_rate": 2.6711642070742564e-05, "loss": 0.0106, "step": 14140 }, { "epoch": 9.937456078706957, "grad_norm": 0.13240742683410645, "learning_rate": 2.6711173576950108e-05, "loss": 0.0185, "step": 14141 }, { "epoch": 9.938158819395642, "grad_norm": 0.11631486564874649, "learning_rate": 2.6710705083157648e-05, "loss": 0.0138, "step": 14142 }, { "epoch": 9.938861560084328, "grad_norm": 0.09177711606025696, "learning_rate": 2.671023658936519e-05, "loss": 0.0192, "step": 14143 }, { "epoch": 9.939564300773014, "grad_norm": 0.0855976939201355, "learning_rate": 2.6709768095572735e-05, "loss": 0.0137, "step": 14144 }, { "epoch": 9.9402670414617, "grad_norm": 0.16943348944187164, "learning_rate": 2.670929960178028e-05, "loss": 0.0356, "step": 14145 }, { "epoch": 9.940969782150386, "grad_norm": 0.18764081597328186, "learning_rate": 2.6708831107987823e-05, "loss": 0.0207, "step": 14146 }, { "epoch": 9.941672522839072, "grad_norm": 0.125188410282135, "learning_rate": 2.6708362614195363e-05, "loss": 0.0268, "step": 14147 }, { "epoch": 9.942375263527758, "grad_norm": 0.2764196991920471, "learning_rate": 2.6707894120402903e-05, "loss": 0.0614, "step": 14148 }, { "epoch": 9.943078004216444, "grad_norm": 0.2546461820602417, "learning_rate": 2.6707425626610447e-05, "loss": 0.0264, "step": 14149 }, { "epoch": 9.94378074490513, "grad_norm": 0.20588532090187073, "learning_rate": 2.670695713281799e-05, "loss": 0.0381, "step": 14150 }, { "epoch": 9.944483485593816, "grad_norm": 0.5306991338729858, "learning_rate": 2.670648863902553e-05, "loss": 0.0536, "step": 14151 }, { "epoch": 9.945186226282502, "grad_norm": 0.40519487857818604, "learning_rate": 2.6706020145233075e-05, "loss": 0.0571, "step": 14152 }, { "epoch": 9.945888966971188, "grad_norm": 0.25529903173446655, "learning_rate": 2.670555165144062e-05, "loss": 0.0722, "step": 14153 }, { "epoch": 9.946591707659874, "grad_norm": 0.525456964969635, "learning_rate": 2.6705083157648162e-05, "loss": 0.1026, "step": 14154 }, { "epoch": 9.94729444834856, "grad_norm": 0.4765070974826813, "learning_rate": 2.6704614663855703e-05, "loss": 0.1415, "step": 14155 }, { "epoch": 9.947997189037245, "grad_norm": 6.672628402709961, "learning_rate": 2.6704146170063247e-05, "loss": 0.1694, "step": 14156 }, { "epoch": 9.948699929725931, "grad_norm": 0.9849568009376526, "learning_rate": 2.670367767627079e-05, "loss": 0.1959, "step": 14157 }, { "epoch": 9.949402670414617, "grad_norm": 0.20210140943527222, "learning_rate": 2.6703209182478334e-05, "loss": 0.0639, "step": 14158 }, { "epoch": 9.950105411103303, "grad_norm": 0.16535021364688873, "learning_rate": 2.6702740688685878e-05, "loss": 0.0258, "step": 14159 }, { "epoch": 9.950808151791989, "grad_norm": 0.33612045645713806, "learning_rate": 2.6702272194893418e-05, "loss": 0.0227, "step": 14160 }, { "epoch": 9.951510892480675, "grad_norm": 0.5200413465499878, "learning_rate": 2.6701803701100962e-05, "loss": 0.0419, "step": 14161 }, { "epoch": 9.952213633169361, "grad_norm": 0.2628518342971802, "learning_rate": 2.6701335207308505e-05, "loss": 0.0255, "step": 14162 }, { "epoch": 9.952916373858047, "grad_norm": 0.20627856254577637, "learning_rate": 2.670086671351605e-05, "loss": 0.0295, "step": 14163 }, { "epoch": 9.953619114546733, "grad_norm": 0.14360842108726501, "learning_rate": 2.670039821972359e-05, "loss": 0.021, "step": 14164 }, { "epoch": 9.954321855235419, "grad_norm": 0.1603914499282837, "learning_rate": 2.6699929725931133e-05, "loss": 0.0079, "step": 14165 }, { "epoch": 9.955024595924105, "grad_norm": 0.1624365895986557, "learning_rate": 2.6699461232138674e-05, "loss": 0.0203, "step": 14166 }, { "epoch": 9.95572733661279, "grad_norm": 0.07607703655958176, "learning_rate": 2.6698992738346217e-05, "loss": 0.0109, "step": 14167 }, { "epoch": 9.956430077301476, "grad_norm": 0.170474573969841, "learning_rate": 2.6698524244553758e-05, "loss": 0.0204, "step": 14168 }, { "epoch": 9.957132817990162, "grad_norm": 0.1910034716129303, "learning_rate": 2.66980557507613e-05, "loss": 0.0131, "step": 14169 }, { "epoch": 9.957835558678848, "grad_norm": 0.14782588183879852, "learning_rate": 2.6697587256968845e-05, "loss": 0.0224, "step": 14170 }, { "epoch": 9.958538299367534, "grad_norm": 0.1446370780467987, "learning_rate": 2.669711876317639e-05, "loss": 0.0163, "step": 14171 }, { "epoch": 9.95924104005622, "grad_norm": 0.18090303242206573, "learning_rate": 2.6696650269383933e-05, "loss": 0.0267, "step": 14172 }, { "epoch": 9.959943780744904, "grad_norm": 0.6206932067871094, "learning_rate": 2.6696181775591473e-05, "loss": 0.0316, "step": 14173 }, { "epoch": 9.96064652143359, "grad_norm": 0.7143435478210449, "learning_rate": 2.6695713281799017e-05, "loss": 0.0191, "step": 14174 }, { "epoch": 9.961349262122276, "grad_norm": 0.15262165665626526, "learning_rate": 2.669524478800656e-05, "loss": 0.0429, "step": 14175 }, { "epoch": 9.962052002810962, "grad_norm": 0.4399622678756714, "learning_rate": 2.6694776294214104e-05, "loss": 0.0463, "step": 14176 }, { "epoch": 9.962754743499648, "grad_norm": 0.26591378450393677, "learning_rate": 2.6694307800421644e-05, "loss": 0.0523, "step": 14177 }, { "epoch": 9.963457484188334, "grad_norm": 0.3852545917034149, "learning_rate": 2.6693839306629188e-05, "loss": 0.0647, "step": 14178 }, { "epoch": 9.96416022487702, "grad_norm": 0.4369857907295227, "learning_rate": 2.6693370812836732e-05, "loss": 0.0974, "step": 14179 }, { "epoch": 9.964862965565706, "grad_norm": 0.5822588801383972, "learning_rate": 2.6692902319044276e-05, "loss": 0.1519, "step": 14180 }, { "epoch": 9.965565706254392, "grad_norm": 0.6051215529441833, "learning_rate": 2.6692433825251816e-05, "loss": 0.1848, "step": 14181 }, { "epoch": 9.966268446943078, "grad_norm": 1.0884592533111572, "learning_rate": 2.669196533145936e-05, "loss": 0.1898, "step": 14182 }, { "epoch": 9.966971187631763, "grad_norm": 0.21626916527748108, "learning_rate": 2.66914968376669e-05, "loss": 0.0666, "step": 14183 }, { "epoch": 9.96767392832045, "grad_norm": 0.3101111352443695, "learning_rate": 2.6691028343874444e-05, "loss": 0.0311, "step": 14184 }, { "epoch": 9.968376669009135, "grad_norm": 0.13736125826835632, "learning_rate": 2.6690559850081987e-05, "loss": 0.0261, "step": 14185 }, { "epoch": 9.969079409697821, "grad_norm": 0.11252866685390472, "learning_rate": 2.6690091356289528e-05, "loss": 0.0123, "step": 14186 }, { "epoch": 9.969782150386507, "grad_norm": 0.1543935388326645, "learning_rate": 2.668962286249707e-05, "loss": 0.0116, "step": 14187 }, { "epoch": 9.970484891075193, "grad_norm": 0.08297087997198105, "learning_rate": 2.6689154368704615e-05, "loss": 0.0109, "step": 14188 }, { "epoch": 9.971187631763879, "grad_norm": 0.1103399395942688, "learning_rate": 2.668868587491216e-05, "loss": 0.0189, "step": 14189 }, { "epoch": 9.971890372452565, "grad_norm": 0.09928370267152786, "learning_rate": 2.66882173811197e-05, "loss": 0.0216, "step": 14190 }, { "epoch": 9.97259311314125, "grad_norm": 0.08736930042505264, "learning_rate": 2.6687748887327243e-05, "loss": 0.0274, "step": 14191 }, { "epoch": 9.973295853829937, "grad_norm": 0.10452046245336533, "learning_rate": 2.6687280393534787e-05, "loss": 0.0327, "step": 14192 }, { "epoch": 9.973998594518623, "grad_norm": 0.1009252741932869, "learning_rate": 2.668681189974233e-05, "loss": 0.0152, "step": 14193 }, { "epoch": 9.974701335207309, "grad_norm": 0.11723939329385757, "learning_rate": 2.668634340594987e-05, "loss": 0.0116, "step": 14194 }, { "epoch": 9.975404075895995, "grad_norm": 0.16588285565376282, "learning_rate": 2.6685874912157415e-05, "loss": 0.0272, "step": 14195 }, { "epoch": 9.97610681658468, "grad_norm": 0.11962948739528656, "learning_rate": 2.6685406418364958e-05, "loss": 0.0157, "step": 14196 }, { "epoch": 9.976809557273366, "grad_norm": 0.1758141666650772, "learning_rate": 2.6684937924572502e-05, "loss": 0.0305, "step": 14197 }, { "epoch": 9.977512297962052, "grad_norm": 0.1235964298248291, "learning_rate": 2.6684469430780046e-05, "loss": 0.0306, "step": 14198 }, { "epoch": 9.978215038650738, "grad_norm": 0.09671978652477264, "learning_rate": 2.6684000936987586e-05, "loss": 0.0209, "step": 14199 }, { "epoch": 9.978917779339424, "grad_norm": 0.2186242640018463, "learning_rate": 2.668353244319513e-05, "loss": 0.0333, "step": 14200 }, { "epoch": 9.97962052002811, "grad_norm": 0.13622091710567474, "learning_rate": 2.668306394940267e-05, "loss": 0.0315, "step": 14201 }, { "epoch": 9.980323260716796, "grad_norm": 0.4964693784713745, "learning_rate": 2.6682595455610214e-05, "loss": 0.0651, "step": 14202 }, { "epoch": 9.981026001405482, "grad_norm": 0.2769075930118561, "learning_rate": 2.6682126961817754e-05, "loss": 0.0768, "step": 14203 }, { "epoch": 9.981728742094168, "grad_norm": 0.3101128339767456, "learning_rate": 2.6681658468025298e-05, "loss": 0.095, "step": 14204 }, { "epoch": 9.982431482782854, "grad_norm": 0.5877214074134827, "learning_rate": 2.668118997423284e-05, "loss": 0.1485, "step": 14205 }, { "epoch": 9.98313422347154, "grad_norm": 0.7825931310653687, "learning_rate": 2.6680721480440385e-05, "loss": 0.1887, "step": 14206 }, { "epoch": 9.983836964160226, "grad_norm": 2.460118055343628, "learning_rate": 2.6680252986647926e-05, "loss": 0.2425, "step": 14207 }, { "epoch": 9.984539704848912, "grad_norm": 0.1572912335395813, "learning_rate": 2.667978449285547e-05, "loss": 0.0596, "step": 14208 }, { "epoch": 9.985242445537597, "grad_norm": 0.12243127822875977, "learning_rate": 2.6679315999063013e-05, "loss": 0.0147, "step": 14209 }, { "epoch": 9.985945186226282, "grad_norm": 0.17314094305038452, "learning_rate": 2.6678847505270557e-05, "loss": 0.0219, "step": 14210 }, { "epoch": 9.98664792691497, "grad_norm": 0.08645942807197571, "learning_rate": 2.66783790114781e-05, "loss": 0.0141, "step": 14211 }, { "epoch": 9.987350667603653, "grad_norm": 0.18373775482177734, "learning_rate": 2.667791051768564e-05, "loss": 0.0077, "step": 14212 }, { "epoch": 9.98805340829234, "grad_norm": 0.1484123170375824, "learning_rate": 2.6677442023893185e-05, "loss": 0.0217, "step": 14213 }, { "epoch": 9.988756148981025, "grad_norm": 0.11747384816408157, "learning_rate": 2.667697353010073e-05, "loss": 0.0165, "step": 14214 }, { "epoch": 9.989458889669711, "grad_norm": 0.16074611246585846, "learning_rate": 2.6676505036308272e-05, "loss": 0.0128, "step": 14215 }, { "epoch": 9.990161630358397, "grad_norm": 0.1323394924402237, "learning_rate": 2.6676036542515812e-05, "loss": 0.0251, "step": 14216 }, { "epoch": 9.990864371047083, "grad_norm": 0.15947258472442627, "learning_rate": 2.6675568048723356e-05, "loss": 0.0254, "step": 14217 }, { "epoch": 9.991567111735769, "grad_norm": 0.9425711631774902, "learning_rate": 2.6675099554930896e-05, "loss": 0.0154, "step": 14218 }, { "epoch": 9.992269852424455, "grad_norm": 0.14589132368564606, "learning_rate": 2.667463106113844e-05, "loss": 0.0323, "step": 14219 }, { "epoch": 9.99297259311314, "grad_norm": 0.11467969417572021, "learning_rate": 2.667416256734598e-05, "loss": 0.018, "step": 14220 }, { "epoch": 9.993675333801827, "grad_norm": 0.24990308284759521, "learning_rate": 2.6673694073553524e-05, "loss": 0.0339, "step": 14221 }, { "epoch": 9.994378074490513, "grad_norm": 0.08499709516763687, "learning_rate": 2.6673225579761068e-05, "loss": 0.0138, "step": 14222 }, { "epoch": 9.995080815179199, "grad_norm": 0.34664788842201233, "learning_rate": 2.6672757085968612e-05, "loss": 0.0449, "step": 14223 }, { "epoch": 9.995783555867884, "grad_norm": 0.357835054397583, "learning_rate": 2.6672288592176155e-05, "loss": 0.0261, "step": 14224 }, { "epoch": 9.99648629655657, "grad_norm": 0.2267388105392456, "learning_rate": 2.6671820098383696e-05, "loss": 0.0484, "step": 14225 }, { "epoch": 9.997189037245256, "grad_norm": 0.4030340611934662, "learning_rate": 2.667135160459124e-05, "loss": 0.0833, "step": 14226 }, { "epoch": 9.997891777933942, "grad_norm": 0.47991231083869934, "learning_rate": 2.6670883110798783e-05, "loss": 0.1088, "step": 14227 }, { "epoch": 9.998594518622628, "grad_norm": 0.545387327671051, "learning_rate": 2.6670414617006327e-05, "loss": 0.1635, "step": 14228 }, { "epoch": 9.999297259311314, "grad_norm": 6.7064924240112305, "learning_rate": 2.6669946123213867e-05, "loss": 0.18, "step": 14229 }, { "epoch": 10.0, "grad_norm": 1.5373810529708862, "learning_rate": 2.666947762942141e-05, "loss": 0.1483, "step": 14230 }, { "epoch": 10.000702740688686, "grad_norm": 0.3200509548187256, "learning_rate": 2.6669009135628955e-05, "loss": 0.0571, "step": 14231 }, { "epoch": 10.001405481377372, "grad_norm": 0.24248827993869781, "learning_rate": 2.66685406418365e-05, "loss": 0.0146, "step": 14232 }, { "epoch": 10.002108222066058, "grad_norm": 0.1071377620100975, "learning_rate": 2.666807214804404e-05, "loss": 0.0204, "step": 14233 }, { "epoch": 10.002810962754744, "grad_norm": 0.1271190494298935, "learning_rate": 2.6667603654251583e-05, "loss": 0.02, "step": 14234 }, { "epoch": 10.00351370344343, "grad_norm": 0.0798720195889473, "learning_rate": 2.6667135160459126e-05, "loss": 0.0169, "step": 14235 }, { "epoch": 10.004216444132116, "grad_norm": 0.1848250776529312, "learning_rate": 2.6666666666666667e-05, "loss": 0.0184, "step": 14236 }, { "epoch": 10.004919184820801, "grad_norm": 0.08533463627099991, "learning_rate": 2.666619817287421e-05, "loss": 0.0067, "step": 14237 }, { "epoch": 10.005621925509487, "grad_norm": 0.10446705669164658, "learning_rate": 2.666572967908175e-05, "loss": 0.0195, "step": 14238 }, { "epoch": 10.006324666198173, "grad_norm": 0.16350825130939484, "learning_rate": 2.6665261185289294e-05, "loss": 0.0213, "step": 14239 }, { "epoch": 10.00702740688686, "grad_norm": 0.18003618717193604, "learning_rate": 2.6664792691496838e-05, "loss": 0.0179, "step": 14240 }, { "epoch": 10.007730147575545, "grad_norm": 0.15848521888256073, "learning_rate": 2.6664324197704382e-05, "loss": 0.0227, "step": 14241 }, { "epoch": 10.008432888264231, "grad_norm": 0.3860994875431061, "learning_rate": 2.6663855703911922e-05, "loss": 0.019, "step": 14242 }, { "epoch": 10.009135628952917, "grad_norm": 0.1611858308315277, "learning_rate": 2.6663387210119466e-05, "loss": 0.025, "step": 14243 }, { "epoch": 10.009838369641603, "grad_norm": 0.15851688385009766, "learning_rate": 2.666291871632701e-05, "loss": 0.0169, "step": 14244 }, { "epoch": 10.010541110330289, "grad_norm": 0.2181805521249771, "learning_rate": 2.6662450222534553e-05, "loss": 0.0311, "step": 14245 }, { "epoch": 10.011243851018975, "grad_norm": 0.23156140744686127, "learning_rate": 2.6661981728742094e-05, "loss": 0.0382, "step": 14246 }, { "epoch": 10.01194659170766, "grad_norm": 0.20822085440158844, "learning_rate": 2.6661513234949637e-05, "loss": 0.0201, "step": 14247 }, { "epoch": 10.012649332396347, "grad_norm": 0.25533148646354675, "learning_rate": 2.666104474115718e-05, "loss": 0.0496, "step": 14248 }, { "epoch": 10.013352073085033, "grad_norm": 0.27985936403274536, "learning_rate": 2.6660576247364725e-05, "loss": 0.0464, "step": 14249 }, { "epoch": 10.014054813773717, "grad_norm": 0.21959243714809418, "learning_rate": 2.666010775357227e-05, "loss": 0.0497, "step": 14250 }, { "epoch": 10.014757554462403, "grad_norm": 0.316466748714447, "learning_rate": 2.665963925977981e-05, "loss": 0.0922, "step": 14251 }, { "epoch": 10.015460295151088, "grad_norm": 0.48529714345932007, "learning_rate": 2.6659170765987353e-05, "loss": 0.1177, "step": 14252 }, { "epoch": 10.016163035839774, "grad_norm": 0.9194666147232056, "learning_rate": 2.6658702272194893e-05, "loss": 0.1484, "step": 14253 }, { "epoch": 10.01686577652846, "grad_norm": 0.6638946533203125, "learning_rate": 2.6658233778402437e-05, "loss": 0.2098, "step": 14254 }, { "epoch": 10.017568517217146, "grad_norm": 0.964684247970581, "learning_rate": 2.6657765284609977e-05, "loss": 0.2047, "step": 14255 }, { "epoch": 10.018271257905832, "grad_norm": 0.2367137223482132, "learning_rate": 2.665729679081752e-05, "loss": 0.0654, "step": 14256 }, { "epoch": 10.018973998594518, "grad_norm": 0.14181353151798248, "learning_rate": 2.6656828297025064e-05, "loss": 0.0254, "step": 14257 }, { "epoch": 10.019676739283204, "grad_norm": 0.1079535186290741, "learning_rate": 2.6656359803232608e-05, "loss": 0.0258, "step": 14258 }, { "epoch": 10.02037947997189, "grad_norm": 0.12036968022584915, "learning_rate": 2.665589130944015e-05, "loss": 0.0147, "step": 14259 }, { "epoch": 10.021082220660576, "grad_norm": 0.1653292030096054, "learning_rate": 2.6655422815647692e-05, "loss": 0.0197, "step": 14260 }, { "epoch": 10.021784961349262, "grad_norm": 0.24487122893333435, "learning_rate": 2.6654954321855236e-05, "loss": 0.0117, "step": 14261 }, { "epoch": 10.022487702037948, "grad_norm": 0.11711530387401581, "learning_rate": 2.665448582806278e-05, "loss": 0.0122, "step": 14262 }, { "epoch": 10.023190442726634, "grad_norm": 0.11724474281072617, "learning_rate": 2.6654017334270323e-05, "loss": 0.0132, "step": 14263 }, { "epoch": 10.02389318341532, "grad_norm": 0.19315233826637268, "learning_rate": 2.6653548840477864e-05, "loss": 0.0233, "step": 14264 }, { "epoch": 10.024595924104005, "grad_norm": 0.07543572038412094, "learning_rate": 2.6653080346685408e-05, "loss": 0.0127, "step": 14265 }, { "epoch": 10.025298664792691, "grad_norm": 0.20423796772956848, "learning_rate": 2.665261185289295e-05, "loss": 0.0248, "step": 14266 }, { "epoch": 10.026001405481377, "grad_norm": 0.08471743017435074, "learning_rate": 2.6652143359100495e-05, "loss": 0.0099, "step": 14267 }, { "epoch": 10.026704146170063, "grad_norm": 0.0797935202717781, "learning_rate": 2.6651674865308035e-05, "loss": 0.0196, "step": 14268 }, { "epoch": 10.02740688685875, "grad_norm": 0.48245903849601746, "learning_rate": 2.665120637151558e-05, "loss": 0.0178, "step": 14269 }, { "epoch": 10.028109627547435, "grad_norm": 0.4418771266937256, "learning_rate": 2.665073787772312e-05, "loss": 0.0234, "step": 14270 }, { "epoch": 10.028812368236121, "grad_norm": 0.2622365951538086, "learning_rate": 2.6650269383930663e-05, "loss": 0.0233, "step": 14271 }, { "epoch": 10.029515108924807, "grad_norm": 0.40569940209388733, "learning_rate": 2.6649800890138203e-05, "loss": 0.0122, "step": 14272 }, { "epoch": 10.030217849613493, "grad_norm": 0.21239084005355835, "learning_rate": 2.6649332396345747e-05, "loss": 0.0324, "step": 14273 }, { "epoch": 10.030920590302179, "grad_norm": 0.3616410493850708, "learning_rate": 2.664886390255329e-05, "loss": 0.0275, "step": 14274 }, { "epoch": 10.031623330990865, "grad_norm": 0.21309125423431396, "learning_rate": 2.6648395408760835e-05, "loss": 0.0591, "step": 14275 }, { "epoch": 10.03232607167955, "grad_norm": 0.26543474197387695, "learning_rate": 2.664792691496838e-05, "loss": 0.0527, "step": 14276 }, { "epoch": 10.033028812368237, "grad_norm": 0.3731556534767151, "learning_rate": 2.664745842117592e-05, "loss": 0.0945, "step": 14277 }, { "epoch": 10.033731553056922, "grad_norm": 0.8008967638015747, "learning_rate": 2.6646989927383462e-05, "loss": 0.14, "step": 14278 }, { "epoch": 10.034434293745608, "grad_norm": 0.50422602891922, "learning_rate": 2.6646521433591006e-05, "loss": 0.1807, "step": 14279 }, { "epoch": 10.035137034434294, "grad_norm": 1.5566165447235107, "learning_rate": 2.664605293979855e-05, "loss": 0.2105, "step": 14280 }, { "epoch": 10.03583977512298, "grad_norm": 0.2896178364753723, "learning_rate": 2.664558444600609e-05, "loss": 0.1111, "step": 14281 }, { "epoch": 10.036542515811666, "grad_norm": 0.3014383316040039, "learning_rate": 2.6645115952213634e-05, "loss": 0.0215, "step": 14282 }, { "epoch": 10.037245256500352, "grad_norm": 0.12714309990406036, "learning_rate": 2.6644647458421178e-05, "loss": 0.018, "step": 14283 }, { "epoch": 10.037947997189038, "grad_norm": 0.1201418861746788, "learning_rate": 2.664417896462872e-05, "loss": 0.0241, "step": 14284 }, { "epoch": 10.038650737877724, "grad_norm": 0.0945052057504654, "learning_rate": 2.664371047083626e-05, "loss": 0.0227, "step": 14285 }, { "epoch": 10.03935347856641, "grad_norm": 0.23553018271923065, "learning_rate": 2.6643241977043805e-05, "loss": 0.0109, "step": 14286 }, { "epoch": 10.040056219255096, "grad_norm": 0.08929014950990677, "learning_rate": 2.664277348325135e-05, "loss": 0.017, "step": 14287 }, { "epoch": 10.04075895994378, "grad_norm": 0.2978444993495941, "learning_rate": 2.664230498945889e-05, "loss": 0.018, "step": 14288 }, { "epoch": 10.041461700632466, "grad_norm": 0.1646863967180252, "learning_rate": 2.6641836495666433e-05, "loss": 0.0266, "step": 14289 }, { "epoch": 10.042164441321152, "grad_norm": 0.08837416768074036, "learning_rate": 2.6641368001873974e-05, "loss": 0.01, "step": 14290 }, { "epoch": 10.042867182009838, "grad_norm": 0.13717204332351685, "learning_rate": 2.6640899508081517e-05, "loss": 0.0137, "step": 14291 }, { "epoch": 10.043569922698524, "grad_norm": 0.202768012881279, "learning_rate": 2.664043101428906e-05, "loss": 0.0155, "step": 14292 }, { "epoch": 10.04427266338721, "grad_norm": 0.2003100961446762, "learning_rate": 2.6639962520496605e-05, "loss": 0.0228, "step": 14293 }, { "epoch": 10.044975404075895, "grad_norm": 0.13191355764865875, "learning_rate": 2.6639494026704145e-05, "loss": 0.0155, "step": 14294 }, { "epoch": 10.045678144764581, "grad_norm": 0.13580124080181122, "learning_rate": 2.663902553291169e-05, "loss": 0.022, "step": 14295 }, { "epoch": 10.046380885453267, "grad_norm": 0.15981996059417725, "learning_rate": 2.6638557039119233e-05, "loss": 0.0288, "step": 14296 }, { "epoch": 10.047083626141953, "grad_norm": 0.11794155091047287, "learning_rate": 2.6638088545326776e-05, "loss": 0.0117, "step": 14297 }, { "epoch": 10.047786366830639, "grad_norm": 0.1642235964536667, "learning_rate": 2.6637620051534317e-05, "loss": 0.0336, "step": 14298 }, { "epoch": 10.048489107519325, "grad_norm": 0.547581136226654, "learning_rate": 2.663715155774186e-05, "loss": 0.0489, "step": 14299 }, { "epoch": 10.049191848208011, "grad_norm": 0.6117194294929504, "learning_rate": 2.6636683063949404e-05, "loss": 0.0522, "step": 14300 }, { "epoch": 10.049894588896697, "grad_norm": 0.2739469110965729, "learning_rate": 2.6636214570156948e-05, "loss": 0.0615, "step": 14301 }, { "epoch": 10.050597329585383, "grad_norm": 0.5234857797622681, "learning_rate": 2.663574607636449e-05, "loss": 0.0918, "step": 14302 }, { "epoch": 10.051300070274069, "grad_norm": 0.9906609058380127, "learning_rate": 2.6635277582572032e-05, "loss": 0.1331, "step": 14303 }, { "epoch": 10.052002810962755, "grad_norm": 0.6814717650413513, "learning_rate": 2.6634809088779576e-05, "loss": 0.1434, "step": 14304 }, { "epoch": 10.05270555165144, "grad_norm": 1.001488208770752, "learning_rate": 2.6634340594987116e-05, "loss": 0.2007, "step": 14305 }, { "epoch": 10.053408292340126, "grad_norm": 0.2732796370983124, "learning_rate": 2.663387210119466e-05, "loss": 0.0683, "step": 14306 }, { "epoch": 10.054111033028812, "grad_norm": 0.19916847348213196, "learning_rate": 2.66334036074022e-05, "loss": 0.0329, "step": 14307 }, { "epoch": 10.054813773717498, "grad_norm": 0.20838399231433868, "learning_rate": 2.6632935113609744e-05, "loss": 0.041, "step": 14308 }, { "epoch": 10.055516514406184, "grad_norm": 0.12396777421236038, "learning_rate": 2.6632466619817287e-05, "loss": 0.0208, "step": 14309 }, { "epoch": 10.05621925509487, "grad_norm": 0.08003176748752594, "learning_rate": 2.663199812602483e-05, "loss": 0.015, "step": 14310 }, { "epoch": 10.056921995783556, "grad_norm": 0.07219037413597107, "learning_rate": 2.6631529632232375e-05, "loss": 0.01, "step": 14311 }, { "epoch": 10.057624736472242, "grad_norm": 0.11718053370714188, "learning_rate": 2.6631061138439915e-05, "loss": 0.0098, "step": 14312 }, { "epoch": 10.058327477160928, "grad_norm": 0.19273178279399872, "learning_rate": 2.663059264464746e-05, "loss": 0.0182, "step": 14313 }, { "epoch": 10.059030217849614, "grad_norm": 0.1930980682373047, "learning_rate": 2.6630124150855003e-05, "loss": 0.0209, "step": 14314 }, { "epoch": 10.0597329585383, "grad_norm": 0.2017589956521988, "learning_rate": 2.6629655657062546e-05, "loss": 0.0153, "step": 14315 }, { "epoch": 10.060435699226986, "grad_norm": 0.340669184923172, "learning_rate": 2.6629187163270087e-05, "loss": 0.0211, "step": 14316 }, { "epoch": 10.061138439915672, "grad_norm": 0.1638372391462326, "learning_rate": 2.662871866947763e-05, "loss": 0.0128, "step": 14317 }, { "epoch": 10.061841180604358, "grad_norm": 0.2910657823085785, "learning_rate": 2.6628250175685174e-05, "loss": 0.0269, "step": 14318 }, { "epoch": 10.062543921293043, "grad_norm": 0.1241535097360611, "learning_rate": 2.6627781681892718e-05, "loss": 0.0112, "step": 14319 }, { "epoch": 10.06324666198173, "grad_norm": 0.13569015264511108, "learning_rate": 2.6627313188100258e-05, "loss": 0.0251, "step": 14320 }, { "epoch": 10.063949402670415, "grad_norm": 0.21738027036190033, "learning_rate": 2.6626844694307802e-05, "loss": 0.0401, "step": 14321 }, { "epoch": 10.064652143359101, "grad_norm": 0.14882630109786987, "learning_rate": 2.6626376200515346e-05, "loss": 0.0175, "step": 14322 }, { "epoch": 10.065354884047787, "grad_norm": 0.298073947429657, "learning_rate": 2.6625907706722886e-05, "loss": 0.0337, "step": 14323 }, { "epoch": 10.066057624736473, "grad_norm": 0.4118204712867737, "learning_rate": 2.662543921293043e-05, "loss": 0.0431, "step": 14324 }, { "epoch": 10.066760365425159, "grad_norm": 0.4772367477416992, "learning_rate": 2.662497071913797e-05, "loss": 0.0487, "step": 14325 }, { "epoch": 10.067463106113845, "grad_norm": 0.352029025554657, "learning_rate": 2.6624502225345514e-05, "loss": 0.0701, "step": 14326 }, { "epoch": 10.068165846802529, "grad_norm": 0.7778975963592529, "learning_rate": 2.6624033731553057e-05, "loss": 0.0968, "step": 14327 }, { "epoch": 10.068868587491215, "grad_norm": 0.6219946146011353, "learning_rate": 2.66235652377606e-05, "loss": 0.172, "step": 14328 }, { "epoch": 10.0695713281799, "grad_norm": 0.680702269077301, "learning_rate": 2.662309674396814e-05, "loss": 0.158, "step": 14329 }, { "epoch": 10.070274068868587, "grad_norm": 1.2014689445495605, "learning_rate": 2.6622628250175685e-05, "loss": 0.1656, "step": 14330 }, { "epoch": 10.070976809557273, "grad_norm": 0.17429591715335846, "learning_rate": 2.662215975638323e-05, "loss": 0.0653, "step": 14331 }, { "epoch": 10.071679550245959, "grad_norm": 0.16860470175743103, "learning_rate": 2.6621691262590773e-05, "loss": 0.0219, "step": 14332 }, { "epoch": 10.072382290934645, "grad_norm": 0.11641786247491837, "learning_rate": 2.6621222768798313e-05, "loss": 0.0192, "step": 14333 }, { "epoch": 10.07308503162333, "grad_norm": 0.13486982882022858, "learning_rate": 2.6620754275005857e-05, "loss": 0.0165, "step": 14334 }, { "epoch": 10.073787772312016, "grad_norm": 0.12297075241804123, "learning_rate": 2.66202857812134e-05, "loss": 0.0229, "step": 14335 }, { "epoch": 10.074490513000702, "grad_norm": 0.12017501890659332, "learning_rate": 2.6619817287420944e-05, "loss": 0.0121, "step": 14336 }, { "epoch": 10.075193253689388, "grad_norm": 0.2208317220211029, "learning_rate": 2.6619348793628488e-05, "loss": 0.0147, "step": 14337 }, { "epoch": 10.075895994378074, "grad_norm": 0.14424531161785126, "learning_rate": 2.6618880299836028e-05, "loss": 0.0167, "step": 14338 }, { "epoch": 10.07659873506676, "grad_norm": 0.12579786777496338, "learning_rate": 2.6618411806043572e-05, "loss": 0.0211, "step": 14339 }, { "epoch": 10.077301475755446, "grad_norm": 0.11862293630838394, "learning_rate": 2.6617943312251112e-05, "loss": 0.0135, "step": 14340 }, { "epoch": 10.078004216444132, "grad_norm": 0.15313062071800232, "learning_rate": 2.6617474818458656e-05, "loss": 0.0213, "step": 14341 }, { "epoch": 10.078706957132818, "grad_norm": 0.13355909287929535, "learning_rate": 2.6617006324666196e-05, "loss": 0.0142, "step": 14342 }, { "epoch": 10.079409697821504, "grad_norm": 0.14195673167705536, "learning_rate": 2.661653783087374e-05, "loss": 0.0315, "step": 14343 }, { "epoch": 10.08011243851019, "grad_norm": 0.46029505133628845, "learning_rate": 2.6616069337081284e-05, "loss": 0.0276, "step": 14344 }, { "epoch": 10.080815179198876, "grad_norm": 0.20661531388759613, "learning_rate": 2.6615600843288828e-05, "loss": 0.0357, "step": 14345 }, { "epoch": 10.081517919887562, "grad_norm": 0.11939366906881332, "learning_rate": 2.6615132349496368e-05, "loss": 0.0275, "step": 14346 }, { "epoch": 10.082220660576247, "grad_norm": 0.14640024304389954, "learning_rate": 2.661466385570391e-05, "loss": 0.0139, "step": 14347 }, { "epoch": 10.082923401264933, "grad_norm": 0.27341318130493164, "learning_rate": 2.6614195361911455e-05, "loss": 0.0526, "step": 14348 }, { "epoch": 10.08362614195362, "grad_norm": 0.16139522194862366, "learning_rate": 2.6613726868119e-05, "loss": 0.0287, "step": 14349 }, { "epoch": 10.084328882642305, "grad_norm": 0.3117160499095917, "learning_rate": 2.6613258374326543e-05, "loss": 0.0531, "step": 14350 }, { "epoch": 10.085031623330991, "grad_norm": 0.26885613799095154, "learning_rate": 2.6612789880534083e-05, "loss": 0.0588, "step": 14351 }, { "epoch": 10.085734364019677, "grad_norm": 0.47411900758743286, "learning_rate": 2.6612321386741627e-05, "loss": 0.0854, "step": 14352 }, { "epoch": 10.086437104708363, "grad_norm": 0.8168596029281616, "learning_rate": 2.661185289294917e-05, "loss": 0.1179, "step": 14353 }, { "epoch": 10.087139845397049, "grad_norm": 0.786307692527771, "learning_rate": 2.6611384399156714e-05, "loss": 0.1742, "step": 14354 }, { "epoch": 10.087842586085735, "grad_norm": 1.569185495376587, "learning_rate": 2.6610915905364255e-05, "loss": 0.1831, "step": 14355 }, { "epoch": 10.08854532677442, "grad_norm": 0.18459449708461761, "learning_rate": 2.66104474115718e-05, "loss": 0.059, "step": 14356 }, { "epoch": 10.089248067463107, "grad_norm": 0.21847017109394073, "learning_rate": 2.660997891777934e-05, "loss": 0.0267, "step": 14357 }, { "epoch": 10.089950808151793, "grad_norm": 0.3161526024341583, "learning_rate": 2.6609510423986882e-05, "loss": 0.0358, "step": 14358 }, { "epoch": 10.090653548840478, "grad_norm": 0.13368965685367584, "learning_rate": 2.6609041930194423e-05, "loss": 0.0144, "step": 14359 }, { "epoch": 10.091356289529164, "grad_norm": 0.16725628077983856, "learning_rate": 2.6608573436401967e-05, "loss": 0.0227, "step": 14360 }, { "epoch": 10.09205903021785, "grad_norm": 0.09625494480133057, "learning_rate": 2.660810494260951e-05, "loss": 0.0116, "step": 14361 }, { "epoch": 10.092761770906536, "grad_norm": 0.09362436085939407, "learning_rate": 2.6607636448817054e-05, "loss": 0.0205, "step": 14362 }, { "epoch": 10.093464511595222, "grad_norm": 0.09659674763679504, "learning_rate": 2.6607167955024598e-05, "loss": 0.0184, "step": 14363 }, { "epoch": 10.094167252283908, "grad_norm": 0.22253870964050293, "learning_rate": 2.6606699461232138e-05, "loss": 0.0146, "step": 14364 }, { "epoch": 10.094869992972592, "grad_norm": 0.14955776929855347, "learning_rate": 2.6606230967439682e-05, "loss": 0.0132, "step": 14365 }, { "epoch": 10.095572733661278, "grad_norm": 0.17586649954319, "learning_rate": 2.6605762473647226e-05, "loss": 0.0209, "step": 14366 }, { "epoch": 10.096275474349964, "grad_norm": 0.19505006074905396, "learning_rate": 2.660529397985477e-05, "loss": 0.0183, "step": 14367 }, { "epoch": 10.09697821503865, "grad_norm": 0.14083002507686615, "learning_rate": 2.660482548606231e-05, "loss": 0.0203, "step": 14368 }, { "epoch": 10.097680955727336, "grad_norm": 0.12459927052259445, "learning_rate": 2.6604356992269853e-05, "loss": 0.0295, "step": 14369 }, { "epoch": 10.098383696416022, "grad_norm": 0.18203122913837433, "learning_rate": 2.6603888498477397e-05, "loss": 0.0374, "step": 14370 }, { "epoch": 10.099086437104708, "grad_norm": 0.6820050477981567, "learning_rate": 2.660342000468494e-05, "loss": 0.0523, "step": 14371 }, { "epoch": 10.099789177793394, "grad_norm": 0.2221817672252655, "learning_rate": 2.660295151089248e-05, "loss": 0.0417, "step": 14372 }, { "epoch": 10.10049191848208, "grad_norm": 0.30123740434646606, "learning_rate": 2.6602483017100025e-05, "loss": 0.0708, "step": 14373 }, { "epoch": 10.101194659170766, "grad_norm": 0.2853740453720093, "learning_rate": 2.660201452330757e-05, "loss": 0.0376, "step": 14374 }, { "epoch": 10.101897399859451, "grad_norm": 0.21219421923160553, "learning_rate": 2.660154602951511e-05, "loss": 0.0562, "step": 14375 }, { "epoch": 10.102600140548137, "grad_norm": 0.31259962916374207, "learning_rate": 2.6601077535722653e-05, "loss": 0.0901, "step": 14376 }, { "epoch": 10.103302881236823, "grad_norm": 0.9922657012939453, "learning_rate": 2.6600609041930193e-05, "loss": 0.0998, "step": 14377 }, { "epoch": 10.10400562192551, "grad_norm": 0.5616621971130371, "learning_rate": 2.6600140548137737e-05, "loss": 0.1699, "step": 14378 }, { "epoch": 10.104708362614195, "grad_norm": 1.1036581993103027, "learning_rate": 2.659967205434528e-05, "loss": 0.1811, "step": 14379 }, { "epoch": 10.105411103302881, "grad_norm": 1.174017071723938, "learning_rate": 2.6599203560552824e-05, "loss": 0.1777, "step": 14380 }, { "epoch": 10.106113843991567, "grad_norm": 0.3497520089149475, "learning_rate": 2.6598735066760364e-05, "loss": 0.0592, "step": 14381 }, { "epoch": 10.106816584680253, "grad_norm": 0.13988369703292847, "learning_rate": 2.6598266572967908e-05, "loss": 0.0267, "step": 14382 }, { "epoch": 10.107519325368939, "grad_norm": 0.12816137075424194, "learning_rate": 2.6597798079175452e-05, "loss": 0.01, "step": 14383 }, { "epoch": 10.108222066057625, "grad_norm": 0.21110108494758606, "learning_rate": 2.6597329585382996e-05, "loss": 0.0208, "step": 14384 }, { "epoch": 10.10892480674631, "grad_norm": 0.1610988974571228, "learning_rate": 2.6596861091590536e-05, "loss": 0.0169, "step": 14385 }, { "epoch": 10.109627547434997, "grad_norm": 0.09068117290735245, "learning_rate": 2.659639259779808e-05, "loss": 0.0076, "step": 14386 }, { "epoch": 10.110330288123683, "grad_norm": 0.23287594318389893, "learning_rate": 2.6595924104005623e-05, "loss": 0.022, "step": 14387 }, { "epoch": 10.111033028812368, "grad_norm": 0.23197340965270996, "learning_rate": 2.6595455610213167e-05, "loss": 0.0298, "step": 14388 }, { "epoch": 10.111735769501054, "grad_norm": 0.393777459859848, "learning_rate": 2.659498711642071e-05, "loss": 0.0273, "step": 14389 }, { "epoch": 10.11243851018974, "grad_norm": 0.23642058670520782, "learning_rate": 2.659451862262825e-05, "loss": 0.0179, "step": 14390 }, { "epoch": 10.113141250878426, "grad_norm": 0.1901732087135315, "learning_rate": 2.6594050128835795e-05, "loss": 0.0212, "step": 14391 }, { "epoch": 10.113843991567112, "grad_norm": 0.0814223662018776, "learning_rate": 2.6593581635043335e-05, "loss": 0.0117, "step": 14392 }, { "epoch": 10.114546732255798, "grad_norm": 0.21370404958724976, "learning_rate": 2.659311314125088e-05, "loss": 0.026, "step": 14393 }, { "epoch": 10.115249472944484, "grad_norm": 0.1284237951040268, "learning_rate": 2.659264464745842e-05, "loss": 0.0292, "step": 14394 }, { "epoch": 10.11595221363317, "grad_norm": 0.30214723944664, "learning_rate": 2.6592176153665963e-05, "loss": 0.0202, "step": 14395 }, { "epoch": 10.116654954321856, "grad_norm": 0.3489038944244385, "learning_rate": 2.6591707659873507e-05, "loss": 0.0197, "step": 14396 }, { "epoch": 10.117357695010542, "grad_norm": 5.244426250457764, "learning_rate": 2.659123916608105e-05, "loss": 0.034, "step": 14397 }, { "epoch": 10.118060435699228, "grad_norm": 0.16237002611160278, "learning_rate": 2.659077067228859e-05, "loss": 0.0405, "step": 14398 }, { "epoch": 10.118763176387914, "grad_norm": 0.2218410074710846, "learning_rate": 2.6590302178496135e-05, "loss": 0.0354, "step": 14399 }, { "epoch": 10.1194659170766, "grad_norm": 0.1865224838256836, "learning_rate": 2.6589833684703678e-05, "loss": 0.0328, "step": 14400 }, { "epoch": 10.120168657765285, "grad_norm": 0.4081604480743408, "learning_rate": 2.6589365190911222e-05, "loss": 0.0765, "step": 14401 }, { "epoch": 10.120871398453971, "grad_norm": 1.0266730785369873, "learning_rate": 2.6588896697118766e-05, "loss": 0.1141, "step": 14402 }, { "epoch": 10.121574139142655, "grad_norm": 0.6994175314903259, "learning_rate": 2.6588428203326306e-05, "loss": 0.1158, "step": 14403 }, { "epoch": 10.122276879831341, "grad_norm": 1.0345797538757324, "learning_rate": 2.658795970953385e-05, "loss": 0.1545, "step": 14404 }, { "epoch": 10.122979620520027, "grad_norm": 0.9060716032981873, "learning_rate": 2.6587491215741394e-05, "loss": 0.1982, "step": 14405 }, { "epoch": 10.123682361208713, "grad_norm": 0.28486940264701843, "learning_rate": 2.6587022721948937e-05, "loss": 0.0904, "step": 14406 }, { "epoch": 10.1243851018974, "grad_norm": 0.15377044677734375, "learning_rate": 2.6586554228156478e-05, "loss": 0.0201, "step": 14407 }, { "epoch": 10.125087842586085, "grad_norm": 0.1617233157157898, "learning_rate": 2.658608573436402e-05, "loss": 0.0251, "step": 14408 }, { "epoch": 10.125790583274771, "grad_norm": 0.1914578378200531, "learning_rate": 2.6585617240571565e-05, "loss": 0.0206, "step": 14409 }, { "epoch": 10.126493323963457, "grad_norm": 0.1513301581144333, "learning_rate": 2.6585148746779105e-05, "loss": 0.0165, "step": 14410 }, { "epoch": 10.127196064652143, "grad_norm": 0.16639231145381927, "learning_rate": 2.6584680252986646e-05, "loss": 0.0129, "step": 14411 }, { "epoch": 10.127898805340829, "grad_norm": 0.0903029814362526, "learning_rate": 2.658421175919419e-05, "loss": 0.0117, "step": 14412 }, { "epoch": 10.128601546029515, "grad_norm": 0.1951969414949417, "learning_rate": 2.6583743265401733e-05, "loss": 0.0144, "step": 14413 }, { "epoch": 10.1293042867182, "grad_norm": 0.13802407681941986, "learning_rate": 2.6583274771609277e-05, "loss": 0.0169, "step": 14414 }, { "epoch": 10.130007027406887, "grad_norm": 0.09093481302261353, "learning_rate": 2.658280627781682e-05, "loss": 0.0086, "step": 14415 }, { "epoch": 10.130709768095572, "grad_norm": 0.1669059842824936, "learning_rate": 2.658233778402436e-05, "loss": 0.0248, "step": 14416 }, { "epoch": 10.131412508784258, "grad_norm": 0.23049871623516083, "learning_rate": 2.6581869290231905e-05, "loss": 0.0214, "step": 14417 }, { "epoch": 10.132115249472944, "grad_norm": 0.17653760313987732, "learning_rate": 2.658140079643945e-05, "loss": 0.0338, "step": 14418 }, { "epoch": 10.13281799016163, "grad_norm": 0.17173956334590912, "learning_rate": 2.6580932302646992e-05, "loss": 0.0233, "step": 14419 }, { "epoch": 10.133520730850316, "grad_norm": 0.14641845226287842, "learning_rate": 2.6580463808854532e-05, "loss": 0.031, "step": 14420 }, { "epoch": 10.134223471539002, "grad_norm": 0.15770603716373444, "learning_rate": 2.6579995315062076e-05, "loss": 0.0322, "step": 14421 }, { "epoch": 10.134926212227688, "grad_norm": 0.1804841309785843, "learning_rate": 2.657952682126962e-05, "loss": 0.0352, "step": 14422 }, { "epoch": 10.135628952916374, "grad_norm": 0.12812232971191406, "learning_rate": 2.6579058327477164e-05, "loss": 0.0296, "step": 14423 }, { "epoch": 10.13633169360506, "grad_norm": 0.47876352071762085, "learning_rate": 2.6578589833684704e-05, "loss": 0.0999, "step": 14424 }, { "epoch": 10.137034434293746, "grad_norm": 0.6733684539794922, "learning_rate": 2.6578121339892248e-05, "loss": 0.0399, "step": 14425 }, { "epoch": 10.137737174982432, "grad_norm": 0.3758637011051178, "learning_rate": 2.657765284609979e-05, "loss": 0.0804, "step": 14426 }, { "epoch": 10.138439915671118, "grad_norm": 0.3723868131637573, "learning_rate": 2.6577184352307332e-05, "loss": 0.0921, "step": 14427 }, { "epoch": 10.139142656359803, "grad_norm": 0.5640697479248047, "learning_rate": 2.6576715858514875e-05, "loss": 0.1536, "step": 14428 }, { "epoch": 10.13984539704849, "grad_norm": 0.7326099872589111, "learning_rate": 2.6576247364722416e-05, "loss": 0.1906, "step": 14429 }, { "epoch": 10.140548137737175, "grad_norm": 0.9994308352470398, "learning_rate": 2.657577887092996e-05, "loss": 0.2024, "step": 14430 }, { "epoch": 10.141250878425861, "grad_norm": 0.16617229580879211, "learning_rate": 2.6575310377137503e-05, "loss": 0.0695, "step": 14431 }, { "epoch": 10.141953619114547, "grad_norm": 0.11413232237100601, "learning_rate": 2.6574841883345047e-05, "loss": 0.0257, "step": 14432 }, { "epoch": 10.142656359803233, "grad_norm": 0.1562420129776001, "learning_rate": 2.6574373389552587e-05, "loss": 0.0337, "step": 14433 }, { "epoch": 10.143359100491919, "grad_norm": 0.15616025030612946, "learning_rate": 2.657390489576013e-05, "loss": 0.0129, "step": 14434 }, { "epoch": 10.144061841180605, "grad_norm": 0.23932260274887085, "learning_rate": 2.6573436401967675e-05, "loss": 0.0272, "step": 14435 }, { "epoch": 10.14476458186929, "grad_norm": 0.1399638056755066, "learning_rate": 2.657296790817522e-05, "loss": 0.0206, "step": 14436 }, { "epoch": 10.145467322557977, "grad_norm": 0.10702480375766754, "learning_rate": 2.657249941438276e-05, "loss": 0.0095, "step": 14437 }, { "epoch": 10.146170063246663, "grad_norm": 0.1655145138502121, "learning_rate": 2.6572030920590303e-05, "loss": 0.0329, "step": 14438 }, { "epoch": 10.146872803935349, "grad_norm": 0.1764053851366043, "learning_rate": 2.6571562426797846e-05, "loss": 0.0321, "step": 14439 }, { "epoch": 10.147575544624035, "grad_norm": 0.08501860499382019, "learning_rate": 2.657109393300539e-05, "loss": 0.0144, "step": 14440 }, { "epoch": 10.14827828531272, "grad_norm": 0.32101359963417053, "learning_rate": 2.6570625439212934e-05, "loss": 0.0306, "step": 14441 }, { "epoch": 10.148981026001405, "grad_norm": 0.13395093381404877, "learning_rate": 2.6570156945420474e-05, "loss": 0.0076, "step": 14442 }, { "epoch": 10.14968376669009, "grad_norm": 0.10427198559045792, "learning_rate": 2.6569688451628018e-05, "loss": 0.0205, "step": 14443 }, { "epoch": 10.150386507378776, "grad_norm": 0.13135051727294922, "learning_rate": 2.656921995783556e-05, "loss": 0.0182, "step": 14444 }, { "epoch": 10.151089248067462, "grad_norm": 0.14667724072933197, "learning_rate": 2.6568751464043102e-05, "loss": 0.0318, "step": 14445 }, { "epoch": 10.151791988756148, "grad_norm": 0.5388857126235962, "learning_rate": 2.6568282970250642e-05, "loss": 0.0321, "step": 14446 }, { "epoch": 10.152494729444834, "grad_norm": 0.10832539200782776, "learning_rate": 2.6567814476458186e-05, "loss": 0.0154, "step": 14447 }, { "epoch": 10.15319747013352, "grad_norm": 0.37185221910476685, "learning_rate": 2.656734598266573e-05, "loss": 0.0731, "step": 14448 }, { "epoch": 10.153900210822206, "grad_norm": 0.28803279995918274, "learning_rate": 2.6566877488873273e-05, "loss": 0.0602, "step": 14449 }, { "epoch": 10.154602951510892, "grad_norm": 0.2350703477859497, "learning_rate": 2.6566408995080814e-05, "loss": 0.0452, "step": 14450 }, { "epoch": 10.155305692199578, "grad_norm": 0.3354501724243164, "learning_rate": 2.6565940501288357e-05, "loss": 0.0705, "step": 14451 }, { "epoch": 10.156008432888264, "grad_norm": 0.3798961937427521, "learning_rate": 2.65654720074959e-05, "loss": 0.0968, "step": 14452 }, { "epoch": 10.15671117357695, "grad_norm": 0.40548571944236755, "learning_rate": 2.6565003513703445e-05, "loss": 0.1612, "step": 14453 }, { "epoch": 10.157413914265636, "grad_norm": 0.6039931774139404, "learning_rate": 2.656453501991099e-05, "loss": 0.168, "step": 14454 }, { "epoch": 10.158116654954322, "grad_norm": 0.9336647987365723, "learning_rate": 2.656406652611853e-05, "loss": 0.1974, "step": 14455 }, { "epoch": 10.158819395643008, "grad_norm": 0.18224917352199554, "learning_rate": 2.6563598032326073e-05, "loss": 0.0598, "step": 14456 }, { "epoch": 10.159522136331693, "grad_norm": 0.12187699973583221, "learning_rate": 2.6563129538533616e-05, "loss": 0.0208, "step": 14457 }, { "epoch": 10.16022487702038, "grad_norm": 0.10725478827953339, "learning_rate": 2.656266104474116e-05, "loss": 0.0146, "step": 14458 }, { "epoch": 10.160927617709065, "grad_norm": 0.1044606864452362, "learning_rate": 2.65621925509487e-05, "loss": 0.025, "step": 14459 }, { "epoch": 10.161630358397751, "grad_norm": 0.08003072440624237, "learning_rate": 2.6561724057156244e-05, "loss": 0.0102, "step": 14460 }, { "epoch": 10.162333099086437, "grad_norm": 0.4684547185897827, "learning_rate": 2.6561255563363788e-05, "loss": 0.0181, "step": 14461 }, { "epoch": 10.163035839775123, "grad_norm": 0.13930629193782806, "learning_rate": 2.6560787069571328e-05, "loss": 0.0195, "step": 14462 }, { "epoch": 10.163738580463809, "grad_norm": 0.08561497181653976, "learning_rate": 2.656031857577887e-05, "loss": 0.0062, "step": 14463 }, { "epoch": 10.164441321152495, "grad_norm": 0.3579188287258148, "learning_rate": 2.6559850081986412e-05, "loss": 0.0168, "step": 14464 }, { "epoch": 10.16514406184118, "grad_norm": 0.07188049703836441, "learning_rate": 2.6559381588193956e-05, "loss": 0.0105, "step": 14465 }, { "epoch": 10.165846802529867, "grad_norm": 0.175230011343956, "learning_rate": 2.65589130944015e-05, "loss": 0.018, "step": 14466 }, { "epoch": 10.166549543218553, "grad_norm": 0.12753652036190033, "learning_rate": 2.6558444600609043e-05, "loss": 0.0137, "step": 14467 }, { "epoch": 10.167252283907239, "grad_norm": 0.19577881693840027, "learning_rate": 2.6557976106816584e-05, "loss": 0.0323, "step": 14468 }, { "epoch": 10.167955024595924, "grad_norm": 0.19023659825325012, "learning_rate": 2.6557507613024128e-05, "loss": 0.0331, "step": 14469 }, { "epoch": 10.16865776528461, "grad_norm": 0.15059641003608704, "learning_rate": 2.655703911923167e-05, "loss": 0.0154, "step": 14470 }, { "epoch": 10.169360505973296, "grad_norm": 0.12190235406160355, "learning_rate": 2.6556570625439215e-05, "loss": 0.0351, "step": 14471 }, { "epoch": 10.170063246661982, "grad_norm": 0.13359475135803223, "learning_rate": 2.6556102131646755e-05, "loss": 0.0205, "step": 14472 }, { "epoch": 10.170765987350668, "grad_norm": 0.4543120563030243, "learning_rate": 2.65556336378543e-05, "loss": 0.0407, "step": 14473 }, { "epoch": 10.171468728039354, "grad_norm": 0.18858903646469116, "learning_rate": 2.6555165144061843e-05, "loss": 0.0288, "step": 14474 }, { "epoch": 10.17217146872804, "grad_norm": 0.18743866682052612, "learning_rate": 2.6554696650269387e-05, "loss": 0.0491, "step": 14475 }, { "epoch": 10.172874209416726, "grad_norm": 0.254373162984848, "learning_rate": 2.6554228156476927e-05, "loss": 0.0779, "step": 14476 }, { "epoch": 10.173576950105412, "grad_norm": 0.34531041979789734, "learning_rate": 2.655375966268447e-05, "loss": 0.0697, "step": 14477 }, { "epoch": 10.174279690794098, "grad_norm": 0.7075573205947876, "learning_rate": 2.6553291168892014e-05, "loss": 0.1533, "step": 14478 }, { "epoch": 10.174982431482784, "grad_norm": 0.9316971302032471, "learning_rate": 2.6552822675099555e-05, "loss": 0.1678, "step": 14479 }, { "epoch": 10.17568517217147, "grad_norm": 1.270798683166504, "learning_rate": 2.65523541813071e-05, "loss": 0.1953, "step": 14480 }, { "epoch": 10.176387912860154, "grad_norm": 0.355264276266098, "learning_rate": 2.655188568751464e-05, "loss": 0.0859, "step": 14481 }, { "epoch": 10.17709065354884, "grad_norm": 0.1483546495437622, "learning_rate": 2.6551417193722182e-05, "loss": 0.018, "step": 14482 }, { "epoch": 10.177793394237526, "grad_norm": 0.10649275034666061, "learning_rate": 2.6550948699929726e-05, "loss": 0.0229, "step": 14483 }, { "epoch": 10.178496134926212, "grad_norm": 0.07402113825082779, "learning_rate": 2.655048020613727e-05, "loss": 0.0125, "step": 14484 }, { "epoch": 10.179198875614897, "grad_norm": 0.20476311445236206, "learning_rate": 2.655001171234481e-05, "loss": 0.017, "step": 14485 }, { "epoch": 10.179901616303583, "grad_norm": 0.07348413020372391, "learning_rate": 2.6549543218552354e-05, "loss": 0.0094, "step": 14486 }, { "epoch": 10.18060435699227, "grad_norm": 0.09455244988203049, "learning_rate": 2.6549074724759898e-05, "loss": 0.0114, "step": 14487 }, { "epoch": 10.181307097680955, "grad_norm": 0.14390432834625244, "learning_rate": 2.654860623096744e-05, "loss": 0.022, "step": 14488 }, { "epoch": 10.182009838369641, "grad_norm": 0.15448740124702454, "learning_rate": 2.6548137737174982e-05, "loss": 0.0153, "step": 14489 }, { "epoch": 10.182712579058327, "grad_norm": 0.10924820601940155, "learning_rate": 2.6547669243382525e-05, "loss": 0.0234, "step": 14490 }, { "epoch": 10.183415319747013, "grad_norm": 0.12225478887557983, "learning_rate": 2.654720074959007e-05, "loss": 0.0264, "step": 14491 }, { "epoch": 10.184118060435699, "grad_norm": 0.06758218258619308, "learning_rate": 2.6546732255797613e-05, "loss": 0.0073, "step": 14492 }, { "epoch": 10.184820801124385, "grad_norm": 0.13442155718803406, "learning_rate": 2.6546263762005157e-05, "loss": 0.0185, "step": 14493 }, { "epoch": 10.18552354181307, "grad_norm": 0.08040422946214676, "learning_rate": 2.6545795268212697e-05, "loss": 0.0153, "step": 14494 }, { "epoch": 10.186226282501757, "grad_norm": 0.1828477829694748, "learning_rate": 2.654532677442024e-05, "loss": 0.0287, "step": 14495 }, { "epoch": 10.186929023190443, "grad_norm": 0.21692630648612976, "learning_rate": 2.6544858280627784e-05, "loss": 0.0392, "step": 14496 }, { "epoch": 10.187631763879128, "grad_norm": 0.14081059396266937, "learning_rate": 2.6544389786835325e-05, "loss": 0.0229, "step": 14497 }, { "epoch": 10.188334504567814, "grad_norm": 0.29250016808509827, "learning_rate": 2.6543921293042865e-05, "loss": 0.0563, "step": 14498 }, { "epoch": 10.1890372452565, "grad_norm": 0.2132779210805893, "learning_rate": 2.654345279925041e-05, "loss": 0.0271, "step": 14499 }, { "epoch": 10.189739985945186, "grad_norm": 0.29927703738212585, "learning_rate": 2.6542984305457953e-05, "loss": 0.0593, "step": 14500 }, { "epoch": 10.190442726633872, "grad_norm": 0.32074397802352905, "learning_rate": 2.6542515811665496e-05, "loss": 0.0526, "step": 14501 }, { "epoch": 10.191145467322558, "grad_norm": 0.3393743634223938, "learning_rate": 2.6542047317873037e-05, "loss": 0.1037, "step": 14502 }, { "epoch": 10.191848208011244, "grad_norm": 0.5990287661552429, "learning_rate": 2.654157882408058e-05, "loss": 0.123, "step": 14503 }, { "epoch": 10.19255094869993, "grad_norm": 0.61993408203125, "learning_rate": 2.6541110330288124e-05, "loss": 0.1792, "step": 14504 }, { "epoch": 10.193253689388616, "grad_norm": 1.843124270439148, "learning_rate": 2.6540641836495668e-05, "loss": 0.1917, "step": 14505 }, { "epoch": 10.193956430077302, "grad_norm": 0.26593589782714844, "learning_rate": 2.654017334270321e-05, "loss": 0.0673, "step": 14506 }, { "epoch": 10.194659170765988, "grad_norm": 0.12630127370357513, "learning_rate": 2.6539704848910752e-05, "loss": 0.0314, "step": 14507 }, { "epoch": 10.195361911454674, "grad_norm": 0.13060809671878815, "learning_rate": 2.6539236355118296e-05, "loss": 0.0314, "step": 14508 }, { "epoch": 10.19606465214336, "grad_norm": 0.09987662732601166, "learning_rate": 2.653876786132584e-05, "loss": 0.017, "step": 14509 }, { "epoch": 10.196767392832045, "grad_norm": 0.1049862802028656, "learning_rate": 2.6538299367533383e-05, "loss": 0.016, "step": 14510 }, { "epoch": 10.197470133520731, "grad_norm": 0.0976593941450119, "learning_rate": 2.6537830873740923e-05, "loss": 0.013, "step": 14511 }, { "epoch": 10.198172874209417, "grad_norm": 0.10453446954488754, "learning_rate": 2.6537362379948467e-05, "loss": 0.0181, "step": 14512 }, { "epoch": 10.198875614898103, "grad_norm": 0.11874160915613174, "learning_rate": 2.653689388615601e-05, "loss": 0.012, "step": 14513 }, { "epoch": 10.19957835558679, "grad_norm": 0.0898841992020607, "learning_rate": 2.653642539236355e-05, "loss": 0.0181, "step": 14514 }, { "epoch": 10.200281096275475, "grad_norm": 0.11819635331630707, "learning_rate": 2.6535956898571095e-05, "loss": 0.0122, "step": 14515 }, { "epoch": 10.200983836964161, "grad_norm": 0.12724296748638153, "learning_rate": 2.6535488404778635e-05, "loss": 0.0229, "step": 14516 }, { "epoch": 10.201686577652847, "grad_norm": 0.14980031549930573, "learning_rate": 2.653501991098618e-05, "loss": 0.0172, "step": 14517 }, { "epoch": 10.202389318341533, "grad_norm": 0.15210874378681183, "learning_rate": 2.6534551417193723e-05, "loss": 0.0191, "step": 14518 }, { "epoch": 10.203092059030217, "grad_norm": 0.17793156206607819, "learning_rate": 2.6534082923401266e-05, "loss": 0.0231, "step": 14519 }, { "epoch": 10.203794799718903, "grad_norm": 0.2912184000015259, "learning_rate": 2.6533614429608807e-05, "loss": 0.0198, "step": 14520 }, { "epoch": 10.204497540407589, "grad_norm": 0.32557469606399536, "learning_rate": 2.653314593581635e-05, "loss": 0.0281, "step": 14521 }, { "epoch": 10.205200281096275, "grad_norm": 0.23719371855258942, "learning_rate": 2.6532677442023894e-05, "loss": 0.0198, "step": 14522 }, { "epoch": 10.20590302178496, "grad_norm": 0.40575718879699707, "learning_rate": 2.6532208948231438e-05, "loss": 0.0383, "step": 14523 }, { "epoch": 10.206605762473647, "grad_norm": 0.13950714468955994, "learning_rate": 2.6531740454438978e-05, "loss": 0.0301, "step": 14524 }, { "epoch": 10.207308503162333, "grad_norm": 0.19410204887390137, "learning_rate": 2.6531271960646522e-05, "loss": 0.0375, "step": 14525 }, { "epoch": 10.208011243851018, "grad_norm": 0.2543860673904419, "learning_rate": 2.6530803466854066e-05, "loss": 0.0655, "step": 14526 }, { "epoch": 10.208713984539704, "grad_norm": 0.837807297706604, "learning_rate": 2.653033497306161e-05, "loss": 0.0961, "step": 14527 }, { "epoch": 10.20941672522839, "grad_norm": 0.5186955332756042, "learning_rate": 2.652986647926915e-05, "loss": 0.1536, "step": 14528 }, { "epoch": 10.210119465917076, "grad_norm": 0.7431923747062683, "learning_rate": 2.6529397985476693e-05, "loss": 0.192, "step": 14529 }, { "epoch": 10.210822206605762, "grad_norm": 1.3852243423461914, "learning_rate": 2.6528929491684237e-05, "loss": 0.1816, "step": 14530 }, { "epoch": 10.211524947294448, "grad_norm": 0.2707407474517822, "learning_rate": 2.652846099789178e-05, "loss": 0.0596, "step": 14531 }, { "epoch": 10.212227687983134, "grad_norm": 0.10724470764398575, "learning_rate": 2.652799250409932e-05, "loss": 0.0196, "step": 14532 }, { "epoch": 10.21293042867182, "grad_norm": 0.309439480304718, "learning_rate": 2.652752401030686e-05, "loss": 0.018, "step": 14533 }, { "epoch": 10.213633169360506, "grad_norm": 0.09098032116889954, "learning_rate": 2.6527055516514405e-05, "loss": 0.0208, "step": 14534 }, { "epoch": 10.214335910049192, "grad_norm": 0.12967875599861145, "learning_rate": 2.652658702272195e-05, "loss": 0.0216, "step": 14535 }, { "epoch": 10.215038650737878, "grad_norm": 0.11117997020483017, "learning_rate": 2.6526118528929493e-05, "loss": 0.0127, "step": 14536 }, { "epoch": 10.215741391426564, "grad_norm": 0.07964574545621872, "learning_rate": 2.6525650035137033e-05, "loss": 0.0146, "step": 14537 }, { "epoch": 10.21644413211525, "grad_norm": 0.0832948237657547, "learning_rate": 2.6525181541344577e-05, "loss": 0.0103, "step": 14538 }, { "epoch": 10.217146872803935, "grad_norm": 0.19575020670890808, "learning_rate": 2.652471304755212e-05, "loss": 0.0535, "step": 14539 }, { "epoch": 10.217849613492621, "grad_norm": 0.13980315625667572, "learning_rate": 2.6524244553759664e-05, "loss": 0.0156, "step": 14540 }, { "epoch": 10.218552354181307, "grad_norm": 0.1530362218618393, "learning_rate": 2.6523776059967208e-05, "loss": 0.0282, "step": 14541 }, { "epoch": 10.219255094869993, "grad_norm": 0.07051673531532288, "learning_rate": 2.652330756617475e-05, "loss": 0.0106, "step": 14542 }, { "epoch": 10.219957835558679, "grad_norm": 0.105568528175354, "learning_rate": 2.6522839072382292e-05, "loss": 0.0201, "step": 14543 }, { "epoch": 10.220660576247365, "grad_norm": 0.1442791372537613, "learning_rate": 2.6522370578589836e-05, "loss": 0.0374, "step": 14544 }, { "epoch": 10.221363316936051, "grad_norm": 0.07255247235298157, "learning_rate": 2.652190208479738e-05, "loss": 0.0118, "step": 14545 }, { "epoch": 10.222066057624737, "grad_norm": 0.21356791257858276, "learning_rate": 2.652143359100492e-05, "loss": 0.0403, "step": 14546 }, { "epoch": 10.222768798313423, "grad_norm": 0.23010584712028503, "learning_rate": 2.6520965097212464e-05, "loss": 0.0324, "step": 14547 }, { "epoch": 10.223471539002109, "grad_norm": 0.20578888058662415, "learning_rate": 2.6520496603420007e-05, "loss": 0.0306, "step": 14548 }, { "epoch": 10.224174279690795, "grad_norm": 0.2679159939289093, "learning_rate": 2.6520028109627548e-05, "loss": 0.0307, "step": 14549 }, { "epoch": 10.22487702037948, "grad_norm": 0.23762644827365875, "learning_rate": 2.6519559615835088e-05, "loss": 0.0354, "step": 14550 }, { "epoch": 10.225579761068166, "grad_norm": 0.515536904335022, "learning_rate": 2.6519091122042632e-05, "loss": 0.0876, "step": 14551 }, { "epoch": 10.226282501756852, "grad_norm": 0.3685443699359894, "learning_rate": 2.6518622628250175e-05, "loss": 0.0983, "step": 14552 }, { "epoch": 10.226985242445538, "grad_norm": 0.7463407516479492, "learning_rate": 2.651815413445772e-05, "loss": 0.1438, "step": 14553 }, { "epoch": 10.227687983134224, "grad_norm": 1.2476171255111694, "learning_rate": 2.6517685640665263e-05, "loss": 0.1806, "step": 14554 }, { "epoch": 10.22839072382291, "grad_norm": 2.1669516563415527, "learning_rate": 2.6517217146872803e-05, "loss": 0.2486, "step": 14555 }, { "epoch": 10.229093464511596, "grad_norm": 0.2060343474149704, "learning_rate": 2.6516748653080347e-05, "loss": 0.0746, "step": 14556 }, { "epoch": 10.22979620520028, "grad_norm": 0.14764194190502167, "learning_rate": 2.651628015928789e-05, "loss": 0.0186, "step": 14557 }, { "epoch": 10.230498945888966, "grad_norm": 0.1591801792383194, "learning_rate": 2.6515811665495434e-05, "loss": 0.0239, "step": 14558 }, { "epoch": 10.231201686577652, "grad_norm": 0.23283344507217407, "learning_rate": 2.6515343171702975e-05, "loss": 0.0366, "step": 14559 }, { "epoch": 10.231904427266338, "grad_norm": 0.3040982484817505, "learning_rate": 2.651487467791052e-05, "loss": 0.0233, "step": 14560 }, { "epoch": 10.232607167955024, "grad_norm": 0.064811572432518, "learning_rate": 2.6514406184118062e-05, "loss": 0.0127, "step": 14561 }, { "epoch": 10.23330990864371, "grad_norm": 0.10143567621707916, "learning_rate": 2.6513937690325606e-05, "loss": 0.0168, "step": 14562 }, { "epoch": 10.234012649332396, "grad_norm": 0.1714969277381897, "learning_rate": 2.6513469196533146e-05, "loss": 0.0253, "step": 14563 }, { "epoch": 10.234715390021082, "grad_norm": 0.062017206102609634, "learning_rate": 2.651300070274069e-05, "loss": 0.0117, "step": 14564 }, { "epoch": 10.235418130709768, "grad_norm": 0.18071532249450684, "learning_rate": 2.6512532208948234e-05, "loss": 0.025, "step": 14565 }, { "epoch": 10.236120871398454, "grad_norm": 0.28109028935432434, "learning_rate": 2.6512063715155774e-05, "loss": 0.0368, "step": 14566 }, { "epoch": 10.23682361208714, "grad_norm": 0.15708638727664948, "learning_rate": 2.6511595221363318e-05, "loss": 0.014, "step": 14567 }, { "epoch": 10.237526352775825, "grad_norm": 0.0754752978682518, "learning_rate": 2.6511126727570858e-05, "loss": 0.009, "step": 14568 }, { "epoch": 10.238229093464511, "grad_norm": 0.09009875357151031, "learning_rate": 2.6510658233778402e-05, "loss": 0.0182, "step": 14569 }, { "epoch": 10.238931834153197, "grad_norm": 0.1719208061695099, "learning_rate": 2.6510189739985946e-05, "loss": 0.0179, "step": 14570 }, { "epoch": 10.239634574841883, "grad_norm": 0.14295944571495056, "learning_rate": 2.650972124619349e-05, "loss": 0.0306, "step": 14571 }, { "epoch": 10.240337315530569, "grad_norm": 0.28708505630493164, "learning_rate": 2.650925275240103e-05, "loss": 0.0302, "step": 14572 }, { "epoch": 10.241040056219255, "grad_norm": 0.14520399272441864, "learning_rate": 2.6508784258608573e-05, "loss": 0.0377, "step": 14573 }, { "epoch": 10.24174279690794, "grad_norm": 0.42689085006713867, "learning_rate": 2.6508315764816117e-05, "loss": 0.0405, "step": 14574 }, { "epoch": 10.242445537596627, "grad_norm": 0.2819131314754486, "learning_rate": 2.650784727102366e-05, "loss": 0.0447, "step": 14575 }, { "epoch": 10.243148278285313, "grad_norm": 0.3710349202156067, "learning_rate": 2.65073787772312e-05, "loss": 0.0875, "step": 14576 }, { "epoch": 10.243851018973999, "grad_norm": 0.3499823808670044, "learning_rate": 2.6506910283438745e-05, "loss": 0.0874, "step": 14577 }, { "epoch": 10.244553759662685, "grad_norm": 0.5811887383460999, "learning_rate": 2.650644178964629e-05, "loss": 0.1366, "step": 14578 }, { "epoch": 10.24525650035137, "grad_norm": 0.7301352620124817, "learning_rate": 2.6505973295853832e-05, "loss": 0.1575, "step": 14579 }, { "epoch": 10.245959241040056, "grad_norm": 1.3282968997955322, "learning_rate": 2.6505504802061376e-05, "loss": 0.1871, "step": 14580 }, { "epoch": 10.246661981728742, "grad_norm": 0.23040702939033508, "learning_rate": 2.6505036308268916e-05, "loss": 0.0616, "step": 14581 }, { "epoch": 10.247364722417428, "grad_norm": 0.3139895498752594, "learning_rate": 2.650456781447646e-05, "loss": 0.0432, "step": 14582 }, { "epoch": 10.248067463106114, "grad_norm": 0.12351097911596298, "learning_rate": 2.6504099320684004e-05, "loss": 0.0166, "step": 14583 }, { "epoch": 10.2487702037948, "grad_norm": 0.1475437730550766, "learning_rate": 2.6503630826891544e-05, "loss": 0.0268, "step": 14584 }, { "epoch": 10.249472944483486, "grad_norm": 0.1805453896522522, "learning_rate": 2.6503162333099084e-05, "loss": 0.0172, "step": 14585 }, { "epoch": 10.250175685172172, "grad_norm": 0.10690691322088242, "learning_rate": 2.6502693839306628e-05, "loss": 0.0148, "step": 14586 }, { "epoch": 10.250878425860858, "grad_norm": 0.13677740097045898, "learning_rate": 2.6502225345514172e-05, "loss": 0.0165, "step": 14587 }, { "epoch": 10.251581166549544, "grad_norm": 0.08833085745573044, "learning_rate": 2.6501756851721716e-05, "loss": 0.0139, "step": 14588 }, { "epoch": 10.25228390723823, "grad_norm": 0.11567298322916031, "learning_rate": 2.6501288357929256e-05, "loss": 0.013, "step": 14589 }, { "epoch": 10.252986647926916, "grad_norm": 0.09687955677509308, "learning_rate": 2.65008198641368e-05, "loss": 0.017, "step": 14590 }, { "epoch": 10.253689388615602, "grad_norm": 0.18408387899398804, "learning_rate": 2.6500351370344343e-05, "loss": 0.0175, "step": 14591 }, { "epoch": 10.254392129304287, "grad_norm": 0.07961820065975189, "learning_rate": 2.6499882876551887e-05, "loss": 0.0105, "step": 14592 }, { "epoch": 10.255094869992973, "grad_norm": 0.12983840703964233, "learning_rate": 2.649941438275943e-05, "loss": 0.0331, "step": 14593 }, { "epoch": 10.25579761068166, "grad_norm": 0.07353535294532776, "learning_rate": 2.649894588896697e-05, "loss": 0.0109, "step": 14594 }, { "epoch": 10.256500351370345, "grad_norm": 0.2371794581413269, "learning_rate": 2.6498477395174515e-05, "loss": 0.0262, "step": 14595 }, { "epoch": 10.25720309205903, "grad_norm": 0.1795428991317749, "learning_rate": 2.649800890138206e-05, "loss": 0.0292, "step": 14596 }, { "epoch": 10.257905832747715, "grad_norm": 0.15857289731502533, "learning_rate": 2.6497540407589602e-05, "loss": 0.0235, "step": 14597 }, { "epoch": 10.258608573436401, "grad_norm": 0.26632994413375854, "learning_rate": 2.6497071913797143e-05, "loss": 0.0221, "step": 14598 }, { "epoch": 10.259311314125087, "grad_norm": 0.29877278208732605, "learning_rate": 2.6496603420004686e-05, "loss": 0.0619, "step": 14599 }, { "epoch": 10.260014054813773, "grad_norm": 0.3039295971393585, "learning_rate": 2.649613492621223e-05, "loss": 0.0638, "step": 14600 }, { "epoch": 10.260716795502459, "grad_norm": 0.33275479078292847, "learning_rate": 2.649566643241977e-05, "loss": 0.0785, "step": 14601 }, { "epoch": 10.261419536191145, "grad_norm": 0.3749484121799469, "learning_rate": 2.649519793862731e-05, "loss": 0.1221, "step": 14602 }, { "epoch": 10.26212227687983, "grad_norm": 0.8814332485198975, "learning_rate": 2.6494729444834855e-05, "loss": 0.1501, "step": 14603 }, { "epoch": 10.262825017568517, "grad_norm": 0.9601648449897766, "learning_rate": 2.64942609510424e-05, "loss": 0.1714, "step": 14604 }, { "epoch": 10.263527758257203, "grad_norm": 1.668975830078125, "learning_rate": 2.6493792457249942e-05, "loss": 0.22, "step": 14605 }, { "epoch": 10.264230498945889, "grad_norm": 0.19524620473384857, "learning_rate": 2.6493323963457486e-05, "loss": 0.0559, "step": 14606 }, { "epoch": 10.264933239634574, "grad_norm": 0.13646911084651947, "learning_rate": 2.6492855469665026e-05, "loss": 0.0213, "step": 14607 }, { "epoch": 10.26563598032326, "grad_norm": 0.07763195782899857, "learning_rate": 2.649238697587257e-05, "loss": 0.0127, "step": 14608 }, { "epoch": 10.266338721011946, "grad_norm": 0.09912469238042831, "learning_rate": 2.6491918482080114e-05, "loss": 0.0172, "step": 14609 }, { "epoch": 10.267041461700632, "grad_norm": 0.18769916892051697, "learning_rate": 2.6491449988287657e-05, "loss": 0.0099, "step": 14610 }, { "epoch": 10.267744202389318, "grad_norm": 0.10334760695695877, "learning_rate": 2.6490981494495198e-05, "loss": 0.0156, "step": 14611 }, { "epoch": 10.268446943078004, "grad_norm": 0.20724698901176453, "learning_rate": 2.649051300070274e-05, "loss": 0.0384, "step": 14612 }, { "epoch": 10.26914968376669, "grad_norm": 0.13481824100017548, "learning_rate": 2.6490044506910285e-05, "loss": 0.023, "step": 14613 }, { "epoch": 10.269852424455376, "grad_norm": 0.1035667359828949, "learning_rate": 2.648957601311783e-05, "loss": 0.0138, "step": 14614 }, { "epoch": 10.270555165144062, "grad_norm": 0.1324407011270523, "learning_rate": 2.648910751932537e-05, "loss": 0.0149, "step": 14615 }, { "epoch": 10.271257905832748, "grad_norm": 0.2751377522945404, "learning_rate": 2.6488639025532913e-05, "loss": 0.0407, "step": 14616 }, { "epoch": 10.271960646521434, "grad_norm": 0.1949094831943512, "learning_rate": 2.6488170531740457e-05, "loss": 0.0194, "step": 14617 }, { "epoch": 10.27266338721012, "grad_norm": 0.12240201234817505, "learning_rate": 2.6487702037948e-05, "loss": 0.0252, "step": 14618 }, { "epoch": 10.273366127898806, "grad_norm": 0.09334330260753632, "learning_rate": 2.648723354415554e-05, "loss": 0.0123, "step": 14619 }, { "epoch": 10.274068868587491, "grad_norm": 0.13389047980308533, "learning_rate": 2.648676505036308e-05, "loss": 0.0264, "step": 14620 }, { "epoch": 10.274771609276177, "grad_norm": 0.22992803156375885, "learning_rate": 2.6486296556570625e-05, "loss": 0.0291, "step": 14621 }, { "epoch": 10.275474349964863, "grad_norm": 0.2280990034341812, "learning_rate": 2.648582806277817e-05, "loss": 0.0229, "step": 14622 }, { "epoch": 10.27617709065355, "grad_norm": 0.4182799756526947, "learning_rate": 2.6485359568985712e-05, "loss": 0.0237, "step": 14623 }, { "epoch": 10.276879831342235, "grad_norm": 0.1744837760925293, "learning_rate": 2.6484891075193252e-05, "loss": 0.0523, "step": 14624 }, { "epoch": 10.277582572030921, "grad_norm": 0.17542091012001038, "learning_rate": 2.6484422581400796e-05, "loss": 0.041, "step": 14625 }, { "epoch": 10.278285312719607, "grad_norm": 0.3984743654727936, "learning_rate": 2.648395408760834e-05, "loss": 0.0594, "step": 14626 }, { "epoch": 10.278988053408293, "grad_norm": 0.25537392497062683, "learning_rate": 2.6483485593815884e-05, "loss": 0.0753, "step": 14627 }, { "epoch": 10.279690794096979, "grad_norm": 0.5572576522827148, "learning_rate": 2.6483017100023424e-05, "loss": 0.1317, "step": 14628 }, { "epoch": 10.280393534785665, "grad_norm": 0.5547968745231628, "learning_rate": 2.6482548606230968e-05, "loss": 0.1949, "step": 14629 }, { "epoch": 10.28109627547435, "grad_norm": 1.7669637203216553, "learning_rate": 2.648208011243851e-05, "loss": 0.2136, "step": 14630 }, { "epoch": 10.281799016163037, "grad_norm": 0.15292485058307648, "learning_rate": 2.6481611618646055e-05, "loss": 0.0569, "step": 14631 }, { "epoch": 10.282501756851723, "grad_norm": 0.15002669394016266, "learning_rate": 2.64811431248536e-05, "loss": 0.0227, "step": 14632 }, { "epoch": 10.283204497540408, "grad_norm": 0.10518378019332886, "learning_rate": 2.648067463106114e-05, "loss": 0.0196, "step": 14633 }, { "epoch": 10.283907238229094, "grad_norm": 0.11695404350757599, "learning_rate": 2.6480206137268683e-05, "loss": 0.0152, "step": 14634 }, { "epoch": 10.284609978917779, "grad_norm": 0.09764941036701202, "learning_rate": 2.6479737643476227e-05, "loss": 0.0145, "step": 14635 }, { "epoch": 10.285312719606464, "grad_norm": 0.38416293263435364, "learning_rate": 2.6479269149683767e-05, "loss": 0.0387, "step": 14636 }, { "epoch": 10.28601546029515, "grad_norm": 0.14381998777389526, "learning_rate": 2.6478800655891307e-05, "loss": 0.014, "step": 14637 }, { "epoch": 10.286718200983836, "grad_norm": 0.1219077855348587, "learning_rate": 2.647833216209885e-05, "loss": 0.0206, "step": 14638 }, { "epoch": 10.287420941672522, "grad_norm": 0.1662501096725464, "learning_rate": 2.6477863668306395e-05, "loss": 0.0285, "step": 14639 }, { "epoch": 10.288123682361208, "grad_norm": 0.07281617820262909, "learning_rate": 2.647739517451394e-05, "loss": 0.012, "step": 14640 }, { "epoch": 10.288826423049894, "grad_norm": 0.16335231065750122, "learning_rate": 2.647692668072148e-05, "loss": 0.0178, "step": 14641 }, { "epoch": 10.28952916373858, "grad_norm": 0.09941022098064423, "learning_rate": 2.6476458186929023e-05, "loss": 0.0125, "step": 14642 }, { "epoch": 10.290231904427266, "grad_norm": 0.12883403897285461, "learning_rate": 2.6475989693136566e-05, "loss": 0.0274, "step": 14643 }, { "epoch": 10.290934645115952, "grad_norm": 0.13824033737182617, "learning_rate": 2.647552119934411e-05, "loss": 0.0216, "step": 14644 }, { "epoch": 10.291637385804638, "grad_norm": 0.17914621531963348, "learning_rate": 2.6475052705551654e-05, "loss": 0.0308, "step": 14645 }, { "epoch": 10.292340126493324, "grad_norm": 0.15958793461322784, "learning_rate": 2.6474584211759194e-05, "loss": 0.0465, "step": 14646 }, { "epoch": 10.29304286718201, "grad_norm": 0.16276034712791443, "learning_rate": 2.6474115717966738e-05, "loss": 0.0272, "step": 14647 }, { "epoch": 10.293745607870695, "grad_norm": 0.45922619104385376, "learning_rate": 2.647364722417428e-05, "loss": 0.043, "step": 14648 }, { "epoch": 10.294448348559381, "grad_norm": 0.7432966828346252, "learning_rate": 2.6473178730381825e-05, "loss": 0.0434, "step": 14649 }, { "epoch": 10.295151089248067, "grad_norm": 0.6245038509368896, "learning_rate": 2.6472710236589366e-05, "loss": 0.0479, "step": 14650 }, { "epoch": 10.295853829936753, "grad_norm": 0.2606273293495178, "learning_rate": 2.647224174279691e-05, "loss": 0.065, "step": 14651 }, { "epoch": 10.29655657062544, "grad_norm": 0.7282556295394897, "learning_rate": 2.6471773249004453e-05, "loss": 0.0987, "step": 14652 }, { "epoch": 10.297259311314125, "grad_norm": 0.5000314116477966, "learning_rate": 2.6471304755211993e-05, "loss": 0.1281, "step": 14653 }, { "epoch": 10.297962052002811, "grad_norm": 1.2405521869659424, "learning_rate": 2.6470836261419534e-05, "loss": 0.1961, "step": 14654 }, { "epoch": 10.298664792691497, "grad_norm": 0.7950864434242249, "learning_rate": 2.6470367767627077e-05, "loss": 0.1957, "step": 14655 }, { "epoch": 10.299367533380183, "grad_norm": 0.2784852683544159, "learning_rate": 2.646989927383462e-05, "loss": 0.0568, "step": 14656 }, { "epoch": 10.300070274068869, "grad_norm": 0.11865301430225372, "learning_rate": 2.6469430780042165e-05, "loss": 0.0228, "step": 14657 }, { "epoch": 10.300773014757555, "grad_norm": 0.09547274559736252, "learning_rate": 2.646896228624971e-05, "loss": 0.0143, "step": 14658 }, { "epoch": 10.30147575544624, "grad_norm": 0.16279762983322144, "learning_rate": 2.646849379245725e-05, "loss": 0.0146, "step": 14659 }, { "epoch": 10.302178496134927, "grad_norm": 0.1497311294078827, "learning_rate": 2.6468025298664793e-05, "loss": 0.0169, "step": 14660 }, { "epoch": 10.302881236823612, "grad_norm": 0.07472943514585495, "learning_rate": 2.6467556804872336e-05, "loss": 0.0075, "step": 14661 }, { "epoch": 10.303583977512298, "grad_norm": 0.17970336973667145, "learning_rate": 2.646708831107988e-05, "loss": 0.0119, "step": 14662 }, { "epoch": 10.304286718200984, "grad_norm": 0.257148802280426, "learning_rate": 2.646661981728742e-05, "loss": 0.0181, "step": 14663 }, { "epoch": 10.30498945888967, "grad_norm": 0.1252472698688507, "learning_rate": 2.6466151323494964e-05, "loss": 0.0181, "step": 14664 }, { "epoch": 10.305692199578356, "grad_norm": 0.18719318509101868, "learning_rate": 2.6465682829702508e-05, "loss": 0.0142, "step": 14665 }, { "epoch": 10.306394940267042, "grad_norm": 0.1835012137889862, "learning_rate": 2.646521433591005e-05, "loss": 0.0226, "step": 14666 }, { "epoch": 10.307097680955728, "grad_norm": 0.12693457305431366, "learning_rate": 2.6464745842117592e-05, "loss": 0.0116, "step": 14667 }, { "epoch": 10.307800421644414, "grad_norm": 0.16908881068229675, "learning_rate": 2.6464277348325136e-05, "loss": 0.0218, "step": 14668 }, { "epoch": 10.3085031623331, "grad_norm": 0.12036165595054626, "learning_rate": 2.646380885453268e-05, "loss": 0.0121, "step": 14669 }, { "epoch": 10.309205903021786, "grad_norm": 0.17434227466583252, "learning_rate": 2.6463340360740223e-05, "loss": 0.0223, "step": 14670 }, { "epoch": 10.309908643710472, "grad_norm": 0.1958373337984085, "learning_rate": 2.6462871866947764e-05, "loss": 0.0439, "step": 14671 }, { "epoch": 10.310611384399156, "grad_norm": 0.08075100183486938, "learning_rate": 2.6462403373155304e-05, "loss": 0.0128, "step": 14672 }, { "epoch": 10.311314125087842, "grad_norm": 0.339944988489151, "learning_rate": 2.6461934879362848e-05, "loss": 0.055, "step": 14673 }, { "epoch": 10.312016865776528, "grad_norm": 0.21372784674167633, "learning_rate": 2.646146638557039e-05, "loss": 0.0277, "step": 14674 }, { "epoch": 10.312719606465214, "grad_norm": 0.2082076072692871, "learning_rate": 2.6460997891777935e-05, "loss": 0.0454, "step": 14675 }, { "epoch": 10.3134223471539, "grad_norm": 0.2663792669773102, "learning_rate": 2.6460529397985475e-05, "loss": 0.0539, "step": 14676 }, { "epoch": 10.314125087842585, "grad_norm": 1.3738294839859009, "learning_rate": 2.646006090419302e-05, "loss": 0.087, "step": 14677 }, { "epoch": 10.314827828531271, "grad_norm": 0.6450603008270264, "learning_rate": 2.6459592410400563e-05, "loss": 0.1454, "step": 14678 }, { "epoch": 10.315530569219957, "grad_norm": 0.6836071014404297, "learning_rate": 2.6459123916608107e-05, "loss": 0.1648, "step": 14679 }, { "epoch": 10.316233309908643, "grad_norm": 1.0463999509811401, "learning_rate": 2.6458655422815647e-05, "loss": 0.21, "step": 14680 }, { "epoch": 10.316936050597329, "grad_norm": 0.22000283002853394, "learning_rate": 2.645818692902319e-05, "loss": 0.0692, "step": 14681 }, { "epoch": 10.317638791286015, "grad_norm": 0.16502045094966888, "learning_rate": 2.6457718435230734e-05, "loss": 0.0161, "step": 14682 }, { "epoch": 10.318341531974701, "grad_norm": 0.17715580761432648, "learning_rate": 2.6457249941438278e-05, "loss": 0.0333, "step": 14683 }, { "epoch": 10.319044272663387, "grad_norm": 0.16105327010154724, "learning_rate": 2.6456781447645822e-05, "loss": 0.0231, "step": 14684 }, { "epoch": 10.319747013352073, "grad_norm": 0.237768292427063, "learning_rate": 2.6456312953853362e-05, "loss": 0.0169, "step": 14685 }, { "epoch": 10.320449754040759, "grad_norm": 3.2616422176361084, "learning_rate": 2.6455844460060906e-05, "loss": 0.0107, "step": 14686 }, { "epoch": 10.321152494729445, "grad_norm": 0.12248166650533676, "learning_rate": 2.645537596626845e-05, "loss": 0.0187, "step": 14687 }, { "epoch": 10.32185523541813, "grad_norm": 0.14514733850955963, "learning_rate": 2.645490747247599e-05, "loss": 0.011, "step": 14688 }, { "epoch": 10.322557976106816, "grad_norm": 0.15223392844200134, "learning_rate": 2.645443897868353e-05, "loss": 0.0201, "step": 14689 }, { "epoch": 10.323260716795502, "grad_norm": 0.11528036743402481, "learning_rate": 2.6453970484891074e-05, "loss": 0.01, "step": 14690 }, { "epoch": 10.323963457484188, "grad_norm": 0.1382710337638855, "learning_rate": 2.6453501991098618e-05, "loss": 0.0257, "step": 14691 }, { "epoch": 10.324666198172874, "grad_norm": 0.259076863527298, "learning_rate": 2.645303349730616e-05, "loss": 0.0353, "step": 14692 }, { "epoch": 10.32536893886156, "grad_norm": 0.23287534713745117, "learning_rate": 2.6452565003513702e-05, "loss": 0.0278, "step": 14693 }, { "epoch": 10.326071679550246, "grad_norm": 0.10104335099458694, "learning_rate": 2.6452096509721245e-05, "loss": 0.014, "step": 14694 }, { "epoch": 10.326774420238932, "grad_norm": 0.20741336047649384, "learning_rate": 2.645162801592879e-05, "loss": 0.0281, "step": 14695 }, { "epoch": 10.327477160927618, "grad_norm": 0.08326876163482666, "learning_rate": 2.6451159522136333e-05, "loss": 0.0171, "step": 14696 }, { "epoch": 10.328179901616304, "grad_norm": 0.20284955203533173, "learning_rate": 2.6450691028343877e-05, "loss": 0.0256, "step": 14697 }, { "epoch": 10.32888264230499, "grad_norm": 0.15614330768585205, "learning_rate": 2.6450222534551417e-05, "loss": 0.0358, "step": 14698 }, { "epoch": 10.329585382993676, "grad_norm": 0.19196514785289764, "learning_rate": 2.644975404075896e-05, "loss": 0.03, "step": 14699 }, { "epoch": 10.330288123682362, "grad_norm": 0.30267333984375, "learning_rate": 2.6449285546966504e-05, "loss": 0.0418, "step": 14700 }, { "epoch": 10.330990864371048, "grad_norm": 0.5310239791870117, "learning_rate": 2.6448817053174048e-05, "loss": 0.0682, "step": 14701 }, { "epoch": 10.331693605059733, "grad_norm": 0.48222339153289795, "learning_rate": 2.644834855938159e-05, "loss": 0.1055, "step": 14702 }, { "epoch": 10.33239634574842, "grad_norm": 0.9254119396209717, "learning_rate": 2.6447880065589132e-05, "loss": 0.1502, "step": 14703 }, { "epoch": 10.333099086437105, "grad_norm": 1.0539828538894653, "learning_rate": 2.6447411571796676e-05, "loss": 0.1514, "step": 14704 }, { "epoch": 10.333801827125791, "grad_norm": 1.3016557693481445, "learning_rate": 2.644694307800422e-05, "loss": 0.2088, "step": 14705 }, { "epoch": 10.334504567814477, "grad_norm": 0.29714080691337585, "learning_rate": 2.644647458421176e-05, "loss": 0.0846, "step": 14706 }, { "epoch": 10.335207308503163, "grad_norm": 0.17356394231319427, "learning_rate": 2.64460060904193e-05, "loss": 0.0359, "step": 14707 }, { "epoch": 10.335910049191849, "grad_norm": 0.11472748219966888, "learning_rate": 2.6445537596626844e-05, "loss": 0.0165, "step": 14708 }, { "epoch": 10.336612789880535, "grad_norm": 0.15518173575401306, "learning_rate": 2.6445069102834388e-05, "loss": 0.0186, "step": 14709 }, { "epoch": 10.33731553056922, "grad_norm": 0.2312704473733902, "learning_rate": 2.644460060904193e-05, "loss": 0.0137, "step": 14710 }, { "epoch": 10.338018271257905, "grad_norm": 0.1168842762708664, "learning_rate": 2.6444132115249472e-05, "loss": 0.0127, "step": 14711 }, { "epoch": 10.33872101194659, "grad_norm": 0.0871841162443161, "learning_rate": 2.6443663621457016e-05, "loss": 0.0125, "step": 14712 }, { "epoch": 10.339423752635277, "grad_norm": 0.2826401889324188, "learning_rate": 2.644319512766456e-05, "loss": 0.0188, "step": 14713 }, { "epoch": 10.340126493323963, "grad_norm": 0.07801128923892975, "learning_rate": 2.6442726633872103e-05, "loss": 0.0189, "step": 14714 }, { "epoch": 10.340829234012649, "grad_norm": 0.08703626692295074, "learning_rate": 2.6442258140079643e-05, "loss": 0.012, "step": 14715 }, { "epoch": 10.341531974701335, "grad_norm": 0.1536606103181839, "learning_rate": 2.6441789646287187e-05, "loss": 0.0383, "step": 14716 }, { "epoch": 10.34223471539002, "grad_norm": 0.2177838683128357, "learning_rate": 2.644132115249473e-05, "loss": 0.016, "step": 14717 }, { "epoch": 10.342937456078706, "grad_norm": 0.15342694520950317, "learning_rate": 2.6440852658702275e-05, "loss": 0.0376, "step": 14718 }, { "epoch": 10.343640196767392, "grad_norm": 0.1647505909204483, "learning_rate": 2.6440384164909815e-05, "loss": 0.0105, "step": 14719 }, { "epoch": 10.344342937456078, "grad_norm": 0.15938174724578857, "learning_rate": 2.643991567111736e-05, "loss": 0.032, "step": 14720 }, { "epoch": 10.345045678144764, "grad_norm": 0.2082729935646057, "learning_rate": 2.6439447177324902e-05, "loss": 0.0435, "step": 14721 }, { "epoch": 10.34574841883345, "grad_norm": 0.16149914264678955, "learning_rate": 2.6438978683532446e-05, "loss": 0.0123, "step": 14722 }, { "epoch": 10.346451159522136, "grad_norm": 0.15180999040603638, "learning_rate": 2.6438510189739986e-05, "loss": 0.031, "step": 14723 }, { "epoch": 10.347153900210822, "grad_norm": 0.333000123500824, "learning_rate": 2.6438041695947527e-05, "loss": 0.0382, "step": 14724 }, { "epoch": 10.347856640899508, "grad_norm": 0.21077735722064972, "learning_rate": 2.643757320215507e-05, "loss": 0.0419, "step": 14725 }, { "epoch": 10.348559381588194, "grad_norm": 0.6283960342407227, "learning_rate": 2.6437104708362614e-05, "loss": 0.0789, "step": 14726 }, { "epoch": 10.34926212227688, "grad_norm": 0.8817362189292908, "learning_rate": 2.6436636214570158e-05, "loss": 0.0892, "step": 14727 }, { "epoch": 10.349964862965566, "grad_norm": 0.8720567226409912, "learning_rate": 2.6436167720777698e-05, "loss": 0.1482, "step": 14728 }, { "epoch": 10.350667603654252, "grad_norm": 0.6330703496932983, "learning_rate": 2.6435699226985242e-05, "loss": 0.1644, "step": 14729 }, { "epoch": 10.351370344342937, "grad_norm": 1.4081368446350098, "learning_rate": 2.6435230733192786e-05, "loss": 0.2675, "step": 14730 }, { "epoch": 10.352073085031623, "grad_norm": 0.22953207790851593, "learning_rate": 2.643476223940033e-05, "loss": 0.0668, "step": 14731 }, { "epoch": 10.35277582572031, "grad_norm": 0.13411809504032135, "learning_rate": 2.6434293745607873e-05, "loss": 0.0195, "step": 14732 }, { "epoch": 10.353478566408995, "grad_norm": 0.2197590172290802, "learning_rate": 2.6433825251815414e-05, "loss": 0.0238, "step": 14733 }, { "epoch": 10.354181307097681, "grad_norm": 0.24024958908557892, "learning_rate": 2.6433356758022957e-05, "loss": 0.0209, "step": 14734 }, { "epoch": 10.354884047786367, "grad_norm": 0.07247685641050339, "learning_rate": 2.64328882642305e-05, "loss": 0.009, "step": 14735 }, { "epoch": 10.355586788475053, "grad_norm": 0.10234438627958298, "learning_rate": 2.6432419770438045e-05, "loss": 0.0111, "step": 14736 }, { "epoch": 10.356289529163739, "grad_norm": 0.17499810457229614, "learning_rate": 2.6431951276645585e-05, "loss": 0.0216, "step": 14737 }, { "epoch": 10.356992269852425, "grad_norm": 0.08341839909553528, "learning_rate": 2.643148278285313e-05, "loss": 0.0061, "step": 14738 }, { "epoch": 10.35769501054111, "grad_norm": 0.1626545637845993, "learning_rate": 2.6431014289060672e-05, "loss": 0.0319, "step": 14739 }, { "epoch": 10.358397751229797, "grad_norm": 0.07040490955114365, "learning_rate": 2.6430545795268216e-05, "loss": 0.0069, "step": 14740 }, { "epoch": 10.359100491918483, "grad_norm": 0.12598244845867157, "learning_rate": 2.6430077301475753e-05, "loss": 0.0292, "step": 14741 }, { "epoch": 10.359803232607169, "grad_norm": 0.26793602108955383, "learning_rate": 2.6429608807683297e-05, "loss": 0.0168, "step": 14742 }, { "epoch": 10.360505973295854, "grad_norm": 0.16510435938835144, "learning_rate": 2.642914031389084e-05, "loss": 0.0206, "step": 14743 }, { "epoch": 10.36120871398454, "grad_norm": 0.10271968692541122, "learning_rate": 2.6428671820098384e-05, "loss": 0.009, "step": 14744 }, { "epoch": 10.361911454673226, "grad_norm": 0.17948321998119354, "learning_rate": 2.6428203326305928e-05, "loss": 0.0201, "step": 14745 }, { "epoch": 10.362614195361912, "grad_norm": 0.22469419240951538, "learning_rate": 2.642773483251347e-05, "loss": 0.0338, "step": 14746 }, { "epoch": 10.363316936050598, "grad_norm": 0.12704095244407654, "learning_rate": 2.6427266338721012e-05, "loss": 0.0171, "step": 14747 }, { "epoch": 10.364019676739284, "grad_norm": 0.23855435848236084, "learning_rate": 2.6426797844928556e-05, "loss": 0.029, "step": 14748 }, { "epoch": 10.36472241742797, "grad_norm": 0.3034380078315735, "learning_rate": 2.64263293511361e-05, "loss": 0.0484, "step": 14749 }, { "epoch": 10.365425158116654, "grad_norm": 1.0072046518325806, "learning_rate": 2.642586085734364e-05, "loss": 0.0479, "step": 14750 }, { "epoch": 10.36612789880534, "grad_norm": 0.628237783908844, "learning_rate": 2.6425392363551184e-05, "loss": 0.092, "step": 14751 }, { "epoch": 10.366830639494026, "grad_norm": 0.8166981339454651, "learning_rate": 2.6424923869758727e-05, "loss": 0.1397, "step": 14752 }, { "epoch": 10.367533380182712, "grad_norm": 0.7374987602233887, "learning_rate": 2.642445537596627e-05, "loss": 0.1553, "step": 14753 }, { "epoch": 10.368236120871398, "grad_norm": 0.8536091446876526, "learning_rate": 2.642398688217381e-05, "loss": 0.1646, "step": 14754 }, { "epoch": 10.368938861560084, "grad_norm": 2.5190610885620117, "learning_rate": 2.6423518388381355e-05, "loss": 0.2023, "step": 14755 }, { "epoch": 10.36964160224877, "grad_norm": 0.24807187914848328, "learning_rate": 2.64230498945889e-05, "loss": 0.0618, "step": 14756 }, { "epoch": 10.370344342937456, "grad_norm": 0.11582975834608078, "learning_rate": 2.6422581400796443e-05, "loss": 0.0243, "step": 14757 }, { "epoch": 10.371047083626141, "grad_norm": 0.10631250590085983, "learning_rate": 2.6422112907003983e-05, "loss": 0.0262, "step": 14758 }, { "epoch": 10.371749824314827, "grad_norm": 0.11830639094114304, "learning_rate": 2.6421644413211523e-05, "loss": 0.0185, "step": 14759 }, { "epoch": 10.372452565003513, "grad_norm": 0.11588204652070999, "learning_rate": 2.6421175919419067e-05, "loss": 0.019, "step": 14760 }, { "epoch": 10.3731553056922, "grad_norm": 0.1254023313522339, "learning_rate": 2.642070742562661e-05, "loss": 0.0103, "step": 14761 }, { "epoch": 10.373858046380885, "grad_norm": 0.16606375575065613, "learning_rate": 2.6420238931834154e-05, "loss": 0.015, "step": 14762 }, { "epoch": 10.374560787069571, "grad_norm": 0.20146037638187408, "learning_rate": 2.6419770438041695e-05, "loss": 0.0355, "step": 14763 }, { "epoch": 10.375263527758257, "grad_norm": 0.11618798226118088, "learning_rate": 2.641930194424924e-05, "loss": 0.0137, "step": 14764 }, { "epoch": 10.375966268446943, "grad_norm": 0.562448263168335, "learning_rate": 2.6418833450456782e-05, "loss": 0.0147, "step": 14765 }, { "epoch": 10.376669009135629, "grad_norm": 0.1798194944858551, "learning_rate": 2.6418364956664326e-05, "loss": 0.0378, "step": 14766 }, { "epoch": 10.377371749824315, "grad_norm": 0.13113290071487427, "learning_rate": 2.6417896462871866e-05, "loss": 0.0226, "step": 14767 }, { "epoch": 10.378074490513, "grad_norm": 0.10308510065078735, "learning_rate": 2.641742796907941e-05, "loss": 0.0208, "step": 14768 }, { "epoch": 10.378777231201687, "grad_norm": 0.08026125282049179, "learning_rate": 2.6416959475286954e-05, "loss": 0.014, "step": 14769 }, { "epoch": 10.379479971890373, "grad_norm": 0.16457830369472504, "learning_rate": 2.6416490981494497e-05, "loss": 0.0338, "step": 14770 }, { "epoch": 10.380182712579058, "grad_norm": 0.3078698217868805, "learning_rate": 2.641602248770204e-05, "loss": 0.0306, "step": 14771 }, { "epoch": 10.380885453267744, "grad_norm": 0.11735519021749496, "learning_rate": 2.641555399390958e-05, "loss": 0.0271, "step": 14772 }, { "epoch": 10.38158819395643, "grad_norm": 0.27772367000579834, "learning_rate": 2.6415085500117125e-05, "loss": 0.066, "step": 14773 }, { "epoch": 10.382290934645116, "grad_norm": 0.15439267456531525, "learning_rate": 2.641461700632467e-05, "loss": 0.0302, "step": 14774 }, { "epoch": 10.382993675333802, "grad_norm": 0.27580973505973816, "learning_rate": 2.641414851253221e-05, "loss": 0.034, "step": 14775 }, { "epoch": 10.383696416022488, "grad_norm": 0.9434952139854431, "learning_rate": 2.641368001873975e-05, "loss": 0.0734, "step": 14776 }, { "epoch": 10.384399156711174, "grad_norm": 0.4034990668296814, "learning_rate": 2.6413211524947293e-05, "loss": 0.0814, "step": 14777 }, { "epoch": 10.38510189739986, "grad_norm": 0.3829081952571869, "learning_rate": 2.6412743031154837e-05, "loss": 0.1307, "step": 14778 }, { "epoch": 10.385804638088546, "grad_norm": 0.8266057968139648, "learning_rate": 2.641227453736238e-05, "loss": 0.1564, "step": 14779 }, { "epoch": 10.386507378777232, "grad_norm": 1.149340271949768, "learning_rate": 2.641180604356992e-05, "loss": 0.1973, "step": 14780 }, { "epoch": 10.387210119465918, "grad_norm": 0.2567138373851776, "learning_rate": 2.6411337549777465e-05, "loss": 0.072, "step": 14781 }, { "epoch": 10.387912860154604, "grad_norm": 0.3086172938346863, "learning_rate": 2.641086905598501e-05, "loss": 0.0377, "step": 14782 }, { "epoch": 10.38861560084329, "grad_norm": 0.25280144810676575, "learning_rate": 2.6410400562192552e-05, "loss": 0.0469, "step": 14783 }, { "epoch": 10.389318341531975, "grad_norm": 0.15840230882167816, "learning_rate": 2.6409932068400096e-05, "loss": 0.0138, "step": 14784 }, { "epoch": 10.390021082220661, "grad_norm": 0.10874143242835999, "learning_rate": 2.6409463574607636e-05, "loss": 0.0135, "step": 14785 }, { "epoch": 10.390723822909347, "grad_norm": 0.08676303923130035, "learning_rate": 2.640899508081518e-05, "loss": 0.0144, "step": 14786 }, { "epoch": 10.391426563598033, "grad_norm": 0.1231534481048584, "learning_rate": 2.6408526587022724e-05, "loss": 0.014, "step": 14787 }, { "epoch": 10.392129304286719, "grad_norm": 0.10188975185155869, "learning_rate": 2.6408058093230268e-05, "loss": 0.0125, "step": 14788 }, { "epoch": 10.392832044975403, "grad_norm": 0.11571352183818817, "learning_rate": 2.6407589599437808e-05, "loss": 0.0192, "step": 14789 }, { "epoch": 10.39353478566409, "grad_norm": 0.1950584352016449, "learning_rate": 2.640712110564535e-05, "loss": 0.0155, "step": 14790 }, { "epoch": 10.394237526352775, "grad_norm": 0.15431201457977295, "learning_rate": 2.6406652611852895e-05, "loss": 0.0221, "step": 14791 }, { "epoch": 10.394940267041461, "grad_norm": 0.19514237344264984, "learning_rate": 2.640618411806044e-05, "loss": 0.0137, "step": 14792 }, { "epoch": 10.395643007730147, "grad_norm": 0.22690576314926147, "learning_rate": 2.6405715624267976e-05, "loss": 0.04, "step": 14793 }, { "epoch": 10.396345748418833, "grad_norm": 0.2991419732570648, "learning_rate": 2.640524713047552e-05, "loss": 0.0159, "step": 14794 }, { "epoch": 10.397048489107519, "grad_norm": 0.17922617495059967, "learning_rate": 2.6404778636683063e-05, "loss": 0.0307, "step": 14795 }, { "epoch": 10.397751229796205, "grad_norm": 0.11921904981136322, "learning_rate": 2.6404310142890607e-05, "loss": 0.019, "step": 14796 }, { "epoch": 10.39845397048489, "grad_norm": 0.1725301891565323, "learning_rate": 2.640384164909815e-05, "loss": 0.0244, "step": 14797 }, { "epoch": 10.399156711173577, "grad_norm": 0.18249163031578064, "learning_rate": 2.640337315530569e-05, "loss": 0.0327, "step": 14798 }, { "epoch": 10.399859451862262, "grad_norm": 0.4564245045185089, "learning_rate": 2.6402904661513235e-05, "loss": 0.0483, "step": 14799 }, { "epoch": 10.400562192550948, "grad_norm": 0.41738763451576233, "learning_rate": 2.640243616772078e-05, "loss": 0.0801, "step": 14800 }, { "epoch": 10.401264933239634, "grad_norm": 0.3082585632801056, "learning_rate": 2.6401967673928322e-05, "loss": 0.053, "step": 14801 }, { "epoch": 10.40196767392832, "grad_norm": 0.4450039863586426, "learning_rate": 2.6401499180135863e-05, "loss": 0.0818, "step": 14802 }, { "epoch": 10.402670414617006, "grad_norm": 0.42517122626304626, "learning_rate": 2.6401030686343407e-05, "loss": 0.1155, "step": 14803 }, { "epoch": 10.403373155305692, "grad_norm": 1.0689541101455688, "learning_rate": 2.640056219255095e-05, "loss": 0.1813, "step": 14804 }, { "epoch": 10.404075895994378, "grad_norm": 1.4022116661071777, "learning_rate": 2.6400093698758494e-05, "loss": 0.198, "step": 14805 }, { "epoch": 10.404778636683064, "grad_norm": 0.23531471192836761, "learning_rate": 2.6399625204966034e-05, "loss": 0.0763, "step": 14806 }, { "epoch": 10.40548137737175, "grad_norm": 0.5107874274253845, "learning_rate": 2.6399156711173578e-05, "loss": 0.0234, "step": 14807 }, { "epoch": 10.406184118060436, "grad_norm": 0.08484824001789093, "learning_rate": 2.6398688217381122e-05, "loss": 0.0172, "step": 14808 }, { "epoch": 10.406886858749122, "grad_norm": 0.1338045448064804, "learning_rate": 2.6398219723588665e-05, "loss": 0.0163, "step": 14809 }, { "epoch": 10.407589599437808, "grad_norm": 0.09686198830604553, "learning_rate": 2.6397751229796206e-05, "loss": 0.0148, "step": 14810 }, { "epoch": 10.408292340126494, "grad_norm": 0.08420535922050476, "learning_rate": 2.6397282736003746e-05, "loss": 0.0141, "step": 14811 }, { "epoch": 10.40899508081518, "grad_norm": 0.11279645562171936, "learning_rate": 2.639681424221129e-05, "loss": 0.0151, "step": 14812 }, { "epoch": 10.409697821503865, "grad_norm": 0.18866296112537384, "learning_rate": 2.6396345748418834e-05, "loss": 0.0253, "step": 14813 }, { "epoch": 10.410400562192551, "grad_norm": 0.1756231188774109, "learning_rate": 2.6395877254626377e-05, "loss": 0.03, "step": 14814 }, { "epoch": 10.411103302881237, "grad_norm": 0.27727240324020386, "learning_rate": 2.6395408760833918e-05, "loss": 0.0158, "step": 14815 }, { "epoch": 10.411806043569923, "grad_norm": 0.09561582654714584, "learning_rate": 2.639494026704146e-05, "loss": 0.0153, "step": 14816 }, { "epoch": 10.412508784258609, "grad_norm": 0.15852661430835724, "learning_rate": 2.6394471773249005e-05, "loss": 0.0228, "step": 14817 }, { "epoch": 10.413211524947295, "grad_norm": 0.21419492363929749, "learning_rate": 2.639400327945655e-05, "loss": 0.0252, "step": 14818 }, { "epoch": 10.41391426563598, "grad_norm": 0.19829945266246796, "learning_rate": 2.639353478566409e-05, "loss": 0.0261, "step": 14819 }, { "epoch": 10.414617006324667, "grad_norm": 0.18746346235275269, "learning_rate": 2.6393066291871633e-05, "loss": 0.0316, "step": 14820 }, { "epoch": 10.415319747013353, "grad_norm": 0.34174996614456177, "learning_rate": 2.6392597798079177e-05, "loss": 0.043, "step": 14821 }, { "epoch": 10.416022487702039, "grad_norm": 0.20636744797229767, "learning_rate": 2.639212930428672e-05, "loss": 0.0238, "step": 14822 }, { "epoch": 10.416725228390725, "grad_norm": 0.11995114386081696, "learning_rate": 2.6391660810494264e-05, "loss": 0.0249, "step": 14823 }, { "epoch": 10.41742796907941, "grad_norm": 0.3133507966995239, "learning_rate": 2.6391192316701804e-05, "loss": 0.0615, "step": 14824 }, { "epoch": 10.418130709768096, "grad_norm": 0.2704182267189026, "learning_rate": 2.6390723822909348e-05, "loss": 0.0551, "step": 14825 }, { "epoch": 10.41883345045678, "grad_norm": 0.32453441619873047, "learning_rate": 2.6390255329116892e-05, "loss": 0.083, "step": 14826 }, { "epoch": 10.419536191145466, "grad_norm": 0.4941439628601074, "learning_rate": 2.6389786835324436e-05, "loss": 0.0869, "step": 14827 }, { "epoch": 10.420238931834152, "grad_norm": 0.9769851565361023, "learning_rate": 2.6389318341531973e-05, "loss": 0.1546, "step": 14828 }, { "epoch": 10.420941672522838, "grad_norm": 0.6585216522216797, "learning_rate": 2.6388849847739516e-05, "loss": 0.1935, "step": 14829 }, { "epoch": 10.421644413211524, "grad_norm": 1.0487784147262573, "learning_rate": 2.638838135394706e-05, "loss": 0.2, "step": 14830 }, { "epoch": 10.42234715390021, "grad_norm": 0.5889256596565247, "learning_rate": 2.6387912860154604e-05, "loss": 0.0543, "step": 14831 }, { "epoch": 10.423049894588896, "grad_norm": 0.19290176033973694, "learning_rate": 2.6387444366362144e-05, "loss": 0.027, "step": 14832 }, { "epoch": 10.423752635277582, "grad_norm": 0.1105276346206665, "learning_rate": 2.6386975872569688e-05, "loss": 0.0203, "step": 14833 }, { "epoch": 10.424455375966268, "grad_norm": 0.16758686304092407, "learning_rate": 2.638650737877723e-05, "loss": 0.0262, "step": 14834 }, { "epoch": 10.425158116654954, "grad_norm": 0.1108354777097702, "learning_rate": 2.6386038884984775e-05, "loss": 0.0126, "step": 14835 }, { "epoch": 10.42586085734364, "grad_norm": 0.12924258410930634, "learning_rate": 2.638557039119232e-05, "loss": 0.0181, "step": 14836 }, { "epoch": 10.426563598032326, "grad_norm": 0.15985645353794098, "learning_rate": 2.638510189739986e-05, "loss": 0.0315, "step": 14837 }, { "epoch": 10.427266338721012, "grad_norm": 0.09083002060651779, "learning_rate": 2.6384633403607403e-05, "loss": 0.0133, "step": 14838 }, { "epoch": 10.427969079409698, "grad_norm": 0.43960484862327576, "learning_rate": 2.6384164909814947e-05, "loss": 0.0239, "step": 14839 }, { "epoch": 10.428671820098383, "grad_norm": 0.10281756520271301, "learning_rate": 2.638369641602249e-05, "loss": 0.007, "step": 14840 }, { "epoch": 10.42937456078707, "grad_norm": 0.17767484486103058, "learning_rate": 2.638322792223003e-05, "loss": 0.0267, "step": 14841 }, { "epoch": 10.430077301475755, "grad_norm": 0.10096414387226105, "learning_rate": 2.6382759428437575e-05, "loss": 0.0193, "step": 14842 }, { "epoch": 10.430780042164441, "grad_norm": 0.12484268099069595, "learning_rate": 2.6382290934645118e-05, "loss": 0.0342, "step": 14843 }, { "epoch": 10.431482782853127, "grad_norm": 0.15723083913326263, "learning_rate": 2.6381822440852662e-05, "loss": 0.0167, "step": 14844 }, { "epoch": 10.432185523541813, "grad_norm": 0.16511619091033936, "learning_rate": 2.63813539470602e-05, "loss": 0.0288, "step": 14845 }, { "epoch": 10.432888264230499, "grad_norm": 0.14354313910007477, "learning_rate": 2.6380885453267743e-05, "loss": 0.0234, "step": 14846 }, { "epoch": 10.433591004919185, "grad_norm": 0.20536111295223236, "learning_rate": 2.6380416959475286e-05, "loss": 0.0131, "step": 14847 }, { "epoch": 10.43429374560787, "grad_norm": 0.21887615323066711, "learning_rate": 2.637994846568283e-05, "loss": 0.0366, "step": 14848 }, { "epoch": 10.434996486296557, "grad_norm": 0.2494448721408844, "learning_rate": 2.6379479971890374e-05, "loss": 0.0357, "step": 14849 }, { "epoch": 10.435699226985243, "grad_norm": 0.3043329119682312, "learning_rate": 2.6379011478097914e-05, "loss": 0.0731, "step": 14850 }, { "epoch": 10.436401967673929, "grad_norm": 0.42852115631103516, "learning_rate": 2.6378542984305458e-05, "loss": 0.0885, "step": 14851 }, { "epoch": 10.437104708362615, "grad_norm": 0.4083644151687622, "learning_rate": 2.6378074490513e-05, "loss": 0.1077, "step": 14852 }, { "epoch": 10.4378074490513, "grad_norm": 0.9999057054519653, "learning_rate": 2.6377605996720545e-05, "loss": 0.1278, "step": 14853 }, { "epoch": 10.438510189739986, "grad_norm": 0.592602550983429, "learning_rate": 2.6377137502928086e-05, "loss": 0.1849, "step": 14854 }, { "epoch": 10.439212930428672, "grad_norm": 1.172469139099121, "learning_rate": 2.637666900913563e-05, "loss": 0.2179, "step": 14855 }, { "epoch": 10.439915671117358, "grad_norm": 0.1725732535123825, "learning_rate": 2.6376200515343173e-05, "loss": 0.0554, "step": 14856 }, { "epoch": 10.440618411806044, "grad_norm": 0.12363734841346741, "learning_rate": 2.6375732021550717e-05, "loss": 0.0231, "step": 14857 }, { "epoch": 10.44132115249473, "grad_norm": 0.2494160383939743, "learning_rate": 2.6375263527758257e-05, "loss": 0.0208, "step": 14858 }, { "epoch": 10.442023893183416, "grad_norm": 0.10287763923406601, "learning_rate": 2.63747950339658e-05, "loss": 0.0122, "step": 14859 }, { "epoch": 10.442726633872102, "grad_norm": 0.20035940408706665, "learning_rate": 2.6374326540173345e-05, "loss": 0.0217, "step": 14860 }, { "epoch": 10.443429374560788, "grad_norm": 0.07832714915275574, "learning_rate": 2.637385804638089e-05, "loss": 0.0067, "step": 14861 }, { "epoch": 10.444132115249474, "grad_norm": 0.1312599927186966, "learning_rate": 2.637338955258843e-05, "loss": 0.0175, "step": 14862 }, { "epoch": 10.44483485593816, "grad_norm": 0.07179433107376099, "learning_rate": 2.637292105879597e-05, "loss": 0.0093, "step": 14863 }, { "epoch": 10.445537596626846, "grad_norm": 0.10909023880958557, "learning_rate": 2.6372452565003513e-05, "loss": 0.014, "step": 14864 }, { "epoch": 10.44624033731553, "grad_norm": 0.2978835701942444, "learning_rate": 2.6371984071211056e-05, "loss": 0.0141, "step": 14865 }, { "epoch": 10.446943078004216, "grad_norm": 0.10125993192195892, "learning_rate": 2.63715155774186e-05, "loss": 0.0217, "step": 14866 }, { "epoch": 10.447645818692902, "grad_norm": 0.1714363843202591, "learning_rate": 2.637104708362614e-05, "loss": 0.0119, "step": 14867 }, { "epoch": 10.448348559381587, "grad_norm": 0.11953901499509811, "learning_rate": 2.6370578589833684e-05, "loss": 0.0219, "step": 14868 }, { "epoch": 10.449051300070273, "grad_norm": 0.1885925829410553, "learning_rate": 2.6370110096041228e-05, "loss": 0.0164, "step": 14869 }, { "epoch": 10.44975404075896, "grad_norm": 0.31504201889038086, "learning_rate": 2.6369641602248772e-05, "loss": 0.0352, "step": 14870 }, { "epoch": 10.450456781447645, "grad_norm": 0.24846160411834717, "learning_rate": 2.6369173108456312e-05, "loss": 0.0403, "step": 14871 }, { "epoch": 10.451159522136331, "grad_norm": 0.20346815884113312, "learning_rate": 2.6368704614663856e-05, "loss": 0.0182, "step": 14872 }, { "epoch": 10.451862262825017, "grad_norm": 0.268852561712265, "learning_rate": 2.63682361208714e-05, "loss": 0.0302, "step": 14873 }, { "epoch": 10.452565003513703, "grad_norm": 0.5395464897155762, "learning_rate": 2.6367767627078943e-05, "loss": 0.0429, "step": 14874 }, { "epoch": 10.453267744202389, "grad_norm": 0.23159533739089966, "learning_rate": 2.6367299133286487e-05, "loss": 0.049, "step": 14875 }, { "epoch": 10.453970484891075, "grad_norm": 0.557664692401886, "learning_rate": 2.6366830639494027e-05, "loss": 0.064, "step": 14876 }, { "epoch": 10.45467322557976, "grad_norm": 0.47741878032684326, "learning_rate": 2.636636214570157e-05, "loss": 0.1045, "step": 14877 }, { "epoch": 10.455375966268447, "grad_norm": 0.5466881394386292, "learning_rate": 2.6365893651909115e-05, "loss": 0.1474, "step": 14878 }, { "epoch": 10.456078706957133, "grad_norm": 0.6301732659339905, "learning_rate": 2.636542515811666e-05, "loss": 0.1702, "step": 14879 }, { "epoch": 10.456781447645819, "grad_norm": 1.093998670578003, "learning_rate": 2.6364956664324195e-05, "loss": 0.202, "step": 14880 }, { "epoch": 10.457484188334504, "grad_norm": 0.2546450197696686, "learning_rate": 2.636448817053174e-05, "loss": 0.066, "step": 14881 }, { "epoch": 10.45818692902319, "grad_norm": 0.3178653120994568, "learning_rate": 2.6364019676739283e-05, "loss": 0.0283, "step": 14882 }, { "epoch": 10.458889669711876, "grad_norm": 0.32870057225227356, "learning_rate": 2.6363551182946827e-05, "loss": 0.0225, "step": 14883 }, { "epoch": 10.459592410400562, "grad_norm": 0.17867767810821533, "learning_rate": 2.6363082689154367e-05, "loss": 0.0183, "step": 14884 }, { "epoch": 10.460295151089248, "grad_norm": 0.2070399820804596, "learning_rate": 2.636261419536191e-05, "loss": 0.0252, "step": 14885 }, { "epoch": 10.460997891777934, "grad_norm": 0.09348434209823608, "learning_rate": 2.6362145701569454e-05, "loss": 0.0106, "step": 14886 }, { "epoch": 10.46170063246662, "grad_norm": 0.10515117645263672, "learning_rate": 2.6361677207776998e-05, "loss": 0.0142, "step": 14887 }, { "epoch": 10.462403373155306, "grad_norm": 0.22555899620056152, "learning_rate": 2.6361208713984542e-05, "loss": 0.0174, "step": 14888 }, { "epoch": 10.463106113843992, "grad_norm": 0.18128551542758942, "learning_rate": 2.6360740220192082e-05, "loss": 0.03, "step": 14889 }, { "epoch": 10.463808854532678, "grad_norm": 0.46803009510040283, "learning_rate": 2.6360271726399626e-05, "loss": 0.0121, "step": 14890 }, { "epoch": 10.464511595221364, "grad_norm": 0.22948527336120605, "learning_rate": 2.635980323260717e-05, "loss": 0.0252, "step": 14891 }, { "epoch": 10.46521433591005, "grad_norm": 0.11853767931461334, "learning_rate": 2.6359334738814713e-05, "loss": 0.0122, "step": 14892 }, { "epoch": 10.465917076598735, "grad_norm": 0.11725768446922302, "learning_rate": 2.6358866245022254e-05, "loss": 0.0243, "step": 14893 }, { "epoch": 10.466619817287421, "grad_norm": 0.10439997911453247, "learning_rate": 2.6358397751229797e-05, "loss": 0.0188, "step": 14894 }, { "epoch": 10.467322557976107, "grad_norm": 0.18596795201301575, "learning_rate": 2.635792925743734e-05, "loss": 0.0204, "step": 14895 }, { "epoch": 10.468025298664793, "grad_norm": 0.3729921877384186, "learning_rate": 2.6357460763644885e-05, "loss": 0.0362, "step": 14896 }, { "epoch": 10.46872803935348, "grad_norm": 0.17067848145961761, "learning_rate": 2.6356992269852422e-05, "loss": 0.0164, "step": 14897 }, { "epoch": 10.469430780042165, "grad_norm": 0.4376174509525299, "learning_rate": 2.6356523776059966e-05, "loss": 0.0295, "step": 14898 }, { "epoch": 10.470133520730851, "grad_norm": 0.41526666283607483, "learning_rate": 2.635605528226751e-05, "loss": 0.0327, "step": 14899 }, { "epoch": 10.470836261419537, "grad_norm": 0.48555243015289307, "learning_rate": 2.6355586788475053e-05, "loss": 0.0514, "step": 14900 }, { "epoch": 10.471539002108223, "grad_norm": 0.47873425483703613, "learning_rate": 2.6355118294682597e-05, "loss": 0.0974, "step": 14901 }, { "epoch": 10.472241742796909, "grad_norm": 0.4515724778175354, "learning_rate": 2.6354649800890137e-05, "loss": 0.1042, "step": 14902 }, { "epoch": 10.472944483485595, "grad_norm": 0.7521071434020996, "learning_rate": 2.635418130709768e-05, "loss": 0.1395, "step": 14903 }, { "epoch": 10.473647224174279, "grad_norm": 0.7739269733428955, "learning_rate": 2.6353712813305224e-05, "loss": 0.1704, "step": 14904 }, { "epoch": 10.474349964862965, "grad_norm": 1.2727289199829102, "learning_rate": 2.6353244319512768e-05, "loss": 0.2005, "step": 14905 }, { "epoch": 10.47505270555165, "grad_norm": 0.1350044161081314, "learning_rate": 2.635277582572031e-05, "loss": 0.0458, "step": 14906 }, { "epoch": 10.475755446240337, "grad_norm": 0.11476792395114899, "learning_rate": 2.6352307331927852e-05, "loss": 0.0248, "step": 14907 }, { "epoch": 10.476458186929023, "grad_norm": 0.10939604789018631, "learning_rate": 2.6351838838135396e-05, "loss": 0.0136, "step": 14908 }, { "epoch": 10.477160927617708, "grad_norm": 0.14078401029109955, "learning_rate": 2.635137034434294e-05, "loss": 0.0212, "step": 14909 }, { "epoch": 10.477863668306394, "grad_norm": 0.3385842442512512, "learning_rate": 2.635090185055048e-05, "loss": 0.025, "step": 14910 }, { "epoch": 10.47856640899508, "grad_norm": 0.08622392266988754, "learning_rate": 2.6350433356758024e-05, "loss": 0.0073, "step": 14911 }, { "epoch": 10.479269149683766, "grad_norm": 0.11900045722723007, "learning_rate": 2.6349964862965568e-05, "loss": 0.0199, "step": 14912 }, { "epoch": 10.479971890372452, "grad_norm": 0.1522575169801712, "learning_rate": 2.634949636917311e-05, "loss": 0.028, "step": 14913 }, { "epoch": 10.480674631061138, "grad_norm": 0.17900201678276062, "learning_rate": 2.6349027875380655e-05, "loss": 0.0209, "step": 14914 }, { "epoch": 10.481377371749824, "grad_norm": 0.1829497069120407, "learning_rate": 2.6348559381588192e-05, "loss": 0.017, "step": 14915 }, { "epoch": 10.48208011243851, "grad_norm": 0.20417292416095734, "learning_rate": 2.6348090887795736e-05, "loss": 0.0249, "step": 14916 }, { "epoch": 10.482782853127196, "grad_norm": 0.1385318487882614, "learning_rate": 2.634762239400328e-05, "loss": 0.0141, "step": 14917 }, { "epoch": 10.483485593815882, "grad_norm": 0.26894500851631165, "learning_rate": 2.6347153900210823e-05, "loss": 0.0452, "step": 14918 }, { "epoch": 10.484188334504568, "grad_norm": 0.11649224162101746, "learning_rate": 2.6346685406418363e-05, "loss": 0.0175, "step": 14919 }, { "epoch": 10.484891075193254, "grad_norm": 0.16903823614120483, "learning_rate": 2.6346216912625907e-05, "loss": 0.0365, "step": 14920 }, { "epoch": 10.48559381588194, "grad_norm": 0.16736240684986115, "learning_rate": 2.634574841883345e-05, "loss": 0.0322, "step": 14921 }, { "epoch": 10.486296556570625, "grad_norm": 0.20127113163471222, "learning_rate": 2.6345279925040995e-05, "loss": 0.0294, "step": 14922 }, { "epoch": 10.486999297259311, "grad_norm": 0.1663280874490738, "learning_rate": 2.634481143124854e-05, "loss": 0.0221, "step": 14923 }, { "epoch": 10.487702037947997, "grad_norm": 0.22057712078094482, "learning_rate": 2.634434293745608e-05, "loss": 0.044, "step": 14924 }, { "epoch": 10.488404778636683, "grad_norm": 0.20662672817707062, "learning_rate": 2.6343874443663622e-05, "loss": 0.0418, "step": 14925 }, { "epoch": 10.489107519325369, "grad_norm": 0.256089448928833, "learning_rate": 2.6343405949871166e-05, "loss": 0.0757, "step": 14926 }, { "epoch": 10.489810260014055, "grad_norm": 0.2897716760635376, "learning_rate": 2.634293745607871e-05, "loss": 0.107, "step": 14927 }, { "epoch": 10.490513000702741, "grad_norm": 0.5601919889450073, "learning_rate": 2.634246896228625e-05, "loss": 0.1558, "step": 14928 }, { "epoch": 10.491215741391427, "grad_norm": 1.1536341905593872, "learning_rate": 2.6342000468493794e-05, "loss": 0.1645, "step": 14929 }, { "epoch": 10.491918482080113, "grad_norm": 1.1296201944351196, "learning_rate": 2.6341531974701338e-05, "loss": 0.2155, "step": 14930 }, { "epoch": 10.492621222768799, "grad_norm": 0.2476913034915924, "learning_rate": 2.634106348090888e-05, "loss": 0.0675, "step": 14931 }, { "epoch": 10.493323963457485, "grad_norm": 0.14613619446754456, "learning_rate": 2.634059498711642e-05, "loss": 0.0196, "step": 14932 }, { "epoch": 10.49402670414617, "grad_norm": 0.12604418396949768, "learning_rate": 2.6340126493323962e-05, "loss": 0.0255, "step": 14933 }, { "epoch": 10.494729444834856, "grad_norm": 0.10067915171384811, "learning_rate": 2.6339657999531506e-05, "loss": 0.0138, "step": 14934 }, { "epoch": 10.495432185523542, "grad_norm": 0.1391066014766693, "learning_rate": 2.633918950573905e-05, "loss": 0.0334, "step": 14935 }, { "epoch": 10.496134926212228, "grad_norm": 0.06143584102392197, "learning_rate": 2.6338721011946593e-05, "loss": 0.0122, "step": 14936 }, { "epoch": 10.496837666900914, "grad_norm": 0.09289709478616714, "learning_rate": 2.6338252518154134e-05, "loss": 0.0108, "step": 14937 }, { "epoch": 10.4975404075896, "grad_norm": 0.29262879490852356, "learning_rate": 2.6337784024361677e-05, "loss": 0.0273, "step": 14938 }, { "epoch": 10.498243148278286, "grad_norm": 0.42520490288734436, "learning_rate": 2.633731553056922e-05, "loss": 0.0238, "step": 14939 }, { "epoch": 10.498945888966972, "grad_norm": 0.1215323880314827, "learning_rate": 2.6336847036776765e-05, "loss": 0.0081, "step": 14940 }, { "epoch": 10.499648629655656, "grad_norm": 0.18126055598258972, "learning_rate": 2.6336378542984305e-05, "loss": 0.0271, "step": 14941 }, { "epoch": 10.500351370344344, "grad_norm": 0.16296899318695068, "learning_rate": 2.633591004919185e-05, "loss": 0.0143, "step": 14942 }, { "epoch": 10.501054111033028, "grad_norm": 0.33011776208877563, "learning_rate": 2.6335441555399393e-05, "loss": 0.0436, "step": 14943 }, { "epoch": 10.501756851721714, "grad_norm": 0.21063341200351715, "learning_rate": 2.6334973061606936e-05, "loss": 0.0193, "step": 14944 }, { "epoch": 10.5024595924104, "grad_norm": 0.1987699270248413, "learning_rate": 2.6334504567814477e-05, "loss": 0.0227, "step": 14945 }, { "epoch": 10.503162333099086, "grad_norm": 0.1607043594121933, "learning_rate": 2.633403607402202e-05, "loss": 0.0405, "step": 14946 }, { "epoch": 10.503865073787772, "grad_norm": 0.19583439826965332, "learning_rate": 2.6333567580229564e-05, "loss": 0.0178, "step": 14947 }, { "epoch": 10.504567814476458, "grad_norm": 0.31449225544929504, "learning_rate": 2.6333099086437108e-05, "loss": 0.0346, "step": 14948 }, { "epoch": 10.505270555165144, "grad_norm": 0.2871105968952179, "learning_rate": 2.633263059264465e-05, "loss": 0.033, "step": 14949 }, { "epoch": 10.50597329585383, "grad_norm": 0.40420493483543396, "learning_rate": 2.633216209885219e-05, "loss": 0.0763, "step": 14950 }, { "epoch": 10.506676036542515, "grad_norm": 0.2996320426464081, "learning_rate": 2.6331693605059732e-05, "loss": 0.0827, "step": 14951 }, { "epoch": 10.507378777231201, "grad_norm": 0.35372912883758545, "learning_rate": 2.6331225111267276e-05, "loss": 0.0655, "step": 14952 }, { "epoch": 10.508081517919887, "grad_norm": 0.48504623770713806, "learning_rate": 2.633075661747482e-05, "loss": 0.1596, "step": 14953 }, { "epoch": 10.508784258608573, "grad_norm": 0.6383324861526489, "learning_rate": 2.633028812368236e-05, "loss": 0.1895, "step": 14954 }, { "epoch": 10.509486999297259, "grad_norm": 1.48545241355896, "learning_rate": 2.6329819629889904e-05, "loss": 0.2047, "step": 14955 }, { "epoch": 10.510189739985945, "grad_norm": 0.24678847193717957, "learning_rate": 2.6329351136097447e-05, "loss": 0.0441, "step": 14956 }, { "epoch": 10.510892480674631, "grad_norm": 0.3008878529071808, "learning_rate": 2.632888264230499e-05, "loss": 0.0154, "step": 14957 }, { "epoch": 10.511595221363317, "grad_norm": 0.27383214235305786, "learning_rate": 2.632841414851253e-05, "loss": 0.0361, "step": 14958 }, { "epoch": 10.512297962052003, "grad_norm": 0.1513180434703827, "learning_rate": 2.6327945654720075e-05, "loss": 0.0177, "step": 14959 }, { "epoch": 10.513000702740689, "grad_norm": 0.16720041632652283, "learning_rate": 2.632747716092762e-05, "loss": 0.0248, "step": 14960 }, { "epoch": 10.513703443429375, "grad_norm": 0.1930844485759735, "learning_rate": 2.6327008667135163e-05, "loss": 0.0141, "step": 14961 }, { "epoch": 10.51440618411806, "grad_norm": 0.1364094316959381, "learning_rate": 2.6326540173342706e-05, "loss": 0.0133, "step": 14962 }, { "epoch": 10.515108924806746, "grad_norm": 0.12904763221740723, "learning_rate": 2.6326071679550247e-05, "loss": 0.0167, "step": 14963 }, { "epoch": 10.515811665495432, "grad_norm": 0.13819937407970428, "learning_rate": 2.632560318575779e-05, "loss": 0.0165, "step": 14964 }, { "epoch": 10.516514406184118, "grad_norm": 0.26356056332588196, "learning_rate": 2.6325134691965334e-05, "loss": 0.0347, "step": 14965 }, { "epoch": 10.517217146872804, "grad_norm": 0.17359061539173126, "learning_rate": 2.6324666198172878e-05, "loss": 0.0169, "step": 14966 }, { "epoch": 10.51791988756149, "grad_norm": 0.09493906050920486, "learning_rate": 2.6324197704380415e-05, "loss": 0.0156, "step": 14967 }, { "epoch": 10.518622628250176, "grad_norm": 0.16859091818332672, "learning_rate": 2.632372921058796e-05, "loss": 0.0325, "step": 14968 }, { "epoch": 10.519325368938862, "grad_norm": 0.11231319606304169, "learning_rate": 2.6323260716795502e-05, "loss": 0.0157, "step": 14969 }, { "epoch": 10.520028109627548, "grad_norm": 0.1534031331539154, "learning_rate": 2.6322792223003046e-05, "loss": 0.0319, "step": 14970 }, { "epoch": 10.520730850316234, "grad_norm": 0.11620374768972397, "learning_rate": 2.6322323729210586e-05, "loss": 0.0119, "step": 14971 }, { "epoch": 10.52143359100492, "grad_norm": 0.2062847763299942, "learning_rate": 2.632185523541813e-05, "loss": 0.0395, "step": 14972 }, { "epoch": 10.522136331693606, "grad_norm": 0.25173163414001465, "learning_rate": 2.6321386741625674e-05, "loss": 0.0278, "step": 14973 }, { "epoch": 10.522839072382292, "grad_norm": 0.2201007753610611, "learning_rate": 2.6320918247833217e-05, "loss": 0.0563, "step": 14974 }, { "epoch": 10.523541813070977, "grad_norm": 0.20475460588932037, "learning_rate": 2.632044975404076e-05, "loss": 0.0374, "step": 14975 }, { "epoch": 10.524244553759663, "grad_norm": 0.5120196342468262, "learning_rate": 2.63199812602483e-05, "loss": 0.0874, "step": 14976 }, { "epoch": 10.52494729444835, "grad_norm": 0.4220435917377472, "learning_rate": 2.6319512766455845e-05, "loss": 0.1033, "step": 14977 }, { "epoch": 10.525650035137035, "grad_norm": 0.6545377373695374, "learning_rate": 2.631904427266339e-05, "loss": 0.1335, "step": 14978 }, { "epoch": 10.526352775825721, "grad_norm": 0.53370600938797, "learning_rate": 2.6318575778870933e-05, "loss": 0.1584, "step": 14979 }, { "epoch": 10.527055516514405, "grad_norm": 1.543756127357483, "learning_rate": 2.6318107285078473e-05, "loss": 0.1997, "step": 14980 }, { "epoch": 10.527758257203093, "grad_norm": 0.1615339070558548, "learning_rate": 2.6317638791286017e-05, "loss": 0.0592, "step": 14981 }, { "epoch": 10.528460997891777, "grad_norm": 0.1064717248082161, "learning_rate": 2.631717029749356e-05, "loss": 0.0221, "step": 14982 }, { "epoch": 10.529163738580463, "grad_norm": 0.11256752163171768, "learning_rate": 2.6316701803701104e-05, "loss": 0.0209, "step": 14983 }, { "epoch": 10.529866479269149, "grad_norm": 0.11609658598899841, "learning_rate": 2.631623330990864e-05, "loss": 0.0162, "step": 14984 }, { "epoch": 10.530569219957835, "grad_norm": 0.09957461804151535, "learning_rate": 2.6315764816116185e-05, "loss": 0.0182, "step": 14985 }, { "epoch": 10.53127196064652, "grad_norm": 0.13296425342559814, "learning_rate": 2.631529632232373e-05, "loss": 0.0115, "step": 14986 }, { "epoch": 10.531974701335207, "grad_norm": 0.15340591967105865, "learning_rate": 2.6314827828531272e-05, "loss": 0.0353, "step": 14987 }, { "epoch": 10.532677442023893, "grad_norm": 0.12995494902133942, "learning_rate": 2.6314359334738816e-05, "loss": 0.0109, "step": 14988 }, { "epoch": 10.533380182712579, "grad_norm": 0.10130122303962708, "learning_rate": 2.6313890840946356e-05, "loss": 0.0199, "step": 14989 }, { "epoch": 10.534082923401265, "grad_norm": 0.3233691155910492, "learning_rate": 2.63134223471539e-05, "loss": 0.0154, "step": 14990 }, { "epoch": 10.53478566408995, "grad_norm": 0.3559563457965851, "learning_rate": 2.6312953853361444e-05, "loss": 0.022, "step": 14991 }, { "epoch": 10.535488404778636, "grad_norm": 0.08986064791679382, "learning_rate": 2.6312485359568988e-05, "loss": 0.0137, "step": 14992 }, { "epoch": 10.536191145467322, "grad_norm": 0.21161971986293793, "learning_rate": 2.6312016865776528e-05, "loss": 0.0269, "step": 14993 }, { "epoch": 10.536893886156008, "grad_norm": 0.09794479608535767, "learning_rate": 2.631154837198407e-05, "loss": 0.0118, "step": 14994 }, { "epoch": 10.537596626844694, "grad_norm": 0.22752246260643005, "learning_rate": 2.6311079878191615e-05, "loss": 0.0307, "step": 14995 }, { "epoch": 10.53829936753338, "grad_norm": 0.20368194580078125, "learning_rate": 2.631061138439916e-05, "loss": 0.0308, "step": 14996 }, { "epoch": 10.539002108222066, "grad_norm": 0.15073734521865845, "learning_rate": 2.63101428906067e-05, "loss": 0.0192, "step": 14997 }, { "epoch": 10.539704848910752, "grad_norm": 0.19169694185256958, "learning_rate": 2.6309674396814243e-05, "loss": 0.0323, "step": 14998 }, { "epoch": 10.540407589599438, "grad_norm": 0.5091800093650818, "learning_rate": 2.6309205903021787e-05, "loss": 0.0543, "step": 14999 }, { "epoch": 10.541110330288124, "grad_norm": 0.22888542711734772, "learning_rate": 2.630873740922933e-05, "loss": 0.0439, "step": 15000 }, { "epoch": 10.541110330288124, "eval_cer": 0.19407438927258247, "eval_loss": 0.26739028096199036, "eval_runtime": 18.753, "eval_samples_per_second": 241.988, "eval_steps_per_second": 0.8, "eval_wer": 0.3484260216195078, "step": 15000 }, { "epoch": 10.54181307097681, "grad_norm": 0.4216839075088501, "learning_rate": 2.6308268915436874e-05, "loss": 0.0776, "step": 15001 }, { "epoch": 10.542515811665496, "grad_norm": 0.3670951724052429, "learning_rate": 2.630780042164441e-05, "loss": 0.0747, "step": 15002 }, { "epoch": 10.543218552354181, "grad_norm": 0.8062164783477783, "learning_rate": 2.6307331927851955e-05, "loss": 0.1559, "step": 15003 }, { "epoch": 10.543921293042867, "grad_norm": 1.5847173929214478, "learning_rate": 2.63068634340595e-05, "loss": 0.1635, "step": 15004 }, { "epoch": 10.544624033731553, "grad_norm": 0.8428027629852295, "learning_rate": 2.6306394940267042e-05, "loss": 0.2381, "step": 15005 }, { "epoch": 10.54532677442024, "grad_norm": 0.23224499821662903, "learning_rate": 2.6305926446474583e-05, "loss": 0.0549, "step": 15006 }, { "epoch": 10.546029515108925, "grad_norm": 0.1404208242893219, "learning_rate": 2.6305457952682127e-05, "loss": 0.0304, "step": 15007 }, { "epoch": 10.546732255797611, "grad_norm": 0.11010179668664932, "learning_rate": 2.630498945888967e-05, "loss": 0.0128, "step": 15008 }, { "epoch": 10.547434996486297, "grad_norm": 0.08524324744939804, "learning_rate": 2.6304520965097214e-05, "loss": 0.01, "step": 15009 }, { "epoch": 10.548137737174983, "grad_norm": 0.09562046825885773, "learning_rate": 2.6304052471304754e-05, "loss": 0.0132, "step": 15010 }, { "epoch": 10.548840477863669, "grad_norm": 0.09593573957681656, "learning_rate": 2.6303583977512298e-05, "loss": 0.0163, "step": 15011 }, { "epoch": 10.549543218552355, "grad_norm": 0.198017418384552, "learning_rate": 2.6303115483719842e-05, "loss": 0.0211, "step": 15012 }, { "epoch": 10.55024595924104, "grad_norm": 0.2397717982530594, "learning_rate": 2.6302646989927386e-05, "loss": 0.0261, "step": 15013 }, { "epoch": 10.550948699929727, "grad_norm": 0.15952032804489136, "learning_rate": 2.630217849613493e-05, "loss": 0.0177, "step": 15014 }, { "epoch": 10.551651440618413, "grad_norm": 0.16521194577217102, "learning_rate": 2.630171000234247e-05, "loss": 0.0195, "step": 15015 }, { "epoch": 10.552354181307098, "grad_norm": 0.49340999126434326, "learning_rate": 2.6301241508550013e-05, "loss": 0.0333, "step": 15016 }, { "epoch": 10.553056921995784, "grad_norm": 0.11150368303060532, "learning_rate": 2.6300773014757557e-05, "loss": 0.018, "step": 15017 }, { "epoch": 10.55375966268447, "grad_norm": 0.2214580625295639, "learning_rate": 2.63003045209651e-05, "loss": 0.0394, "step": 15018 }, { "epoch": 10.554462403373154, "grad_norm": 0.47342783212661743, "learning_rate": 2.6299836027172638e-05, "loss": 0.02, "step": 15019 }, { "epoch": 10.55516514406184, "grad_norm": 0.21392227709293365, "learning_rate": 2.629936753338018e-05, "loss": 0.0254, "step": 15020 }, { "epoch": 10.555867884750526, "grad_norm": 0.2055204212665558, "learning_rate": 2.6298899039587725e-05, "loss": 0.0448, "step": 15021 }, { "epoch": 10.556570625439212, "grad_norm": 0.10467799007892609, "learning_rate": 2.629843054579527e-05, "loss": 0.0191, "step": 15022 }, { "epoch": 10.557273366127898, "grad_norm": 0.1326936036348343, "learning_rate": 2.629796205200281e-05, "loss": 0.0248, "step": 15023 }, { "epoch": 10.557976106816584, "grad_norm": 0.12732680141925812, "learning_rate": 2.6297493558210353e-05, "loss": 0.0237, "step": 15024 }, { "epoch": 10.55867884750527, "grad_norm": 0.45239198207855225, "learning_rate": 2.6297025064417897e-05, "loss": 0.0395, "step": 15025 }, { "epoch": 10.559381588193956, "grad_norm": 0.5078778266906738, "learning_rate": 2.629655657062544e-05, "loss": 0.0701, "step": 15026 }, { "epoch": 10.560084328882642, "grad_norm": 0.38910090923309326, "learning_rate": 2.6296088076832984e-05, "loss": 0.1085, "step": 15027 }, { "epoch": 10.560787069571328, "grad_norm": 0.7549901604652405, "learning_rate": 2.6295619583040524e-05, "loss": 0.1243, "step": 15028 }, { "epoch": 10.561489810260014, "grad_norm": 0.4788980484008789, "learning_rate": 2.6295151089248068e-05, "loss": 0.1566, "step": 15029 }, { "epoch": 10.5621925509487, "grad_norm": 0.6947950720787048, "learning_rate": 2.6294682595455612e-05, "loss": 0.2246, "step": 15030 }, { "epoch": 10.562895291637385, "grad_norm": 0.23500151932239532, "learning_rate": 2.6294214101663156e-05, "loss": 0.0663, "step": 15031 }, { "epoch": 10.563598032326071, "grad_norm": 0.3311825096607208, "learning_rate": 2.6293745607870696e-05, "loss": 0.028, "step": 15032 }, { "epoch": 10.564300773014757, "grad_norm": 0.09813155233860016, "learning_rate": 2.629327711407824e-05, "loss": 0.0257, "step": 15033 }, { "epoch": 10.565003513703443, "grad_norm": 0.10953304171562195, "learning_rate": 2.6292808620285783e-05, "loss": 0.0205, "step": 15034 }, { "epoch": 10.56570625439213, "grad_norm": 0.08937166631221771, "learning_rate": 2.6292340126493327e-05, "loss": 0.0162, "step": 15035 }, { "epoch": 10.566408995080815, "grad_norm": 0.11953681707382202, "learning_rate": 2.6291871632700867e-05, "loss": 0.0099, "step": 15036 }, { "epoch": 10.567111735769501, "grad_norm": 0.18948259949684143, "learning_rate": 2.6291403138908408e-05, "loss": 0.0139, "step": 15037 }, { "epoch": 10.567814476458187, "grad_norm": 0.1385185271501541, "learning_rate": 2.629093464511595e-05, "loss": 0.0236, "step": 15038 }, { "epoch": 10.568517217146873, "grad_norm": 0.13085035979747772, "learning_rate": 2.6290466151323495e-05, "loss": 0.0252, "step": 15039 }, { "epoch": 10.569219957835559, "grad_norm": 0.09954393655061722, "learning_rate": 2.628999765753104e-05, "loss": 0.0092, "step": 15040 }, { "epoch": 10.569922698524245, "grad_norm": 0.12004972249269485, "learning_rate": 2.628952916373858e-05, "loss": 0.0291, "step": 15041 }, { "epoch": 10.57062543921293, "grad_norm": 0.13431811332702637, "learning_rate": 2.6289060669946123e-05, "loss": 0.0233, "step": 15042 }, { "epoch": 10.571328179901617, "grad_norm": 0.1536276638507843, "learning_rate": 2.6288592176153667e-05, "loss": 0.0221, "step": 15043 }, { "epoch": 10.572030920590302, "grad_norm": 0.21658490598201752, "learning_rate": 2.628812368236121e-05, "loss": 0.0139, "step": 15044 }, { "epoch": 10.572733661278988, "grad_norm": 0.17863312363624573, "learning_rate": 2.628765518856875e-05, "loss": 0.0376, "step": 15045 }, { "epoch": 10.573436401967674, "grad_norm": 0.7246469855308533, "learning_rate": 2.6287186694776295e-05, "loss": 0.0323, "step": 15046 }, { "epoch": 10.57413914265636, "grad_norm": 0.1361309438943863, "learning_rate": 2.6286718200983838e-05, "loss": 0.0184, "step": 15047 }, { "epoch": 10.574841883345046, "grad_norm": 0.11694975942373276, "learning_rate": 2.6286249707191382e-05, "loss": 0.0338, "step": 15048 }, { "epoch": 10.575544624033732, "grad_norm": 0.196768119931221, "learning_rate": 2.6285781213398922e-05, "loss": 0.0435, "step": 15049 }, { "epoch": 10.576247364722418, "grad_norm": 0.2381603866815567, "learning_rate": 2.6285312719606466e-05, "loss": 0.0409, "step": 15050 }, { "epoch": 10.576950105411104, "grad_norm": 0.46806800365448, "learning_rate": 2.628484422581401e-05, "loss": 0.0821, "step": 15051 }, { "epoch": 10.57765284609979, "grad_norm": 0.9343453645706177, "learning_rate": 2.6284375732021554e-05, "loss": 0.1106, "step": 15052 }, { "epoch": 10.578355586788476, "grad_norm": 0.39038702845573425, "learning_rate": 2.6283907238229097e-05, "loss": 0.1318, "step": 15053 }, { "epoch": 10.579058327477162, "grad_norm": 1.4349911212921143, "learning_rate": 2.6283438744436634e-05, "loss": 0.1652, "step": 15054 }, { "epoch": 10.579761068165848, "grad_norm": 0.9720922708511353, "learning_rate": 2.6282970250644178e-05, "loss": 0.1982, "step": 15055 }, { "epoch": 10.580463808854532, "grad_norm": 0.3817174732685089, "learning_rate": 2.628250175685172e-05, "loss": 0.0551, "step": 15056 }, { "epoch": 10.58116654954322, "grad_norm": 0.13933320343494415, "learning_rate": 2.6282033263059265e-05, "loss": 0.0399, "step": 15057 }, { "epoch": 10.581869290231904, "grad_norm": 0.19427242875099182, "learning_rate": 2.6281564769266806e-05, "loss": 0.0285, "step": 15058 }, { "epoch": 10.58257203092059, "grad_norm": 0.242465078830719, "learning_rate": 2.628109627547435e-05, "loss": 0.0101, "step": 15059 }, { "epoch": 10.583274771609275, "grad_norm": 0.17618080973625183, "learning_rate": 2.6280627781681893e-05, "loss": 0.024, "step": 15060 }, { "epoch": 10.583977512297961, "grad_norm": 0.10496082901954651, "learning_rate": 2.6280159287889437e-05, "loss": 0.0125, "step": 15061 }, { "epoch": 10.584680252986647, "grad_norm": 0.10452792048454285, "learning_rate": 2.6279690794096977e-05, "loss": 0.0189, "step": 15062 }, { "epoch": 10.585382993675333, "grad_norm": 0.07731204479932785, "learning_rate": 2.627922230030452e-05, "loss": 0.0116, "step": 15063 }, { "epoch": 10.58608573436402, "grad_norm": 0.11687331646680832, "learning_rate": 2.6278753806512065e-05, "loss": 0.0327, "step": 15064 }, { "epoch": 10.586788475052705, "grad_norm": 0.12901560962200165, "learning_rate": 2.627828531271961e-05, "loss": 0.0142, "step": 15065 }, { "epoch": 10.587491215741391, "grad_norm": 0.12821918725967407, "learning_rate": 2.6277816818927152e-05, "loss": 0.0185, "step": 15066 }, { "epoch": 10.588193956430077, "grad_norm": 0.10523518919944763, "learning_rate": 2.6277348325134692e-05, "loss": 0.0151, "step": 15067 }, { "epoch": 10.588896697118763, "grad_norm": 0.15286406874656677, "learning_rate": 2.6276879831342236e-05, "loss": 0.0249, "step": 15068 }, { "epoch": 10.589599437807449, "grad_norm": 0.1649860143661499, "learning_rate": 2.627641133754978e-05, "loss": 0.0154, "step": 15069 }, { "epoch": 10.590302178496135, "grad_norm": 0.8139630556106567, "learning_rate": 2.6275942843757324e-05, "loss": 0.0465, "step": 15070 }, { "epoch": 10.59100491918482, "grad_norm": 0.1451079398393631, "learning_rate": 2.627547434996486e-05, "loss": 0.026, "step": 15071 }, { "epoch": 10.591707659873506, "grad_norm": 0.21251071989536285, "learning_rate": 2.6275005856172404e-05, "loss": 0.0273, "step": 15072 }, { "epoch": 10.592410400562192, "grad_norm": 0.5636089444160461, "learning_rate": 2.6274537362379948e-05, "loss": 0.0386, "step": 15073 }, { "epoch": 10.593113141250878, "grad_norm": 0.22701820731163025, "learning_rate": 2.6274068868587492e-05, "loss": 0.0383, "step": 15074 }, { "epoch": 10.593815881939564, "grad_norm": 0.3732399642467499, "learning_rate": 2.6273600374795032e-05, "loss": 0.0748, "step": 15075 }, { "epoch": 10.59451862262825, "grad_norm": 0.7114240527153015, "learning_rate": 2.6273131881002576e-05, "loss": 0.0735, "step": 15076 }, { "epoch": 10.595221363316936, "grad_norm": 0.3554910123348236, "learning_rate": 2.627266338721012e-05, "loss": 0.0935, "step": 15077 }, { "epoch": 10.595924104005622, "grad_norm": 0.42959049344062805, "learning_rate": 2.6272194893417663e-05, "loss": 0.1446, "step": 15078 }, { "epoch": 10.596626844694308, "grad_norm": 0.5550584197044373, "learning_rate": 2.6271726399625207e-05, "loss": 0.19, "step": 15079 }, { "epoch": 10.597329585382994, "grad_norm": 1.5984337329864502, "learning_rate": 2.6271257905832747e-05, "loss": 0.206, "step": 15080 }, { "epoch": 10.59803232607168, "grad_norm": 0.25886037945747375, "learning_rate": 2.627078941204029e-05, "loss": 0.067, "step": 15081 }, { "epoch": 10.598735066760366, "grad_norm": 0.14397454261779785, "learning_rate": 2.6270320918247835e-05, "loss": 0.0255, "step": 15082 }, { "epoch": 10.599437807449052, "grad_norm": 0.09771396219730377, "learning_rate": 2.626985242445538e-05, "loss": 0.0136, "step": 15083 }, { "epoch": 10.600140548137738, "grad_norm": 0.09047213196754456, "learning_rate": 2.626938393066292e-05, "loss": 0.015, "step": 15084 }, { "epoch": 10.600843288826423, "grad_norm": 0.1620907336473465, "learning_rate": 2.6268915436870463e-05, "loss": 0.0185, "step": 15085 }, { "epoch": 10.60154602951511, "grad_norm": 0.085394948720932, "learning_rate": 2.6268446943078006e-05, "loss": 0.0137, "step": 15086 }, { "epoch": 10.602248770203795, "grad_norm": 0.10785558074712753, "learning_rate": 2.626797844928555e-05, "loss": 0.0174, "step": 15087 }, { "epoch": 10.602951510892481, "grad_norm": 0.1616887003183365, "learning_rate": 2.626750995549309e-05, "loss": 0.0254, "step": 15088 }, { "epoch": 10.603654251581167, "grad_norm": 0.18301668763160706, "learning_rate": 2.626704146170063e-05, "loss": 0.0348, "step": 15089 }, { "epoch": 10.604356992269853, "grad_norm": 0.1329621523618698, "learning_rate": 2.6266572967908174e-05, "loss": 0.021, "step": 15090 }, { "epoch": 10.605059732958539, "grad_norm": 0.21847711503505707, "learning_rate": 2.6266104474115718e-05, "loss": 0.0236, "step": 15091 }, { "epoch": 10.605762473647225, "grad_norm": 0.09764324128627777, "learning_rate": 2.6265635980323262e-05, "loss": 0.0102, "step": 15092 }, { "epoch": 10.60646521433591, "grad_norm": 0.1586291790008545, "learning_rate": 2.6265167486530802e-05, "loss": 0.0244, "step": 15093 }, { "epoch": 10.607167955024597, "grad_norm": 0.13268643617630005, "learning_rate": 2.6264698992738346e-05, "loss": 0.0172, "step": 15094 }, { "epoch": 10.607870695713281, "grad_norm": 0.16628114879131317, "learning_rate": 2.626423049894589e-05, "loss": 0.0294, "step": 15095 }, { "epoch": 10.608573436401969, "grad_norm": 0.22815407812595367, "learning_rate": 2.6263762005153433e-05, "loss": 0.0432, "step": 15096 }, { "epoch": 10.609276177090653, "grad_norm": 0.17949278652668, "learning_rate": 2.6263293511360974e-05, "loss": 0.0337, "step": 15097 }, { "epoch": 10.609978917779339, "grad_norm": 0.2263576090335846, "learning_rate": 2.6262825017568517e-05, "loss": 0.0291, "step": 15098 }, { "epoch": 10.610681658468025, "grad_norm": 0.17197509109973907, "learning_rate": 2.626235652377606e-05, "loss": 0.0393, "step": 15099 }, { "epoch": 10.61138439915671, "grad_norm": 0.2430265247821808, "learning_rate": 2.6261888029983605e-05, "loss": 0.0471, "step": 15100 }, { "epoch": 10.612087139845396, "grad_norm": 0.3217657208442688, "learning_rate": 2.6261419536191145e-05, "loss": 0.0885, "step": 15101 }, { "epoch": 10.612789880534082, "grad_norm": 0.3551521599292755, "learning_rate": 2.626095104239869e-05, "loss": 0.1142, "step": 15102 }, { "epoch": 10.613492621222768, "grad_norm": 0.7599896192550659, "learning_rate": 2.6260482548606233e-05, "loss": 0.1416, "step": 15103 }, { "epoch": 10.614195361911454, "grad_norm": 0.6897962093353271, "learning_rate": 2.6260014054813776e-05, "loss": 0.1879, "step": 15104 }, { "epoch": 10.61489810260014, "grad_norm": 0.7143858671188354, "learning_rate": 2.625954556102132e-05, "loss": 0.1935, "step": 15105 }, { "epoch": 10.615600843288826, "grad_norm": 0.16750076413154602, "learning_rate": 2.6259077067228857e-05, "loss": 0.0662, "step": 15106 }, { "epoch": 10.616303583977512, "grad_norm": 0.16672679781913757, "learning_rate": 2.62586085734364e-05, "loss": 0.0199, "step": 15107 }, { "epoch": 10.617006324666198, "grad_norm": 0.08780549466609955, "learning_rate": 2.6258140079643945e-05, "loss": 0.0153, "step": 15108 }, { "epoch": 10.617709065354884, "grad_norm": 0.18917062878608704, "learning_rate": 2.6257671585851488e-05, "loss": 0.0142, "step": 15109 }, { "epoch": 10.61841180604357, "grad_norm": 0.12149914354085922, "learning_rate": 2.625720309205903e-05, "loss": 0.0177, "step": 15110 }, { "epoch": 10.619114546732256, "grad_norm": 0.09318628162145615, "learning_rate": 2.6256734598266572e-05, "loss": 0.0111, "step": 15111 }, { "epoch": 10.619817287420942, "grad_norm": 0.19137391448020935, "learning_rate": 2.6256266104474116e-05, "loss": 0.0319, "step": 15112 }, { "epoch": 10.620520028109627, "grad_norm": 0.10033038258552551, "learning_rate": 2.625579761068166e-05, "loss": 0.0227, "step": 15113 }, { "epoch": 10.621222768798313, "grad_norm": 0.17504645884037018, "learning_rate": 2.62553291168892e-05, "loss": 0.0412, "step": 15114 }, { "epoch": 10.621925509487, "grad_norm": 0.1847381293773651, "learning_rate": 2.6254860623096744e-05, "loss": 0.0105, "step": 15115 }, { "epoch": 10.622628250175685, "grad_norm": 0.20577380061149597, "learning_rate": 2.6254392129304288e-05, "loss": 0.0196, "step": 15116 }, { "epoch": 10.623330990864371, "grad_norm": 0.12032070755958557, "learning_rate": 2.625392363551183e-05, "loss": 0.0113, "step": 15117 }, { "epoch": 10.624033731553057, "grad_norm": 0.21473316848278046, "learning_rate": 2.6253455141719375e-05, "loss": 0.0229, "step": 15118 }, { "epoch": 10.624736472241743, "grad_norm": 0.16965769231319427, "learning_rate": 2.6252986647926915e-05, "loss": 0.025, "step": 15119 }, { "epoch": 10.625439212930429, "grad_norm": 0.23494777083396912, "learning_rate": 2.625251815413446e-05, "loss": 0.0518, "step": 15120 }, { "epoch": 10.626141953619115, "grad_norm": 0.272213876247406, "learning_rate": 2.6252049660342003e-05, "loss": 0.0297, "step": 15121 }, { "epoch": 10.6268446943078, "grad_norm": 0.22017936408519745, "learning_rate": 2.6251581166549547e-05, "loss": 0.0254, "step": 15122 }, { "epoch": 10.627547434996487, "grad_norm": 0.09178847074508667, "learning_rate": 2.6251112672757087e-05, "loss": 0.0134, "step": 15123 }, { "epoch": 10.628250175685173, "grad_norm": 0.16956017911434174, "learning_rate": 2.6250644178964627e-05, "loss": 0.0474, "step": 15124 }, { "epoch": 10.628952916373859, "grad_norm": 0.20258761942386627, "learning_rate": 2.625017568517217e-05, "loss": 0.0386, "step": 15125 }, { "epoch": 10.629655657062544, "grad_norm": 0.3488331139087677, "learning_rate": 2.6249707191379715e-05, "loss": 0.0754, "step": 15126 }, { "epoch": 10.63035839775123, "grad_norm": 0.709814727306366, "learning_rate": 2.624923869758726e-05, "loss": 0.1384, "step": 15127 }, { "epoch": 10.631061138439916, "grad_norm": 1.0742520093917847, "learning_rate": 2.62487702037948e-05, "loss": 0.1423, "step": 15128 }, { "epoch": 10.631763879128602, "grad_norm": 1.1033105850219727, "learning_rate": 2.6248301710002342e-05, "loss": 0.1959, "step": 15129 }, { "epoch": 10.632466619817288, "grad_norm": 0.8582461476325989, "learning_rate": 2.6247833216209886e-05, "loss": 0.2105, "step": 15130 }, { "epoch": 10.633169360505974, "grad_norm": 0.1392917037010193, "learning_rate": 2.624736472241743e-05, "loss": 0.0674, "step": 15131 }, { "epoch": 10.63387210119466, "grad_norm": 0.11945382505655289, "learning_rate": 2.624689622862497e-05, "loss": 0.0339, "step": 15132 }, { "epoch": 10.634574841883346, "grad_norm": 0.08853767812252045, "learning_rate": 2.6246427734832514e-05, "loss": 0.0181, "step": 15133 }, { "epoch": 10.63527758257203, "grad_norm": 0.19593964517116547, "learning_rate": 2.6245959241040058e-05, "loss": 0.022, "step": 15134 }, { "epoch": 10.635980323260716, "grad_norm": 0.08174977451562881, "learning_rate": 2.62454907472476e-05, "loss": 0.0079, "step": 15135 }, { "epoch": 10.636683063949402, "grad_norm": 0.09397299587726593, "learning_rate": 2.6245022253455142e-05, "loss": 0.0125, "step": 15136 }, { "epoch": 10.637385804638088, "grad_norm": 0.2532731890678406, "learning_rate": 2.6244553759662685e-05, "loss": 0.0211, "step": 15137 }, { "epoch": 10.638088545326774, "grad_norm": 0.09650919586420059, "learning_rate": 2.624408526587023e-05, "loss": 0.0234, "step": 15138 }, { "epoch": 10.63879128601546, "grad_norm": 0.25271865725517273, "learning_rate": 2.6243616772077773e-05, "loss": 0.0194, "step": 15139 }, { "epoch": 10.639494026704146, "grad_norm": 0.2990124225616455, "learning_rate": 2.6243148278285313e-05, "loss": 0.012, "step": 15140 }, { "epoch": 10.640196767392831, "grad_norm": 0.19617082178592682, "learning_rate": 2.6242679784492854e-05, "loss": 0.0298, "step": 15141 }, { "epoch": 10.640899508081517, "grad_norm": 0.08807153254747391, "learning_rate": 2.6242211290700397e-05, "loss": 0.0104, "step": 15142 }, { "epoch": 10.641602248770203, "grad_norm": 0.12500198185443878, "learning_rate": 2.624174279690794e-05, "loss": 0.0219, "step": 15143 }, { "epoch": 10.64230498945889, "grad_norm": 0.14589136838912964, "learning_rate": 2.6241274303115485e-05, "loss": 0.0126, "step": 15144 }, { "epoch": 10.643007730147575, "grad_norm": 0.12798786163330078, "learning_rate": 2.6240805809323025e-05, "loss": 0.0283, "step": 15145 }, { "epoch": 10.643710470836261, "grad_norm": 0.2222437560558319, "learning_rate": 2.624033731553057e-05, "loss": 0.036, "step": 15146 }, { "epoch": 10.644413211524947, "grad_norm": 0.16485436260700226, "learning_rate": 2.6239868821738113e-05, "loss": 0.0308, "step": 15147 }, { "epoch": 10.645115952213633, "grad_norm": 0.273113489151001, "learning_rate": 2.6239400327945656e-05, "loss": 0.0423, "step": 15148 }, { "epoch": 10.645818692902319, "grad_norm": 0.23795686662197113, "learning_rate": 2.6238931834153197e-05, "loss": 0.0597, "step": 15149 }, { "epoch": 10.646521433591005, "grad_norm": 0.1988760232925415, "learning_rate": 2.623846334036074e-05, "loss": 0.0504, "step": 15150 }, { "epoch": 10.64722417427969, "grad_norm": 0.3312396705150604, "learning_rate": 2.6237994846568284e-05, "loss": 0.0717, "step": 15151 }, { "epoch": 10.647926914968377, "grad_norm": 0.4493972957134247, "learning_rate": 2.6237526352775828e-05, "loss": 0.0849, "step": 15152 }, { "epoch": 10.648629655657063, "grad_norm": 0.6927344799041748, "learning_rate": 2.623705785898337e-05, "loss": 0.1651, "step": 15153 }, { "epoch": 10.649332396345748, "grad_norm": 0.8620419502258301, "learning_rate": 2.6236589365190912e-05, "loss": 0.1572, "step": 15154 }, { "epoch": 10.650035137034434, "grad_norm": 0.741184651851654, "learning_rate": 2.6236120871398456e-05, "loss": 0.1976, "step": 15155 }, { "epoch": 10.65073787772312, "grad_norm": 0.5197039246559143, "learning_rate": 2.6235652377606e-05, "loss": 0.0866, "step": 15156 }, { "epoch": 10.651440618411806, "grad_norm": 0.08138106763362885, "learning_rate": 2.6235183883813543e-05, "loss": 0.0178, "step": 15157 }, { "epoch": 10.652143359100492, "grad_norm": 0.1297615021467209, "learning_rate": 2.6234715390021083e-05, "loss": 0.0341, "step": 15158 }, { "epoch": 10.652846099789178, "grad_norm": 0.11846351623535156, "learning_rate": 2.6234246896228624e-05, "loss": 0.022, "step": 15159 }, { "epoch": 10.653548840477864, "grad_norm": 0.1990063488483429, "learning_rate": 2.6233778402436167e-05, "loss": 0.0286, "step": 15160 }, { "epoch": 10.65425158116655, "grad_norm": 0.0860581323504448, "learning_rate": 2.623330990864371e-05, "loss": 0.0113, "step": 15161 }, { "epoch": 10.654954321855236, "grad_norm": 0.10658320039510727, "learning_rate": 2.623284141485125e-05, "loss": 0.01, "step": 15162 }, { "epoch": 10.655657062543922, "grad_norm": 0.15445944666862488, "learning_rate": 2.6232372921058795e-05, "loss": 0.0235, "step": 15163 }, { "epoch": 10.656359803232608, "grad_norm": 0.14622396230697632, "learning_rate": 2.623190442726634e-05, "loss": 0.0139, "step": 15164 }, { "epoch": 10.657062543921294, "grad_norm": 0.5653480887413025, "learning_rate": 2.6231435933473883e-05, "loss": 0.0084, "step": 15165 }, { "epoch": 10.65776528460998, "grad_norm": 0.08532214164733887, "learning_rate": 2.6230967439681426e-05, "loss": 0.0148, "step": 15166 }, { "epoch": 10.658468025298665, "grad_norm": 0.3281303942203522, "learning_rate": 2.6230498945888967e-05, "loss": 0.0227, "step": 15167 }, { "epoch": 10.659170765987351, "grad_norm": 0.21855202317237854, "learning_rate": 2.623003045209651e-05, "loss": 0.0228, "step": 15168 }, { "epoch": 10.659873506676037, "grad_norm": 0.10823952406644821, "learning_rate": 2.6229561958304054e-05, "loss": 0.0183, "step": 15169 }, { "epoch": 10.660576247364723, "grad_norm": 0.15962599217891693, "learning_rate": 2.6229093464511598e-05, "loss": 0.0261, "step": 15170 }, { "epoch": 10.66127898805341, "grad_norm": 0.8227232098579407, "learning_rate": 2.6228624970719138e-05, "loss": 0.0393, "step": 15171 }, { "epoch": 10.661981728742095, "grad_norm": 0.23387667536735535, "learning_rate": 2.6228156476926682e-05, "loss": 0.0273, "step": 15172 }, { "epoch": 10.66268446943078, "grad_norm": 0.2713663876056671, "learning_rate": 2.6227687983134226e-05, "loss": 0.0292, "step": 15173 }, { "epoch": 10.663387210119465, "grad_norm": 0.33099445700645447, "learning_rate": 2.622721948934177e-05, "loss": 0.068, "step": 15174 }, { "epoch": 10.664089950808151, "grad_norm": 0.25154200196266174, "learning_rate": 2.622675099554931e-05, "loss": 0.0523, "step": 15175 }, { "epoch": 10.664792691496837, "grad_norm": 0.6432788372039795, "learning_rate": 2.622628250175685e-05, "loss": 0.0815, "step": 15176 }, { "epoch": 10.665495432185523, "grad_norm": 0.6383109092712402, "learning_rate": 2.6225814007964394e-05, "loss": 0.1122, "step": 15177 }, { "epoch": 10.666198172874209, "grad_norm": 0.44778019189834595, "learning_rate": 2.6225345514171938e-05, "loss": 0.1394, "step": 15178 }, { "epoch": 10.666900913562895, "grad_norm": 0.7345669865608215, "learning_rate": 2.622487702037948e-05, "loss": 0.2122, "step": 15179 }, { "epoch": 10.66760365425158, "grad_norm": 1.3156192302703857, "learning_rate": 2.622440852658702e-05, "loss": 0.2209, "step": 15180 }, { "epoch": 10.668306394940267, "grad_norm": 0.5362129807472229, "learning_rate": 2.6223940032794565e-05, "loss": 0.0673, "step": 15181 }, { "epoch": 10.669009135628952, "grad_norm": 0.16469287872314453, "learning_rate": 2.622347153900211e-05, "loss": 0.0253, "step": 15182 }, { "epoch": 10.669711876317638, "grad_norm": 0.32256194949150085, "learning_rate": 2.6223003045209653e-05, "loss": 0.0296, "step": 15183 }, { "epoch": 10.670414617006324, "grad_norm": 0.1193162053823471, "learning_rate": 2.6222534551417193e-05, "loss": 0.0177, "step": 15184 }, { "epoch": 10.67111735769501, "grad_norm": 0.08866117894649506, "learning_rate": 2.6222066057624737e-05, "loss": 0.0134, "step": 15185 }, { "epoch": 10.671820098383696, "grad_norm": 0.10707518458366394, "learning_rate": 2.622159756383228e-05, "loss": 0.0089, "step": 15186 }, { "epoch": 10.672522839072382, "grad_norm": 0.4148761034011841, "learning_rate": 2.6221129070039824e-05, "loss": 0.0235, "step": 15187 }, { "epoch": 10.673225579761068, "grad_norm": 0.1888945996761322, "learning_rate": 2.6220660576247365e-05, "loss": 0.0276, "step": 15188 }, { "epoch": 10.673928320449754, "grad_norm": 0.1233275905251503, "learning_rate": 2.622019208245491e-05, "loss": 0.0182, "step": 15189 }, { "epoch": 10.67463106113844, "grad_norm": 0.17590981721878052, "learning_rate": 2.6219723588662452e-05, "loss": 0.0242, "step": 15190 }, { "epoch": 10.675333801827126, "grad_norm": 0.09787794947624207, "learning_rate": 2.6219255094869996e-05, "loss": 0.0169, "step": 15191 }, { "epoch": 10.676036542515812, "grad_norm": 0.1529238075017929, "learning_rate": 2.621878660107754e-05, "loss": 0.0233, "step": 15192 }, { "epoch": 10.676739283204498, "grad_norm": 0.22453416883945465, "learning_rate": 2.6218318107285076e-05, "loss": 0.0281, "step": 15193 }, { "epoch": 10.677442023893184, "grad_norm": 0.1316273957490921, "learning_rate": 2.621784961349262e-05, "loss": 0.0242, "step": 15194 }, { "epoch": 10.67814476458187, "grad_norm": 0.228190079331398, "learning_rate": 2.6217381119700164e-05, "loss": 0.0574, "step": 15195 }, { "epoch": 10.678847505270555, "grad_norm": 0.1787363737821579, "learning_rate": 2.6216912625907708e-05, "loss": 0.0279, "step": 15196 }, { "epoch": 10.679550245959241, "grad_norm": 0.10343178361654282, "learning_rate": 2.6216444132115248e-05, "loss": 0.0146, "step": 15197 }, { "epoch": 10.680252986647927, "grad_norm": 0.2104295939207077, "learning_rate": 2.6215975638322792e-05, "loss": 0.0365, "step": 15198 }, { "epoch": 10.680955727336613, "grad_norm": 0.160671666264534, "learning_rate": 2.6215507144530335e-05, "loss": 0.0326, "step": 15199 }, { "epoch": 10.681658468025299, "grad_norm": 0.25192591547966003, "learning_rate": 2.621503865073788e-05, "loss": 0.0719, "step": 15200 }, { "epoch": 10.682361208713985, "grad_norm": 0.30207204818725586, "learning_rate": 2.621457015694542e-05, "loss": 0.0492, "step": 15201 }, { "epoch": 10.683063949402671, "grad_norm": 0.43889135122299194, "learning_rate": 2.6214101663152963e-05, "loss": 0.1146, "step": 15202 }, { "epoch": 10.683766690091357, "grad_norm": 0.49977293610572815, "learning_rate": 2.6213633169360507e-05, "loss": 0.1341, "step": 15203 }, { "epoch": 10.684469430780043, "grad_norm": 1.0218180418014526, "learning_rate": 2.621316467556805e-05, "loss": 0.1787, "step": 15204 }, { "epoch": 10.685172171468729, "grad_norm": 2.022468090057373, "learning_rate": 2.6212696181775594e-05, "loss": 0.1986, "step": 15205 }, { "epoch": 10.685874912157415, "grad_norm": 0.29602399468421936, "learning_rate": 2.6212227687983135e-05, "loss": 0.0811, "step": 15206 }, { "epoch": 10.6865776528461, "grad_norm": 0.2041058987379074, "learning_rate": 2.621175919419068e-05, "loss": 0.0284, "step": 15207 }, { "epoch": 10.687280393534786, "grad_norm": 0.10071547329425812, "learning_rate": 2.6211290700398222e-05, "loss": 0.0217, "step": 15208 }, { "epoch": 10.687983134223472, "grad_norm": 0.3129004240036011, "learning_rate": 2.6210822206605766e-05, "loss": 0.0182, "step": 15209 }, { "epoch": 10.688685874912156, "grad_norm": 0.1252792775630951, "learning_rate": 2.6210353712813306e-05, "loss": 0.0164, "step": 15210 }, { "epoch": 10.689388615600844, "grad_norm": 0.07280606776475906, "learning_rate": 2.6209885219020847e-05, "loss": 0.0105, "step": 15211 }, { "epoch": 10.690091356289528, "grad_norm": 0.08324643224477768, "learning_rate": 2.620941672522839e-05, "loss": 0.0147, "step": 15212 }, { "epoch": 10.690794096978214, "grad_norm": 0.2950451672077179, "learning_rate": 2.6208948231435934e-05, "loss": 0.0172, "step": 15213 }, { "epoch": 10.6914968376669, "grad_norm": 0.1268540769815445, "learning_rate": 2.6208479737643474e-05, "loss": 0.0188, "step": 15214 }, { "epoch": 10.692199578355586, "grad_norm": 0.1495225876569748, "learning_rate": 2.6208011243851018e-05, "loss": 0.0348, "step": 15215 }, { "epoch": 10.692902319044272, "grad_norm": 0.11092962324619293, "learning_rate": 2.6207542750058562e-05, "loss": 0.0182, "step": 15216 }, { "epoch": 10.693605059732958, "grad_norm": 0.12810535728931427, "learning_rate": 2.6207074256266106e-05, "loss": 0.0117, "step": 15217 }, { "epoch": 10.694307800421644, "grad_norm": 0.11750584095716476, "learning_rate": 2.620660576247365e-05, "loss": 0.0182, "step": 15218 }, { "epoch": 10.69501054111033, "grad_norm": 0.11076610535383224, "learning_rate": 2.620613726868119e-05, "loss": 0.0143, "step": 15219 }, { "epoch": 10.695713281799016, "grad_norm": 0.49270766973495483, "learning_rate": 2.6205668774888733e-05, "loss": 0.0349, "step": 15220 }, { "epoch": 10.696416022487702, "grad_norm": 0.174774631857872, "learning_rate": 2.6205200281096277e-05, "loss": 0.0371, "step": 15221 }, { "epoch": 10.697118763176388, "grad_norm": 0.09284406155347824, "learning_rate": 2.620473178730382e-05, "loss": 0.0154, "step": 15222 }, { "epoch": 10.697821503865073, "grad_norm": 0.1718774139881134, "learning_rate": 2.620426329351136e-05, "loss": 0.0329, "step": 15223 }, { "epoch": 10.69852424455376, "grad_norm": 0.4906969368457794, "learning_rate": 2.6203794799718905e-05, "loss": 0.0406, "step": 15224 }, { "epoch": 10.699226985242445, "grad_norm": 0.2126922607421875, "learning_rate": 2.620332630592645e-05, "loss": 0.0557, "step": 15225 }, { "epoch": 10.699929725931131, "grad_norm": 0.22959187626838684, "learning_rate": 2.6202857812133992e-05, "loss": 0.0577, "step": 15226 }, { "epoch": 10.700632466619817, "grad_norm": 0.7023213505744934, "learning_rate": 2.6202389318341533e-05, "loss": 0.1266, "step": 15227 }, { "epoch": 10.701335207308503, "grad_norm": 0.41440725326538086, "learning_rate": 2.6201920824549073e-05, "loss": 0.1188, "step": 15228 }, { "epoch": 10.702037947997189, "grad_norm": 1.0862464904785156, "learning_rate": 2.6201452330756617e-05, "loss": 0.1949, "step": 15229 }, { "epoch": 10.702740688685875, "grad_norm": 0.798689603805542, "learning_rate": 2.620098383696416e-05, "loss": 0.1795, "step": 15230 }, { "epoch": 10.70344342937456, "grad_norm": 0.23112863302230835, "learning_rate": 2.6200515343171704e-05, "loss": 0.0712, "step": 15231 }, { "epoch": 10.704146170063247, "grad_norm": 0.142582505941391, "learning_rate": 2.6200046849379244e-05, "loss": 0.0272, "step": 15232 }, { "epoch": 10.704848910751933, "grad_norm": 0.10546908527612686, "learning_rate": 2.6199578355586788e-05, "loss": 0.0179, "step": 15233 }, { "epoch": 10.705551651440619, "grad_norm": 0.10510038584470749, "learning_rate": 2.6199109861794332e-05, "loss": 0.0177, "step": 15234 }, { "epoch": 10.706254392129305, "grad_norm": 0.10683274269104004, "learning_rate": 2.6198641368001876e-05, "loss": 0.0171, "step": 15235 }, { "epoch": 10.70695713281799, "grad_norm": 0.07265741378068924, "learning_rate": 2.6198172874209416e-05, "loss": 0.0054, "step": 15236 }, { "epoch": 10.707659873506676, "grad_norm": 0.19737407565116882, "learning_rate": 2.619770438041696e-05, "loss": 0.0133, "step": 15237 }, { "epoch": 10.708362614195362, "grad_norm": 0.061972327530384064, "learning_rate": 2.6197235886624503e-05, "loss": 0.0103, "step": 15238 }, { "epoch": 10.709065354884048, "grad_norm": 0.3297988176345825, "learning_rate": 2.6196767392832047e-05, "loss": 0.0351, "step": 15239 }, { "epoch": 10.709768095572734, "grad_norm": 0.09648102521896362, "learning_rate": 2.6196298899039588e-05, "loss": 0.0083, "step": 15240 }, { "epoch": 10.71047083626142, "grad_norm": 0.2408931851387024, "learning_rate": 2.619583040524713e-05, "loss": 0.0226, "step": 15241 }, { "epoch": 10.711173576950106, "grad_norm": 0.4725772440433502, "learning_rate": 2.6195361911454675e-05, "loss": 0.0128, "step": 15242 }, { "epoch": 10.711876317638792, "grad_norm": 0.22812940180301666, "learning_rate": 2.619489341766222e-05, "loss": 0.0235, "step": 15243 }, { "epoch": 10.712579058327478, "grad_norm": 0.08717326074838638, "learning_rate": 2.6194424923869762e-05, "loss": 0.012, "step": 15244 }, { "epoch": 10.713281799016164, "grad_norm": 0.21876311302185059, "learning_rate": 2.6193956430077303e-05, "loss": 0.0269, "step": 15245 }, { "epoch": 10.71398453970485, "grad_norm": 0.14645777642726898, "learning_rate": 2.6193487936284843e-05, "loss": 0.0467, "step": 15246 }, { "epoch": 10.714687280393536, "grad_norm": 0.17809158563613892, "learning_rate": 2.6193019442492387e-05, "loss": 0.0208, "step": 15247 }, { "epoch": 10.715390021082221, "grad_norm": 0.184430330991745, "learning_rate": 2.619255094869993e-05, "loss": 0.0387, "step": 15248 }, { "epoch": 10.716092761770906, "grad_norm": 0.6697099804878235, "learning_rate": 2.619208245490747e-05, "loss": 0.0358, "step": 15249 }, { "epoch": 10.716795502459593, "grad_norm": 0.20962491631507874, "learning_rate": 2.6191613961115015e-05, "loss": 0.0599, "step": 15250 }, { "epoch": 10.717498243148277, "grad_norm": 0.3967507481575012, "learning_rate": 2.619114546732256e-05, "loss": 0.0653, "step": 15251 }, { "epoch": 10.718200983836963, "grad_norm": 0.6295823454856873, "learning_rate": 2.6190676973530102e-05, "loss": 0.1428, "step": 15252 }, { "epoch": 10.71890372452565, "grad_norm": 1.3790431022644043, "learning_rate": 2.6190208479737642e-05, "loss": 0.1726, "step": 15253 }, { "epoch": 10.719606465214335, "grad_norm": 0.7124723196029663, "learning_rate": 2.6189739985945186e-05, "loss": 0.1805, "step": 15254 }, { "epoch": 10.720309205903021, "grad_norm": 1.5367584228515625, "learning_rate": 2.618927149215273e-05, "loss": 0.2031, "step": 15255 }, { "epoch": 10.721011946591707, "grad_norm": 0.19522356986999512, "learning_rate": 2.6188802998360274e-05, "loss": 0.054, "step": 15256 }, { "epoch": 10.721714687280393, "grad_norm": 0.25233539938926697, "learning_rate": 2.6188334504567817e-05, "loss": 0.0222, "step": 15257 }, { "epoch": 10.722417427969079, "grad_norm": 0.10702356696128845, "learning_rate": 2.6187866010775358e-05, "loss": 0.0259, "step": 15258 }, { "epoch": 10.723120168657765, "grad_norm": 0.21673248708248138, "learning_rate": 2.61873975169829e-05, "loss": 0.0167, "step": 15259 }, { "epoch": 10.72382290934645, "grad_norm": 0.1731489896774292, "learning_rate": 2.6186929023190445e-05, "loss": 0.0173, "step": 15260 }, { "epoch": 10.724525650035137, "grad_norm": 0.17779268324375153, "learning_rate": 2.618646052939799e-05, "loss": 0.0197, "step": 15261 }, { "epoch": 10.725228390723823, "grad_norm": 0.1731007844209671, "learning_rate": 2.618599203560553e-05, "loss": 0.0103, "step": 15262 }, { "epoch": 10.725931131412509, "grad_norm": 0.1657629907131195, "learning_rate": 2.618552354181307e-05, "loss": 0.0159, "step": 15263 }, { "epoch": 10.726633872101194, "grad_norm": 0.19134673476219177, "learning_rate": 2.6185055048020613e-05, "loss": 0.0269, "step": 15264 }, { "epoch": 10.72733661278988, "grad_norm": 0.06324949115514755, "learning_rate": 2.6184586554228157e-05, "loss": 0.0085, "step": 15265 }, { "epoch": 10.728039353478566, "grad_norm": 0.4278852939605713, "learning_rate": 2.6184118060435697e-05, "loss": 0.0199, "step": 15266 }, { "epoch": 10.728742094167252, "grad_norm": 0.22265613079071045, "learning_rate": 2.618364956664324e-05, "loss": 0.0149, "step": 15267 }, { "epoch": 10.729444834855938, "grad_norm": 0.19572806358337402, "learning_rate": 2.6183181072850785e-05, "loss": 0.0384, "step": 15268 }, { "epoch": 10.730147575544624, "grad_norm": 0.10062634944915771, "learning_rate": 2.618271257905833e-05, "loss": 0.0144, "step": 15269 }, { "epoch": 10.73085031623331, "grad_norm": 0.2207050323486328, "learning_rate": 2.6182244085265872e-05, "loss": 0.0277, "step": 15270 }, { "epoch": 10.731553056921996, "grad_norm": 0.30312249064445496, "learning_rate": 2.6181775591473412e-05, "loss": 0.0468, "step": 15271 }, { "epoch": 10.732255797610682, "grad_norm": 0.12993519008159637, "learning_rate": 2.6181307097680956e-05, "loss": 0.0234, "step": 15272 }, { "epoch": 10.732958538299368, "grad_norm": 0.19125863909721375, "learning_rate": 2.61808386038885e-05, "loss": 0.0266, "step": 15273 }, { "epoch": 10.733661278988054, "grad_norm": 0.29068389534950256, "learning_rate": 2.6180370110096044e-05, "loss": 0.0549, "step": 15274 }, { "epoch": 10.73436401967674, "grad_norm": 0.28728801012039185, "learning_rate": 2.6179901616303584e-05, "loss": 0.0596, "step": 15275 }, { "epoch": 10.735066760365426, "grad_norm": 0.6950199007987976, "learning_rate": 2.6179433122511128e-05, "loss": 0.0766, "step": 15276 }, { "epoch": 10.735769501054111, "grad_norm": 0.48607268929481506, "learning_rate": 2.617896462871867e-05, "loss": 0.0912, "step": 15277 }, { "epoch": 10.736472241742797, "grad_norm": 0.5314457416534424, "learning_rate": 2.6178496134926215e-05, "loss": 0.1576, "step": 15278 }, { "epoch": 10.737174982431483, "grad_norm": 1.2875529527664185, "learning_rate": 2.6178027641133756e-05, "loss": 0.2082, "step": 15279 }, { "epoch": 10.73787772312017, "grad_norm": 1.1747779846191406, "learning_rate": 2.6177559147341296e-05, "loss": 0.1879, "step": 15280 }, { "epoch": 10.738580463808855, "grad_norm": 0.49963322281837463, "learning_rate": 2.617709065354884e-05, "loss": 0.0809, "step": 15281 }, { "epoch": 10.739283204497541, "grad_norm": 0.2048996537923813, "learning_rate": 2.6176622159756383e-05, "loss": 0.0249, "step": 15282 }, { "epoch": 10.739985945186227, "grad_norm": 0.13331614434719086, "learning_rate": 2.6176153665963927e-05, "loss": 0.0399, "step": 15283 }, { "epoch": 10.740688685874913, "grad_norm": 0.19072334468364716, "learning_rate": 2.6175685172171467e-05, "loss": 0.0283, "step": 15284 }, { "epoch": 10.741391426563599, "grad_norm": 0.1371651589870453, "learning_rate": 2.617521667837901e-05, "loss": 0.0223, "step": 15285 }, { "epoch": 10.742094167252285, "grad_norm": 0.09545108675956726, "learning_rate": 2.6174748184586555e-05, "loss": 0.007, "step": 15286 }, { "epoch": 10.74279690794097, "grad_norm": 0.2283615916967392, "learning_rate": 2.61742796907941e-05, "loss": 0.0223, "step": 15287 }, { "epoch": 10.743499648629655, "grad_norm": 0.1663561761379242, "learning_rate": 2.617381119700164e-05, "loss": 0.0178, "step": 15288 }, { "epoch": 10.74420238931834, "grad_norm": 0.2061925083398819, "learning_rate": 2.6173342703209183e-05, "loss": 0.0241, "step": 15289 }, { "epoch": 10.744905130007027, "grad_norm": 0.13798192143440247, "learning_rate": 2.6172874209416726e-05, "loss": 0.0128, "step": 15290 }, { "epoch": 10.745607870695713, "grad_norm": 0.20306949317455292, "learning_rate": 2.617240571562427e-05, "loss": 0.0251, "step": 15291 }, { "epoch": 10.746310611384398, "grad_norm": 0.18798844516277313, "learning_rate": 2.617193722183181e-05, "loss": 0.0116, "step": 15292 }, { "epoch": 10.747013352073084, "grad_norm": 0.18130573630332947, "learning_rate": 2.6171468728039354e-05, "loss": 0.0291, "step": 15293 }, { "epoch": 10.74771609276177, "grad_norm": 0.16180266439914703, "learning_rate": 2.6171000234246898e-05, "loss": 0.016, "step": 15294 }, { "epoch": 10.748418833450456, "grad_norm": 0.17995567619800568, "learning_rate": 2.617053174045444e-05, "loss": 0.0291, "step": 15295 }, { "epoch": 10.749121574139142, "grad_norm": 0.16711419820785522, "learning_rate": 2.6170063246661985e-05, "loss": 0.0225, "step": 15296 }, { "epoch": 10.749824314827828, "grad_norm": 0.22667038440704346, "learning_rate": 2.6169594752869526e-05, "loss": 0.0358, "step": 15297 }, { "epoch": 10.750527055516514, "grad_norm": 0.17859041690826416, "learning_rate": 2.6169126259077066e-05, "loss": 0.0185, "step": 15298 }, { "epoch": 10.7512297962052, "grad_norm": 0.3929392397403717, "learning_rate": 2.616865776528461e-05, "loss": 0.0659, "step": 15299 }, { "epoch": 10.751932536893886, "grad_norm": 0.2818388342857361, "learning_rate": 2.6168189271492153e-05, "loss": 0.0605, "step": 15300 }, { "epoch": 10.752635277582572, "grad_norm": 0.4065715968608856, "learning_rate": 2.6167720777699694e-05, "loss": 0.0763, "step": 15301 }, { "epoch": 10.753338018271258, "grad_norm": 0.5069671869277954, "learning_rate": 2.6167252283907237e-05, "loss": 0.1276, "step": 15302 }, { "epoch": 10.754040758959944, "grad_norm": 0.4974959194660187, "learning_rate": 2.616678379011478e-05, "loss": 0.1349, "step": 15303 }, { "epoch": 10.75474349964863, "grad_norm": 0.5986671447753906, "learning_rate": 2.6166315296322325e-05, "loss": 0.1616, "step": 15304 }, { "epoch": 10.755446240337315, "grad_norm": 0.8787641525268555, "learning_rate": 2.6165846802529865e-05, "loss": 0.2014, "step": 15305 }, { "epoch": 10.756148981026001, "grad_norm": 0.21404439210891724, "learning_rate": 2.616537830873741e-05, "loss": 0.0749, "step": 15306 }, { "epoch": 10.756851721714687, "grad_norm": 0.21339036524295807, "learning_rate": 2.6164909814944953e-05, "loss": 0.0456, "step": 15307 }, { "epoch": 10.757554462403373, "grad_norm": 0.10784685611724854, "learning_rate": 2.6164441321152496e-05, "loss": 0.015, "step": 15308 }, { "epoch": 10.75825720309206, "grad_norm": 0.08139056712388992, "learning_rate": 2.616397282736004e-05, "loss": 0.015, "step": 15309 }, { "epoch": 10.758959943780745, "grad_norm": 0.10426808893680573, "learning_rate": 2.616350433356758e-05, "loss": 0.0115, "step": 15310 }, { "epoch": 10.759662684469431, "grad_norm": 0.09910184144973755, "learning_rate": 2.6163035839775124e-05, "loss": 0.0168, "step": 15311 }, { "epoch": 10.760365425158117, "grad_norm": 0.08868807554244995, "learning_rate": 2.6162567345982668e-05, "loss": 0.0178, "step": 15312 }, { "epoch": 10.761068165846803, "grad_norm": 0.128025084733963, "learning_rate": 2.616209885219021e-05, "loss": 0.0209, "step": 15313 }, { "epoch": 10.761770906535489, "grad_norm": 0.1314845234155655, "learning_rate": 2.6161630358397752e-05, "loss": 0.0206, "step": 15314 }, { "epoch": 10.762473647224175, "grad_norm": 0.08005359768867493, "learning_rate": 2.6161161864605292e-05, "loss": 0.0069, "step": 15315 }, { "epoch": 10.76317638791286, "grad_norm": 0.17685848474502563, "learning_rate": 2.6160693370812836e-05, "loss": 0.014, "step": 15316 }, { "epoch": 10.763879128601546, "grad_norm": 0.13402347266674042, "learning_rate": 2.616022487702038e-05, "loss": 0.0139, "step": 15317 }, { "epoch": 10.764581869290232, "grad_norm": 0.26338303089141846, "learning_rate": 2.6159756383227924e-05, "loss": 0.0184, "step": 15318 }, { "epoch": 10.765284609978918, "grad_norm": 0.2114400863647461, "learning_rate": 2.6159287889435464e-05, "loss": 0.0177, "step": 15319 }, { "epoch": 10.765987350667604, "grad_norm": 0.17297714948654175, "learning_rate": 2.6158819395643008e-05, "loss": 0.0388, "step": 15320 }, { "epoch": 10.76669009135629, "grad_norm": 0.25572365522384644, "learning_rate": 2.615835090185055e-05, "loss": 0.0231, "step": 15321 }, { "epoch": 10.767392832044976, "grad_norm": 0.4172373116016388, "learning_rate": 2.6157882408058095e-05, "loss": 0.0267, "step": 15322 }, { "epoch": 10.768095572733662, "grad_norm": 0.1980050951242447, "learning_rate": 2.6157413914265635e-05, "loss": 0.0313, "step": 15323 }, { "epoch": 10.768798313422348, "grad_norm": 0.21884475648403168, "learning_rate": 2.615694542047318e-05, "loss": 0.0447, "step": 15324 }, { "epoch": 10.769501054111032, "grad_norm": 0.38350003957748413, "learning_rate": 2.6156476926680723e-05, "loss": 0.058, "step": 15325 }, { "epoch": 10.77020379479972, "grad_norm": 0.7410509586334229, "learning_rate": 2.6156008432888267e-05, "loss": 0.0568, "step": 15326 }, { "epoch": 10.770906535488404, "grad_norm": 0.9079629778862, "learning_rate": 2.6155539939095807e-05, "loss": 0.1289, "step": 15327 }, { "epoch": 10.77160927617709, "grad_norm": 0.44968342781066895, "learning_rate": 2.615507144530335e-05, "loss": 0.129, "step": 15328 }, { "epoch": 10.772312016865776, "grad_norm": 0.475864052772522, "learning_rate": 2.6154602951510894e-05, "loss": 0.1548, "step": 15329 }, { "epoch": 10.773014757554462, "grad_norm": 1.594181776046753, "learning_rate": 2.6154134457718438e-05, "loss": 0.1867, "step": 15330 }, { "epoch": 10.773717498243148, "grad_norm": 0.2667117118835449, "learning_rate": 2.615366596392598e-05, "loss": 0.0911, "step": 15331 }, { "epoch": 10.774420238931834, "grad_norm": 0.17851455509662628, "learning_rate": 2.6153197470133522e-05, "loss": 0.0506, "step": 15332 }, { "epoch": 10.77512297962052, "grad_norm": 0.15232692658901215, "learning_rate": 2.6152728976341062e-05, "loss": 0.0154, "step": 15333 }, { "epoch": 10.775825720309205, "grad_norm": 0.14845533668994904, "learning_rate": 2.6152260482548606e-05, "loss": 0.0166, "step": 15334 }, { "epoch": 10.776528460997891, "grad_norm": 0.06421695649623871, "learning_rate": 2.615179198875615e-05, "loss": 0.0136, "step": 15335 }, { "epoch": 10.777231201686577, "grad_norm": 0.521028995513916, "learning_rate": 2.615132349496369e-05, "loss": 0.0112, "step": 15336 }, { "epoch": 10.777933942375263, "grad_norm": 0.14116567373275757, "learning_rate": 2.6150855001171234e-05, "loss": 0.0231, "step": 15337 }, { "epoch": 10.778636683063949, "grad_norm": 0.22918587923049927, "learning_rate": 2.6150386507378778e-05, "loss": 0.0193, "step": 15338 }, { "epoch": 10.779339423752635, "grad_norm": 0.13566909730434418, "learning_rate": 2.614991801358632e-05, "loss": 0.0209, "step": 15339 }, { "epoch": 10.780042164441321, "grad_norm": 0.07986178994178772, "learning_rate": 2.6149449519793862e-05, "loss": 0.0123, "step": 15340 }, { "epoch": 10.780744905130007, "grad_norm": 0.12586426734924316, "learning_rate": 2.6148981026001405e-05, "loss": 0.0272, "step": 15341 }, { "epoch": 10.781447645818693, "grad_norm": 0.12848594784736633, "learning_rate": 2.614851253220895e-05, "loss": 0.0193, "step": 15342 }, { "epoch": 10.782150386507379, "grad_norm": 0.12597209215164185, "learning_rate": 2.6148044038416493e-05, "loss": 0.0224, "step": 15343 }, { "epoch": 10.782853127196065, "grad_norm": 0.09703409671783447, "learning_rate": 2.6147575544624037e-05, "loss": 0.0135, "step": 15344 }, { "epoch": 10.78355586788475, "grad_norm": 0.1435108333826065, "learning_rate": 2.6147107050831577e-05, "loss": 0.0288, "step": 15345 }, { "epoch": 10.784258608573436, "grad_norm": 0.19651798903942108, "learning_rate": 2.614663855703912e-05, "loss": 0.0282, "step": 15346 }, { "epoch": 10.784961349262122, "grad_norm": 1.5693421363830566, "learning_rate": 2.6146170063246664e-05, "loss": 0.0215, "step": 15347 }, { "epoch": 10.785664089950808, "grad_norm": 0.11872405558824539, "learning_rate": 2.6145701569454208e-05, "loss": 0.0224, "step": 15348 }, { "epoch": 10.786366830639494, "grad_norm": 0.23642641305923462, "learning_rate": 2.614523307566175e-05, "loss": 0.0491, "step": 15349 }, { "epoch": 10.78706957132818, "grad_norm": 0.19945469498634338, "learning_rate": 2.614476458186929e-05, "loss": 0.0619, "step": 15350 }, { "epoch": 10.787772312016866, "grad_norm": 0.46192750334739685, "learning_rate": 2.6144296088076833e-05, "loss": 0.079, "step": 15351 }, { "epoch": 10.788475052705552, "grad_norm": 0.330684095621109, "learning_rate": 2.6143827594284376e-05, "loss": 0.0996, "step": 15352 }, { "epoch": 10.789177793394238, "grad_norm": 0.8248141407966614, "learning_rate": 2.6143359100491917e-05, "loss": 0.1259, "step": 15353 }, { "epoch": 10.789880534082924, "grad_norm": 1.0290172100067139, "learning_rate": 2.614289060669946e-05, "loss": 0.1794, "step": 15354 }, { "epoch": 10.79058327477161, "grad_norm": 1.6973087787628174, "learning_rate": 2.6142422112907004e-05, "loss": 0.2225, "step": 15355 }, { "epoch": 10.791286015460296, "grad_norm": 0.212325319647789, "learning_rate": 2.6141953619114548e-05, "loss": 0.0683, "step": 15356 }, { "epoch": 10.791988756148982, "grad_norm": 0.11046416312456131, "learning_rate": 2.614148512532209e-05, "loss": 0.0279, "step": 15357 }, { "epoch": 10.792691496837667, "grad_norm": 0.1994098424911499, "learning_rate": 2.6141016631529632e-05, "loss": 0.0225, "step": 15358 }, { "epoch": 10.793394237526353, "grad_norm": 0.17162559926509857, "learning_rate": 2.6140548137737176e-05, "loss": 0.0155, "step": 15359 }, { "epoch": 10.79409697821504, "grad_norm": 0.1382276862859726, "learning_rate": 2.614007964394472e-05, "loss": 0.015, "step": 15360 }, { "epoch": 10.794799718903725, "grad_norm": 0.20770806074142456, "learning_rate": 2.6139611150152263e-05, "loss": 0.0222, "step": 15361 }, { "epoch": 10.795502459592411, "grad_norm": 0.1505495011806488, "learning_rate": 2.6139142656359803e-05, "loss": 0.0098, "step": 15362 }, { "epoch": 10.796205200281097, "grad_norm": 0.10568428784608841, "learning_rate": 2.6138674162567347e-05, "loss": 0.008, "step": 15363 }, { "epoch": 10.796907940969781, "grad_norm": 0.20851072669029236, "learning_rate": 2.613820566877489e-05, "loss": 0.0568, "step": 15364 }, { "epoch": 10.797610681658469, "grad_norm": 0.09502649307250977, "learning_rate": 2.6137737174982435e-05, "loss": 0.0172, "step": 15365 }, { "epoch": 10.798313422347153, "grad_norm": 0.29943782091140747, "learning_rate": 2.6137268681189975e-05, "loss": 0.0136, "step": 15366 }, { "epoch": 10.799016163035839, "grad_norm": 0.14481176435947418, "learning_rate": 2.613680018739752e-05, "loss": 0.0174, "step": 15367 }, { "epoch": 10.799718903724525, "grad_norm": 0.14080049097537994, "learning_rate": 2.613633169360506e-05, "loss": 0.0306, "step": 15368 }, { "epoch": 10.80042164441321, "grad_norm": 0.12424690276384354, "learning_rate": 2.6135863199812603e-05, "loss": 0.0166, "step": 15369 }, { "epoch": 10.801124385101897, "grad_norm": 0.28062379360198975, "learning_rate": 2.6135394706020146e-05, "loss": 0.0274, "step": 15370 }, { "epoch": 10.801827125790583, "grad_norm": 0.16785942018032074, "learning_rate": 2.6134926212227687e-05, "loss": 0.0317, "step": 15371 }, { "epoch": 10.802529866479269, "grad_norm": 0.16058717668056488, "learning_rate": 2.613445771843523e-05, "loss": 0.0217, "step": 15372 }, { "epoch": 10.803232607167955, "grad_norm": 0.28043460845947266, "learning_rate": 2.6133989224642774e-05, "loss": 0.0278, "step": 15373 }, { "epoch": 10.80393534785664, "grad_norm": 0.28275689482688904, "learning_rate": 2.6133520730850318e-05, "loss": 0.0466, "step": 15374 }, { "epoch": 10.804638088545326, "grad_norm": 0.4297564625740051, "learning_rate": 2.6133052237057858e-05, "loss": 0.0523, "step": 15375 }, { "epoch": 10.805340829234012, "grad_norm": 0.2641294300556183, "learning_rate": 2.6132583743265402e-05, "loss": 0.0581, "step": 15376 }, { "epoch": 10.806043569922698, "grad_norm": 0.4566677212715149, "learning_rate": 2.6132115249472946e-05, "loss": 0.1078, "step": 15377 }, { "epoch": 10.806746310611384, "grad_norm": 0.6651168465614319, "learning_rate": 2.613164675568049e-05, "loss": 0.1568, "step": 15378 }, { "epoch": 10.80744905130007, "grad_norm": 0.7645195126533508, "learning_rate": 2.613117826188803e-05, "loss": 0.1488, "step": 15379 }, { "epoch": 10.808151791988756, "grad_norm": 1.038184404373169, "learning_rate": 2.6130709768095574e-05, "loss": 0.1775, "step": 15380 }, { "epoch": 10.808854532677442, "grad_norm": 0.20606862008571625, "learning_rate": 2.6130241274303117e-05, "loss": 0.0651, "step": 15381 }, { "epoch": 10.809557273366128, "grad_norm": 0.1844870150089264, "learning_rate": 2.612977278051066e-05, "loss": 0.0233, "step": 15382 }, { "epoch": 10.810260014054814, "grad_norm": 0.15378372371196747, "learning_rate": 2.6129304286718205e-05, "loss": 0.0232, "step": 15383 }, { "epoch": 10.8109627547435, "grad_norm": 0.19772480428218842, "learning_rate": 2.6128835792925745e-05, "loss": 0.0102, "step": 15384 }, { "epoch": 10.811665495432186, "grad_norm": 0.15247094631195068, "learning_rate": 2.6128367299133285e-05, "loss": 0.0254, "step": 15385 }, { "epoch": 10.812368236120872, "grad_norm": 0.188715398311615, "learning_rate": 2.612789880534083e-05, "loss": 0.0203, "step": 15386 }, { "epoch": 10.813070976809557, "grad_norm": 0.2971233129501343, "learning_rate": 2.6127430311548373e-05, "loss": 0.0293, "step": 15387 }, { "epoch": 10.813773717498243, "grad_norm": 0.3403303623199463, "learning_rate": 2.6126961817755913e-05, "loss": 0.0226, "step": 15388 }, { "epoch": 10.81447645818693, "grad_norm": 0.14739841222763062, "learning_rate": 2.6126493323963457e-05, "loss": 0.026, "step": 15389 }, { "epoch": 10.815179198875615, "grad_norm": 0.15895995497703552, "learning_rate": 2.6126024830171e-05, "loss": 0.0104, "step": 15390 }, { "epoch": 10.815881939564301, "grad_norm": 0.27577123045921326, "learning_rate": 2.6125556336378544e-05, "loss": 0.03, "step": 15391 }, { "epoch": 10.816584680252987, "grad_norm": 0.18332624435424805, "learning_rate": 2.6125087842586085e-05, "loss": 0.0147, "step": 15392 }, { "epoch": 10.817287420941673, "grad_norm": 0.20024806261062622, "learning_rate": 2.612461934879363e-05, "loss": 0.0255, "step": 15393 }, { "epoch": 10.817990161630359, "grad_norm": 0.09416691213846207, "learning_rate": 2.6124150855001172e-05, "loss": 0.0123, "step": 15394 }, { "epoch": 10.818692902319045, "grad_norm": 0.1845335215330124, "learning_rate": 2.6123682361208716e-05, "loss": 0.0249, "step": 15395 }, { "epoch": 10.81939564300773, "grad_norm": 0.17824605107307434, "learning_rate": 2.612321386741626e-05, "loss": 0.0424, "step": 15396 }, { "epoch": 10.820098383696417, "grad_norm": 0.17800748348236084, "learning_rate": 2.61227453736238e-05, "loss": 0.0263, "step": 15397 }, { "epoch": 10.820801124385103, "grad_norm": 0.44558727741241455, "learning_rate": 2.6122276879831344e-05, "loss": 0.0224, "step": 15398 }, { "epoch": 10.821503865073788, "grad_norm": 0.8846772909164429, "learning_rate": 2.6121808386038887e-05, "loss": 0.0695, "step": 15399 }, { "epoch": 10.822206605762474, "grad_norm": 0.30290448665618896, "learning_rate": 2.612133989224643e-05, "loss": 0.0433, "step": 15400 }, { "epoch": 10.82290934645116, "grad_norm": 0.5949831008911133, "learning_rate": 2.612087139845397e-05, "loss": 0.0858, "step": 15401 }, { "epoch": 10.823612087139846, "grad_norm": 0.7457334399223328, "learning_rate": 2.6120402904661512e-05, "loss": 0.1042, "step": 15402 }, { "epoch": 10.82431482782853, "grad_norm": 0.47201302647590637, "learning_rate": 2.6119934410869055e-05, "loss": 0.1351, "step": 15403 }, { "epoch": 10.825017568517218, "grad_norm": 0.7705894112586975, "learning_rate": 2.61194659170766e-05, "loss": 0.1398, "step": 15404 }, { "epoch": 10.825720309205902, "grad_norm": 1.214746356010437, "learning_rate": 2.611899742328414e-05, "loss": 0.2522, "step": 15405 }, { "epoch": 10.826423049894588, "grad_norm": 0.5030608177185059, "learning_rate": 2.6118528929491683e-05, "loss": 0.0537, "step": 15406 }, { "epoch": 10.827125790583274, "grad_norm": 0.11325526982545853, "learning_rate": 2.6118060435699227e-05, "loss": 0.0199, "step": 15407 }, { "epoch": 10.82782853127196, "grad_norm": 0.1196441799402237, "learning_rate": 2.611759194190677e-05, "loss": 0.0305, "step": 15408 }, { "epoch": 10.828531271960646, "grad_norm": 0.12801222503185272, "learning_rate": 2.6117123448114314e-05, "loss": 0.02, "step": 15409 }, { "epoch": 10.829234012649332, "grad_norm": 0.15538036823272705, "learning_rate": 2.6116654954321855e-05, "loss": 0.0183, "step": 15410 }, { "epoch": 10.829936753338018, "grad_norm": 0.12356353551149368, "learning_rate": 2.61161864605294e-05, "loss": 0.0121, "step": 15411 }, { "epoch": 10.830639494026704, "grad_norm": 0.12348341941833496, "learning_rate": 2.6115717966736942e-05, "loss": 0.0189, "step": 15412 }, { "epoch": 10.83134223471539, "grad_norm": 0.1558862030506134, "learning_rate": 2.6115249472944486e-05, "loss": 0.0325, "step": 15413 }, { "epoch": 10.832044975404076, "grad_norm": 0.09635324776172638, "learning_rate": 2.6114780979152026e-05, "loss": 0.0143, "step": 15414 }, { "epoch": 10.832747716092761, "grad_norm": 0.18176612257957458, "learning_rate": 2.611431248535957e-05, "loss": 0.047, "step": 15415 }, { "epoch": 10.833450456781447, "grad_norm": 0.24698933959007263, "learning_rate": 2.6113843991567114e-05, "loss": 0.0224, "step": 15416 }, { "epoch": 10.834153197470133, "grad_norm": 0.06903154402971268, "learning_rate": 2.6113375497774657e-05, "loss": 0.0134, "step": 15417 }, { "epoch": 10.83485593815882, "grad_norm": 0.1594376564025879, "learning_rate": 2.6112907003982198e-05, "loss": 0.0367, "step": 15418 }, { "epoch": 10.835558678847505, "grad_norm": 0.25496020913124084, "learning_rate": 2.611243851018974e-05, "loss": 0.0133, "step": 15419 }, { "epoch": 10.836261419536191, "grad_norm": 0.19558623433113098, "learning_rate": 2.6111970016397282e-05, "loss": 0.0338, "step": 15420 }, { "epoch": 10.836964160224877, "grad_norm": 0.3600279986858368, "learning_rate": 2.6111501522604826e-05, "loss": 0.0257, "step": 15421 }, { "epoch": 10.837666900913563, "grad_norm": 0.0925767794251442, "learning_rate": 2.611103302881237e-05, "loss": 0.0162, "step": 15422 }, { "epoch": 10.838369641602249, "grad_norm": 0.15536822378635406, "learning_rate": 2.611056453501991e-05, "loss": 0.0294, "step": 15423 }, { "epoch": 10.839072382290935, "grad_norm": 0.2124621421098709, "learning_rate": 2.6110096041227453e-05, "loss": 0.0394, "step": 15424 }, { "epoch": 10.83977512297962, "grad_norm": 0.2676282823085785, "learning_rate": 2.6109627547434997e-05, "loss": 0.0478, "step": 15425 }, { "epoch": 10.840477863668307, "grad_norm": 0.37312808632850647, "learning_rate": 2.610915905364254e-05, "loss": 0.0622, "step": 15426 }, { "epoch": 10.841180604356992, "grad_norm": 0.461544007062912, "learning_rate": 2.610869055985008e-05, "loss": 0.1257, "step": 15427 }, { "epoch": 10.841883345045678, "grad_norm": 1.2396193742752075, "learning_rate": 2.6108222066057625e-05, "loss": 0.1434, "step": 15428 }, { "epoch": 10.842586085734364, "grad_norm": 0.5394777059555054, "learning_rate": 2.610775357226517e-05, "loss": 0.1537, "step": 15429 }, { "epoch": 10.84328882642305, "grad_norm": 0.9669746160507202, "learning_rate": 2.6107285078472712e-05, "loss": 0.2425, "step": 15430 }, { "epoch": 10.843991567111736, "grad_norm": 0.2559802234172821, "learning_rate": 2.6106816584680253e-05, "loss": 0.0637, "step": 15431 }, { "epoch": 10.844694307800422, "grad_norm": 0.21649901568889618, "learning_rate": 2.6106348090887796e-05, "loss": 0.0222, "step": 15432 }, { "epoch": 10.845397048489108, "grad_norm": 0.1076745018362999, "learning_rate": 2.610587959709534e-05, "loss": 0.0157, "step": 15433 }, { "epoch": 10.846099789177794, "grad_norm": 0.17011553049087524, "learning_rate": 2.6105411103302884e-05, "loss": 0.0205, "step": 15434 }, { "epoch": 10.84680252986648, "grad_norm": 0.12330841273069382, "learning_rate": 2.6104942609510428e-05, "loss": 0.0171, "step": 15435 }, { "epoch": 10.847505270555166, "grad_norm": 0.17974606156349182, "learning_rate": 2.6104474115717968e-05, "loss": 0.0101, "step": 15436 }, { "epoch": 10.848208011243852, "grad_norm": 0.09066842496395111, "learning_rate": 2.6104005621925508e-05, "loss": 0.0102, "step": 15437 }, { "epoch": 10.848910751932538, "grad_norm": 0.12859345972537994, "learning_rate": 2.6103537128133052e-05, "loss": 0.0178, "step": 15438 }, { "epoch": 10.849613492621224, "grad_norm": 0.11463732272386551, "learning_rate": 2.6103068634340596e-05, "loss": 0.0308, "step": 15439 }, { "epoch": 10.85031623330991, "grad_norm": 0.10040530562400818, "learning_rate": 2.6102600140548136e-05, "loss": 0.0117, "step": 15440 }, { "epoch": 10.851018973998595, "grad_norm": 0.4268132746219635, "learning_rate": 2.610213164675568e-05, "loss": 0.0219, "step": 15441 }, { "epoch": 10.85172171468728, "grad_norm": 0.11715878546237946, "learning_rate": 2.6101663152963223e-05, "loss": 0.0228, "step": 15442 }, { "epoch": 10.852424455375965, "grad_norm": 0.46936845779418945, "learning_rate": 2.6101194659170767e-05, "loss": 0.0298, "step": 15443 }, { "epoch": 10.853127196064651, "grad_norm": 0.10120007395744324, "learning_rate": 2.6100726165378308e-05, "loss": 0.0178, "step": 15444 }, { "epoch": 10.853829936753337, "grad_norm": 0.12209081649780273, "learning_rate": 2.610025767158585e-05, "loss": 0.0158, "step": 15445 }, { "epoch": 10.854532677442023, "grad_norm": 0.20526088774204254, "learning_rate": 2.6099789177793395e-05, "loss": 0.0414, "step": 15446 }, { "epoch": 10.85523541813071, "grad_norm": 0.23063673079013824, "learning_rate": 2.609932068400094e-05, "loss": 0.0206, "step": 15447 }, { "epoch": 10.855938158819395, "grad_norm": 0.28465935587882996, "learning_rate": 2.6098852190208482e-05, "loss": 0.0263, "step": 15448 }, { "epoch": 10.856640899508081, "grad_norm": 0.7437887787818909, "learning_rate": 2.6098383696416023e-05, "loss": 0.0394, "step": 15449 }, { "epoch": 10.857343640196767, "grad_norm": 0.29220050573349, "learning_rate": 2.6097915202623567e-05, "loss": 0.0762, "step": 15450 }, { "epoch": 10.858046380885453, "grad_norm": 0.289145290851593, "learning_rate": 2.609744670883111e-05, "loss": 0.045, "step": 15451 }, { "epoch": 10.858749121574139, "grad_norm": 1.3330296277999878, "learning_rate": 2.6096978215038654e-05, "loss": 0.1323, "step": 15452 }, { "epoch": 10.859451862262825, "grad_norm": 0.6348970532417297, "learning_rate": 2.6096509721246194e-05, "loss": 0.1608, "step": 15453 }, { "epoch": 10.86015460295151, "grad_norm": 1.5695326328277588, "learning_rate": 2.6096041227453738e-05, "loss": 0.2012, "step": 15454 }, { "epoch": 10.860857343640197, "grad_norm": 0.9338257312774658, "learning_rate": 2.609557273366128e-05, "loss": 0.1738, "step": 15455 }, { "epoch": 10.861560084328882, "grad_norm": 0.32569923996925354, "learning_rate": 2.6095104239868822e-05, "loss": 0.0858, "step": 15456 }, { "epoch": 10.862262825017568, "grad_norm": 0.11245137453079224, "learning_rate": 2.6094635746076362e-05, "loss": 0.0211, "step": 15457 }, { "epoch": 10.862965565706254, "grad_norm": 0.20243483781814575, "learning_rate": 2.6094167252283906e-05, "loss": 0.0481, "step": 15458 }, { "epoch": 10.86366830639494, "grad_norm": 0.09060635417699814, "learning_rate": 2.609369875849145e-05, "loss": 0.0148, "step": 15459 }, { "epoch": 10.864371047083626, "grad_norm": 0.1290655881166458, "learning_rate": 2.6093230264698994e-05, "loss": 0.0145, "step": 15460 }, { "epoch": 10.865073787772312, "grad_norm": 0.11948718875646591, "learning_rate": 2.6092761770906537e-05, "loss": 0.0157, "step": 15461 }, { "epoch": 10.865776528460998, "grad_norm": 0.1348450928926468, "learning_rate": 2.6092293277114078e-05, "loss": 0.0097, "step": 15462 }, { "epoch": 10.866479269149684, "grad_norm": 0.10806598514318466, "learning_rate": 2.609182478332162e-05, "loss": 0.0197, "step": 15463 }, { "epoch": 10.86718200983837, "grad_norm": 0.2643935978412628, "learning_rate": 2.6091356289529165e-05, "loss": 0.0202, "step": 15464 }, { "epoch": 10.867884750527056, "grad_norm": 0.9739620089530945, "learning_rate": 2.609088779573671e-05, "loss": 0.0113, "step": 15465 }, { "epoch": 10.868587491215742, "grad_norm": 0.12973786890506744, "learning_rate": 2.609041930194425e-05, "loss": 0.0173, "step": 15466 }, { "epoch": 10.869290231904428, "grad_norm": 0.13615037500858307, "learning_rate": 2.6089950808151793e-05, "loss": 0.0122, "step": 15467 }, { "epoch": 10.869992972593113, "grad_norm": 0.19130973517894745, "learning_rate": 2.6089482314359337e-05, "loss": 0.0343, "step": 15468 }, { "epoch": 10.8706957132818, "grad_norm": 0.07690874487161636, "learning_rate": 2.608901382056688e-05, "loss": 0.0113, "step": 15469 }, { "epoch": 10.871398453970485, "grad_norm": 0.2020830661058426, "learning_rate": 2.608854532677442e-05, "loss": 0.0243, "step": 15470 }, { "epoch": 10.872101194659171, "grad_norm": 0.24156005680561066, "learning_rate": 2.6088076832981964e-05, "loss": 0.0331, "step": 15471 }, { "epoch": 10.872803935347857, "grad_norm": 0.16077974438667297, "learning_rate": 2.6087608339189505e-05, "loss": 0.0134, "step": 15472 }, { "epoch": 10.873506676036543, "grad_norm": 0.22239568829536438, "learning_rate": 2.608713984539705e-05, "loss": 0.0332, "step": 15473 }, { "epoch": 10.874209416725229, "grad_norm": 0.9847629070281982, "learning_rate": 2.6086671351604592e-05, "loss": 0.0556, "step": 15474 }, { "epoch": 10.874912157413915, "grad_norm": 0.21823851764202118, "learning_rate": 2.6086202857812133e-05, "loss": 0.0529, "step": 15475 }, { "epoch": 10.8756148981026, "grad_norm": 0.24713866412639618, "learning_rate": 2.6085734364019676e-05, "loss": 0.0562, "step": 15476 }, { "epoch": 10.876317638791287, "grad_norm": 0.44018757343292236, "learning_rate": 2.608526587022722e-05, "loss": 0.1052, "step": 15477 }, { "epoch": 10.877020379479973, "grad_norm": 0.5780289769172668, "learning_rate": 2.6084797376434764e-05, "loss": 0.1463, "step": 15478 }, { "epoch": 10.877723120168657, "grad_norm": 0.7125159502029419, "learning_rate": 2.6084328882642304e-05, "loss": 0.1563, "step": 15479 }, { "epoch": 10.878425860857345, "grad_norm": Infinity, "learning_rate": 2.6084328882642304e-05, "loss": 0.2229, "step": 15480 }, { "epoch": 10.879128601546029, "grad_norm": 0.20385335385799408, "learning_rate": 2.6083860388849848e-05, "loss": 0.0696, "step": 15481 }, { "epoch": 10.879831342234715, "grad_norm": 0.22583910822868347, "learning_rate": 2.608339189505739e-05, "loss": 0.0304, "step": 15482 }, { "epoch": 10.8805340829234, "grad_norm": 0.14903362095355988, "learning_rate": 2.6082923401264935e-05, "loss": 0.019, "step": 15483 }, { "epoch": 10.881236823612086, "grad_norm": 0.09633819013834, "learning_rate": 2.6082454907472476e-05, "loss": 0.0198, "step": 15484 }, { "epoch": 10.881939564300772, "grad_norm": 0.1292886584997177, "learning_rate": 2.608198641368002e-05, "loss": 0.0138, "step": 15485 }, { "epoch": 10.882642304989458, "grad_norm": 0.1028335690498352, "learning_rate": 2.6081517919887563e-05, "loss": 0.0082, "step": 15486 }, { "epoch": 10.883345045678144, "grad_norm": 0.07477099448442459, "learning_rate": 2.6081049426095107e-05, "loss": 0.0111, "step": 15487 }, { "epoch": 10.88404778636683, "grad_norm": 0.1507098525762558, "learning_rate": 2.608058093230265e-05, "loss": 0.0303, "step": 15488 }, { "epoch": 10.884750527055516, "grad_norm": 0.2023901641368866, "learning_rate": 2.608011243851019e-05, "loss": 0.0295, "step": 15489 }, { "epoch": 10.885453267744202, "grad_norm": 0.13202014565467834, "learning_rate": 2.607964394471773e-05, "loss": 0.0139, "step": 15490 }, { "epoch": 10.886156008432888, "grad_norm": 0.11852947622537613, "learning_rate": 2.6079175450925275e-05, "loss": 0.0199, "step": 15491 }, { "epoch": 10.886858749121574, "grad_norm": 0.12193844467401505, "learning_rate": 2.607870695713282e-05, "loss": 0.0155, "step": 15492 }, { "epoch": 10.88756148981026, "grad_norm": 0.24168898165225983, "learning_rate": 2.607823846334036e-05, "loss": 0.0268, "step": 15493 }, { "epoch": 10.888264230498946, "grad_norm": 0.09386929124593735, "learning_rate": 2.6077769969547903e-05, "loss": 0.0112, "step": 15494 }, { "epoch": 10.888966971187632, "grad_norm": 0.1343718320131302, "learning_rate": 2.6077301475755446e-05, "loss": 0.0367, "step": 15495 }, { "epoch": 10.889669711876317, "grad_norm": 0.12296479195356369, "learning_rate": 2.607683298196299e-05, "loss": 0.0251, "step": 15496 }, { "epoch": 10.890372452565003, "grad_norm": 0.25111669301986694, "learning_rate": 2.607636448817053e-05, "loss": 0.0315, "step": 15497 }, { "epoch": 10.89107519325369, "grad_norm": 0.2573074400424957, "learning_rate": 2.6075895994378074e-05, "loss": 0.027, "step": 15498 }, { "epoch": 10.891777933942375, "grad_norm": 0.2961517870426178, "learning_rate": 2.6075427500585618e-05, "loss": 0.0506, "step": 15499 }, { "epoch": 10.892480674631061, "grad_norm": 0.39118292927742004, "learning_rate": 2.607495900679316e-05, "loss": 0.0591, "step": 15500 }, { "epoch": 10.893183415319747, "grad_norm": 0.40577206015586853, "learning_rate": 2.6074490513000705e-05, "loss": 0.0871, "step": 15501 }, { "epoch": 10.893886156008433, "grad_norm": 0.7412991523742676, "learning_rate": 2.6074022019208246e-05, "loss": 0.0963, "step": 15502 }, { "epoch": 10.894588896697119, "grad_norm": 0.6649042963981628, "learning_rate": 2.607355352541579e-05, "loss": 0.1585, "step": 15503 }, { "epoch": 10.895291637385805, "grad_norm": 0.8683761358261108, "learning_rate": 2.6073085031623333e-05, "loss": 0.1724, "step": 15504 }, { "epoch": 10.89599437807449, "grad_norm": 0.8302462100982666, "learning_rate": 2.6072616537830877e-05, "loss": 0.2176, "step": 15505 }, { "epoch": 10.896697118763177, "grad_norm": 0.20422397553920746, "learning_rate": 2.6072148044038417e-05, "loss": 0.0548, "step": 15506 }, { "epoch": 10.897399859451863, "grad_norm": 0.23662441968917847, "learning_rate": 2.607167955024596e-05, "loss": 0.0303, "step": 15507 }, { "epoch": 10.898102600140549, "grad_norm": 0.12197016924619675, "learning_rate": 2.60712110564535e-05, "loss": 0.0193, "step": 15508 }, { "epoch": 10.898805340829234, "grad_norm": 0.13029518723487854, "learning_rate": 2.6070742562661045e-05, "loss": 0.0255, "step": 15509 }, { "epoch": 10.89950808151792, "grad_norm": 0.07875574380159378, "learning_rate": 2.6070274068868585e-05, "loss": 0.0133, "step": 15510 }, { "epoch": 10.900210822206606, "grad_norm": 0.07626311480998993, "learning_rate": 2.606980557507613e-05, "loss": 0.0128, "step": 15511 }, { "epoch": 10.900913562895292, "grad_norm": 0.1466478407382965, "learning_rate": 2.6069337081283673e-05, "loss": 0.0138, "step": 15512 }, { "epoch": 10.901616303583978, "grad_norm": 0.0936673954129219, "learning_rate": 2.6068868587491216e-05, "loss": 0.015, "step": 15513 }, { "epoch": 10.902319044272664, "grad_norm": 0.190620556473732, "learning_rate": 2.606840009369876e-05, "loss": 0.0211, "step": 15514 }, { "epoch": 10.90302178496135, "grad_norm": 0.07485909014940262, "learning_rate": 2.60679315999063e-05, "loss": 0.0098, "step": 15515 }, { "epoch": 10.903724525650036, "grad_norm": 0.0901532992720604, "learning_rate": 2.6067463106113844e-05, "loss": 0.0232, "step": 15516 }, { "epoch": 10.904427266338722, "grad_norm": 0.11062455922365189, "learning_rate": 2.6066994612321388e-05, "loss": 0.0164, "step": 15517 }, { "epoch": 10.905130007027406, "grad_norm": 0.1491369605064392, "learning_rate": 2.6066526118528932e-05, "loss": 0.0209, "step": 15518 }, { "epoch": 10.905832747716094, "grad_norm": 0.09562111645936966, "learning_rate": 2.6066057624736472e-05, "loss": 0.0127, "step": 15519 }, { "epoch": 10.906535488404778, "grad_norm": 0.08799675107002258, "learning_rate": 2.6065589130944016e-05, "loss": 0.0192, "step": 15520 }, { "epoch": 10.907238229093464, "grad_norm": 0.12635140120983124, "learning_rate": 2.606512063715156e-05, "loss": 0.0346, "step": 15521 }, { "epoch": 10.90794096978215, "grad_norm": 0.07870133966207504, "learning_rate": 2.6064652143359103e-05, "loss": 0.0136, "step": 15522 }, { "epoch": 10.908643710470836, "grad_norm": 0.17924991250038147, "learning_rate": 2.6064183649566644e-05, "loss": 0.0264, "step": 15523 }, { "epoch": 10.909346451159522, "grad_norm": 0.11035677045583725, "learning_rate": 2.6063715155774187e-05, "loss": 0.0231, "step": 15524 }, { "epoch": 10.910049191848207, "grad_norm": 0.22730283439159393, "learning_rate": 2.6063246661981728e-05, "loss": 0.0351, "step": 15525 }, { "epoch": 10.910751932536893, "grad_norm": 0.3690074384212494, "learning_rate": 2.606277816818927e-05, "loss": 0.0876, "step": 15526 }, { "epoch": 10.91145467322558, "grad_norm": 0.4890909790992737, "learning_rate": 2.6062309674396815e-05, "loss": 0.0983, "step": 15527 }, { "epoch": 10.912157413914265, "grad_norm": 0.48736241459846497, "learning_rate": 2.6061841180604355e-05, "loss": 0.1746, "step": 15528 }, { "epoch": 10.912860154602951, "grad_norm": 0.8302334547042847, "learning_rate": 2.60613726868119e-05, "loss": 0.1972, "step": 15529 }, { "epoch": 10.913562895291637, "grad_norm": 0.7415242195129395, "learning_rate": 2.6060904193019443e-05, "loss": 0.1766, "step": 15530 }, { "epoch": 10.914265635980323, "grad_norm": 0.18651309609413147, "learning_rate": 2.6060435699226987e-05, "loss": 0.054, "step": 15531 }, { "epoch": 10.914968376669009, "grad_norm": 0.1503438800573349, "learning_rate": 2.6059967205434527e-05, "loss": 0.0216, "step": 15532 }, { "epoch": 10.915671117357695, "grad_norm": 0.1168084591627121, "learning_rate": 2.605949871164207e-05, "loss": 0.0244, "step": 15533 }, { "epoch": 10.91637385804638, "grad_norm": 0.10434335470199585, "learning_rate": 2.6059030217849614e-05, "loss": 0.0154, "step": 15534 }, { "epoch": 10.917076598735067, "grad_norm": 0.2456035017967224, "learning_rate": 2.6058561724057158e-05, "loss": 0.022, "step": 15535 }, { "epoch": 10.917779339423753, "grad_norm": 0.1674378216266632, "learning_rate": 2.60580932302647e-05, "loss": 0.0083, "step": 15536 }, { "epoch": 10.918482080112438, "grad_norm": 0.25740212202072144, "learning_rate": 2.6057624736472242e-05, "loss": 0.0339, "step": 15537 }, { "epoch": 10.919184820801124, "grad_norm": 0.10639616847038269, "learning_rate": 2.6057156242679786e-05, "loss": 0.0126, "step": 15538 }, { "epoch": 10.91988756148981, "grad_norm": 0.18964119255542755, "learning_rate": 2.605668774888733e-05, "loss": 0.0162, "step": 15539 }, { "epoch": 10.920590302178496, "grad_norm": 0.10899554938077927, "learning_rate": 2.6056219255094873e-05, "loss": 0.0232, "step": 15540 }, { "epoch": 10.921293042867182, "grad_norm": 0.11486808210611343, "learning_rate": 2.6055750761302414e-05, "loss": 0.0246, "step": 15541 }, { "epoch": 10.921995783555868, "grad_norm": 0.2071787267923355, "learning_rate": 2.6055282267509957e-05, "loss": 0.0158, "step": 15542 }, { "epoch": 10.922698524244554, "grad_norm": 0.13965922594070435, "learning_rate": 2.6054813773717498e-05, "loss": 0.0249, "step": 15543 }, { "epoch": 10.92340126493324, "grad_norm": 0.15452155470848083, "learning_rate": 2.605434527992504e-05, "loss": 0.007, "step": 15544 }, { "epoch": 10.924104005621926, "grad_norm": 0.26270973682403564, "learning_rate": 2.6053876786132582e-05, "loss": 0.0345, "step": 15545 }, { "epoch": 10.924806746310612, "grad_norm": 0.3286009728908539, "learning_rate": 2.6053408292340126e-05, "loss": 0.0297, "step": 15546 }, { "epoch": 10.925509486999298, "grad_norm": 0.12669432163238525, "learning_rate": 2.605293979854767e-05, "loss": 0.0179, "step": 15547 }, { "epoch": 10.926212227687984, "grad_norm": 0.23595349490642548, "learning_rate": 2.6052471304755213e-05, "loss": 0.0436, "step": 15548 }, { "epoch": 10.92691496837667, "grad_norm": 0.22935040295124054, "learning_rate": 2.6052002810962757e-05, "loss": 0.0467, "step": 15549 }, { "epoch": 10.927617709065355, "grad_norm": 0.28142639994621277, "learning_rate": 2.6051534317170297e-05, "loss": 0.0346, "step": 15550 }, { "epoch": 10.928320449754041, "grad_norm": 0.35940125584602356, "learning_rate": 2.605106582337784e-05, "loss": 0.0975, "step": 15551 }, { "epoch": 10.929023190442727, "grad_norm": 0.41679227352142334, "learning_rate": 2.6050597329585384e-05, "loss": 0.1275, "step": 15552 }, { "epoch": 10.929725931131413, "grad_norm": 0.43774542212486267, "learning_rate": 2.6050128835792928e-05, "loss": 0.15, "step": 15553 }, { "epoch": 10.9304286718201, "grad_norm": 0.8370504975318909, "learning_rate": 2.604966034200047e-05, "loss": 0.1744, "step": 15554 }, { "epoch": 10.931131412508785, "grad_norm": 0.8740219473838806, "learning_rate": 2.6049191848208012e-05, "loss": 0.208, "step": 15555 }, { "epoch": 10.931834153197471, "grad_norm": 0.46728718280792236, "learning_rate": 2.6048723354415556e-05, "loss": 0.101, "step": 15556 }, { "epoch": 10.932536893886155, "grad_norm": 0.20143893361091614, "learning_rate": 2.60482548606231e-05, "loss": 0.0346, "step": 15557 }, { "epoch": 10.933239634574843, "grad_norm": 0.11707780510187149, "learning_rate": 2.604778636683064e-05, "loss": 0.0168, "step": 15558 }, { "epoch": 10.933942375263527, "grad_norm": 0.13646280765533447, "learning_rate": 2.6047317873038184e-05, "loss": 0.0197, "step": 15559 }, { "epoch": 10.934645115952213, "grad_norm": 0.0869923010468483, "learning_rate": 2.6046849379245724e-05, "loss": 0.0093, "step": 15560 }, { "epoch": 10.935347856640899, "grad_norm": 0.1555773913860321, "learning_rate": 2.6046380885453268e-05, "loss": 0.01, "step": 15561 }, { "epoch": 10.936050597329585, "grad_norm": 0.11398069560527802, "learning_rate": 2.604591239166081e-05, "loss": 0.0222, "step": 15562 }, { "epoch": 10.93675333801827, "grad_norm": 0.09239926189184189, "learning_rate": 2.6045443897868352e-05, "loss": 0.0177, "step": 15563 }, { "epoch": 10.937456078706957, "grad_norm": 0.12328463047742844, "learning_rate": 2.6044975404075896e-05, "loss": 0.0156, "step": 15564 }, { "epoch": 10.938158819395642, "grad_norm": 0.12823179364204407, "learning_rate": 2.604450691028344e-05, "loss": 0.0252, "step": 15565 }, { "epoch": 10.938861560084328, "grad_norm": 0.10857800394296646, "learning_rate": 2.6044038416490983e-05, "loss": 0.0281, "step": 15566 }, { "epoch": 10.939564300773014, "grad_norm": 0.12364639341831207, "learning_rate": 2.6043569922698523e-05, "loss": 0.0064, "step": 15567 }, { "epoch": 10.9402670414617, "grad_norm": 0.1497732549905777, "learning_rate": 2.6043101428906067e-05, "loss": 0.0122, "step": 15568 }, { "epoch": 10.940969782150386, "grad_norm": 0.13652849197387695, "learning_rate": 2.604263293511361e-05, "loss": 0.0185, "step": 15569 }, { "epoch": 10.941672522839072, "grad_norm": 0.28466251492500305, "learning_rate": 2.6042164441321155e-05, "loss": 0.0195, "step": 15570 }, { "epoch": 10.942375263527758, "grad_norm": 0.15199768543243408, "learning_rate": 2.6041695947528695e-05, "loss": 0.0274, "step": 15571 }, { "epoch": 10.943078004216444, "grad_norm": 0.11771194636821747, "learning_rate": 2.604122745373624e-05, "loss": 0.016, "step": 15572 }, { "epoch": 10.94378074490513, "grad_norm": 0.14262889325618744, "learning_rate": 2.6040758959943782e-05, "loss": 0.0317, "step": 15573 }, { "epoch": 10.944483485593816, "grad_norm": 0.1439865529537201, "learning_rate": 2.6040290466151326e-05, "loss": 0.0243, "step": 15574 }, { "epoch": 10.945186226282502, "grad_norm": 0.1783788800239563, "learning_rate": 2.603982197235887e-05, "loss": 0.038, "step": 15575 }, { "epoch": 10.945888966971188, "grad_norm": 0.37389281392097473, "learning_rate": 2.603935347856641e-05, "loss": 0.07, "step": 15576 }, { "epoch": 10.946591707659874, "grad_norm": 0.42374828457832336, "learning_rate": 2.603888498477395e-05, "loss": 0.083, "step": 15577 }, { "epoch": 10.94729444834856, "grad_norm": 0.6381624341011047, "learning_rate": 2.6038416490981494e-05, "loss": 0.1353, "step": 15578 }, { "epoch": 10.947997189037245, "grad_norm": 1.1448051929473877, "learning_rate": 2.6037947997189038e-05, "loss": 0.1851, "step": 15579 }, { "epoch": 10.948699929725931, "grad_norm": 1.3101478815078735, "learning_rate": 2.603747950339658e-05, "loss": 0.1777, "step": 15580 }, { "epoch": 10.949402670414617, "grad_norm": 0.2567567229270935, "learning_rate": 2.6037011009604122e-05, "loss": 0.0585, "step": 15581 }, { "epoch": 10.950105411103303, "grad_norm": 0.3073594868183136, "learning_rate": 2.6036542515811666e-05, "loss": 0.04, "step": 15582 }, { "epoch": 10.950808151791989, "grad_norm": 0.17369136214256287, "learning_rate": 2.603607402201921e-05, "loss": 0.0272, "step": 15583 }, { "epoch": 10.951510892480675, "grad_norm": 0.11962531507015228, "learning_rate": 2.603560552822675e-05, "loss": 0.0225, "step": 15584 }, { "epoch": 10.952213633169361, "grad_norm": 0.14781798422336578, "learning_rate": 2.6035137034434294e-05, "loss": 0.035, "step": 15585 }, { "epoch": 10.952916373858047, "grad_norm": 0.2445562481880188, "learning_rate": 2.6034668540641837e-05, "loss": 0.0227, "step": 15586 }, { "epoch": 10.953619114546733, "grad_norm": 0.12009379267692566, "learning_rate": 2.603420004684938e-05, "loss": 0.0194, "step": 15587 }, { "epoch": 10.954321855235419, "grad_norm": 0.5095903873443604, "learning_rate": 2.6033731553056925e-05, "loss": 0.0181, "step": 15588 }, { "epoch": 10.955024595924105, "grad_norm": 0.1288839429616928, "learning_rate": 2.6033263059264465e-05, "loss": 0.0179, "step": 15589 }, { "epoch": 10.95572733661279, "grad_norm": 0.15052619576454163, "learning_rate": 2.603279456547201e-05, "loss": 0.0258, "step": 15590 }, { "epoch": 10.956430077301476, "grad_norm": 0.23907887935638428, "learning_rate": 2.6032326071679553e-05, "loss": 0.018, "step": 15591 }, { "epoch": 10.957132817990162, "grad_norm": 0.14520381391048431, "learning_rate": 2.6031857577887096e-05, "loss": 0.0144, "step": 15592 }, { "epoch": 10.957835558678848, "grad_norm": 0.130891814827919, "learning_rate": 2.6031389084094637e-05, "loss": 0.0238, "step": 15593 }, { "epoch": 10.958538299367534, "grad_norm": 0.11307504028081894, "learning_rate": 2.603092059030218e-05, "loss": 0.0179, "step": 15594 }, { "epoch": 10.95924104005622, "grad_norm": 0.14959470927715302, "learning_rate": 2.603045209650972e-05, "loss": 0.0268, "step": 15595 }, { "epoch": 10.959943780744904, "grad_norm": 0.6208428144454956, "learning_rate": 2.6029983602717264e-05, "loss": 0.0254, "step": 15596 }, { "epoch": 10.96064652143359, "grad_norm": 0.18153555691242218, "learning_rate": 2.6029515108924805e-05, "loss": 0.023, "step": 15597 }, { "epoch": 10.961349262122276, "grad_norm": 0.20049284398555756, "learning_rate": 2.602904661513235e-05, "loss": 0.0351, "step": 15598 }, { "epoch": 10.962052002810962, "grad_norm": 0.4871216118335724, "learning_rate": 2.6028578121339892e-05, "loss": 0.0342, "step": 15599 }, { "epoch": 10.962754743499648, "grad_norm": 0.3559231758117676, "learning_rate": 2.6028109627547436e-05, "loss": 0.0593, "step": 15600 }, { "epoch": 10.963457484188334, "grad_norm": 0.5435808300971985, "learning_rate": 2.602764113375498e-05, "loss": 0.1073, "step": 15601 }, { "epoch": 10.96416022487702, "grad_norm": 1.705763578414917, "learning_rate": 2.602717263996252e-05, "loss": 0.0982, "step": 15602 }, { "epoch": 10.964862965565706, "grad_norm": 0.9512278437614441, "learning_rate": 2.6026704146170064e-05, "loss": 0.1656, "step": 15603 }, { "epoch": 10.965565706254392, "grad_norm": 0.6008925437927246, "learning_rate": 2.6026235652377607e-05, "loss": 0.1482, "step": 15604 }, { "epoch": 10.966268446943078, "grad_norm": 1.142792820930481, "learning_rate": 2.602576715858515e-05, "loss": 0.1778, "step": 15605 }, { "epoch": 10.966971187631763, "grad_norm": 0.6287229657173157, "learning_rate": 2.602529866479269e-05, "loss": 0.0567, "step": 15606 }, { "epoch": 10.96767392832045, "grad_norm": 0.19153304398059845, "learning_rate": 2.6024830171000235e-05, "loss": 0.036, "step": 15607 }, { "epoch": 10.968376669009135, "grad_norm": 0.2437935471534729, "learning_rate": 2.602436167720778e-05, "loss": 0.025, "step": 15608 }, { "epoch": 10.969079409697821, "grad_norm": 0.14532172679901123, "learning_rate": 2.6023893183415323e-05, "loss": 0.0264, "step": 15609 }, { "epoch": 10.969782150386507, "grad_norm": 0.14606404304504395, "learning_rate": 2.6023424689622863e-05, "loss": 0.0171, "step": 15610 }, { "epoch": 10.970484891075193, "grad_norm": 0.16351141035556793, "learning_rate": 2.6022956195830407e-05, "loss": 0.0165, "step": 15611 }, { "epoch": 10.971187631763879, "grad_norm": 0.15399861335754395, "learning_rate": 2.6022487702037947e-05, "loss": 0.0216, "step": 15612 }, { "epoch": 10.971890372452565, "grad_norm": 0.07929738610982895, "learning_rate": 2.602201920824549e-05, "loss": 0.0122, "step": 15613 }, { "epoch": 10.97259311314125, "grad_norm": 0.1528872549533844, "learning_rate": 2.6021550714453034e-05, "loss": 0.0253, "step": 15614 }, { "epoch": 10.973295853829937, "grad_norm": 0.09182653576135635, "learning_rate": 2.6021082220660575e-05, "loss": 0.0086, "step": 15615 }, { "epoch": 10.973998594518623, "grad_norm": 0.2569291591644287, "learning_rate": 2.602061372686812e-05, "loss": 0.03, "step": 15616 }, { "epoch": 10.974701335207309, "grad_norm": 0.17063795030117035, "learning_rate": 2.6020145233075662e-05, "loss": 0.0198, "step": 15617 }, { "epoch": 10.975404075895995, "grad_norm": 0.2014821618795395, "learning_rate": 2.6019676739283206e-05, "loss": 0.0444, "step": 15618 }, { "epoch": 10.97610681658468, "grad_norm": 0.12679794430732727, "learning_rate": 2.6019208245490746e-05, "loss": 0.0165, "step": 15619 }, { "epoch": 10.976809557273366, "grad_norm": 0.14080491662025452, "learning_rate": 2.601873975169829e-05, "loss": 0.0201, "step": 15620 }, { "epoch": 10.977512297962052, "grad_norm": 0.27731671929359436, "learning_rate": 2.6018271257905834e-05, "loss": 0.0297, "step": 15621 }, { "epoch": 10.978215038650738, "grad_norm": 0.21133653819561005, "learning_rate": 2.6017802764113378e-05, "loss": 0.0207, "step": 15622 }, { "epoch": 10.978917779339424, "grad_norm": 0.1660260111093521, "learning_rate": 2.6017334270320918e-05, "loss": 0.0414, "step": 15623 }, { "epoch": 10.97962052002811, "grad_norm": 0.13204169273376465, "learning_rate": 2.601686577652846e-05, "loss": 0.0251, "step": 15624 }, { "epoch": 10.980323260716796, "grad_norm": 0.28402969241142273, "learning_rate": 2.6016397282736005e-05, "loss": 0.0264, "step": 15625 }, { "epoch": 10.981026001405482, "grad_norm": 0.302285760641098, "learning_rate": 2.601592878894355e-05, "loss": 0.058, "step": 15626 }, { "epoch": 10.981728742094168, "grad_norm": 0.32472628355026245, "learning_rate": 2.6015460295151093e-05, "loss": 0.127, "step": 15627 }, { "epoch": 10.982431482782854, "grad_norm": 1.2328708171844482, "learning_rate": 2.6014991801358633e-05, "loss": 0.143, "step": 15628 }, { "epoch": 10.98313422347154, "grad_norm": 0.953971266746521, "learning_rate": 2.6014523307566177e-05, "loss": 0.1556, "step": 15629 }, { "epoch": 10.983836964160226, "grad_norm": 1.1234796047210693, "learning_rate": 2.6014054813773717e-05, "loss": 0.2102, "step": 15630 }, { "epoch": 10.984539704848912, "grad_norm": 0.18547682464122772, "learning_rate": 2.601358631998126e-05, "loss": 0.0569, "step": 15631 }, { "epoch": 10.985242445537597, "grad_norm": 0.15109403431415558, "learning_rate": 2.60131178261888e-05, "loss": 0.0201, "step": 15632 }, { "epoch": 10.985945186226282, "grad_norm": 0.10614565014839172, "learning_rate": 2.6012649332396345e-05, "loss": 0.0196, "step": 15633 }, { "epoch": 10.98664792691497, "grad_norm": 0.2532556354999542, "learning_rate": 2.601218083860389e-05, "loss": 0.0418, "step": 15634 }, { "epoch": 10.987350667603653, "grad_norm": 0.12898705899715424, "learning_rate": 2.6011712344811432e-05, "loss": 0.0134, "step": 15635 }, { "epoch": 10.98805340829234, "grad_norm": 0.13510631024837494, "learning_rate": 2.6011243851018973e-05, "loss": 0.0223, "step": 15636 }, { "epoch": 10.988756148981025, "grad_norm": 0.5438981056213379, "learning_rate": 2.6010775357226516e-05, "loss": 0.0134, "step": 15637 }, { "epoch": 10.989458889669711, "grad_norm": 0.5512681007385254, "learning_rate": 2.601030686343406e-05, "loss": 0.0177, "step": 15638 }, { "epoch": 10.990161630358397, "grad_norm": 0.21830162405967712, "learning_rate": 2.6009838369641604e-05, "loss": 0.0109, "step": 15639 }, { "epoch": 10.990864371047083, "grad_norm": 0.20249930024147034, "learning_rate": 2.6009369875849148e-05, "loss": 0.0226, "step": 15640 }, { "epoch": 10.991567111735769, "grad_norm": 0.17132613062858582, "learning_rate": 2.6008901382056688e-05, "loss": 0.021, "step": 15641 }, { "epoch": 10.992269852424455, "grad_norm": 0.19337914884090424, "learning_rate": 2.600843288826423e-05, "loss": 0.0333, "step": 15642 }, { "epoch": 10.99297259311314, "grad_norm": 0.07567064464092255, "learning_rate": 2.6007964394471775e-05, "loss": 0.0158, "step": 15643 }, { "epoch": 10.993675333801827, "grad_norm": 0.2437915802001953, "learning_rate": 2.600749590067932e-05, "loss": 0.0199, "step": 15644 }, { "epoch": 10.994378074490513, "grad_norm": 0.45769259333610535, "learning_rate": 2.600702740688686e-05, "loss": 0.0153, "step": 15645 }, { "epoch": 10.995080815179199, "grad_norm": 0.14882147312164307, "learning_rate": 2.6006558913094403e-05, "loss": 0.0316, "step": 15646 }, { "epoch": 10.995783555867884, "grad_norm": 0.21347665786743164, "learning_rate": 2.6006090419301944e-05, "loss": 0.0383, "step": 15647 }, { "epoch": 10.99648629655657, "grad_norm": 0.45094215869903564, "learning_rate": 2.6005621925509487e-05, "loss": 0.0639, "step": 15648 }, { "epoch": 10.997189037245256, "grad_norm": 0.3254939913749695, "learning_rate": 2.6005153431717028e-05, "loss": 0.068, "step": 15649 }, { "epoch": 10.997891777933942, "grad_norm": 0.42950892448425293, "learning_rate": 2.600468493792457e-05, "loss": 0.137, "step": 15650 }, { "epoch": 10.998594518622628, "grad_norm": 0.5933679938316345, "learning_rate": 2.6004216444132115e-05, "loss": 0.1463, "step": 15651 }, { "epoch": 10.999297259311314, "grad_norm": 0.9166581034660339, "learning_rate": 2.600374795033966e-05, "loss": 0.2283, "step": 15652 }, { "epoch": 11.0, "grad_norm": 0.46049070358276367, "learning_rate": 2.6003279456547202e-05, "loss": 0.1168, "step": 15653 }, { "epoch": 11.000702740688686, "grad_norm": 0.3804231584072113, "learning_rate": 2.6002810962754743e-05, "loss": 0.0623, "step": 15654 }, { "epoch": 11.001405481377372, "grad_norm": 0.10023080557584763, "learning_rate": 2.6002342468962287e-05, "loss": 0.0193, "step": 15655 }, { "epoch": 11.002108222066058, "grad_norm": 0.21430540084838867, "learning_rate": 2.600187397516983e-05, "loss": 0.0303, "step": 15656 }, { "epoch": 11.002810962754744, "grad_norm": 0.12100803107023239, "learning_rate": 2.6001405481377374e-05, "loss": 0.0168, "step": 15657 }, { "epoch": 11.00351370344343, "grad_norm": 0.14075656235218048, "learning_rate": 2.6000936987584914e-05, "loss": 0.0272, "step": 15658 }, { "epoch": 11.004216444132116, "grad_norm": 0.3579648733139038, "learning_rate": 2.6000468493792458e-05, "loss": 0.0117, "step": 15659 }, { "epoch": 11.004919184820801, "grad_norm": 0.09537724405527115, "learning_rate": 2.6000000000000002e-05, "loss": 0.0116, "step": 15660 }, { "epoch": 11.005621925509487, "grad_norm": 0.1273154765367508, "learning_rate": 2.5999531506207546e-05, "loss": 0.0172, "step": 15661 }, { "epoch": 11.006324666198173, "grad_norm": 0.11217910051345825, "learning_rate": 2.5999063012415086e-05, "loss": 0.0129, "step": 15662 }, { "epoch": 11.00702740688686, "grad_norm": 0.11687246710062027, "learning_rate": 2.599859451862263e-05, "loss": 0.0114, "step": 15663 }, { "epoch": 11.007730147575545, "grad_norm": 0.19657780230045319, "learning_rate": 2.5998126024830173e-05, "loss": 0.0206, "step": 15664 }, { "epoch": 11.008432888264231, "grad_norm": 0.11888603121042252, "learning_rate": 2.5997657531037714e-05, "loss": 0.0107, "step": 15665 }, { "epoch": 11.009135628952917, "grad_norm": 0.10738997161388397, "learning_rate": 2.5997189037245257e-05, "loss": 0.033, "step": 15666 }, { "epoch": 11.009838369641603, "grad_norm": 0.12346199154853821, "learning_rate": 2.5996720543452798e-05, "loss": 0.0133, "step": 15667 }, { "epoch": 11.010541110330289, "grad_norm": 0.24868988990783691, "learning_rate": 2.599625204966034e-05, "loss": 0.0288, "step": 15668 }, { "epoch": 11.011243851018975, "grad_norm": 0.2029963582754135, "learning_rate": 2.5995783555867885e-05, "loss": 0.024, "step": 15669 }, { "epoch": 11.01194659170766, "grad_norm": 0.15760160982608795, "learning_rate": 2.599531506207543e-05, "loss": 0.0278, "step": 15670 }, { "epoch": 11.012649332396347, "grad_norm": 0.22064337134361267, "learning_rate": 2.599484656828297e-05, "loss": 0.0398, "step": 15671 }, { "epoch": 11.013352073085033, "grad_norm": 0.12558893859386444, "learning_rate": 2.5994378074490513e-05, "loss": 0.0365, "step": 15672 }, { "epoch": 11.014054813773717, "grad_norm": 0.22800855338573456, "learning_rate": 2.5993909580698057e-05, "loss": 0.0436, "step": 15673 }, { "epoch": 11.014757554462403, "grad_norm": 0.21749860048294067, "learning_rate": 2.59934410869056e-05, "loss": 0.0595, "step": 15674 }, { "epoch": 11.015460295151088, "grad_norm": 0.565197765827179, "learning_rate": 2.599297259311314e-05, "loss": 0.0921, "step": 15675 }, { "epoch": 11.016163035839774, "grad_norm": 0.7759774923324585, "learning_rate": 2.5992504099320684e-05, "loss": 0.1209, "step": 15676 }, { "epoch": 11.01686577652846, "grad_norm": 0.6350467205047607, "learning_rate": 2.5992035605528228e-05, "loss": 0.1658, "step": 15677 }, { "epoch": 11.017568517217146, "grad_norm": 1.2806320190429688, "learning_rate": 2.5991567111735772e-05, "loss": 0.1754, "step": 15678 }, { "epoch": 11.018271257905832, "grad_norm": 0.3086373805999756, "learning_rate": 2.5991098617943316e-05, "loss": 0.0647, "step": 15679 }, { "epoch": 11.018973998594518, "grad_norm": 0.18513451516628265, "learning_rate": 2.5990630124150856e-05, "loss": 0.0265, "step": 15680 }, { "epoch": 11.019676739283204, "grad_norm": 0.12083390355110168, "learning_rate": 2.59901616303584e-05, "loss": 0.0133, "step": 15681 }, { "epoch": 11.02037947997189, "grad_norm": 0.09767550975084305, "learning_rate": 2.598969313656594e-05, "loss": 0.0133, "step": 15682 }, { "epoch": 11.021082220660576, "grad_norm": 0.2274075746536255, "learning_rate": 2.5989224642773484e-05, "loss": 0.0214, "step": 15683 }, { "epoch": 11.021784961349262, "grad_norm": 0.04614641144871712, "learning_rate": 2.5988756148981024e-05, "loss": 0.006, "step": 15684 }, { "epoch": 11.022487702037948, "grad_norm": 0.11495564132928848, "learning_rate": 2.5988287655188568e-05, "loss": 0.0269, "step": 15685 }, { "epoch": 11.023190442726634, "grad_norm": 0.11179263889789581, "learning_rate": 2.598781916139611e-05, "loss": 0.0165, "step": 15686 }, { "epoch": 11.02389318341532, "grad_norm": 0.19640085101127625, "learning_rate": 2.5987350667603655e-05, "loss": 0.0188, "step": 15687 }, { "epoch": 11.024595924104005, "grad_norm": 0.10660363733768463, "learning_rate": 2.5986882173811196e-05, "loss": 0.0177, "step": 15688 }, { "epoch": 11.025298664792691, "grad_norm": 0.14276526868343353, "learning_rate": 2.598641368001874e-05, "loss": 0.0236, "step": 15689 }, { "epoch": 11.026001405481377, "grad_norm": 0.14753131568431854, "learning_rate": 2.5985945186226283e-05, "loss": 0.0191, "step": 15690 }, { "epoch": 11.026704146170063, "grad_norm": 0.3548637330532074, "learning_rate": 2.5985476692433827e-05, "loss": 0.0174, "step": 15691 }, { "epoch": 11.02740688685875, "grad_norm": 0.14140821993350983, "learning_rate": 2.598500819864137e-05, "loss": 0.0145, "step": 15692 }, { "epoch": 11.028109627547435, "grad_norm": 0.2615775167942047, "learning_rate": 2.598453970484891e-05, "loss": 0.0274, "step": 15693 }, { "epoch": 11.028812368236121, "grad_norm": 0.13940639793872833, "learning_rate": 2.5984071211056455e-05, "loss": 0.0301, "step": 15694 }, { "epoch": 11.029515108924807, "grad_norm": 0.13880668580532074, "learning_rate": 2.5983602717263998e-05, "loss": 0.0117, "step": 15695 }, { "epoch": 11.030217849613493, "grad_norm": 0.18969160318374634, "learning_rate": 2.5983134223471542e-05, "loss": 0.0438, "step": 15696 }, { "epoch": 11.030920590302179, "grad_norm": 0.36474817991256714, "learning_rate": 2.5982665729679082e-05, "loss": 0.0308, "step": 15697 }, { "epoch": 11.031623330990865, "grad_norm": 0.20893988013267517, "learning_rate": 2.5982197235886626e-05, "loss": 0.0393, "step": 15698 }, { "epoch": 11.03232607167955, "grad_norm": 0.47666072845458984, "learning_rate": 2.5981728742094166e-05, "loss": 0.0772, "step": 15699 }, { "epoch": 11.033028812368237, "grad_norm": 0.4604697823524475, "learning_rate": 2.598126024830171e-05, "loss": 0.0836, "step": 15700 }, { "epoch": 11.033731553056922, "grad_norm": 1.1060556173324585, "learning_rate": 2.598079175450925e-05, "loss": 0.1847, "step": 15701 }, { "epoch": 11.034434293745608, "grad_norm": 0.6482527256011963, "learning_rate": 2.5980323260716794e-05, "loss": 0.1571, "step": 15702 }, { "epoch": 11.035137034434294, "grad_norm": 0.7303730249404907, "learning_rate": 2.5979854766924338e-05, "loss": 0.1498, "step": 15703 }, { "epoch": 11.03583977512298, "grad_norm": 0.57386314868927, "learning_rate": 2.597938627313188e-05, "loss": 0.0761, "step": 15704 }, { "epoch": 11.036542515811666, "grad_norm": 0.11156611889600754, "learning_rate": 2.5978917779339425e-05, "loss": 0.0186, "step": 15705 }, { "epoch": 11.037245256500352, "grad_norm": 0.2274051308631897, "learning_rate": 2.5978449285546966e-05, "loss": 0.0297, "step": 15706 }, { "epoch": 11.037947997189038, "grad_norm": 0.08256752789020538, "learning_rate": 2.597798079175451e-05, "loss": 0.0176, "step": 15707 }, { "epoch": 11.038650737877724, "grad_norm": 0.08324193954467773, "learning_rate": 2.5977512297962053e-05, "loss": 0.0193, "step": 15708 }, { "epoch": 11.03935347856641, "grad_norm": 0.10293512046337128, "learning_rate": 2.5977043804169597e-05, "loss": 0.0099, "step": 15709 }, { "epoch": 11.040056219255096, "grad_norm": 0.12819181382656097, "learning_rate": 2.5976575310377137e-05, "loss": 0.023, "step": 15710 }, { "epoch": 11.04075895994378, "grad_norm": 0.43945321440696716, "learning_rate": 2.597610681658468e-05, "loss": 0.0262, "step": 15711 }, { "epoch": 11.041461700632466, "grad_norm": 0.6121578812599182, "learning_rate": 2.5975638322792225e-05, "loss": 0.0482, "step": 15712 }, { "epoch": 11.042164441321152, "grad_norm": 0.1016765758395195, "learning_rate": 2.597516982899977e-05, "loss": 0.0093, "step": 15713 }, { "epoch": 11.042867182009838, "grad_norm": 0.15872372686862946, "learning_rate": 2.597470133520731e-05, "loss": 0.0268, "step": 15714 }, { "epoch": 11.043569922698524, "grad_norm": 0.16187995672225952, "learning_rate": 2.5974232841414852e-05, "loss": 0.0109, "step": 15715 }, { "epoch": 11.04427266338721, "grad_norm": 0.18191282451152802, "learning_rate": 2.5973764347622396e-05, "loss": 0.0246, "step": 15716 }, { "epoch": 11.044975404075895, "grad_norm": 0.08155781775712967, "learning_rate": 2.5973295853829937e-05, "loss": 0.0085, "step": 15717 }, { "epoch": 11.045678144764581, "grad_norm": 0.2253909409046173, "learning_rate": 2.597282736003748e-05, "loss": 0.0349, "step": 15718 }, { "epoch": 11.046380885453267, "grad_norm": 0.19770832359790802, "learning_rate": 2.597235886624502e-05, "loss": 0.0344, "step": 15719 }, { "epoch": 11.047083626141953, "grad_norm": 0.12007368355989456, "learning_rate": 2.5971890372452564e-05, "loss": 0.0176, "step": 15720 }, { "epoch": 11.047786366830639, "grad_norm": 0.1862945556640625, "learning_rate": 2.5971421878660108e-05, "loss": 0.0208, "step": 15721 }, { "epoch": 11.048489107519325, "grad_norm": 0.29298314452171326, "learning_rate": 2.5970953384867652e-05, "loss": 0.0488, "step": 15722 }, { "epoch": 11.049191848208011, "grad_norm": 0.2838098704814911, "learning_rate": 2.5970484891075192e-05, "loss": 0.0539, "step": 15723 }, { "epoch": 11.049894588896697, "grad_norm": 0.33389583230018616, "learning_rate": 2.5970016397282736e-05, "loss": 0.0477, "step": 15724 }, { "epoch": 11.050597329585383, "grad_norm": 0.9146403074264526, "learning_rate": 2.596954790349028e-05, "loss": 0.1245, "step": 15725 }, { "epoch": 11.051300070274069, "grad_norm": 0.8261399269104004, "learning_rate": 2.5969079409697823e-05, "loss": 0.1531, "step": 15726 }, { "epoch": 11.052002810962755, "grad_norm": 0.9238146543502808, "learning_rate": 2.5968610915905364e-05, "loss": 0.207, "step": 15727 }, { "epoch": 11.05270555165144, "grad_norm": 1.3879432678222656, "learning_rate": 2.5968142422112907e-05, "loss": 0.1854, "step": 15728 }, { "epoch": 11.053408292340126, "grad_norm": 0.4158121347427368, "learning_rate": 2.596767392832045e-05, "loss": 0.0644, "step": 15729 }, { "epoch": 11.054111033028812, "grad_norm": 0.14374065399169922, "learning_rate": 2.5967205434527995e-05, "loss": 0.027, "step": 15730 }, { "epoch": 11.054813773717498, "grad_norm": 0.19260872900485992, "learning_rate": 2.596673694073554e-05, "loss": 0.0264, "step": 15731 }, { "epoch": 11.055516514406184, "grad_norm": 0.32385021448135376, "learning_rate": 2.596626844694308e-05, "loss": 0.0161, "step": 15732 }, { "epoch": 11.05621925509487, "grad_norm": 0.20422609150409698, "learning_rate": 2.5965799953150623e-05, "loss": 0.0145, "step": 15733 }, { "epoch": 11.056921995783556, "grad_norm": 0.17414915561676025, "learning_rate": 2.5965331459358163e-05, "loss": 0.0162, "step": 15734 }, { "epoch": 11.057624736472242, "grad_norm": 0.09561280161142349, "learning_rate": 2.5964862965565707e-05, "loss": 0.0115, "step": 15735 }, { "epoch": 11.058327477160928, "grad_norm": 0.09707546979188919, "learning_rate": 2.5964394471773247e-05, "loss": 0.0091, "step": 15736 }, { "epoch": 11.059030217849614, "grad_norm": 0.8161218166351318, "learning_rate": 2.596392597798079e-05, "loss": 0.0151, "step": 15737 }, { "epoch": 11.0597329585383, "grad_norm": 0.1347489207983017, "learning_rate": 2.5963457484188334e-05, "loss": 0.0243, "step": 15738 }, { "epoch": 11.060435699226986, "grad_norm": 0.5100241899490356, "learning_rate": 2.5962988990395878e-05, "loss": 0.0189, "step": 15739 }, { "epoch": 11.061138439915672, "grad_norm": 0.07634419947862625, "learning_rate": 2.5962520496603422e-05, "loss": 0.011, "step": 15740 }, { "epoch": 11.061841180604358, "grad_norm": 0.103687584400177, "learning_rate": 2.5962052002810962e-05, "loss": 0.0199, "step": 15741 }, { "epoch": 11.062543921293043, "grad_norm": 0.18516768515110016, "learning_rate": 2.5961583509018506e-05, "loss": 0.0156, "step": 15742 }, { "epoch": 11.06324666198173, "grad_norm": 0.47636792063713074, "learning_rate": 2.596111501522605e-05, "loss": 0.036, "step": 15743 }, { "epoch": 11.063949402670415, "grad_norm": 0.38380423188209534, "learning_rate": 2.5960646521433593e-05, "loss": 0.0516, "step": 15744 }, { "epoch": 11.064652143359101, "grad_norm": 0.10965104401111603, "learning_rate": 2.5960178027641134e-05, "loss": 0.0106, "step": 15745 }, { "epoch": 11.065354884047787, "grad_norm": 0.2218974232673645, "learning_rate": 2.5959709533848677e-05, "loss": 0.0257, "step": 15746 }, { "epoch": 11.066057624736473, "grad_norm": 0.2090042531490326, "learning_rate": 2.595924104005622e-05, "loss": 0.0593, "step": 15747 }, { "epoch": 11.066760365425159, "grad_norm": 0.9372153878211975, "learning_rate": 2.5958772546263765e-05, "loss": 0.0503, "step": 15748 }, { "epoch": 11.067463106113845, "grad_norm": 0.4413531720638275, "learning_rate": 2.5958304052471305e-05, "loss": 0.0984, "step": 15749 }, { "epoch": 11.068165846802529, "grad_norm": 0.4366624057292938, "learning_rate": 2.595783555867885e-05, "loss": 0.0976, "step": 15750 }, { "epoch": 11.068868587491215, "grad_norm": 1.2072919607162476, "learning_rate": 2.5957367064886393e-05, "loss": 0.1453, "step": 15751 }, { "epoch": 11.0695713281799, "grad_norm": 0.46870121359825134, "learning_rate": 2.5956898571093933e-05, "loss": 0.1634, "step": 15752 }, { "epoch": 11.070274068868587, "grad_norm": 1.2471128702163696, "learning_rate": 2.5956430077301477e-05, "loss": 0.1946, "step": 15753 }, { "epoch": 11.070976809557273, "grad_norm": 0.2685888707637787, "learning_rate": 2.5955961583509017e-05, "loss": 0.0448, "step": 15754 }, { "epoch": 11.071679550245959, "grad_norm": 0.1925748586654663, "learning_rate": 2.595549308971656e-05, "loss": 0.0238, "step": 15755 }, { "epoch": 11.072382290934645, "grad_norm": 0.09976422041654587, "learning_rate": 2.5955024595924105e-05, "loss": 0.0268, "step": 15756 }, { "epoch": 11.07308503162333, "grad_norm": 0.1034979596734047, "learning_rate": 2.5954556102131648e-05, "loss": 0.0107, "step": 15757 }, { "epoch": 11.073787772312016, "grad_norm": 0.09328779578208923, "learning_rate": 2.595408760833919e-05, "loss": 0.0132, "step": 15758 }, { "epoch": 11.074490513000702, "grad_norm": 0.1137966737151146, "learning_rate": 2.5953619114546732e-05, "loss": 0.011, "step": 15759 }, { "epoch": 11.075193253689388, "grad_norm": 0.09787911921739578, "learning_rate": 2.5953150620754276e-05, "loss": 0.0161, "step": 15760 }, { "epoch": 11.075895994378074, "grad_norm": 0.1199483647942543, "learning_rate": 2.595268212696182e-05, "loss": 0.0299, "step": 15761 }, { "epoch": 11.07659873506676, "grad_norm": 0.13421691954135895, "learning_rate": 2.595221363316936e-05, "loss": 0.018, "step": 15762 }, { "epoch": 11.077301475755446, "grad_norm": 0.22441215813159943, "learning_rate": 2.5951745139376904e-05, "loss": 0.0121, "step": 15763 }, { "epoch": 11.078004216444132, "grad_norm": 0.36007246375083923, "learning_rate": 2.5951276645584448e-05, "loss": 0.0149, "step": 15764 }, { "epoch": 11.078706957132818, "grad_norm": 0.11621629446744919, "learning_rate": 2.595080815179199e-05, "loss": 0.016, "step": 15765 }, { "epoch": 11.079409697821504, "grad_norm": 0.2336779087781906, "learning_rate": 2.5950339657999535e-05, "loss": 0.036, "step": 15766 }, { "epoch": 11.08011243851019, "grad_norm": 0.14128628373146057, "learning_rate": 2.5949871164207075e-05, "loss": 0.0153, "step": 15767 }, { "epoch": 11.080815179198876, "grad_norm": 0.2506658136844635, "learning_rate": 2.594940267041462e-05, "loss": 0.0499, "step": 15768 }, { "epoch": 11.081517919887562, "grad_norm": 0.18480151891708374, "learning_rate": 2.594893417662216e-05, "loss": 0.016, "step": 15769 }, { "epoch": 11.082220660576247, "grad_norm": 0.10590897500514984, "learning_rate": 2.5948465682829703e-05, "loss": 0.0195, "step": 15770 }, { "epoch": 11.082923401264933, "grad_norm": 0.9189013838768005, "learning_rate": 2.5947997189037243e-05, "loss": 0.0328, "step": 15771 }, { "epoch": 11.08362614195362, "grad_norm": 0.19208429753780365, "learning_rate": 2.5947528695244787e-05, "loss": 0.0301, "step": 15772 }, { "epoch": 11.084328882642305, "grad_norm": 0.3715018928050995, "learning_rate": 2.594706020145233e-05, "loss": 0.0456, "step": 15773 }, { "epoch": 11.085031623330991, "grad_norm": 0.2993139922618866, "learning_rate": 2.5946591707659875e-05, "loss": 0.0574, "step": 15774 }, { "epoch": 11.085734364019677, "grad_norm": 0.6204224228858948, "learning_rate": 2.5946123213867415e-05, "loss": 0.102, "step": 15775 }, { "epoch": 11.086437104708363, "grad_norm": 0.49405837059020996, "learning_rate": 2.594565472007496e-05, "loss": 0.123, "step": 15776 }, { "epoch": 11.087139845397049, "grad_norm": 0.7390380501747131, "learning_rate": 2.5945186226282502e-05, "loss": 0.1519, "step": 15777 }, { "epoch": 11.087842586085735, "grad_norm": 1.5483996868133545, "learning_rate": 2.5944717732490046e-05, "loss": 0.1759, "step": 15778 }, { "epoch": 11.08854532677442, "grad_norm": 0.21202827990055084, "learning_rate": 2.594424923869759e-05, "loss": 0.0693, "step": 15779 }, { "epoch": 11.089248067463107, "grad_norm": 0.2418929785490036, "learning_rate": 2.594378074490513e-05, "loss": 0.0281, "step": 15780 }, { "epoch": 11.089950808151793, "grad_norm": 0.24300062656402588, "learning_rate": 2.5943312251112674e-05, "loss": 0.0338, "step": 15781 }, { "epoch": 11.090653548840478, "grad_norm": 0.11263938993215561, "learning_rate": 2.5942843757320218e-05, "loss": 0.0126, "step": 15782 }, { "epoch": 11.091356289529164, "grad_norm": 0.14473584294319153, "learning_rate": 2.594237526352776e-05, "loss": 0.0131, "step": 15783 }, { "epoch": 11.09205903021785, "grad_norm": 0.11402153968811035, "learning_rate": 2.5941906769735302e-05, "loss": 0.0142, "step": 15784 }, { "epoch": 11.092761770906536, "grad_norm": 0.08208277821540833, "learning_rate": 2.5941438275942845e-05, "loss": 0.0141, "step": 15785 }, { "epoch": 11.093464511595222, "grad_norm": 0.14094457030296326, "learning_rate": 2.5940969782150386e-05, "loss": 0.0275, "step": 15786 }, { "epoch": 11.094167252283908, "grad_norm": 0.1236240342259407, "learning_rate": 2.594050128835793e-05, "loss": 0.0194, "step": 15787 }, { "epoch": 11.094869992972592, "grad_norm": 0.21031859517097473, "learning_rate": 2.594003279456547e-05, "loss": 0.0066, "step": 15788 }, { "epoch": 11.095572733661278, "grad_norm": 0.16982337832450867, "learning_rate": 2.5939564300773014e-05, "loss": 0.0218, "step": 15789 }, { "epoch": 11.096275474349964, "grad_norm": 0.128143310546875, "learning_rate": 2.5939095806980557e-05, "loss": 0.0123, "step": 15790 }, { "epoch": 11.09697821503865, "grad_norm": 0.22485998272895813, "learning_rate": 2.59386273131881e-05, "loss": 0.0232, "step": 15791 }, { "epoch": 11.097680955727336, "grad_norm": 0.117161326110363, "learning_rate": 2.5938158819395645e-05, "loss": 0.0277, "step": 15792 }, { "epoch": 11.098383696416022, "grad_norm": 0.11914310604333878, "learning_rate": 2.5937690325603185e-05, "loss": 0.0172, "step": 15793 }, { "epoch": 11.099086437104708, "grad_norm": 0.18047469854354858, "learning_rate": 2.593722183181073e-05, "loss": 0.0313, "step": 15794 }, { "epoch": 11.099789177793394, "grad_norm": 0.11052944511175156, "learning_rate": 2.5936753338018273e-05, "loss": 0.0163, "step": 15795 }, { "epoch": 11.10049191848208, "grad_norm": 0.2209535837173462, "learning_rate": 2.5936284844225816e-05, "loss": 0.0335, "step": 15796 }, { "epoch": 11.101194659170766, "grad_norm": 0.16880127787590027, "learning_rate": 2.5935816350433357e-05, "loss": 0.038, "step": 15797 }, { "epoch": 11.101897399859451, "grad_norm": 0.33844390511512756, "learning_rate": 2.59353478566409e-05, "loss": 0.0666, "step": 15798 }, { "epoch": 11.102600140548137, "grad_norm": 0.26342591643333435, "learning_rate": 2.5934879362848444e-05, "loss": 0.0766, "step": 15799 }, { "epoch": 11.103302881236823, "grad_norm": 0.810843825340271, "learning_rate": 2.5934410869055988e-05, "loss": 0.117, "step": 15800 }, { "epoch": 11.10400562192551, "grad_norm": 0.5221264958381653, "learning_rate": 2.5933942375263528e-05, "loss": 0.1499, "step": 15801 }, { "epoch": 11.104708362614195, "grad_norm": 0.544469952583313, "learning_rate": 2.5933473881471072e-05, "loss": 0.1523, "step": 15802 }, { "epoch": 11.105411103302881, "grad_norm": 1.6807359457015991, "learning_rate": 2.5933005387678616e-05, "loss": 0.1904, "step": 15803 }, { "epoch": 11.106113843991567, "grad_norm": 0.32100802659988403, "learning_rate": 2.5932536893886156e-05, "loss": 0.0783, "step": 15804 }, { "epoch": 11.106816584680253, "grad_norm": 0.17802542448043823, "learning_rate": 2.59320684000937e-05, "loss": 0.0335, "step": 15805 }, { "epoch": 11.107519325368939, "grad_norm": 0.14151105284690857, "learning_rate": 2.593159990630124e-05, "loss": 0.0181, "step": 15806 }, { "epoch": 11.108222066057625, "grad_norm": 0.0812540128827095, "learning_rate": 2.5931131412508784e-05, "loss": 0.0124, "step": 15807 }, { "epoch": 11.10892480674631, "grad_norm": 0.10401836782693863, "learning_rate": 2.5930662918716327e-05, "loss": 0.016, "step": 15808 }, { "epoch": 11.109627547434997, "grad_norm": 0.2927757501602173, "learning_rate": 2.593019442492387e-05, "loss": 0.0094, "step": 15809 }, { "epoch": 11.110330288123683, "grad_norm": 0.10923131555318832, "learning_rate": 2.592972593113141e-05, "loss": 0.015, "step": 15810 }, { "epoch": 11.111033028812368, "grad_norm": 0.09824118763208389, "learning_rate": 2.5929257437338955e-05, "loss": 0.0154, "step": 15811 }, { "epoch": 11.111735769501054, "grad_norm": 0.16852480173110962, "learning_rate": 2.59287889435465e-05, "loss": 0.0129, "step": 15812 }, { "epoch": 11.11243851018974, "grad_norm": 0.08097822219133377, "learning_rate": 2.5928320449754043e-05, "loss": 0.0101, "step": 15813 }, { "epoch": 11.113141250878426, "grad_norm": 0.17782311141490936, "learning_rate": 2.5927851955961583e-05, "loss": 0.0268, "step": 15814 }, { "epoch": 11.113843991567112, "grad_norm": 0.16584353148937225, "learning_rate": 2.5927383462169127e-05, "loss": 0.0101, "step": 15815 }, { "epoch": 11.114546732255798, "grad_norm": 0.22462685406208038, "learning_rate": 2.592691496837667e-05, "loss": 0.0268, "step": 15816 }, { "epoch": 11.115249472944484, "grad_norm": 0.19459721446037292, "learning_rate": 2.5926446474584214e-05, "loss": 0.0202, "step": 15817 }, { "epoch": 11.11595221363317, "grad_norm": 0.7295082807540894, "learning_rate": 2.5925977980791758e-05, "loss": 0.0554, "step": 15818 }, { "epoch": 11.116654954321856, "grad_norm": 0.14372262358665466, "learning_rate": 2.5925509486999298e-05, "loss": 0.0143, "step": 15819 }, { "epoch": 11.117357695010542, "grad_norm": 0.3459244966506958, "learning_rate": 2.5925040993206842e-05, "loss": 0.0258, "step": 15820 }, { "epoch": 11.118060435699228, "grad_norm": 0.26123103499412537, "learning_rate": 2.5924572499414382e-05, "loss": 0.0256, "step": 15821 }, { "epoch": 11.118763176387914, "grad_norm": 0.23199868202209473, "learning_rate": 2.5924104005621926e-05, "loss": 0.0453, "step": 15822 }, { "epoch": 11.1194659170766, "grad_norm": 0.2638886868953705, "learning_rate": 2.5923635511829466e-05, "loss": 0.0619, "step": 15823 }, { "epoch": 11.120168657765285, "grad_norm": 0.751242995262146, "learning_rate": 2.592316701803701e-05, "loss": 0.0625, "step": 15824 }, { "epoch": 11.120871398453971, "grad_norm": 0.41099414229393005, "learning_rate": 2.5922698524244554e-05, "loss": 0.103, "step": 15825 }, { "epoch": 11.121574139142655, "grad_norm": 0.45479822158813477, "learning_rate": 2.5922230030452098e-05, "loss": 0.1336, "step": 15826 }, { "epoch": 11.122276879831341, "grad_norm": 0.6036891937255859, "learning_rate": 2.5921761536659638e-05, "loss": 0.167, "step": 15827 }, { "epoch": 11.122979620520027, "grad_norm": 4.077118873596191, "learning_rate": 2.592129304286718e-05, "loss": 0.1866, "step": 15828 }, { "epoch": 11.123682361208713, "grad_norm": 0.21361194550991058, "learning_rate": 2.5920824549074725e-05, "loss": 0.0598, "step": 15829 }, { "epoch": 11.1243851018974, "grad_norm": 0.17410780489444733, "learning_rate": 2.592035605528227e-05, "loss": 0.0275, "step": 15830 }, { "epoch": 11.125087842586085, "grad_norm": 0.10609543323516846, "learning_rate": 2.5919887561489813e-05, "loss": 0.0166, "step": 15831 }, { "epoch": 11.125790583274771, "grad_norm": 0.09549566358327866, "learning_rate": 2.5919419067697353e-05, "loss": 0.014, "step": 15832 }, { "epoch": 11.126493323963457, "grad_norm": 0.11632303148508072, "learning_rate": 2.5918950573904897e-05, "loss": 0.0284, "step": 15833 }, { "epoch": 11.127196064652143, "grad_norm": 0.12252406775951385, "learning_rate": 2.591848208011244e-05, "loss": 0.0184, "step": 15834 }, { "epoch": 11.127898805340829, "grad_norm": 0.06826808303594589, "learning_rate": 2.5918013586319984e-05, "loss": 0.0123, "step": 15835 }, { "epoch": 11.128601546029515, "grad_norm": 0.1483389139175415, "learning_rate": 2.5917545092527525e-05, "loss": 0.0178, "step": 15836 }, { "epoch": 11.1293042867182, "grad_norm": 0.08748821169137955, "learning_rate": 2.591707659873507e-05, "loss": 0.021, "step": 15837 }, { "epoch": 11.130007027406887, "grad_norm": 0.08114025741815567, "learning_rate": 2.5916608104942612e-05, "loss": 0.0127, "step": 15838 }, { "epoch": 11.130709768095572, "grad_norm": 0.19096477329730988, "learning_rate": 2.5916139611150152e-05, "loss": 0.028, "step": 15839 }, { "epoch": 11.131412508784258, "grad_norm": 0.1331137865781784, "learning_rate": 2.5915671117357693e-05, "loss": 0.0151, "step": 15840 }, { "epoch": 11.132115249472944, "grad_norm": 0.13803312182426453, "learning_rate": 2.5915202623565236e-05, "loss": 0.029, "step": 15841 }, { "epoch": 11.13281799016163, "grad_norm": 0.11615875363349915, "learning_rate": 2.591473412977278e-05, "loss": 0.0166, "step": 15842 }, { "epoch": 11.133520730850316, "grad_norm": 0.2337808459997177, "learning_rate": 2.5914265635980324e-05, "loss": 0.0334, "step": 15843 }, { "epoch": 11.134223471539002, "grad_norm": 0.14043192565441132, "learning_rate": 2.5913797142187868e-05, "loss": 0.0253, "step": 15844 }, { "epoch": 11.134926212227688, "grad_norm": 0.10997147858142853, "learning_rate": 2.5913328648395408e-05, "loss": 0.0231, "step": 15845 }, { "epoch": 11.135628952916374, "grad_norm": 0.3733355402946472, "learning_rate": 2.5912860154602952e-05, "loss": 0.0326, "step": 15846 }, { "epoch": 11.13633169360506, "grad_norm": 0.1702602505683899, "learning_rate": 2.5912391660810495e-05, "loss": 0.0338, "step": 15847 }, { "epoch": 11.137034434293746, "grad_norm": 0.22532682120800018, "learning_rate": 2.591192316701804e-05, "loss": 0.064, "step": 15848 }, { "epoch": 11.137737174982432, "grad_norm": 0.3274182081222534, "learning_rate": 2.591145467322558e-05, "loss": 0.0616, "step": 15849 }, { "epoch": 11.138439915671118, "grad_norm": 0.47460269927978516, "learning_rate": 2.5910986179433123e-05, "loss": 0.0793, "step": 15850 }, { "epoch": 11.139142656359803, "grad_norm": 0.5483234524726868, "learning_rate": 2.5910517685640667e-05, "loss": 0.1415, "step": 15851 }, { "epoch": 11.13984539704849, "grad_norm": 0.7649657130241394, "learning_rate": 2.591004919184821e-05, "loss": 0.164, "step": 15852 }, { "epoch": 11.140548137737175, "grad_norm": 2.3016607761383057, "learning_rate": 2.590958069805575e-05, "loss": 0.1882, "step": 15853 }, { "epoch": 11.141250878425861, "grad_norm": 0.19854103028774261, "learning_rate": 2.5909112204263295e-05, "loss": 0.0696, "step": 15854 }, { "epoch": 11.141953619114547, "grad_norm": 0.12661749124526978, "learning_rate": 2.590864371047084e-05, "loss": 0.0289, "step": 15855 }, { "epoch": 11.142656359803233, "grad_norm": 0.13825543224811554, "learning_rate": 2.590817521667838e-05, "loss": 0.0204, "step": 15856 }, { "epoch": 11.143359100491919, "grad_norm": 0.27804049849510193, "learning_rate": 2.5907706722885923e-05, "loss": 0.0222, "step": 15857 }, { "epoch": 11.144061841180605, "grad_norm": 0.11449802666902542, "learning_rate": 2.5907238229093463e-05, "loss": 0.0104, "step": 15858 }, { "epoch": 11.14476458186929, "grad_norm": 0.1682385802268982, "learning_rate": 2.5906769735301007e-05, "loss": 0.0138, "step": 15859 }, { "epoch": 11.145467322557977, "grad_norm": 0.1072593480348587, "learning_rate": 2.590630124150855e-05, "loss": 0.0157, "step": 15860 }, { "epoch": 11.146170063246663, "grad_norm": 0.1505059152841568, "learning_rate": 2.5905832747716094e-05, "loss": 0.0212, "step": 15861 }, { "epoch": 11.146872803935349, "grad_norm": 0.21223749220371246, "learning_rate": 2.5905364253923634e-05, "loss": 0.0154, "step": 15862 }, { "epoch": 11.147575544624035, "grad_norm": 0.13995827734470367, "learning_rate": 2.5904895760131178e-05, "loss": 0.0187, "step": 15863 }, { "epoch": 11.14827828531272, "grad_norm": 0.11452001333236694, "learning_rate": 2.5904427266338722e-05, "loss": 0.0313, "step": 15864 }, { "epoch": 11.148981026001405, "grad_norm": 0.4524368345737457, "learning_rate": 2.5903958772546266e-05, "loss": 0.0173, "step": 15865 }, { "epoch": 11.14968376669009, "grad_norm": 1.061469554901123, "learning_rate": 2.5903490278753806e-05, "loss": 0.0176, "step": 15866 }, { "epoch": 11.150386507378776, "grad_norm": 0.1765846610069275, "learning_rate": 2.590302178496135e-05, "loss": 0.0157, "step": 15867 }, { "epoch": 11.151089248067462, "grad_norm": 0.27101820707321167, "learning_rate": 2.5902553291168893e-05, "loss": 0.0373, "step": 15868 }, { "epoch": 11.151791988756148, "grad_norm": 0.11280860006809235, "learning_rate": 2.5902084797376437e-05, "loss": 0.0199, "step": 15869 }, { "epoch": 11.152494729444834, "grad_norm": 0.6165058016777039, "learning_rate": 2.590161630358398e-05, "loss": 0.0133, "step": 15870 }, { "epoch": 11.15319747013352, "grad_norm": 0.29225239157676697, "learning_rate": 2.590114780979152e-05, "loss": 0.0325, "step": 15871 }, { "epoch": 11.153900210822206, "grad_norm": 0.27625662088394165, "learning_rate": 2.5900679315999065e-05, "loss": 0.0284, "step": 15872 }, { "epoch": 11.154602951510892, "grad_norm": 0.7073192596435547, "learning_rate": 2.590021082220661e-05, "loss": 0.0432, "step": 15873 }, { "epoch": 11.155305692199578, "grad_norm": 0.44498589634895325, "learning_rate": 2.589974232841415e-05, "loss": 0.0589, "step": 15874 }, { "epoch": 11.156008432888264, "grad_norm": 0.8837130069732666, "learning_rate": 2.589927383462169e-05, "loss": 0.0988, "step": 15875 }, { "epoch": 11.15671117357695, "grad_norm": 0.9031681418418884, "learning_rate": 2.5898805340829233e-05, "loss": 0.1107, "step": 15876 }, { "epoch": 11.157413914265636, "grad_norm": 0.7831636071205139, "learning_rate": 2.5898336847036777e-05, "loss": 0.166, "step": 15877 }, { "epoch": 11.158116654954322, "grad_norm": 0.9664334058761597, "learning_rate": 2.589786835324432e-05, "loss": 0.1649, "step": 15878 }, { "epoch": 11.158819395643008, "grad_norm": 0.4291076064109802, "learning_rate": 2.589739985945186e-05, "loss": 0.0956, "step": 15879 }, { "epoch": 11.159522136331693, "grad_norm": 0.246519535779953, "learning_rate": 2.5896931365659404e-05, "loss": 0.0278, "step": 15880 }, { "epoch": 11.16022487702038, "grad_norm": 0.08866525441408157, "learning_rate": 2.5896462871866948e-05, "loss": 0.0192, "step": 15881 }, { "epoch": 11.160927617709065, "grad_norm": 0.21052347123622894, "learning_rate": 2.5895994378074492e-05, "loss": 0.022, "step": 15882 }, { "epoch": 11.161630358397751, "grad_norm": 0.12976036965847015, "learning_rate": 2.5895525884282036e-05, "loss": 0.0171, "step": 15883 }, { "epoch": 11.162333099086437, "grad_norm": 0.3154918849468231, "learning_rate": 2.5895057390489576e-05, "loss": 0.0102, "step": 15884 }, { "epoch": 11.163035839775123, "grad_norm": 0.0977422371506691, "learning_rate": 2.589458889669712e-05, "loss": 0.0154, "step": 15885 }, { "epoch": 11.163738580463809, "grad_norm": 0.19070975482463837, "learning_rate": 2.5894120402904663e-05, "loss": 0.0241, "step": 15886 }, { "epoch": 11.164441321152495, "grad_norm": 0.22295239567756653, "learning_rate": 2.5893651909112207e-05, "loss": 0.0339, "step": 15887 }, { "epoch": 11.16514406184118, "grad_norm": 0.10262317955493927, "learning_rate": 2.5893183415319748e-05, "loss": 0.0128, "step": 15888 }, { "epoch": 11.165846802529867, "grad_norm": 0.17969858646392822, "learning_rate": 2.589271492152729e-05, "loss": 0.0236, "step": 15889 }, { "epoch": 11.166549543218553, "grad_norm": 0.17821356654167175, "learning_rate": 2.5892246427734835e-05, "loss": 0.0096, "step": 15890 }, { "epoch": 11.167252283907239, "grad_norm": 0.2020745873451233, "learning_rate": 2.5891777933942375e-05, "loss": 0.0179, "step": 15891 }, { "epoch": 11.167955024595924, "grad_norm": 0.10459079593420029, "learning_rate": 2.5891309440149916e-05, "loss": 0.014, "step": 15892 }, { "epoch": 11.16865776528461, "grad_norm": 0.24620689451694489, "learning_rate": 2.589084094635746e-05, "loss": 0.025, "step": 15893 }, { "epoch": 11.169360505973296, "grad_norm": 0.3065463602542877, "learning_rate": 2.5890372452565003e-05, "loss": 0.0244, "step": 15894 }, { "epoch": 11.170063246661982, "grad_norm": 0.19950595498085022, "learning_rate": 2.5889903958772547e-05, "loss": 0.0503, "step": 15895 }, { "epoch": 11.170765987350668, "grad_norm": 0.1616927683353424, "learning_rate": 2.588943546498009e-05, "loss": 0.0141, "step": 15896 }, { "epoch": 11.171468728039354, "grad_norm": 0.22198542952537537, "learning_rate": 2.588896697118763e-05, "loss": 0.0501, "step": 15897 }, { "epoch": 11.17217146872804, "grad_norm": 0.19701436161994934, "learning_rate": 2.5888498477395175e-05, "loss": 0.0331, "step": 15898 }, { "epoch": 11.172874209416726, "grad_norm": 0.3250613510608673, "learning_rate": 2.588802998360272e-05, "loss": 0.0717, "step": 15899 }, { "epoch": 11.173576950105412, "grad_norm": 0.5619009733200073, "learning_rate": 2.5887561489810262e-05, "loss": 0.1105, "step": 15900 }, { "epoch": 11.174279690794098, "grad_norm": 0.8863487839698792, "learning_rate": 2.5887092996017802e-05, "loss": 0.1231, "step": 15901 }, { "epoch": 11.174982431482784, "grad_norm": 0.7687501907348633, "learning_rate": 2.5886624502225346e-05, "loss": 0.1728, "step": 15902 }, { "epoch": 11.17568517217147, "grad_norm": 1.5816529989242554, "learning_rate": 2.588615600843289e-05, "loss": 0.1799, "step": 15903 }, { "epoch": 11.176387912860154, "grad_norm": 0.20208323001861572, "learning_rate": 2.5885687514640434e-05, "loss": 0.0594, "step": 15904 }, { "epoch": 11.17709065354884, "grad_norm": 0.1515141874551773, "learning_rate": 2.5885219020847974e-05, "loss": 0.0295, "step": 15905 }, { "epoch": 11.177793394237526, "grad_norm": 0.11784360557794571, "learning_rate": 2.5884750527055518e-05, "loss": 0.0152, "step": 15906 }, { "epoch": 11.178496134926212, "grad_norm": 0.14615418016910553, "learning_rate": 2.588428203326306e-05, "loss": 0.0196, "step": 15907 }, { "epoch": 11.179198875614897, "grad_norm": 0.09420890361070633, "learning_rate": 2.58838135394706e-05, "loss": 0.0173, "step": 15908 }, { "epoch": 11.179901616303583, "grad_norm": 0.10994136333465576, "learning_rate": 2.5883345045678145e-05, "loss": 0.0139, "step": 15909 }, { "epoch": 11.18060435699227, "grad_norm": 0.2511065900325775, "learning_rate": 2.5882876551885686e-05, "loss": 0.017, "step": 15910 }, { "epoch": 11.181307097680955, "grad_norm": 0.08402150124311447, "learning_rate": 2.588240805809323e-05, "loss": 0.0161, "step": 15911 }, { "epoch": 11.182009838369641, "grad_norm": 0.1254037767648697, "learning_rate": 2.5881939564300773e-05, "loss": 0.016, "step": 15912 }, { "epoch": 11.182712579058327, "grad_norm": 0.07389258593320847, "learning_rate": 2.5881471070508317e-05, "loss": 0.005, "step": 15913 }, { "epoch": 11.183415319747013, "grad_norm": 0.13879500329494476, "learning_rate": 2.5881002576715857e-05, "loss": 0.0257, "step": 15914 }, { "epoch": 11.184118060435699, "grad_norm": 0.23030021786689758, "learning_rate": 2.58805340829234e-05, "loss": 0.0063, "step": 15915 }, { "epoch": 11.184820801124385, "grad_norm": 0.4968957304954529, "learning_rate": 2.5880065589130945e-05, "loss": 0.0436, "step": 15916 }, { "epoch": 11.18552354181307, "grad_norm": 0.09729976207017899, "learning_rate": 2.587959709533849e-05, "loss": 0.0122, "step": 15917 }, { "epoch": 11.186226282501757, "grad_norm": 0.11700654029846191, "learning_rate": 2.587912860154603e-05, "loss": 0.0228, "step": 15918 }, { "epoch": 11.186929023190443, "grad_norm": 0.23253454267978668, "learning_rate": 2.5878660107753573e-05, "loss": 0.023, "step": 15919 }, { "epoch": 11.187631763879128, "grad_norm": 0.23572108149528503, "learning_rate": 2.5878191613961116e-05, "loss": 0.0136, "step": 15920 }, { "epoch": 11.188334504567814, "grad_norm": 0.21937860548496246, "learning_rate": 2.587772312016866e-05, "loss": 0.0356, "step": 15921 }, { "epoch": 11.1890372452565, "grad_norm": 0.1593315154314041, "learning_rate": 2.5877254626376204e-05, "loss": 0.0286, "step": 15922 }, { "epoch": 11.189739985945186, "grad_norm": 0.8579145073890686, "learning_rate": 2.5876786132583744e-05, "loss": 0.0521, "step": 15923 }, { "epoch": 11.190442726633872, "grad_norm": 1.7464566230773926, "learning_rate": 2.5876317638791288e-05, "loss": 0.0826, "step": 15924 }, { "epoch": 11.191145467322558, "grad_norm": 0.3533075153827667, "learning_rate": 2.587584914499883e-05, "loss": 0.0983, "step": 15925 }, { "epoch": 11.191848208011244, "grad_norm": 2.765531063079834, "learning_rate": 2.5875380651206372e-05, "loss": 0.1714, "step": 15926 }, { "epoch": 11.19255094869993, "grad_norm": 0.9086959362030029, "learning_rate": 2.5874912157413912e-05, "loss": 0.1741, "step": 15927 }, { "epoch": 11.193253689388616, "grad_norm": 1.3421484231948853, "learning_rate": 2.5874443663621456e-05, "loss": 0.2027, "step": 15928 }, { "epoch": 11.193956430077302, "grad_norm": 0.4201631546020508, "learning_rate": 2.5873975169829e-05, "loss": 0.0589, "step": 15929 }, { "epoch": 11.194659170765988, "grad_norm": 0.23230569064617157, "learning_rate": 2.5873506676036543e-05, "loss": 0.0361, "step": 15930 }, { "epoch": 11.195361911454674, "grad_norm": 0.22036947309970856, "learning_rate": 2.5873038182244087e-05, "loss": 0.0219, "step": 15931 }, { "epoch": 11.19606465214336, "grad_norm": 0.11924993991851807, "learning_rate": 2.5872569688451627e-05, "loss": 0.0169, "step": 15932 }, { "epoch": 11.196767392832045, "grad_norm": 0.20933377742767334, "learning_rate": 2.587210119465917e-05, "loss": 0.0172, "step": 15933 }, { "epoch": 11.197470133520731, "grad_norm": 0.09502159804105759, "learning_rate": 2.5871632700866715e-05, "loss": 0.0186, "step": 15934 }, { "epoch": 11.198172874209417, "grad_norm": 0.153798446059227, "learning_rate": 2.587116420707426e-05, "loss": 0.0134, "step": 15935 }, { "epoch": 11.198875614898103, "grad_norm": 0.10776864737272263, "learning_rate": 2.58706957132818e-05, "loss": 0.0159, "step": 15936 }, { "epoch": 11.19957835558679, "grad_norm": 0.15385310351848602, "learning_rate": 2.5870227219489343e-05, "loss": 0.022, "step": 15937 }, { "epoch": 11.200281096275475, "grad_norm": 0.1430337280035019, "learning_rate": 2.5869758725696886e-05, "loss": 0.0143, "step": 15938 }, { "epoch": 11.200983836964161, "grad_norm": 0.14577938616275787, "learning_rate": 2.586929023190443e-05, "loss": 0.0255, "step": 15939 }, { "epoch": 11.201686577652847, "grad_norm": 0.19845271110534668, "learning_rate": 2.586882173811197e-05, "loss": 0.011, "step": 15940 }, { "epoch": 11.202389318341533, "grad_norm": 0.2324439287185669, "learning_rate": 2.5868353244319514e-05, "loss": 0.0375, "step": 15941 }, { "epoch": 11.203092059030217, "grad_norm": 0.6183469295501709, "learning_rate": 2.5867884750527058e-05, "loss": 0.0219, "step": 15942 }, { "epoch": 11.203794799718903, "grad_norm": 0.16823557019233704, "learning_rate": 2.5867416256734598e-05, "loss": 0.0232, "step": 15943 }, { "epoch": 11.204497540407589, "grad_norm": 0.40704867243766785, "learning_rate": 2.5866947762942142e-05, "loss": 0.0396, "step": 15944 }, { "epoch": 11.205200281096275, "grad_norm": 0.1705973893404007, "learning_rate": 2.5866479269149682e-05, "loss": 0.013, "step": 15945 }, { "epoch": 11.20590302178496, "grad_norm": 0.18924476206302643, "learning_rate": 2.5866010775357226e-05, "loss": 0.034, "step": 15946 }, { "epoch": 11.206605762473647, "grad_norm": 0.3400486409664154, "learning_rate": 2.586554228156477e-05, "loss": 0.0337, "step": 15947 }, { "epoch": 11.207308503162333, "grad_norm": 0.5174731016159058, "learning_rate": 2.5865073787772313e-05, "loss": 0.0534, "step": 15948 }, { "epoch": 11.208011243851018, "grad_norm": 0.24908092617988586, "learning_rate": 2.5864605293979854e-05, "loss": 0.0645, "step": 15949 }, { "epoch": 11.208713984539704, "grad_norm": 0.5035071969032288, "learning_rate": 2.5864136800187397e-05, "loss": 0.1091, "step": 15950 }, { "epoch": 11.20941672522839, "grad_norm": 0.9215637445449829, "learning_rate": 2.586366830639494e-05, "loss": 0.1413, "step": 15951 }, { "epoch": 11.210119465917076, "grad_norm": 0.4559909999370575, "learning_rate": 2.5863199812602485e-05, "loss": 0.1458, "step": 15952 }, { "epoch": 11.210822206605762, "grad_norm": 1.0493218898773193, "learning_rate": 2.5862731318810025e-05, "loss": 0.1867, "step": 15953 }, { "epoch": 11.211524947294448, "grad_norm": 0.19113466143608093, "learning_rate": 2.586226282501757e-05, "loss": 0.0629, "step": 15954 }, { "epoch": 11.212227687983134, "grad_norm": 0.14397044479846954, "learning_rate": 2.5861794331225113e-05, "loss": 0.023, "step": 15955 }, { "epoch": 11.21293042867182, "grad_norm": 0.27631324529647827, "learning_rate": 2.5861325837432656e-05, "loss": 0.026, "step": 15956 }, { "epoch": 11.213633169360506, "grad_norm": 0.1492926925420761, "learning_rate": 2.58608573436402e-05, "loss": 0.0176, "step": 15957 }, { "epoch": 11.214335910049192, "grad_norm": 0.10149207711219788, "learning_rate": 2.586038884984774e-05, "loss": 0.0203, "step": 15958 }, { "epoch": 11.215038650737878, "grad_norm": 0.06051555275917053, "learning_rate": 2.5859920356055284e-05, "loss": 0.0099, "step": 15959 }, { "epoch": 11.215741391426564, "grad_norm": 0.1180640235543251, "learning_rate": 2.5859451862262828e-05, "loss": 0.0116, "step": 15960 }, { "epoch": 11.21644413211525, "grad_norm": 0.16525278985500336, "learning_rate": 2.5858983368470368e-05, "loss": 0.013, "step": 15961 }, { "epoch": 11.217146872803935, "grad_norm": 0.1090078130364418, "learning_rate": 2.585851487467791e-05, "loss": 0.0188, "step": 15962 }, { "epoch": 11.217849613492621, "grad_norm": 0.07833588123321533, "learning_rate": 2.5858046380885452e-05, "loss": 0.0097, "step": 15963 }, { "epoch": 11.218552354181307, "grad_norm": 0.2189439982175827, "learning_rate": 2.5857577887092996e-05, "loss": 0.0238, "step": 15964 }, { "epoch": 11.219255094869993, "grad_norm": 0.0743776336312294, "learning_rate": 2.585710939330054e-05, "loss": 0.0113, "step": 15965 }, { "epoch": 11.219957835558679, "grad_norm": 0.2431115210056305, "learning_rate": 2.585664089950808e-05, "loss": 0.0187, "step": 15966 }, { "epoch": 11.220660576247365, "grad_norm": 0.15142634510993958, "learning_rate": 2.5856172405715624e-05, "loss": 0.0086, "step": 15967 }, { "epoch": 11.221363316936051, "grad_norm": 0.36875712871551514, "learning_rate": 2.5855703911923168e-05, "loss": 0.0262, "step": 15968 }, { "epoch": 11.222066057624737, "grad_norm": 0.2891145348548889, "learning_rate": 2.585523541813071e-05, "loss": 0.0468, "step": 15969 }, { "epoch": 11.222768798313423, "grad_norm": 0.24349744617938995, "learning_rate": 2.5854766924338255e-05, "loss": 0.0174, "step": 15970 }, { "epoch": 11.223471539002109, "grad_norm": 0.33376652002334595, "learning_rate": 2.5854298430545795e-05, "loss": 0.0478, "step": 15971 }, { "epoch": 11.224174279690795, "grad_norm": 0.2642844021320343, "learning_rate": 2.585382993675334e-05, "loss": 0.0312, "step": 15972 }, { "epoch": 11.22487702037948, "grad_norm": 0.4858895242214203, "learning_rate": 2.5853361442960883e-05, "loss": 0.0401, "step": 15973 }, { "epoch": 11.225579761068166, "grad_norm": 0.3406163454055786, "learning_rate": 2.5852892949168427e-05, "loss": 0.0654, "step": 15974 }, { "epoch": 11.226282501756852, "grad_norm": 1.0428377389907837, "learning_rate": 2.5852424455375967e-05, "loss": 0.1286, "step": 15975 }, { "epoch": 11.226985242445538, "grad_norm": 0.9669190645217896, "learning_rate": 2.585195596158351e-05, "loss": 0.135, "step": 15976 }, { "epoch": 11.227687983134224, "grad_norm": 0.692253828048706, "learning_rate": 2.5851487467791054e-05, "loss": 0.1553, "step": 15977 }, { "epoch": 11.22839072382291, "grad_norm": 0.7211858034133911, "learning_rate": 2.5851018973998595e-05, "loss": 0.1796, "step": 15978 }, { "epoch": 11.229093464511596, "grad_norm": 0.16203290224075317, "learning_rate": 2.5850550480206135e-05, "loss": 0.0617, "step": 15979 }, { "epoch": 11.22979620520028, "grad_norm": 0.33420246839523315, "learning_rate": 2.585008198641368e-05, "loss": 0.0252, "step": 15980 }, { "epoch": 11.230498945888966, "grad_norm": 0.12914182245731354, "learning_rate": 2.5849613492621222e-05, "loss": 0.0296, "step": 15981 }, { "epoch": 11.231201686577652, "grad_norm": 0.15243081748485565, "learning_rate": 2.5849144998828766e-05, "loss": 0.0156, "step": 15982 }, { "epoch": 11.231904427266338, "grad_norm": 0.1662488728761673, "learning_rate": 2.584867650503631e-05, "loss": 0.0149, "step": 15983 }, { "epoch": 11.232607167955024, "grad_norm": 0.12549255788326263, "learning_rate": 2.584820801124385e-05, "loss": 0.0117, "step": 15984 }, { "epoch": 11.23330990864371, "grad_norm": 0.17582963407039642, "learning_rate": 2.5847739517451394e-05, "loss": 0.0151, "step": 15985 }, { "epoch": 11.234012649332396, "grad_norm": 0.2514265179634094, "learning_rate": 2.5847271023658938e-05, "loss": 0.0255, "step": 15986 }, { "epoch": 11.234715390021082, "grad_norm": 0.1809973269701004, "learning_rate": 2.584680252986648e-05, "loss": 0.02, "step": 15987 }, { "epoch": 11.235418130709768, "grad_norm": 0.13110040128231049, "learning_rate": 2.5846334036074022e-05, "loss": 0.0167, "step": 15988 }, { "epoch": 11.236120871398454, "grad_norm": 0.2174709588289261, "learning_rate": 2.5845865542281566e-05, "loss": 0.0207, "step": 15989 }, { "epoch": 11.23682361208714, "grad_norm": 0.12257987260818481, "learning_rate": 2.584539704848911e-05, "loss": 0.0072, "step": 15990 }, { "epoch": 11.237526352775825, "grad_norm": 0.5391083359718323, "learning_rate": 2.5844928554696653e-05, "loss": 0.0254, "step": 15991 }, { "epoch": 11.238229093464511, "grad_norm": 0.1726672500371933, "learning_rate": 2.5844460060904193e-05, "loss": 0.0126, "step": 15992 }, { "epoch": 11.238931834153197, "grad_norm": 1.0719624757766724, "learning_rate": 2.5843991567111737e-05, "loss": 0.0409, "step": 15993 }, { "epoch": 11.239634574841883, "grad_norm": 0.24347001314163208, "learning_rate": 2.584352307331928e-05, "loss": 0.041, "step": 15994 }, { "epoch": 11.240337315530569, "grad_norm": 1.7880522012710571, "learning_rate": 2.584305457952682e-05, "loss": 0.019, "step": 15995 }, { "epoch": 11.241040056219255, "grad_norm": 0.2915912866592407, "learning_rate": 2.5842586085734365e-05, "loss": 0.0459, "step": 15996 }, { "epoch": 11.24174279690794, "grad_norm": 0.4263696074485779, "learning_rate": 2.5842117591941905e-05, "loss": 0.0579, "step": 15997 }, { "epoch": 11.242445537596627, "grad_norm": 0.2745542824268341, "learning_rate": 2.584164909814945e-05, "loss": 0.0458, "step": 15998 }, { "epoch": 11.243148278285313, "grad_norm": 0.6190388798713684, "learning_rate": 2.5841180604356993e-05, "loss": 0.0944, "step": 15999 }, { "epoch": 11.243851018973999, "grad_norm": 0.49445387721061707, "learning_rate": 2.5840712110564536e-05, "loss": 0.0998, "step": 16000 }, { "epoch": 11.243851018973999, "eval_cer": 0.19460355907608043, "eval_loss": 0.2659454345703125, "eval_runtime": 18.9541, "eval_samples_per_second": 239.421, "eval_steps_per_second": 0.791, "eval_wer": 0.3474763274768861, "step": 16000 }, { "epoch": 11.244553759662685, "grad_norm": 0.6366737484931946, "learning_rate": 2.5840243616772077e-05, "loss": 0.1293, "step": 16001 }, { "epoch": 11.24525650035137, "grad_norm": 0.9305012822151184, "learning_rate": 2.583977512297962e-05, "loss": 0.1668, "step": 16002 }, { "epoch": 11.245959241040056, "grad_norm": 1.1806811094284058, "learning_rate": 2.5839306629187164e-05, "loss": 0.1879, "step": 16003 }, { "epoch": 11.246661981728742, "grad_norm": 0.1836925745010376, "learning_rate": 2.5838838135394708e-05, "loss": 0.0674, "step": 16004 }, { "epoch": 11.247364722417428, "grad_norm": 0.19796423614025116, "learning_rate": 2.5838369641602248e-05, "loss": 0.0187, "step": 16005 }, { "epoch": 11.248067463106114, "grad_norm": 0.12312128394842148, "learning_rate": 2.5837901147809792e-05, "loss": 0.0192, "step": 16006 }, { "epoch": 11.2487702037948, "grad_norm": 1.1914730072021484, "learning_rate": 2.5837432654017336e-05, "loss": 0.0176, "step": 16007 }, { "epoch": 11.249472944483486, "grad_norm": 0.11230229586362839, "learning_rate": 2.583696416022488e-05, "loss": 0.0199, "step": 16008 }, { "epoch": 11.250175685172172, "grad_norm": 0.22946511209011078, "learning_rate": 2.5836495666432423e-05, "loss": 0.0182, "step": 16009 }, { "epoch": 11.250878425860858, "grad_norm": 0.1935587227344513, "learning_rate": 2.5836027172639963e-05, "loss": 0.0381, "step": 16010 }, { "epoch": 11.251581166549544, "grad_norm": 0.17602956295013428, "learning_rate": 2.5835558678847507e-05, "loss": 0.0232, "step": 16011 }, { "epoch": 11.25228390723823, "grad_norm": 0.14836883544921875, "learning_rate": 2.583509018505505e-05, "loss": 0.0266, "step": 16012 }, { "epoch": 11.252986647926916, "grad_norm": 0.1376647651195526, "learning_rate": 2.583462169126259e-05, "loss": 0.0192, "step": 16013 }, { "epoch": 11.253689388615602, "grad_norm": 0.4110453426837921, "learning_rate": 2.583415319747013e-05, "loss": 0.0401, "step": 16014 }, { "epoch": 11.254392129304287, "grad_norm": 0.10364297032356262, "learning_rate": 2.5833684703677675e-05, "loss": 0.0126, "step": 16015 }, { "epoch": 11.255094869992973, "grad_norm": 0.1604071855545044, "learning_rate": 2.583321620988522e-05, "loss": 0.041, "step": 16016 }, { "epoch": 11.25579761068166, "grad_norm": 0.09131594747304916, "learning_rate": 2.5832747716092763e-05, "loss": 0.0176, "step": 16017 }, { "epoch": 11.256500351370345, "grad_norm": 0.30426687002182007, "learning_rate": 2.5832279222300303e-05, "loss": 0.0293, "step": 16018 }, { "epoch": 11.25720309205903, "grad_norm": 0.19078169763088226, "learning_rate": 2.5831810728507847e-05, "loss": 0.0425, "step": 16019 }, { "epoch": 11.257905832747715, "grad_norm": 0.1562803089618683, "learning_rate": 2.583134223471539e-05, "loss": 0.0238, "step": 16020 }, { "epoch": 11.258608573436401, "grad_norm": 0.1346282809972763, "learning_rate": 2.5830873740922934e-05, "loss": 0.021, "step": 16021 }, { "epoch": 11.259311314125087, "grad_norm": 0.16028185188770294, "learning_rate": 2.5830405247130478e-05, "loss": 0.0419, "step": 16022 }, { "epoch": 11.260014054813773, "grad_norm": 0.4053652584552765, "learning_rate": 2.5829936753338018e-05, "loss": 0.0356, "step": 16023 }, { "epoch": 11.260716795502459, "grad_norm": 0.487538605928421, "learning_rate": 2.5829468259545562e-05, "loss": 0.0549, "step": 16024 }, { "epoch": 11.261419536191145, "grad_norm": 0.512993335723877, "learning_rate": 2.5828999765753106e-05, "loss": 0.0879, "step": 16025 }, { "epoch": 11.26212227687983, "grad_norm": 1.0521725416183472, "learning_rate": 2.582853127196065e-05, "loss": 0.175, "step": 16026 }, { "epoch": 11.262825017568517, "grad_norm": 1.1392028331756592, "learning_rate": 2.582806277816819e-05, "loss": 0.1907, "step": 16027 }, { "epoch": 11.263527758257203, "grad_norm": 0.8393661975860596, "learning_rate": 2.5827594284375734e-05, "loss": 0.2015, "step": 16028 }, { "epoch": 11.264230498945889, "grad_norm": 0.21525739133358002, "learning_rate": 2.5827125790583277e-05, "loss": 0.0547, "step": 16029 }, { "epoch": 11.264933239634574, "grad_norm": 0.1433022916316986, "learning_rate": 2.5826657296790818e-05, "loss": 0.0212, "step": 16030 }, { "epoch": 11.26563598032326, "grad_norm": 0.1433505266904831, "learning_rate": 2.5826188802998358e-05, "loss": 0.0191, "step": 16031 }, { "epoch": 11.266338721011946, "grad_norm": 0.1981021761894226, "learning_rate": 2.58257203092059e-05, "loss": 0.0322, "step": 16032 }, { "epoch": 11.267041461700632, "grad_norm": 0.09400516003370285, "learning_rate": 2.5825251815413445e-05, "loss": 0.0178, "step": 16033 }, { "epoch": 11.267744202389318, "grad_norm": 0.11786644905805588, "learning_rate": 2.582478332162099e-05, "loss": 0.0124, "step": 16034 }, { "epoch": 11.268446943078004, "grad_norm": 0.18916399776935577, "learning_rate": 2.5824314827828533e-05, "loss": 0.0194, "step": 16035 }, { "epoch": 11.26914968376669, "grad_norm": 0.07988081872463226, "learning_rate": 2.5823846334036073e-05, "loss": 0.0151, "step": 16036 }, { "epoch": 11.269852424455376, "grad_norm": 0.18906888365745544, "learning_rate": 2.5823377840243617e-05, "loss": 0.0176, "step": 16037 }, { "epoch": 11.270555165144062, "grad_norm": 0.11944394558668137, "learning_rate": 2.582290934645116e-05, "loss": 0.0136, "step": 16038 }, { "epoch": 11.271257905832748, "grad_norm": 0.16497448086738586, "learning_rate": 2.5822440852658704e-05, "loss": 0.0348, "step": 16039 }, { "epoch": 11.271960646521434, "grad_norm": 0.07850155979394913, "learning_rate": 2.5821972358866245e-05, "loss": 0.0145, "step": 16040 }, { "epoch": 11.27266338721012, "grad_norm": 0.11801543831825256, "learning_rate": 2.582150386507379e-05, "loss": 0.0273, "step": 16041 }, { "epoch": 11.273366127898806, "grad_norm": 0.10829149931669235, "learning_rate": 2.5821035371281332e-05, "loss": 0.0137, "step": 16042 }, { "epoch": 11.274068868587491, "grad_norm": 0.22911052405834198, "learning_rate": 2.5820566877488876e-05, "loss": 0.0352, "step": 16043 }, { "epoch": 11.274771609276177, "grad_norm": 0.23670098185539246, "learning_rate": 2.5820098383696416e-05, "loss": 0.0351, "step": 16044 }, { "epoch": 11.275474349964863, "grad_norm": 0.15396010875701904, "learning_rate": 2.581962988990396e-05, "loss": 0.0165, "step": 16045 }, { "epoch": 11.27617709065355, "grad_norm": 0.18623457849025726, "learning_rate": 2.5819161396111504e-05, "loss": 0.0486, "step": 16046 }, { "epoch": 11.276879831342235, "grad_norm": 0.20682533085346222, "learning_rate": 2.5818692902319047e-05, "loss": 0.0402, "step": 16047 }, { "epoch": 11.277582572030921, "grad_norm": 0.2987406849861145, "learning_rate": 2.5818224408526588e-05, "loss": 0.0525, "step": 16048 }, { "epoch": 11.278285312719607, "grad_norm": 0.43036702275276184, "learning_rate": 2.5817755914734128e-05, "loss": 0.0705, "step": 16049 }, { "epoch": 11.278988053408293, "grad_norm": 0.6861766576766968, "learning_rate": 2.5817287420941672e-05, "loss": 0.1141, "step": 16050 }, { "epoch": 11.279690794096979, "grad_norm": 0.560857355594635, "learning_rate": 2.5816818927149215e-05, "loss": 0.1479, "step": 16051 }, { "epoch": 11.280393534785665, "grad_norm": 1.2372772693634033, "learning_rate": 2.581635043335676e-05, "loss": 0.1719, "step": 16052 }, { "epoch": 11.28109627547435, "grad_norm": 1.0635768175125122, "learning_rate": 2.58158819395643e-05, "loss": 0.1998, "step": 16053 }, { "epoch": 11.281799016163037, "grad_norm": 0.4729604125022888, "learning_rate": 2.5815413445771843e-05, "loss": 0.0681, "step": 16054 }, { "epoch": 11.282501756851723, "grad_norm": 0.4600253105163574, "learning_rate": 2.5814944951979387e-05, "loss": 0.0224, "step": 16055 }, { "epoch": 11.283204497540408, "grad_norm": 0.620872437953949, "learning_rate": 2.581447645818693e-05, "loss": 0.0415, "step": 16056 }, { "epoch": 11.283907238229094, "grad_norm": 0.0860792025923729, "learning_rate": 2.581400796439447e-05, "loss": 0.0118, "step": 16057 }, { "epoch": 11.284609978917779, "grad_norm": 0.0826149508357048, "learning_rate": 2.5813539470602015e-05, "loss": 0.0103, "step": 16058 }, { "epoch": 11.285312719606464, "grad_norm": 0.23500265181064606, "learning_rate": 2.581307097680956e-05, "loss": 0.0266, "step": 16059 }, { "epoch": 11.28601546029515, "grad_norm": 0.11884333938360214, "learning_rate": 2.5812602483017102e-05, "loss": 0.0168, "step": 16060 }, { "epoch": 11.286718200983836, "grad_norm": 0.09176183491945267, "learning_rate": 2.5812133989224646e-05, "loss": 0.0099, "step": 16061 }, { "epoch": 11.287420941672522, "grad_norm": 0.10510531067848206, "learning_rate": 2.5811665495432186e-05, "loss": 0.0188, "step": 16062 }, { "epoch": 11.288123682361208, "grad_norm": 0.17167861759662628, "learning_rate": 2.581119700163973e-05, "loss": 0.0236, "step": 16063 }, { "epoch": 11.288826423049894, "grad_norm": 0.10506107658147812, "learning_rate": 2.5810728507847274e-05, "loss": 0.0169, "step": 16064 }, { "epoch": 11.28952916373858, "grad_norm": 0.2180965393781662, "learning_rate": 2.5810260014054814e-05, "loss": 0.0209, "step": 16065 }, { "epoch": 11.290231904427266, "grad_norm": 0.15000370144844055, "learning_rate": 2.5809791520262354e-05, "loss": 0.0198, "step": 16066 }, { "epoch": 11.290934645115952, "grad_norm": 0.1596466302871704, "learning_rate": 2.5809323026469898e-05, "loss": 0.0233, "step": 16067 }, { "epoch": 11.291637385804638, "grad_norm": 0.19471397995948792, "learning_rate": 2.5808854532677442e-05, "loss": 0.0231, "step": 16068 }, { "epoch": 11.292340126493324, "grad_norm": 0.2498645782470703, "learning_rate": 2.5808386038884986e-05, "loss": 0.0234, "step": 16069 }, { "epoch": 11.29304286718201, "grad_norm": 0.17201389372348785, "learning_rate": 2.5807917545092526e-05, "loss": 0.015, "step": 16070 }, { "epoch": 11.293745607870695, "grad_norm": 0.20128844678401947, "learning_rate": 2.580744905130007e-05, "loss": 0.024, "step": 16071 }, { "epoch": 11.294448348559381, "grad_norm": 0.17426063120365143, "learning_rate": 2.5806980557507613e-05, "loss": 0.0383, "step": 16072 }, { "epoch": 11.295151089248067, "grad_norm": 0.42713311314582825, "learning_rate": 2.5806512063715157e-05, "loss": 0.0498, "step": 16073 }, { "epoch": 11.295853829936753, "grad_norm": 1.040771722793579, "learning_rate": 2.58060435699227e-05, "loss": 0.0938, "step": 16074 }, { "epoch": 11.29655657062544, "grad_norm": 0.2944212257862091, "learning_rate": 2.580557507613024e-05, "loss": 0.1055, "step": 16075 }, { "epoch": 11.297259311314125, "grad_norm": 0.6960487365722656, "learning_rate": 2.5805106582337785e-05, "loss": 0.1843, "step": 16076 }, { "epoch": 11.297962052002811, "grad_norm": 0.6558912992477417, "learning_rate": 2.580463808854533e-05, "loss": 0.164, "step": 16077 }, { "epoch": 11.298664792691497, "grad_norm": 0.918648362159729, "learning_rate": 2.5804169594752872e-05, "loss": 0.2104, "step": 16078 }, { "epoch": 11.299367533380183, "grad_norm": 0.2874893248081207, "learning_rate": 2.5803701100960413e-05, "loss": 0.0595, "step": 16079 }, { "epoch": 11.300070274068869, "grad_norm": 0.18043462932109833, "learning_rate": 2.5803232607167956e-05, "loss": 0.0446, "step": 16080 }, { "epoch": 11.300773014757555, "grad_norm": 0.1054200828075409, "learning_rate": 2.58027641133755e-05, "loss": 0.0213, "step": 16081 }, { "epoch": 11.30147575544624, "grad_norm": 0.08802516013383865, "learning_rate": 2.5802295619583044e-05, "loss": 0.0104, "step": 16082 }, { "epoch": 11.302178496134927, "grad_norm": 0.4133031666278839, "learning_rate": 2.580182712579058e-05, "loss": 0.0301, "step": 16083 }, { "epoch": 11.302881236823612, "grad_norm": 0.08184796571731567, "learning_rate": 2.5801358631998125e-05, "loss": 0.013, "step": 16084 }, { "epoch": 11.303583977512298, "grad_norm": 0.07803966850042343, "learning_rate": 2.5800890138205668e-05, "loss": 0.0123, "step": 16085 }, { "epoch": 11.304286718200984, "grad_norm": 0.10774646699428558, "learning_rate": 2.5800421644413212e-05, "loss": 0.0169, "step": 16086 }, { "epoch": 11.30498945888967, "grad_norm": 0.23395049571990967, "learning_rate": 2.5799953150620756e-05, "loss": 0.0335, "step": 16087 }, { "epoch": 11.305692199578356, "grad_norm": 0.07134299725294113, "learning_rate": 2.5799484656828296e-05, "loss": 0.007, "step": 16088 }, { "epoch": 11.306394940267042, "grad_norm": 0.17812702059745789, "learning_rate": 2.579901616303584e-05, "loss": 0.0189, "step": 16089 }, { "epoch": 11.307097680955728, "grad_norm": 0.08544829487800598, "learning_rate": 2.5798547669243383e-05, "loss": 0.0079, "step": 16090 }, { "epoch": 11.307800421644414, "grad_norm": 0.2141302227973938, "learning_rate": 2.5798079175450927e-05, "loss": 0.0206, "step": 16091 }, { "epoch": 11.3085031623331, "grad_norm": 0.4615943729877472, "learning_rate": 2.5797610681658468e-05, "loss": 0.0346, "step": 16092 }, { "epoch": 11.309205903021786, "grad_norm": 0.08411211520433426, "learning_rate": 2.579714218786601e-05, "loss": 0.0141, "step": 16093 }, { "epoch": 11.309908643710472, "grad_norm": 0.37221306562423706, "learning_rate": 2.5796673694073555e-05, "loss": 0.0219, "step": 16094 }, { "epoch": 11.310611384399156, "grad_norm": 0.4034770727157593, "learning_rate": 2.57962052002811e-05, "loss": 0.0309, "step": 16095 }, { "epoch": 11.311314125087842, "grad_norm": 0.17315009236335754, "learning_rate": 2.579573670648864e-05, "loss": 0.0427, "step": 16096 }, { "epoch": 11.312016865776528, "grad_norm": 0.29760387539863586, "learning_rate": 2.5795268212696183e-05, "loss": 0.0338, "step": 16097 }, { "epoch": 11.312719606465214, "grad_norm": 0.49994155764579773, "learning_rate": 2.5794799718903727e-05, "loss": 0.0417, "step": 16098 }, { "epoch": 11.3134223471539, "grad_norm": 0.2596731185913086, "learning_rate": 2.579433122511127e-05, "loss": 0.0529, "step": 16099 }, { "epoch": 11.314125087842585, "grad_norm": 0.36511653661727905, "learning_rate": 2.579386273131881e-05, "loss": 0.0918, "step": 16100 }, { "epoch": 11.314827828531271, "grad_norm": 0.7243281006813049, "learning_rate": 2.579339423752635e-05, "loss": 0.1041, "step": 16101 }, { "epoch": 11.315530569219957, "grad_norm": 0.5701380968093872, "learning_rate": 2.5792925743733895e-05, "loss": 0.1296, "step": 16102 }, { "epoch": 11.316233309908643, "grad_norm": 0.9949381351470947, "learning_rate": 2.579245724994144e-05, "loss": 0.1863, "step": 16103 }, { "epoch": 11.316936050597329, "grad_norm": 0.4048618674278259, "learning_rate": 2.5791988756148982e-05, "loss": 0.0684, "step": 16104 }, { "epoch": 11.317638791286015, "grad_norm": 0.2911911904811859, "learning_rate": 2.5791520262356522e-05, "loss": 0.03, "step": 16105 }, { "epoch": 11.318341531974701, "grad_norm": 0.1842225044965744, "learning_rate": 2.5791051768564066e-05, "loss": 0.0207, "step": 16106 }, { "epoch": 11.319044272663387, "grad_norm": 0.22140158712863922, "learning_rate": 2.579058327477161e-05, "loss": 0.0112, "step": 16107 }, { "epoch": 11.319747013352073, "grad_norm": 0.1493179202079773, "learning_rate": 2.5790114780979154e-05, "loss": 0.0212, "step": 16108 }, { "epoch": 11.320449754040759, "grad_norm": 0.1637410819530487, "learning_rate": 2.5789646287186694e-05, "loss": 0.0301, "step": 16109 }, { "epoch": 11.321152494729445, "grad_norm": 0.09014491736888885, "learning_rate": 2.5789177793394238e-05, "loss": 0.01, "step": 16110 }, { "epoch": 11.32185523541813, "grad_norm": 0.3018991947174072, "learning_rate": 2.578870929960178e-05, "loss": 0.0359, "step": 16111 }, { "epoch": 11.322557976106816, "grad_norm": 0.10462036728858948, "learning_rate": 2.5788240805809325e-05, "loss": 0.0111, "step": 16112 }, { "epoch": 11.323260716795502, "grad_norm": 0.26298266649246216, "learning_rate": 2.578777231201687e-05, "loss": 0.021, "step": 16113 }, { "epoch": 11.323963457484188, "grad_norm": 0.11819769442081451, "learning_rate": 2.578730381822441e-05, "loss": 0.0122, "step": 16114 }, { "epoch": 11.324666198172874, "grad_norm": 0.12906388938426971, "learning_rate": 2.5786835324431953e-05, "loss": 0.0225, "step": 16115 }, { "epoch": 11.32536893886156, "grad_norm": 0.9714913368225098, "learning_rate": 2.5786366830639497e-05, "loss": 0.036, "step": 16116 }, { "epoch": 11.326071679550246, "grad_norm": 0.5767734050750732, "learning_rate": 2.5785898336847037e-05, "loss": 0.0207, "step": 16117 }, { "epoch": 11.326774420238932, "grad_norm": 0.20893679559230804, "learning_rate": 2.5785429843054577e-05, "loss": 0.0272, "step": 16118 }, { "epoch": 11.327477160927618, "grad_norm": 1.0468695163726807, "learning_rate": 2.578496134926212e-05, "loss": 0.0198, "step": 16119 }, { "epoch": 11.328179901616304, "grad_norm": 0.7956557869911194, "learning_rate": 2.5784492855469665e-05, "loss": 0.0232, "step": 16120 }, { "epoch": 11.32888264230499, "grad_norm": 0.38637393712997437, "learning_rate": 2.578402436167721e-05, "loss": 0.0455, "step": 16121 }, { "epoch": 11.329585382993676, "grad_norm": 0.21753108501434326, "learning_rate": 2.578355586788475e-05, "loss": 0.0424, "step": 16122 }, { "epoch": 11.330288123682362, "grad_norm": 0.23131880164146423, "learning_rate": 2.5783087374092293e-05, "loss": 0.047, "step": 16123 }, { "epoch": 11.330990864371048, "grad_norm": 0.3145088851451874, "learning_rate": 2.5782618880299836e-05, "loss": 0.0713, "step": 16124 }, { "epoch": 11.331693605059733, "grad_norm": 0.30714181065559387, "learning_rate": 2.578215038650738e-05, "loss": 0.1044, "step": 16125 }, { "epoch": 11.33239634574842, "grad_norm": 1.577102780342102, "learning_rate": 2.5781681892714924e-05, "loss": 0.1412, "step": 16126 }, { "epoch": 11.333099086437105, "grad_norm": 1.4576764106750488, "learning_rate": 2.5781213398922464e-05, "loss": 0.1563, "step": 16127 }, { "epoch": 11.333801827125791, "grad_norm": 1.8575705289840698, "learning_rate": 2.5780744905130008e-05, "loss": 0.2035, "step": 16128 }, { "epoch": 11.334504567814477, "grad_norm": 0.19599300622940063, "learning_rate": 2.578027641133755e-05, "loss": 0.0696, "step": 16129 }, { "epoch": 11.335207308503163, "grad_norm": 0.12990537285804749, "learning_rate": 2.5779807917545095e-05, "loss": 0.0189, "step": 16130 }, { "epoch": 11.335910049191849, "grad_norm": 0.11217876523733139, "learning_rate": 2.5779339423752636e-05, "loss": 0.0252, "step": 16131 }, { "epoch": 11.336612789880535, "grad_norm": 0.15882644057273865, "learning_rate": 2.577887092996018e-05, "loss": 0.0188, "step": 16132 }, { "epoch": 11.33731553056922, "grad_norm": 0.16235852241516113, "learning_rate": 2.5778402436167723e-05, "loss": 0.0107, "step": 16133 }, { "epoch": 11.338018271257905, "grad_norm": 0.10462409257888794, "learning_rate": 2.5777933942375267e-05, "loss": 0.0198, "step": 16134 }, { "epoch": 11.33872101194659, "grad_norm": 0.3017890453338623, "learning_rate": 2.5777465448582807e-05, "loss": 0.0153, "step": 16135 }, { "epoch": 11.339423752635277, "grad_norm": 0.14278629422187805, "learning_rate": 2.5776996954790347e-05, "loss": 0.0076, "step": 16136 }, { "epoch": 11.340126493323963, "grad_norm": 0.1851678192615509, "learning_rate": 2.577652846099789e-05, "loss": 0.0248, "step": 16137 }, { "epoch": 11.340829234012649, "grad_norm": 0.0796043872833252, "learning_rate": 2.5776059967205435e-05, "loss": 0.0089, "step": 16138 }, { "epoch": 11.341531974701335, "grad_norm": 0.3291192352771759, "learning_rate": 2.577559147341298e-05, "loss": 0.0221, "step": 16139 }, { "epoch": 11.34223471539002, "grad_norm": 0.09441528469324112, "learning_rate": 2.577512297962052e-05, "loss": 0.013, "step": 16140 }, { "epoch": 11.342937456078706, "grad_norm": 0.7006756067276001, "learning_rate": 2.5774654485828063e-05, "loss": 0.0319, "step": 16141 }, { "epoch": 11.343640196767392, "grad_norm": 0.14778830111026764, "learning_rate": 2.5774185992035606e-05, "loss": 0.0143, "step": 16142 }, { "epoch": 11.344342937456078, "grad_norm": 0.18648168444633484, "learning_rate": 2.577371749824315e-05, "loss": 0.0288, "step": 16143 }, { "epoch": 11.345045678144764, "grad_norm": 0.23272301256656647, "learning_rate": 2.577324900445069e-05, "loss": 0.0491, "step": 16144 }, { "epoch": 11.34574841883345, "grad_norm": 0.13523271679878235, "learning_rate": 2.5772780510658234e-05, "loss": 0.0124, "step": 16145 }, { "epoch": 11.346451159522136, "grad_norm": 0.11986823379993439, "learning_rate": 2.5772312016865778e-05, "loss": 0.0252, "step": 16146 }, { "epoch": 11.347153900210822, "grad_norm": 0.18376395106315613, "learning_rate": 2.577184352307332e-05, "loss": 0.0457, "step": 16147 }, { "epoch": 11.347856640899508, "grad_norm": 0.48126518726348877, "learning_rate": 2.5771375029280862e-05, "loss": 0.0571, "step": 16148 }, { "epoch": 11.348559381588194, "grad_norm": 0.23323377966880798, "learning_rate": 2.5770906535488406e-05, "loss": 0.0539, "step": 16149 }, { "epoch": 11.34926212227688, "grad_norm": 0.37510648369789124, "learning_rate": 2.577043804169595e-05, "loss": 0.1235, "step": 16150 }, { "epoch": 11.349964862965566, "grad_norm": 0.5261026620864868, "learning_rate": 2.5769969547903493e-05, "loss": 0.1401, "step": 16151 }, { "epoch": 11.350667603654252, "grad_norm": 0.7973678708076477, "learning_rate": 2.5769501054111033e-05, "loss": 0.1677, "step": 16152 }, { "epoch": 11.351370344342937, "grad_norm": 1.6451432704925537, "learning_rate": 2.5769032560318574e-05, "loss": 0.1896, "step": 16153 }, { "epoch": 11.352073085031623, "grad_norm": 0.9059488773345947, "learning_rate": 2.5768564066526118e-05, "loss": 0.0711, "step": 16154 }, { "epoch": 11.35277582572031, "grad_norm": 0.26307910680770874, "learning_rate": 2.576809557273366e-05, "loss": 0.0344, "step": 16155 }, { "epoch": 11.353478566408995, "grad_norm": 0.28835970163345337, "learning_rate": 2.5767627078941205e-05, "loss": 0.0183, "step": 16156 }, { "epoch": 11.354181307097681, "grad_norm": 0.13635526597499847, "learning_rate": 2.5767158585148745e-05, "loss": 0.0262, "step": 16157 }, { "epoch": 11.354884047786367, "grad_norm": 0.4883035719394684, "learning_rate": 2.576669009135629e-05, "loss": 0.0234, "step": 16158 }, { "epoch": 11.355586788475053, "grad_norm": 0.09135358035564423, "learning_rate": 2.5766221597563833e-05, "loss": 0.0094, "step": 16159 }, { "epoch": 11.356289529163739, "grad_norm": 0.2822313904762268, "learning_rate": 2.5765753103771376e-05, "loss": 0.0222, "step": 16160 }, { "epoch": 11.356992269852425, "grad_norm": 0.1771659106016159, "learning_rate": 2.576528460997892e-05, "loss": 0.0189, "step": 16161 }, { "epoch": 11.35769501054111, "grad_norm": 0.14177675545215607, "learning_rate": 2.576481611618646e-05, "loss": 0.0188, "step": 16162 }, { "epoch": 11.358397751229797, "grad_norm": 0.13134868443012238, "learning_rate": 2.5764347622394004e-05, "loss": 0.015, "step": 16163 }, { "epoch": 11.359100491918483, "grad_norm": 0.24477046728134155, "learning_rate": 2.5763879128601548e-05, "loss": 0.0171, "step": 16164 }, { "epoch": 11.359803232607169, "grad_norm": 0.15262003242969513, "learning_rate": 2.5763410634809092e-05, "loss": 0.0167, "step": 16165 }, { "epoch": 11.360505973295854, "grad_norm": 0.23951934278011322, "learning_rate": 2.5762942141016632e-05, "loss": 0.0161, "step": 16166 }, { "epoch": 11.36120871398454, "grad_norm": 0.4160345792770386, "learning_rate": 2.5762473647224176e-05, "loss": 0.013, "step": 16167 }, { "epoch": 11.361911454673226, "grad_norm": 0.3049655556678772, "learning_rate": 2.576200515343172e-05, "loss": 0.0229, "step": 16168 }, { "epoch": 11.362614195361912, "grad_norm": 0.4402464032173157, "learning_rate": 2.5761536659639263e-05, "loss": 0.0462, "step": 16169 }, { "epoch": 11.363316936050598, "grad_norm": 0.2990477383136749, "learning_rate": 2.57610681658468e-05, "loss": 0.0216, "step": 16170 }, { "epoch": 11.364019676739284, "grad_norm": 0.2620486915111542, "learning_rate": 2.5760599672054344e-05, "loss": 0.0596, "step": 16171 }, { "epoch": 11.36472241742797, "grad_norm": 0.3294700086116791, "learning_rate": 2.5760131178261888e-05, "loss": 0.0334, "step": 16172 }, { "epoch": 11.365425158116654, "grad_norm": 0.2368878722190857, "learning_rate": 2.575966268446943e-05, "loss": 0.0497, "step": 16173 }, { "epoch": 11.36612789880534, "grad_norm": 0.3747401237487793, "learning_rate": 2.5759194190676975e-05, "loss": 0.1033, "step": 16174 }, { "epoch": 11.366830639494026, "grad_norm": 0.4525534212589264, "learning_rate": 2.5758725696884515e-05, "loss": 0.1154, "step": 16175 }, { "epoch": 11.367533380182712, "grad_norm": 1.0840024948120117, "learning_rate": 2.575825720309206e-05, "loss": 0.149, "step": 16176 }, { "epoch": 11.368236120871398, "grad_norm": 1.3045337200164795, "learning_rate": 2.5757788709299603e-05, "loss": 0.1538, "step": 16177 }, { "epoch": 11.368938861560084, "grad_norm": NaN, "learning_rate": 2.5757788709299603e-05, "loss": 0.155, "step": 16178 }, { "epoch": 11.36964160224877, "grad_norm": 0.44105231761932373, "learning_rate": 2.5757320215507147e-05, "loss": 0.0679, "step": 16179 }, { "epoch": 11.370344342937456, "grad_norm": 0.15932974219322205, "learning_rate": 2.5756851721714687e-05, "loss": 0.0209, "step": 16180 }, { "epoch": 11.371047083626141, "grad_norm": 0.2151062935590744, "learning_rate": 2.575638322792223e-05, "loss": 0.038, "step": 16181 }, { "epoch": 11.371749824314827, "grad_norm": 0.3778969347476959, "learning_rate": 2.5755914734129774e-05, "loss": 0.0147, "step": 16182 }, { "epoch": 11.372452565003513, "grad_norm": 0.13804450631141663, "learning_rate": 2.5755446240337318e-05, "loss": 0.016, "step": 16183 }, { "epoch": 11.3731553056922, "grad_norm": 0.10461218655109406, "learning_rate": 2.575497774654486e-05, "loss": 0.0132, "step": 16184 }, { "epoch": 11.373858046380885, "grad_norm": 0.0621640719473362, "learning_rate": 2.5754509252752402e-05, "loss": 0.0101, "step": 16185 }, { "epoch": 11.374560787069571, "grad_norm": 0.1821889877319336, "learning_rate": 2.5754040758959946e-05, "loss": 0.0242, "step": 16186 }, { "epoch": 11.375263527758257, "grad_norm": 0.1737959384918213, "learning_rate": 2.575357226516749e-05, "loss": 0.0199, "step": 16187 }, { "epoch": 11.375966268446943, "grad_norm": 0.1932356208562851, "learning_rate": 2.575310377137503e-05, "loss": 0.0198, "step": 16188 }, { "epoch": 11.376669009135629, "grad_norm": 0.1101432740688324, "learning_rate": 2.575263527758257e-05, "loss": 0.0155, "step": 16189 }, { "epoch": 11.377371749824315, "grad_norm": 0.15205375850200653, "learning_rate": 2.5752166783790114e-05, "loss": 0.0146, "step": 16190 }, { "epoch": 11.378074490513, "grad_norm": 0.3145904242992401, "learning_rate": 2.5751698289997658e-05, "loss": 0.022, "step": 16191 }, { "epoch": 11.378777231201687, "grad_norm": 0.1567273885011673, "learning_rate": 2.57512297962052e-05, "loss": 0.0187, "step": 16192 }, { "epoch": 11.379479971890373, "grad_norm": 0.20943865180015564, "learning_rate": 2.5750761302412742e-05, "loss": 0.0195, "step": 16193 }, { "epoch": 11.380182712579058, "grad_norm": 0.2013184130191803, "learning_rate": 2.5750292808620286e-05, "loss": 0.0347, "step": 16194 }, { "epoch": 11.380885453267744, "grad_norm": 0.15397492051124573, "learning_rate": 2.574982431482783e-05, "loss": 0.0259, "step": 16195 }, { "epoch": 11.38158819395643, "grad_norm": 0.519894540309906, "learning_rate": 2.5749355821035373e-05, "loss": 0.0488, "step": 16196 }, { "epoch": 11.382290934645116, "grad_norm": 0.3339831829071045, "learning_rate": 2.5748887327242913e-05, "loss": 0.0644, "step": 16197 }, { "epoch": 11.382993675333802, "grad_norm": 0.31141290068626404, "learning_rate": 2.5748418833450457e-05, "loss": 0.0423, "step": 16198 }, { "epoch": 11.383696416022488, "grad_norm": 0.2617904543876648, "learning_rate": 2.5747950339658e-05, "loss": 0.0707, "step": 16199 }, { "epoch": 11.384399156711174, "grad_norm": 0.6613979339599609, "learning_rate": 2.5747481845865545e-05, "loss": 0.109, "step": 16200 }, { "epoch": 11.38510189739986, "grad_norm": 0.6784820556640625, "learning_rate": 2.5747013352073088e-05, "loss": 0.1504, "step": 16201 }, { "epoch": 11.385804638088546, "grad_norm": 0.8314385414123535, "learning_rate": 2.574654485828063e-05, "loss": 0.1743, "step": 16202 }, { "epoch": 11.386507378777232, "grad_norm": 1.6405055522918701, "learning_rate": 2.5746076364488172e-05, "loss": 0.1741, "step": 16203 }, { "epoch": 11.387210119465918, "grad_norm": 0.3004257380962372, "learning_rate": 2.5745607870695716e-05, "loss": 0.0497, "step": 16204 }, { "epoch": 11.387912860154604, "grad_norm": 0.15407098829746246, "learning_rate": 2.5745139376903256e-05, "loss": 0.0223, "step": 16205 }, { "epoch": 11.38861560084329, "grad_norm": 0.22511950135231018, "learning_rate": 2.5744670883110797e-05, "loss": 0.0244, "step": 16206 }, { "epoch": 11.389318341531975, "grad_norm": 0.0881001204252243, "learning_rate": 2.574420238931834e-05, "loss": 0.0188, "step": 16207 }, { "epoch": 11.390021082220661, "grad_norm": 0.3466561436653137, "learning_rate": 2.5743733895525884e-05, "loss": 0.0165, "step": 16208 }, { "epoch": 11.390723822909347, "grad_norm": 0.23634885251522064, "learning_rate": 2.5743265401733428e-05, "loss": 0.0165, "step": 16209 }, { "epoch": 11.391426563598033, "grad_norm": 0.09540320932865143, "learning_rate": 2.5742796907940968e-05, "loss": 0.0127, "step": 16210 }, { "epoch": 11.392129304286719, "grad_norm": 0.39751699566841125, "learning_rate": 2.5742328414148512e-05, "loss": 0.0266, "step": 16211 }, { "epoch": 11.392832044975403, "grad_norm": 0.1404104232788086, "learning_rate": 2.5741859920356056e-05, "loss": 0.0135, "step": 16212 }, { "epoch": 11.39353478566409, "grad_norm": 0.1494590789079666, "learning_rate": 2.57413914265636e-05, "loss": 0.0352, "step": 16213 }, { "epoch": 11.394237526352775, "grad_norm": 0.1824062019586563, "learning_rate": 2.5740922932771143e-05, "loss": 0.0226, "step": 16214 }, { "epoch": 11.394940267041461, "grad_norm": 0.09512615948915482, "learning_rate": 2.5740454438978683e-05, "loss": 0.0149, "step": 16215 }, { "epoch": 11.395643007730147, "grad_norm": 0.16444122791290283, "learning_rate": 2.5739985945186227e-05, "loss": 0.0225, "step": 16216 }, { "epoch": 11.396345748418833, "grad_norm": 0.20416447520256042, "learning_rate": 2.573951745139377e-05, "loss": 0.0182, "step": 16217 }, { "epoch": 11.397048489107519, "grad_norm": 1.783867359161377, "learning_rate": 2.5739048957601315e-05, "loss": 0.0148, "step": 16218 }, { "epoch": 11.397751229796205, "grad_norm": 0.2129853367805481, "learning_rate": 2.5738580463808855e-05, "loss": 0.0242, "step": 16219 }, { "epoch": 11.39845397048489, "grad_norm": 0.579470157623291, "learning_rate": 2.57381119700164e-05, "loss": 0.0184, "step": 16220 }, { "epoch": 11.399156711173577, "grad_norm": 0.23165978491306305, "learning_rate": 2.5737643476223942e-05, "loss": 0.0359, "step": 16221 }, { "epoch": 11.399859451862262, "grad_norm": 0.45772585272789, "learning_rate": 2.5737174982431486e-05, "loss": 0.0454, "step": 16222 }, { "epoch": 11.400562192550948, "grad_norm": 0.30304181575775146, "learning_rate": 2.5736706488639023e-05, "loss": 0.0623, "step": 16223 }, { "epoch": 11.401264933239634, "grad_norm": 0.6772919297218323, "learning_rate": 2.5736237994846567e-05, "loss": 0.0705, "step": 16224 }, { "epoch": 11.40196767392832, "grad_norm": 0.41709622740745544, "learning_rate": 2.573576950105411e-05, "loss": 0.1124, "step": 16225 }, { "epoch": 11.402670414617006, "grad_norm": 0.46390607953071594, "learning_rate": 2.5735301007261654e-05, "loss": 0.1178, "step": 16226 }, { "epoch": 11.403373155305692, "grad_norm": 0.6586401462554932, "learning_rate": 2.5734832513469198e-05, "loss": 0.1647, "step": 16227 }, { "epoch": 11.404075895994378, "grad_norm": 1.1453777551651, "learning_rate": 2.573436401967674e-05, "loss": 0.1559, "step": 16228 }, { "epoch": 11.404778636683064, "grad_norm": 0.2572493553161621, "learning_rate": 2.5733895525884282e-05, "loss": 0.0599, "step": 16229 }, { "epoch": 11.40548137737175, "grad_norm": 0.21344418823719025, "learning_rate": 2.5733427032091826e-05, "loss": 0.016, "step": 16230 }, { "epoch": 11.406184118060436, "grad_norm": 0.2239261418581009, "learning_rate": 2.573295853829937e-05, "loss": 0.0211, "step": 16231 }, { "epoch": 11.406886858749122, "grad_norm": 0.15337848663330078, "learning_rate": 2.573249004450691e-05, "loss": 0.0132, "step": 16232 }, { "epoch": 11.407589599437808, "grad_norm": 0.08079562336206436, "learning_rate": 2.5732021550714454e-05, "loss": 0.0153, "step": 16233 }, { "epoch": 11.408292340126494, "grad_norm": 0.13011294603347778, "learning_rate": 2.5731553056921997e-05, "loss": 0.0106, "step": 16234 }, { "epoch": 11.40899508081518, "grad_norm": 0.13185730576515198, "learning_rate": 2.573108456312954e-05, "loss": 0.012, "step": 16235 }, { "epoch": 11.409697821503865, "grad_norm": 0.3242087960243225, "learning_rate": 2.573061606933708e-05, "loss": 0.0141, "step": 16236 }, { "epoch": 11.410400562192551, "grad_norm": 0.15875351428985596, "learning_rate": 2.5730147575544625e-05, "loss": 0.0194, "step": 16237 }, { "epoch": 11.411103302881237, "grad_norm": 0.16290904581546783, "learning_rate": 2.572967908175217e-05, "loss": 0.0096, "step": 16238 }, { "epoch": 11.411806043569923, "grad_norm": 0.12139333784580231, "learning_rate": 2.5729210587959713e-05, "loss": 0.0194, "step": 16239 }, { "epoch": 11.412508784258609, "grad_norm": 0.1914108395576477, "learning_rate": 2.5728742094167253e-05, "loss": 0.0196, "step": 16240 }, { "epoch": 11.413211524947295, "grad_norm": 0.4308459162712097, "learning_rate": 2.5728273600374793e-05, "loss": 0.0295, "step": 16241 }, { "epoch": 11.41391426563598, "grad_norm": 0.09395384788513184, "learning_rate": 2.5727805106582337e-05, "loss": 0.0108, "step": 16242 }, { "epoch": 11.414617006324667, "grad_norm": 0.21932575106620789, "learning_rate": 2.572733661278988e-05, "loss": 0.0263, "step": 16243 }, { "epoch": 11.415319747013353, "grad_norm": 0.14270350337028503, "learning_rate": 2.5726868118997424e-05, "loss": 0.0335, "step": 16244 }, { "epoch": 11.416022487702039, "grad_norm": 0.17826801538467407, "learning_rate": 2.5726399625204965e-05, "loss": 0.0248, "step": 16245 }, { "epoch": 11.416725228390725, "grad_norm": 0.18485614657402039, "learning_rate": 2.572593113141251e-05, "loss": 0.0249, "step": 16246 }, { "epoch": 11.41742796907941, "grad_norm": 0.2096198946237564, "learning_rate": 2.5725462637620052e-05, "loss": 0.0349, "step": 16247 }, { "epoch": 11.418130709768096, "grad_norm": 0.36525315046310425, "learning_rate": 2.5724994143827596e-05, "loss": 0.0525, "step": 16248 }, { "epoch": 11.41883345045678, "grad_norm": 0.41248422861099243, "learning_rate": 2.5724525650035136e-05, "loss": 0.0755, "step": 16249 }, { "epoch": 11.419536191145466, "grad_norm": 0.9783048033714294, "learning_rate": 2.572405715624268e-05, "loss": 0.0907, "step": 16250 }, { "epoch": 11.420238931834152, "grad_norm": 0.6738077998161316, "learning_rate": 2.5723588662450224e-05, "loss": 0.1367, "step": 16251 }, { "epoch": 11.420941672522838, "grad_norm": 1.5120848417282104, "learning_rate": 2.5723120168657767e-05, "loss": 0.1799, "step": 16252 }, { "epoch": 11.421644413211524, "grad_norm": 2.1335339546203613, "learning_rate": 2.572265167486531e-05, "loss": 0.2064, "step": 16253 }, { "epoch": 11.42234715390021, "grad_norm": 0.3867167532444, "learning_rate": 2.572218318107285e-05, "loss": 0.0527, "step": 16254 }, { "epoch": 11.423049894588896, "grad_norm": 0.28984764218330383, "learning_rate": 2.5721714687280395e-05, "loss": 0.0215, "step": 16255 }, { "epoch": 11.423752635277582, "grad_norm": 0.20905299484729767, "learning_rate": 2.572124619348794e-05, "loss": 0.0302, "step": 16256 }, { "epoch": 11.424455375966268, "grad_norm": 0.09485333412885666, "learning_rate": 2.5720777699695483e-05, "loss": 0.0137, "step": 16257 }, { "epoch": 11.425158116654954, "grad_norm": 0.1985849142074585, "learning_rate": 2.572030920590302e-05, "loss": 0.019, "step": 16258 }, { "epoch": 11.42586085734364, "grad_norm": 0.18340647220611572, "learning_rate": 2.5719840712110563e-05, "loss": 0.0181, "step": 16259 }, { "epoch": 11.426563598032326, "grad_norm": 0.15868215262889862, "learning_rate": 2.5719372218318107e-05, "loss": 0.0074, "step": 16260 }, { "epoch": 11.427266338721012, "grad_norm": 0.1502072513103485, "learning_rate": 2.571890372452565e-05, "loss": 0.0365, "step": 16261 }, { "epoch": 11.427969079409698, "grad_norm": 0.14386960864067078, "learning_rate": 2.571843523073319e-05, "loss": 0.0288, "step": 16262 }, { "epoch": 11.428671820098383, "grad_norm": 0.07421227544546127, "learning_rate": 2.5717966736940735e-05, "loss": 0.0094, "step": 16263 }, { "epoch": 11.42937456078707, "grad_norm": 0.24354350566864014, "learning_rate": 2.571749824314828e-05, "loss": 0.0356, "step": 16264 }, { "epoch": 11.430077301475755, "grad_norm": 0.1728586107492447, "learning_rate": 2.5717029749355822e-05, "loss": 0.0075, "step": 16265 }, { "epoch": 11.430780042164441, "grad_norm": 0.08303987234830856, "learning_rate": 2.5716561255563366e-05, "loss": 0.0161, "step": 16266 }, { "epoch": 11.431482782853127, "grad_norm": 0.10726417601108551, "learning_rate": 2.5716092761770906e-05, "loss": 0.013, "step": 16267 }, { "epoch": 11.432185523541813, "grad_norm": 0.1720738410949707, "learning_rate": 2.571562426797845e-05, "loss": 0.0302, "step": 16268 }, { "epoch": 11.432888264230499, "grad_norm": 0.17541684210300446, "learning_rate": 2.5715155774185994e-05, "loss": 0.0273, "step": 16269 }, { "epoch": 11.433591004919185, "grad_norm": 0.1595301479101181, "learning_rate": 2.5714687280393538e-05, "loss": 0.0405, "step": 16270 }, { "epoch": 11.43429374560787, "grad_norm": 0.19958995282649994, "learning_rate": 2.5714218786601078e-05, "loss": 0.0391, "step": 16271 }, { "epoch": 11.434996486296557, "grad_norm": 0.427558034658432, "learning_rate": 2.571375029280862e-05, "loss": 0.0293, "step": 16272 }, { "epoch": 11.435699226985243, "grad_norm": 0.21983666718006134, "learning_rate": 2.5713281799016165e-05, "loss": 0.0462, "step": 16273 }, { "epoch": 11.436401967673929, "grad_norm": 1.1759942770004272, "learning_rate": 2.571281330522371e-05, "loss": 0.0873, "step": 16274 }, { "epoch": 11.437104708362615, "grad_norm": 0.5860216021537781, "learning_rate": 2.5712344811431246e-05, "loss": 0.1135, "step": 16275 }, { "epoch": 11.4378074490513, "grad_norm": 0.587195873260498, "learning_rate": 2.571187631763879e-05, "loss": 0.1206, "step": 16276 }, { "epoch": 11.438510189739986, "grad_norm": 1.012074589729309, "learning_rate": 2.5711407823846333e-05, "loss": 0.1759, "step": 16277 }, { "epoch": 11.439212930428672, "grad_norm": 2.0635581016540527, "learning_rate": 2.5710939330053877e-05, "loss": 0.1945, "step": 16278 }, { "epoch": 11.439915671117358, "grad_norm": 0.3985772430896759, "learning_rate": 2.571047083626142e-05, "loss": 0.0589, "step": 16279 }, { "epoch": 11.440618411806044, "grad_norm": 0.13859568536281586, "learning_rate": 2.571000234246896e-05, "loss": 0.0202, "step": 16280 }, { "epoch": 11.44132115249473, "grad_norm": 0.08274310827255249, "learning_rate": 2.5709533848676505e-05, "loss": 0.0139, "step": 16281 }, { "epoch": 11.442023893183416, "grad_norm": 0.14794079959392548, "learning_rate": 2.570906535488405e-05, "loss": 0.0182, "step": 16282 }, { "epoch": 11.442726633872102, "grad_norm": 0.10600265115499496, "learning_rate": 2.5708596861091592e-05, "loss": 0.0146, "step": 16283 }, { "epoch": 11.443429374560788, "grad_norm": 0.20290571451187134, "learning_rate": 2.5708128367299133e-05, "loss": 0.0266, "step": 16284 }, { "epoch": 11.444132115249474, "grad_norm": 0.28320926427841187, "learning_rate": 2.5707659873506676e-05, "loss": 0.0175, "step": 16285 }, { "epoch": 11.44483485593816, "grad_norm": 0.17918238043785095, "learning_rate": 2.570719137971422e-05, "loss": 0.0115, "step": 16286 }, { "epoch": 11.445537596626846, "grad_norm": 0.160713329911232, "learning_rate": 2.5706722885921764e-05, "loss": 0.0249, "step": 16287 }, { "epoch": 11.44624033731553, "grad_norm": 0.11033681780099869, "learning_rate": 2.5706254392129304e-05, "loss": 0.0143, "step": 16288 }, { "epoch": 11.446943078004216, "grad_norm": 0.2339053750038147, "learning_rate": 2.5705785898336848e-05, "loss": 0.0365, "step": 16289 }, { "epoch": 11.447645818692902, "grad_norm": 0.18425898253917694, "learning_rate": 2.570531740454439e-05, "loss": 0.0215, "step": 16290 }, { "epoch": 11.448348559381587, "grad_norm": 0.19723807275295258, "learning_rate": 2.5704848910751935e-05, "loss": 0.0231, "step": 16291 }, { "epoch": 11.449051300070273, "grad_norm": 0.19383831322193146, "learning_rate": 2.570438041695948e-05, "loss": 0.0249, "step": 16292 }, { "epoch": 11.44975404075896, "grad_norm": 0.23288366198539734, "learning_rate": 2.5703911923167016e-05, "loss": 0.042, "step": 16293 }, { "epoch": 11.450456781447645, "grad_norm": 0.33463311195373535, "learning_rate": 2.570344342937456e-05, "loss": 0.0405, "step": 16294 }, { "epoch": 11.451159522136331, "grad_norm": 0.11786237359046936, "learning_rate": 2.5702974935582104e-05, "loss": 0.0213, "step": 16295 }, { "epoch": 11.451862262825017, "grad_norm": 0.1349155455827713, "learning_rate": 2.5702506441789647e-05, "loss": 0.0465, "step": 16296 }, { "epoch": 11.452565003513703, "grad_norm": 0.27348706126213074, "learning_rate": 2.5702037947997188e-05, "loss": 0.0521, "step": 16297 }, { "epoch": 11.453267744202389, "grad_norm": 0.20702765882015228, "learning_rate": 2.570156945420473e-05, "loss": 0.0423, "step": 16298 }, { "epoch": 11.453970484891075, "grad_norm": 0.2883358299732208, "learning_rate": 2.5701100960412275e-05, "loss": 0.0624, "step": 16299 }, { "epoch": 11.45467322557976, "grad_norm": 1.0923783779144287, "learning_rate": 2.570063246661982e-05, "loss": 0.0837, "step": 16300 }, { "epoch": 11.455375966268447, "grad_norm": 0.9378703832626343, "learning_rate": 2.570016397282736e-05, "loss": 0.1309, "step": 16301 }, { "epoch": 11.456078706957133, "grad_norm": 0.5386337637901306, "learning_rate": 2.5699695479034903e-05, "loss": 0.1544, "step": 16302 }, { "epoch": 11.456781447645819, "grad_norm": 1.2674278020858765, "learning_rate": 2.5699226985242447e-05, "loss": 0.1824, "step": 16303 }, { "epoch": 11.457484188334504, "grad_norm": 0.27805837988853455, "learning_rate": 2.569875849144999e-05, "loss": 0.059, "step": 16304 }, { "epoch": 11.45818692902319, "grad_norm": 0.10906299203634262, "learning_rate": 2.5698289997657534e-05, "loss": 0.0295, "step": 16305 }, { "epoch": 11.458889669711876, "grad_norm": 0.1320357620716095, "learning_rate": 2.5697821503865074e-05, "loss": 0.0391, "step": 16306 }, { "epoch": 11.459592410400562, "grad_norm": 0.2017102688550949, "learning_rate": 2.5697353010072618e-05, "loss": 0.0229, "step": 16307 }, { "epoch": 11.460295151089248, "grad_norm": 0.08928820490837097, "learning_rate": 2.5696884516280162e-05, "loss": 0.0128, "step": 16308 }, { "epoch": 11.460997891777934, "grad_norm": 0.11542549729347229, "learning_rate": 2.5696416022487706e-05, "loss": 0.0157, "step": 16309 }, { "epoch": 11.46170063246662, "grad_norm": 0.13404464721679688, "learning_rate": 2.5695947528695242e-05, "loss": 0.019, "step": 16310 }, { "epoch": 11.462403373155306, "grad_norm": 0.170725017786026, "learning_rate": 2.5695479034902786e-05, "loss": 0.0225, "step": 16311 }, { "epoch": 11.463106113843992, "grad_norm": 0.13932238519191742, "learning_rate": 2.569501054111033e-05, "loss": 0.0224, "step": 16312 }, { "epoch": 11.463808854532678, "grad_norm": 0.05507487431168556, "learning_rate": 2.5694542047317874e-05, "loss": 0.0087, "step": 16313 }, { "epoch": 11.464511595221364, "grad_norm": 0.1862335503101349, "learning_rate": 2.5694073553525414e-05, "loss": 0.0211, "step": 16314 }, { "epoch": 11.46521433591005, "grad_norm": 0.06678564101457596, "learning_rate": 2.5693605059732958e-05, "loss": 0.0103, "step": 16315 }, { "epoch": 11.465917076598735, "grad_norm": 0.10191679000854492, "learning_rate": 2.56931365659405e-05, "loss": 0.0195, "step": 16316 }, { "epoch": 11.466619817287421, "grad_norm": 0.1895732432603836, "learning_rate": 2.5692668072148045e-05, "loss": 0.0241, "step": 16317 }, { "epoch": 11.467322557976107, "grad_norm": 0.1972806602716446, "learning_rate": 2.569219957835559e-05, "loss": 0.0288, "step": 16318 }, { "epoch": 11.468025298664793, "grad_norm": 0.5005550384521484, "learning_rate": 2.569173108456313e-05, "loss": 0.0396, "step": 16319 }, { "epoch": 11.46872803935348, "grad_norm": 0.2931790053844452, "learning_rate": 2.5691262590770673e-05, "loss": 0.0205, "step": 16320 }, { "epoch": 11.469430780042165, "grad_norm": 0.22133088111877441, "learning_rate": 2.5690794096978217e-05, "loss": 0.0219, "step": 16321 }, { "epoch": 11.470133520730851, "grad_norm": 1.5798026323318481, "learning_rate": 2.569032560318576e-05, "loss": 0.0499, "step": 16322 }, { "epoch": 11.470836261419537, "grad_norm": 0.24501994252204895, "learning_rate": 2.56898571093933e-05, "loss": 0.0373, "step": 16323 }, { "epoch": 11.471539002108223, "grad_norm": 1.0365980863571167, "learning_rate": 2.5689388615600844e-05, "loss": 0.0593, "step": 16324 }, { "epoch": 11.472241742796909, "grad_norm": 0.5040557980537415, "learning_rate": 2.5688920121808388e-05, "loss": 0.1061, "step": 16325 }, { "epoch": 11.472944483485595, "grad_norm": 0.5789747834205627, "learning_rate": 2.5688451628015932e-05, "loss": 0.1678, "step": 16326 }, { "epoch": 11.473647224174279, "grad_norm": 0.5286425948143005, "learning_rate": 2.5687983134223472e-05, "loss": 0.1632, "step": 16327 }, { "epoch": 11.474349964862965, "grad_norm": 1.3531138896942139, "learning_rate": 2.5687514640431013e-05, "loss": 0.2117, "step": 16328 }, { "epoch": 11.47505270555165, "grad_norm": 0.5313993692398071, "learning_rate": 2.5687046146638556e-05, "loss": 0.057, "step": 16329 }, { "epoch": 11.475755446240337, "grad_norm": 0.14201822876930237, "learning_rate": 2.56865776528461e-05, "loss": 0.0194, "step": 16330 }, { "epoch": 11.476458186929023, "grad_norm": 0.20359137654304504, "learning_rate": 2.5686109159053644e-05, "loss": 0.0269, "step": 16331 }, { "epoch": 11.477160927617708, "grad_norm": 0.09889937937259674, "learning_rate": 2.5685640665261184e-05, "loss": 0.012, "step": 16332 }, { "epoch": 11.477863668306394, "grad_norm": 0.1352437138557434, "learning_rate": 2.5685172171468728e-05, "loss": 0.0165, "step": 16333 }, { "epoch": 11.47856640899508, "grad_norm": 0.33243879675865173, "learning_rate": 2.568470367767627e-05, "loss": 0.0129, "step": 16334 }, { "epoch": 11.479269149683766, "grad_norm": 0.1532072126865387, "learning_rate": 2.5684235183883815e-05, "loss": 0.0179, "step": 16335 }, { "epoch": 11.479971890372452, "grad_norm": 0.24636587500572205, "learning_rate": 2.5683766690091356e-05, "loss": 0.0205, "step": 16336 }, { "epoch": 11.480674631061138, "grad_norm": 0.15653881430625916, "learning_rate": 2.56832981962989e-05, "loss": 0.0207, "step": 16337 }, { "epoch": 11.481377371749824, "grad_norm": 0.0909980908036232, "learning_rate": 2.5682829702506443e-05, "loss": 0.0161, "step": 16338 }, { "epoch": 11.48208011243851, "grad_norm": 0.14650987088680267, "learning_rate": 2.5682361208713987e-05, "loss": 0.0152, "step": 16339 }, { "epoch": 11.482782853127196, "grad_norm": 0.14097236096858978, "learning_rate": 2.5681892714921527e-05, "loss": 0.0177, "step": 16340 }, { "epoch": 11.483485593815882, "grad_norm": 0.15827743709087372, "learning_rate": 2.568142422112907e-05, "loss": 0.0261, "step": 16341 }, { "epoch": 11.484188334504568, "grad_norm": 0.18858221173286438, "learning_rate": 2.5680955727336615e-05, "loss": 0.0178, "step": 16342 }, { "epoch": 11.484891075193254, "grad_norm": 0.17295818030834198, "learning_rate": 2.5680487233544158e-05, "loss": 0.021, "step": 16343 }, { "epoch": 11.48559381588194, "grad_norm": 0.17941507697105408, "learning_rate": 2.5680018739751702e-05, "loss": 0.0308, "step": 16344 }, { "epoch": 11.486296556570625, "grad_norm": 0.12000437080860138, "learning_rate": 2.567955024595924e-05, "loss": 0.0163, "step": 16345 }, { "epoch": 11.486999297259311, "grad_norm": 0.40659886598587036, "learning_rate": 2.5679081752166783e-05, "loss": 0.0397, "step": 16346 }, { "epoch": 11.487702037947997, "grad_norm": 0.1997835487127304, "learning_rate": 2.5678613258374326e-05, "loss": 0.0428, "step": 16347 }, { "epoch": 11.488404778636683, "grad_norm": 0.2017471343278885, "learning_rate": 2.567814476458187e-05, "loss": 0.0564, "step": 16348 }, { "epoch": 11.489107519325369, "grad_norm": 0.5972229838371277, "learning_rate": 2.567767627078941e-05, "loss": 0.0559, "step": 16349 }, { "epoch": 11.489810260014055, "grad_norm": 0.45379209518432617, "learning_rate": 2.5677207776996954e-05, "loss": 0.1019, "step": 16350 }, { "epoch": 11.490513000702741, "grad_norm": 0.4582245945930481, "learning_rate": 2.5676739283204498e-05, "loss": 0.1154, "step": 16351 }, { "epoch": 11.491215741391427, "grad_norm": 0.9772191047668457, "learning_rate": 2.567627078941204e-05, "loss": 0.2097, "step": 16352 }, { "epoch": 11.491918482080113, "grad_norm": 0.8761594891548157, "learning_rate": 2.5675802295619585e-05, "loss": 0.1721, "step": 16353 }, { "epoch": 11.492621222768799, "grad_norm": 0.2983866333961487, "learning_rate": 2.5675333801827126e-05, "loss": 0.0844, "step": 16354 }, { "epoch": 11.493323963457485, "grad_norm": 0.10525001585483551, "learning_rate": 2.567486530803467e-05, "loss": 0.0219, "step": 16355 }, { "epoch": 11.49402670414617, "grad_norm": 0.19900314509868622, "learning_rate": 2.5674396814242213e-05, "loss": 0.0442, "step": 16356 }, { "epoch": 11.494729444834856, "grad_norm": 0.15244944393634796, "learning_rate": 2.5673928320449757e-05, "loss": 0.0218, "step": 16357 }, { "epoch": 11.495432185523542, "grad_norm": 0.08239564299583435, "learning_rate": 2.5673459826657297e-05, "loss": 0.0198, "step": 16358 }, { "epoch": 11.496134926212228, "grad_norm": 0.09646471589803696, "learning_rate": 2.567299133286484e-05, "loss": 0.0125, "step": 16359 }, { "epoch": 11.496837666900914, "grad_norm": 0.48491770029067993, "learning_rate": 2.5672522839072385e-05, "loss": 0.0131, "step": 16360 }, { "epoch": 11.4975404075896, "grad_norm": 0.11898701637983322, "learning_rate": 2.567205434527993e-05, "loss": 0.0183, "step": 16361 }, { "epoch": 11.498243148278286, "grad_norm": 0.19448637962341309, "learning_rate": 2.5671585851487465e-05, "loss": 0.0177, "step": 16362 }, { "epoch": 11.498945888966972, "grad_norm": 0.14730960130691528, "learning_rate": 2.567111735769501e-05, "loss": 0.0167, "step": 16363 }, { "epoch": 11.499648629655656, "grad_norm": 0.11554165184497833, "learning_rate": 2.5670648863902553e-05, "loss": 0.0151, "step": 16364 }, { "epoch": 11.500351370344344, "grad_norm": 0.10029266029596329, "learning_rate": 2.5670180370110097e-05, "loss": 0.0128, "step": 16365 }, { "epoch": 11.501054111033028, "grad_norm": 0.42243945598602295, "learning_rate": 2.566971187631764e-05, "loss": 0.0303, "step": 16366 }, { "epoch": 11.501756851721714, "grad_norm": 0.107813261449337, "learning_rate": 2.566924338252518e-05, "loss": 0.0173, "step": 16367 }, { "epoch": 11.5024595924104, "grad_norm": 0.13169340789318085, "learning_rate": 2.5668774888732724e-05, "loss": 0.0243, "step": 16368 }, { "epoch": 11.503162333099086, "grad_norm": 0.1291537880897522, "learning_rate": 2.5668306394940268e-05, "loss": 0.0371, "step": 16369 }, { "epoch": 11.503865073787772, "grad_norm": 0.10007589310407639, "learning_rate": 2.5667837901147812e-05, "loss": 0.0128, "step": 16370 }, { "epoch": 11.504567814476458, "grad_norm": 0.21027785539627075, "learning_rate": 2.5667369407355352e-05, "loss": 0.0295, "step": 16371 }, { "epoch": 11.505270555165144, "grad_norm": 0.40074411034584045, "learning_rate": 2.5666900913562896e-05, "loss": 0.043, "step": 16372 }, { "epoch": 11.50597329585383, "grad_norm": 0.18413229286670685, "learning_rate": 2.566643241977044e-05, "loss": 0.0515, "step": 16373 }, { "epoch": 11.506676036542515, "grad_norm": 0.5090696811676025, "learning_rate": 2.5665963925977983e-05, "loss": 0.0934, "step": 16374 }, { "epoch": 11.507378777231201, "grad_norm": 0.4446251094341278, "learning_rate": 2.5665495432185524e-05, "loss": 0.0905, "step": 16375 }, { "epoch": 11.508081517919887, "grad_norm": 0.9265231490135193, "learning_rate": 2.5665026938393067e-05, "loss": 0.1549, "step": 16376 }, { "epoch": 11.508784258608573, "grad_norm": 0.7372732162475586, "learning_rate": 2.566455844460061e-05, "loss": 0.1848, "step": 16377 }, { "epoch": 11.509486999297259, "grad_norm": 1.2728633880615234, "learning_rate": 2.5664089950808155e-05, "loss": 0.2181, "step": 16378 }, { "epoch": 11.510189739985945, "grad_norm": 0.4292519986629486, "learning_rate": 2.56636214570157e-05, "loss": 0.0826, "step": 16379 }, { "epoch": 11.510892480674631, "grad_norm": 0.102399542927742, "learning_rate": 2.5663152963223235e-05, "loss": 0.0192, "step": 16380 }, { "epoch": 11.511595221363317, "grad_norm": 0.20049723982810974, "learning_rate": 2.566268446943078e-05, "loss": 0.0245, "step": 16381 }, { "epoch": 11.512297962052003, "grad_norm": 0.14102432131767273, "learning_rate": 2.5662215975638323e-05, "loss": 0.0139, "step": 16382 }, { "epoch": 11.513000702740689, "grad_norm": 0.1618206650018692, "learning_rate": 2.5661747481845867e-05, "loss": 0.0201, "step": 16383 }, { "epoch": 11.513703443429375, "grad_norm": 0.10613162815570831, "learning_rate": 2.5661278988053407e-05, "loss": 0.0096, "step": 16384 }, { "epoch": 11.51440618411806, "grad_norm": 0.10940945148468018, "learning_rate": 2.566081049426095e-05, "loss": 0.0139, "step": 16385 }, { "epoch": 11.515108924806746, "grad_norm": 0.19106760621070862, "learning_rate": 2.5660342000468494e-05, "loss": 0.0346, "step": 16386 }, { "epoch": 11.515811665495432, "grad_norm": 0.10795541107654572, "learning_rate": 2.5659873506676038e-05, "loss": 0.0173, "step": 16387 }, { "epoch": 11.516514406184118, "grad_norm": 0.11575505882501602, "learning_rate": 2.565940501288358e-05, "loss": 0.0154, "step": 16388 }, { "epoch": 11.517217146872804, "grad_norm": 0.10756494104862213, "learning_rate": 2.5658936519091122e-05, "loss": 0.0226, "step": 16389 }, { "epoch": 11.51791988756149, "grad_norm": 0.0687396302819252, "learning_rate": 2.5658468025298666e-05, "loss": 0.0127, "step": 16390 }, { "epoch": 11.518622628250176, "grad_norm": 0.20728671550750732, "learning_rate": 2.565799953150621e-05, "loss": 0.0217, "step": 16391 }, { "epoch": 11.519325368938862, "grad_norm": 0.17580784857273102, "learning_rate": 2.5657531037713753e-05, "loss": 0.0175, "step": 16392 }, { "epoch": 11.520028109627548, "grad_norm": 0.24687474966049194, "learning_rate": 2.5657062543921294e-05, "loss": 0.0332, "step": 16393 }, { "epoch": 11.520730850316234, "grad_norm": 0.17039431631565094, "learning_rate": 2.5656594050128837e-05, "loss": 0.0306, "step": 16394 }, { "epoch": 11.52143359100492, "grad_norm": 0.20591378211975098, "learning_rate": 2.565612555633638e-05, "loss": 0.0183, "step": 16395 }, { "epoch": 11.522136331693606, "grad_norm": 0.13299177587032318, "learning_rate": 2.5655657062543925e-05, "loss": 0.0262, "step": 16396 }, { "epoch": 11.522839072382292, "grad_norm": 0.1659124493598938, "learning_rate": 2.5655188568751462e-05, "loss": 0.0491, "step": 16397 }, { "epoch": 11.523541813070977, "grad_norm": 0.24192216992378235, "learning_rate": 2.5654720074959006e-05, "loss": 0.0486, "step": 16398 }, { "epoch": 11.524244553759663, "grad_norm": 0.8924581408500671, "learning_rate": 2.565425158116655e-05, "loss": 0.0657, "step": 16399 }, { "epoch": 11.52494729444835, "grad_norm": 0.7667932510375977, "learning_rate": 2.5653783087374093e-05, "loss": 0.1003, "step": 16400 }, { "epoch": 11.525650035137035, "grad_norm": 0.4878385066986084, "learning_rate": 2.5653314593581633e-05, "loss": 0.1172, "step": 16401 }, { "epoch": 11.526352775825721, "grad_norm": 1.3111340999603271, "learning_rate": 2.5652846099789177e-05, "loss": 0.18, "step": 16402 }, { "epoch": 11.527055516514405, "grad_norm": 1.6545330286026, "learning_rate": 2.565237760599672e-05, "loss": 0.1823, "step": 16403 }, { "epoch": 11.527758257203093, "grad_norm": 0.23098249733448029, "learning_rate": 2.5651909112204265e-05, "loss": 0.068, "step": 16404 }, { "epoch": 11.528460997891777, "grad_norm": 0.11800672113895416, "learning_rate": 2.5651440618411808e-05, "loss": 0.0282, "step": 16405 }, { "epoch": 11.529163738580463, "grad_norm": 0.1264648735523224, "learning_rate": 2.565097212461935e-05, "loss": 0.026, "step": 16406 }, { "epoch": 11.529866479269149, "grad_norm": 0.10689437389373779, "learning_rate": 2.5650503630826892e-05, "loss": 0.0231, "step": 16407 }, { "epoch": 11.530569219957835, "grad_norm": 0.17587141692638397, "learning_rate": 2.5650035137034436e-05, "loss": 0.012, "step": 16408 }, { "epoch": 11.53127196064652, "grad_norm": 0.2562389671802521, "learning_rate": 2.564956664324198e-05, "loss": 0.0156, "step": 16409 }, { "epoch": 11.531974701335207, "grad_norm": 0.07347512245178223, "learning_rate": 2.564909814944952e-05, "loss": 0.0143, "step": 16410 }, { "epoch": 11.532677442023893, "grad_norm": 0.062270790338516235, "learning_rate": 2.5648629655657064e-05, "loss": 0.0106, "step": 16411 }, { "epoch": 11.533380182712579, "grad_norm": 0.12183790653944016, "learning_rate": 2.5648161161864608e-05, "loss": 0.0216, "step": 16412 }, { "epoch": 11.534082923401265, "grad_norm": 0.10557231307029724, "learning_rate": 2.564769266807215e-05, "loss": 0.0256, "step": 16413 }, { "epoch": 11.53478566408995, "grad_norm": 0.11350446194410324, "learning_rate": 2.5647224174279688e-05, "loss": 0.0202, "step": 16414 }, { "epoch": 11.535488404778636, "grad_norm": 0.16564519703388214, "learning_rate": 2.5646755680487232e-05, "loss": 0.019, "step": 16415 }, { "epoch": 11.536191145467322, "grad_norm": 0.3671376705169678, "learning_rate": 2.5646287186694776e-05, "loss": 0.017, "step": 16416 }, { "epoch": 11.536893886156008, "grad_norm": 0.13184022903442383, "learning_rate": 2.564581869290232e-05, "loss": 0.0119, "step": 16417 }, { "epoch": 11.537596626844694, "grad_norm": 0.21382315456867218, "learning_rate": 2.5645350199109863e-05, "loss": 0.0252, "step": 16418 }, { "epoch": 11.53829936753338, "grad_norm": 0.23565734922885895, "learning_rate": 2.5644881705317403e-05, "loss": 0.0211, "step": 16419 }, { "epoch": 11.539002108222066, "grad_norm": 0.2358631193637848, "learning_rate": 2.5644413211524947e-05, "loss": 0.0127, "step": 16420 }, { "epoch": 11.539704848910752, "grad_norm": 0.27751776576042175, "learning_rate": 2.564394471773249e-05, "loss": 0.0463, "step": 16421 }, { "epoch": 11.540407589599438, "grad_norm": 1.6984590291976929, "learning_rate": 2.5643476223940035e-05, "loss": 0.0425, "step": 16422 }, { "epoch": 11.541110330288124, "grad_norm": 0.3205741047859192, "learning_rate": 2.5643007730147575e-05, "loss": 0.0448, "step": 16423 }, { "epoch": 11.54181307097681, "grad_norm": 0.9725870490074158, "learning_rate": 2.564253923635512e-05, "loss": 0.1234, "step": 16424 }, { "epoch": 11.542515811665496, "grad_norm": 0.3796507716178894, "learning_rate": 2.5642070742562662e-05, "loss": 0.0804, "step": 16425 }, { "epoch": 11.543218552354181, "grad_norm": 0.9023538827896118, "learning_rate": 2.5641602248770206e-05, "loss": 0.1539, "step": 16426 }, { "epoch": 11.543921293042867, "grad_norm": 0.9438847303390503, "learning_rate": 2.5641133754977747e-05, "loss": 0.159, "step": 16427 }, { "epoch": 11.544624033731553, "grad_norm": 2.696584939956665, "learning_rate": 2.564066526118529e-05, "loss": 0.2205, "step": 16428 }, { "epoch": 11.54532677442024, "grad_norm": 0.3161773085594177, "learning_rate": 2.5640196767392834e-05, "loss": 0.0682, "step": 16429 }, { "epoch": 11.546029515108925, "grad_norm": 0.2696634531021118, "learning_rate": 2.5639728273600378e-05, "loss": 0.0194, "step": 16430 }, { "epoch": 11.546732255797611, "grad_norm": 0.2206355184316635, "learning_rate": 2.563925977980792e-05, "loss": 0.0211, "step": 16431 }, { "epoch": 11.547434996486297, "grad_norm": 0.26181116700172424, "learning_rate": 2.563879128601546e-05, "loss": 0.014, "step": 16432 }, { "epoch": 11.548137737174983, "grad_norm": 0.11850892752408981, "learning_rate": 2.5638322792223002e-05, "loss": 0.016, "step": 16433 }, { "epoch": 11.548840477863669, "grad_norm": 0.055644724518060684, "learning_rate": 2.5637854298430546e-05, "loss": 0.0044, "step": 16434 }, { "epoch": 11.549543218552355, "grad_norm": 0.28046876192092896, "learning_rate": 2.563738580463809e-05, "loss": 0.0204, "step": 16435 }, { "epoch": 11.55024595924104, "grad_norm": 0.917905867099762, "learning_rate": 2.563691731084563e-05, "loss": 0.0268, "step": 16436 }, { "epoch": 11.550948699929727, "grad_norm": 0.13708947598934174, "learning_rate": 2.5636448817053174e-05, "loss": 0.0219, "step": 16437 }, { "epoch": 11.551651440618413, "grad_norm": 0.09551902860403061, "learning_rate": 2.5635980323260717e-05, "loss": 0.0092, "step": 16438 }, { "epoch": 11.552354181307098, "grad_norm": 0.13097059726715088, "learning_rate": 2.563551182946826e-05, "loss": 0.0176, "step": 16439 }, { "epoch": 11.553056921995784, "grad_norm": 0.10207587480545044, "learning_rate": 2.56350433356758e-05, "loss": 0.0131, "step": 16440 }, { "epoch": 11.55375966268447, "grad_norm": 0.32237544655799866, "learning_rate": 2.5634574841883345e-05, "loss": 0.0272, "step": 16441 }, { "epoch": 11.554462403373154, "grad_norm": 0.1285579949617386, "learning_rate": 2.563410634809089e-05, "loss": 0.0165, "step": 16442 }, { "epoch": 11.55516514406184, "grad_norm": 0.16012071073055267, "learning_rate": 2.5633637854298433e-05, "loss": 0.034, "step": 16443 }, { "epoch": 11.555867884750526, "grad_norm": 0.19219864904880524, "learning_rate": 2.5633169360505976e-05, "loss": 0.0427, "step": 16444 }, { "epoch": 11.556570625439212, "grad_norm": 0.19800063967704773, "learning_rate": 2.5632700866713517e-05, "loss": 0.0202, "step": 16445 }, { "epoch": 11.557273366127898, "grad_norm": 0.16694283485412598, "learning_rate": 2.563223237292106e-05, "loss": 0.0421, "step": 16446 }, { "epoch": 11.557976106816584, "grad_norm": 0.4981203079223633, "learning_rate": 2.5631763879128604e-05, "loss": 0.0276, "step": 16447 }, { "epoch": 11.55867884750527, "grad_norm": 0.40790805220603943, "learning_rate": 2.5631295385336148e-05, "loss": 0.0477, "step": 16448 }, { "epoch": 11.559381588193956, "grad_norm": 0.8329806923866272, "learning_rate": 2.5630826891543685e-05, "loss": 0.1009, "step": 16449 }, { "epoch": 11.560084328882642, "grad_norm": 0.7705890536308289, "learning_rate": 2.563035839775123e-05, "loss": 0.1125, "step": 16450 }, { "epoch": 11.560787069571328, "grad_norm": 0.428926557302475, "learning_rate": 2.5629889903958772e-05, "loss": 0.1075, "step": 16451 }, { "epoch": 11.561489810260014, "grad_norm": 0.5761129856109619, "learning_rate": 2.5629421410166316e-05, "loss": 0.1586, "step": 16452 }, { "epoch": 11.5621925509487, "grad_norm": 0.7899782657623291, "learning_rate": 2.5628952916373856e-05, "loss": 0.2056, "step": 16453 }, { "epoch": 11.562895291637385, "grad_norm": 0.2611497640609741, "learning_rate": 2.56284844225814e-05, "loss": 0.074, "step": 16454 }, { "epoch": 11.563598032326071, "grad_norm": 0.13338583707809448, "learning_rate": 2.5628015928788944e-05, "loss": 0.0315, "step": 16455 }, { "epoch": 11.564300773014757, "grad_norm": 0.13950952887535095, "learning_rate": 2.5627547434996487e-05, "loss": 0.018, "step": 16456 }, { "epoch": 11.565003513703443, "grad_norm": 0.18562698364257812, "learning_rate": 2.562707894120403e-05, "loss": 0.0234, "step": 16457 }, { "epoch": 11.56570625439213, "grad_norm": 0.11515728384256363, "learning_rate": 2.562661044741157e-05, "loss": 0.0147, "step": 16458 }, { "epoch": 11.566408995080815, "grad_norm": 0.07143287360668182, "learning_rate": 2.5626141953619115e-05, "loss": 0.0093, "step": 16459 }, { "epoch": 11.567111735769501, "grad_norm": 0.0924251452088356, "learning_rate": 2.562567345982666e-05, "loss": 0.0091, "step": 16460 }, { "epoch": 11.567814476458187, "grad_norm": 0.6207988262176514, "learning_rate": 2.5625204966034203e-05, "loss": 0.0404, "step": 16461 }, { "epoch": 11.568517217146873, "grad_norm": 0.17476744949817657, "learning_rate": 2.5624736472241743e-05, "loss": 0.035, "step": 16462 }, { "epoch": 11.569219957835559, "grad_norm": 0.07224521040916443, "learning_rate": 2.5624267978449287e-05, "loss": 0.0089, "step": 16463 }, { "epoch": 11.569922698524245, "grad_norm": 0.2169305831193924, "learning_rate": 2.562379948465683e-05, "loss": 0.0218, "step": 16464 }, { "epoch": 11.57062543921293, "grad_norm": 0.09629765897989273, "learning_rate": 2.5623330990864374e-05, "loss": 0.0113, "step": 16465 }, { "epoch": 11.571328179901617, "grad_norm": 0.1539878100156784, "learning_rate": 2.5622862497071915e-05, "loss": 0.0297, "step": 16466 }, { "epoch": 11.572030920590302, "grad_norm": 0.20498113334178925, "learning_rate": 2.5622394003279455e-05, "loss": 0.0143, "step": 16467 }, { "epoch": 11.572733661278988, "grad_norm": 0.12293940037488937, "learning_rate": 2.5621925509487e-05, "loss": 0.0219, "step": 16468 }, { "epoch": 11.573436401967674, "grad_norm": 0.10544416308403015, "learning_rate": 2.5621457015694542e-05, "loss": 0.0191, "step": 16469 }, { "epoch": 11.57413914265636, "grad_norm": 0.15575551986694336, "learning_rate": 2.5620988521902086e-05, "loss": 0.0179, "step": 16470 }, { "epoch": 11.574841883345046, "grad_norm": 0.20527386665344238, "learning_rate": 2.5620520028109626e-05, "loss": 0.0359, "step": 16471 }, { "epoch": 11.575544624033732, "grad_norm": 0.306662380695343, "learning_rate": 2.562005153431717e-05, "loss": 0.0296, "step": 16472 }, { "epoch": 11.576247364722418, "grad_norm": 0.3985600471496582, "learning_rate": 2.5619583040524714e-05, "loss": 0.0578, "step": 16473 }, { "epoch": 11.576950105411104, "grad_norm": 0.36972132325172424, "learning_rate": 2.5619114546732258e-05, "loss": 0.0624, "step": 16474 }, { "epoch": 11.57765284609979, "grad_norm": 0.5249558091163635, "learning_rate": 2.5618646052939798e-05, "loss": 0.0879, "step": 16475 }, { "epoch": 11.578355586788476, "grad_norm": 0.5673834681510925, "learning_rate": 2.561817755914734e-05, "loss": 0.1316, "step": 16476 }, { "epoch": 11.579058327477162, "grad_norm": 0.8303595781326294, "learning_rate": 2.5617709065354885e-05, "loss": 0.1472, "step": 16477 }, { "epoch": 11.579761068165848, "grad_norm": 0.9156123399734497, "learning_rate": 2.561724057156243e-05, "loss": 0.1952, "step": 16478 }, { "epoch": 11.580463808854532, "grad_norm": 0.23252186179161072, "learning_rate": 2.561677207776997e-05, "loss": 0.0726, "step": 16479 }, { "epoch": 11.58116654954322, "grad_norm": 0.22793914377689362, "learning_rate": 2.5616303583977513e-05, "loss": 0.022, "step": 16480 }, { "epoch": 11.581869290231904, "grad_norm": 0.15786074101924896, "learning_rate": 2.5615835090185057e-05, "loss": 0.0143, "step": 16481 }, { "epoch": 11.58257203092059, "grad_norm": 0.12412440031766891, "learning_rate": 2.56153665963926e-05, "loss": 0.0178, "step": 16482 }, { "epoch": 11.583274771609275, "grad_norm": 0.2862984836101532, "learning_rate": 2.5614898102600144e-05, "loss": 0.0244, "step": 16483 }, { "epoch": 11.583977512297961, "grad_norm": 0.175899937748909, "learning_rate": 2.561442960880768e-05, "loss": 0.0215, "step": 16484 }, { "epoch": 11.584680252986647, "grad_norm": 0.10423661023378372, "learning_rate": 2.5613961115015225e-05, "loss": 0.0099, "step": 16485 }, { "epoch": 11.585382993675333, "grad_norm": 0.15851455926895142, "learning_rate": 2.561349262122277e-05, "loss": 0.0278, "step": 16486 }, { "epoch": 11.58608573436402, "grad_norm": 0.11252180486917496, "learning_rate": 2.5613024127430312e-05, "loss": 0.0213, "step": 16487 }, { "epoch": 11.586788475052705, "grad_norm": 0.08689210563898087, "learning_rate": 2.5612555633637853e-05, "loss": 0.0113, "step": 16488 }, { "epoch": 11.587491215741391, "grad_norm": 0.10953543335199356, "learning_rate": 2.5612087139845396e-05, "loss": 0.0172, "step": 16489 }, { "epoch": 11.588193956430077, "grad_norm": 0.11650247126817703, "learning_rate": 2.561161864605294e-05, "loss": 0.015, "step": 16490 }, { "epoch": 11.588896697118763, "grad_norm": 0.2293199896812439, "learning_rate": 2.5611150152260484e-05, "loss": 0.0323, "step": 16491 }, { "epoch": 11.589599437807449, "grad_norm": 0.1287347674369812, "learning_rate": 2.5610681658468024e-05, "loss": 0.0123, "step": 16492 }, { "epoch": 11.590302178496135, "grad_norm": 0.13239651918411255, "learning_rate": 2.5610213164675568e-05, "loss": 0.0168, "step": 16493 }, { "epoch": 11.59100491918482, "grad_norm": 0.50284343957901, "learning_rate": 2.5609744670883112e-05, "loss": 0.0336, "step": 16494 }, { "epoch": 11.591707659873506, "grad_norm": 0.13400360941886902, "learning_rate": 2.5609276177090655e-05, "loss": 0.0303, "step": 16495 }, { "epoch": 11.592410400562192, "grad_norm": 0.5395850539207458, "learning_rate": 2.56088076832982e-05, "loss": 0.048, "step": 16496 }, { "epoch": 11.593113141250878, "grad_norm": 0.2004842907190323, "learning_rate": 2.560833918950574e-05, "loss": 0.0298, "step": 16497 }, { "epoch": 11.593815881939564, "grad_norm": 0.4367239475250244, "learning_rate": 2.5607870695713283e-05, "loss": 0.0572, "step": 16498 }, { "epoch": 11.59451862262825, "grad_norm": 2.815063238143921, "learning_rate": 2.5607402201920827e-05, "loss": 0.0956, "step": 16499 }, { "epoch": 11.595221363316936, "grad_norm": 0.33307892084121704, "learning_rate": 2.560693370812837e-05, "loss": 0.093, "step": 16500 }, { "epoch": 11.595924104005622, "grad_norm": 0.778972864151001, "learning_rate": 2.5606465214335908e-05, "loss": 0.1592, "step": 16501 }, { "epoch": 11.596626844694308, "grad_norm": 0.9948810935020447, "learning_rate": 2.560599672054345e-05, "loss": 0.1273, "step": 16502 }, { "epoch": 11.597329585382994, "grad_norm": 1.8558225631713867, "learning_rate": 2.5605528226750995e-05, "loss": 0.1929, "step": 16503 }, { "epoch": 11.59803232607168, "grad_norm": 0.20164598524570465, "learning_rate": 2.560505973295854e-05, "loss": 0.0616, "step": 16504 }, { "epoch": 11.598735066760366, "grad_norm": 0.17669539153575897, "learning_rate": 2.560459123916608e-05, "loss": 0.0239, "step": 16505 }, { "epoch": 11.599437807449052, "grad_norm": 0.11607842892408371, "learning_rate": 2.5604122745373623e-05, "loss": 0.0265, "step": 16506 }, { "epoch": 11.600140548137738, "grad_norm": 0.08951877057552338, "learning_rate": 2.5603654251581167e-05, "loss": 0.013, "step": 16507 }, { "epoch": 11.600843288826423, "grad_norm": 0.27474674582481384, "learning_rate": 2.560318575778871e-05, "loss": 0.03, "step": 16508 }, { "epoch": 11.60154602951511, "grad_norm": 0.07840131223201752, "learning_rate": 2.5602717263996254e-05, "loss": 0.007, "step": 16509 }, { "epoch": 11.602248770203795, "grad_norm": 0.9138235449790955, "learning_rate": 2.5602248770203794e-05, "loss": 0.0097, "step": 16510 }, { "epoch": 11.602951510892481, "grad_norm": 0.13166239857673645, "learning_rate": 2.5601780276411338e-05, "loss": 0.0193, "step": 16511 }, { "epoch": 11.603654251581167, "grad_norm": 0.11946721374988556, "learning_rate": 2.5601311782618882e-05, "loss": 0.0125, "step": 16512 }, { "epoch": 11.604356992269853, "grad_norm": 0.11065477877855301, "learning_rate": 2.5600843288826426e-05, "loss": 0.0138, "step": 16513 }, { "epoch": 11.605059732958539, "grad_norm": 0.31868278980255127, "learning_rate": 2.5600374795033966e-05, "loss": 0.0314, "step": 16514 }, { "epoch": 11.605762473647225, "grad_norm": 0.09761884063482285, "learning_rate": 2.559990630124151e-05, "loss": 0.0129, "step": 16515 }, { "epoch": 11.60646521433591, "grad_norm": 0.18361014127731323, "learning_rate": 2.5599437807449053e-05, "loss": 0.0393, "step": 16516 }, { "epoch": 11.607167955024597, "grad_norm": 0.08559095859527588, "learning_rate": 2.5598969313656597e-05, "loss": 0.01, "step": 16517 }, { "epoch": 11.607870695713281, "grad_norm": 0.18634866178035736, "learning_rate": 2.5598500819864137e-05, "loss": 0.0405, "step": 16518 }, { "epoch": 11.608573436401969, "grad_norm": 0.41973793506622314, "learning_rate": 2.5598032326071678e-05, "loss": 0.0451, "step": 16519 }, { "epoch": 11.609276177090653, "grad_norm": 0.6907607913017273, "learning_rate": 2.559756383227922e-05, "loss": 0.0183, "step": 16520 }, { "epoch": 11.609978917779339, "grad_norm": 0.27388423681259155, "learning_rate": 2.5597095338486765e-05, "loss": 0.0329, "step": 16521 }, { "epoch": 11.610681658468025, "grad_norm": 0.17363527417182922, "learning_rate": 2.559662684469431e-05, "loss": 0.037, "step": 16522 }, { "epoch": 11.61138439915671, "grad_norm": 0.3691949248313904, "learning_rate": 2.559615835090185e-05, "loss": 0.059, "step": 16523 }, { "epoch": 11.612087139845396, "grad_norm": 0.34585633873939514, "learning_rate": 2.5595689857109393e-05, "loss": 0.0537, "step": 16524 }, { "epoch": 11.612789880534082, "grad_norm": 0.5213842988014221, "learning_rate": 2.5595221363316937e-05, "loss": 0.105, "step": 16525 }, { "epoch": 11.613492621222768, "grad_norm": 0.5863708853721619, "learning_rate": 2.559475286952448e-05, "loss": 0.1488, "step": 16526 }, { "epoch": 11.614195361911454, "grad_norm": 0.6770254373550415, "learning_rate": 2.559428437573202e-05, "loss": 0.1853, "step": 16527 }, { "epoch": 11.61489810260014, "grad_norm": 1.0588419437408447, "learning_rate": 2.5593815881939564e-05, "loss": 0.196, "step": 16528 }, { "epoch": 11.615600843288826, "grad_norm": 0.22797216475009918, "learning_rate": 2.5593347388147108e-05, "loss": 0.0824, "step": 16529 }, { "epoch": 11.616303583977512, "grad_norm": 0.19699551165103912, "learning_rate": 2.5592878894354652e-05, "loss": 0.023, "step": 16530 }, { "epoch": 11.617006324666198, "grad_norm": 0.13739433884620667, "learning_rate": 2.5592410400562192e-05, "loss": 0.0148, "step": 16531 }, { "epoch": 11.617709065354884, "grad_norm": 0.08040064573287964, "learning_rate": 2.5591941906769736e-05, "loss": 0.009, "step": 16532 }, { "epoch": 11.61841180604357, "grad_norm": 0.19355571269989014, "learning_rate": 2.559147341297728e-05, "loss": 0.0237, "step": 16533 }, { "epoch": 11.619114546732256, "grad_norm": 1.4960556030273438, "learning_rate": 2.5591004919184823e-05, "loss": 0.0046, "step": 16534 }, { "epoch": 11.619817287420942, "grad_norm": 0.12996694445610046, "learning_rate": 2.5590536425392367e-05, "loss": 0.0181, "step": 16535 }, { "epoch": 11.620520028109627, "grad_norm": 0.13326485455036163, "learning_rate": 2.5590067931599904e-05, "loss": 0.0184, "step": 16536 }, { "epoch": 11.621222768798313, "grad_norm": 0.1272953301668167, "learning_rate": 2.5589599437807448e-05, "loss": 0.0183, "step": 16537 }, { "epoch": 11.621925509487, "grad_norm": 0.6092647314071655, "learning_rate": 2.558913094401499e-05, "loss": 0.0204, "step": 16538 }, { "epoch": 11.622628250175685, "grad_norm": 0.1415330469608307, "learning_rate": 2.5588662450222535e-05, "loss": 0.0164, "step": 16539 }, { "epoch": 11.623330990864371, "grad_norm": 0.07869536429643631, "learning_rate": 2.5588193956430076e-05, "loss": 0.0092, "step": 16540 }, { "epoch": 11.624033731553057, "grad_norm": 0.2344488948583603, "learning_rate": 2.558772546263762e-05, "loss": 0.0311, "step": 16541 }, { "epoch": 11.624736472241743, "grad_norm": 0.22076554596424103, "learning_rate": 2.5587256968845163e-05, "loss": 0.0172, "step": 16542 }, { "epoch": 11.625439212930429, "grad_norm": 0.3774627149105072, "learning_rate": 2.5586788475052707e-05, "loss": 0.0382, "step": 16543 }, { "epoch": 11.626141953619115, "grad_norm": 0.28675276041030884, "learning_rate": 2.5586319981260247e-05, "loss": 0.0497, "step": 16544 }, { "epoch": 11.6268446943078, "grad_norm": 0.27806490659713745, "learning_rate": 2.558585148746779e-05, "loss": 0.0207, "step": 16545 }, { "epoch": 11.627547434996487, "grad_norm": 0.1338081955909729, "learning_rate": 2.5585382993675335e-05, "loss": 0.0308, "step": 16546 }, { "epoch": 11.628250175685173, "grad_norm": 0.4359043538570404, "learning_rate": 2.558491449988288e-05, "loss": 0.0704, "step": 16547 }, { "epoch": 11.628952916373859, "grad_norm": 0.23946578800678253, "learning_rate": 2.5584446006090422e-05, "loss": 0.0565, "step": 16548 }, { "epoch": 11.629655657062544, "grad_norm": 0.30952274799346924, "learning_rate": 2.5583977512297962e-05, "loss": 0.0587, "step": 16549 }, { "epoch": 11.63035839775123, "grad_norm": 0.7822174429893494, "learning_rate": 2.5583509018505506e-05, "loss": 0.1002, "step": 16550 }, { "epoch": 11.631061138439916, "grad_norm": 0.740456223487854, "learning_rate": 2.558304052471305e-05, "loss": 0.1374, "step": 16551 }, { "epoch": 11.631763879128602, "grad_norm": 1.1178065538406372, "learning_rate": 2.5582572030920594e-05, "loss": 0.179, "step": 16552 }, { "epoch": 11.632466619817288, "grad_norm": 1.0406501293182373, "learning_rate": 2.5582103537128134e-05, "loss": 0.1649, "step": 16553 }, { "epoch": 11.633169360505974, "grad_norm": 0.17454935610294342, "learning_rate": 2.5581635043335674e-05, "loss": 0.0676, "step": 16554 }, { "epoch": 11.63387210119466, "grad_norm": 0.09571774303913116, "learning_rate": 2.5581166549543218e-05, "loss": 0.0314, "step": 16555 }, { "epoch": 11.634574841883346, "grad_norm": 0.2299884557723999, "learning_rate": 2.5580698055750762e-05, "loss": 0.0177, "step": 16556 }, { "epoch": 11.63527758257203, "grad_norm": 0.1271054744720459, "learning_rate": 2.5580229561958305e-05, "loss": 0.0187, "step": 16557 }, { "epoch": 11.635980323260716, "grad_norm": 0.17530031502246857, "learning_rate": 2.5579761068165846e-05, "loss": 0.016, "step": 16558 }, { "epoch": 11.636683063949402, "grad_norm": 0.13659049570560455, "learning_rate": 2.557929257437339e-05, "loss": 0.0136, "step": 16559 }, { "epoch": 11.637385804638088, "grad_norm": 0.26105430722236633, "learning_rate": 2.5578824080580933e-05, "loss": 0.0157, "step": 16560 }, { "epoch": 11.638088545326774, "grad_norm": 0.16895529627799988, "learning_rate": 2.5578355586788477e-05, "loss": 0.0139, "step": 16561 }, { "epoch": 11.63879128601546, "grad_norm": 0.09861885756254196, "learning_rate": 2.5577887092996017e-05, "loss": 0.0137, "step": 16562 }, { "epoch": 11.639494026704146, "grad_norm": 0.09625107795000076, "learning_rate": 2.557741859920356e-05, "loss": 0.0078, "step": 16563 }, { "epoch": 11.640196767392831, "grad_norm": 0.20129621028900146, "learning_rate": 2.5576950105411105e-05, "loss": 0.018, "step": 16564 }, { "epoch": 11.640899508081517, "grad_norm": 0.10454443842172623, "learning_rate": 2.557648161161865e-05, "loss": 0.014, "step": 16565 }, { "epoch": 11.641602248770203, "grad_norm": 0.136106476187706, "learning_rate": 2.557601311782619e-05, "loss": 0.0206, "step": 16566 }, { "epoch": 11.64230498945889, "grad_norm": 0.6157243251800537, "learning_rate": 2.5575544624033733e-05, "loss": 0.0234, "step": 16567 }, { "epoch": 11.643007730147575, "grad_norm": 0.26326191425323486, "learning_rate": 2.5575076130241276e-05, "loss": 0.0249, "step": 16568 }, { "epoch": 11.643710470836261, "grad_norm": 0.26708984375, "learning_rate": 2.557460763644882e-05, "loss": 0.0284, "step": 16569 }, { "epoch": 11.644413211524947, "grad_norm": 0.1834126114845276, "learning_rate": 2.5574139142656364e-05, "loss": 0.0279, "step": 16570 }, { "epoch": 11.645115952213633, "grad_norm": 0.2489696741104126, "learning_rate": 2.55736706488639e-05, "loss": 0.0337, "step": 16571 }, { "epoch": 11.645818692902319, "grad_norm": 0.33067718148231506, "learning_rate": 2.5573202155071444e-05, "loss": 0.0388, "step": 16572 }, { "epoch": 11.646521433591005, "grad_norm": 0.36753690242767334, "learning_rate": 2.5572733661278988e-05, "loss": 0.0528, "step": 16573 }, { "epoch": 11.64722417427969, "grad_norm": 0.42756444215774536, "learning_rate": 2.5572265167486532e-05, "loss": 0.0769, "step": 16574 }, { "epoch": 11.647926914968377, "grad_norm": 0.38739097118377686, "learning_rate": 2.5571796673694072e-05, "loss": 0.1222, "step": 16575 }, { "epoch": 11.648629655657063, "grad_norm": 0.6488527059555054, "learning_rate": 2.5571328179901616e-05, "loss": 0.1403, "step": 16576 }, { "epoch": 11.649332396345748, "grad_norm": 0.5999335050582886, "learning_rate": 2.557085968610916e-05, "loss": 0.1436, "step": 16577 }, { "epoch": 11.650035137034434, "grad_norm": 3.299678325653076, "learning_rate": 2.5570391192316703e-05, "loss": 0.2066, "step": 16578 }, { "epoch": 11.65073787772312, "grad_norm": 0.18674489855766296, "learning_rate": 2.5569922698524244e-05, "loss": 0.0527, "step": 16579 }, { "epoch": 11.651440618411806, "grad_norm": 0.12167452275753021, "learning_rate": 2.5569454204731787e-05, "loss": 0.0422, "step": 16580 }, { "epoch": 11.652143359100492, "grad_norm": 0.6288358569145203, "learning_rate": 2.556898571093933e-05, "loss": 0.0207, "step": 16581 }, { "epoch": 11.652846099789178, "grad_norm": 0.38257962465286255, "learning_rate": 2.5568517217146875e-05, "loss": 0.0217, "step": 16582 }, { "epoch": 11.653548840477864, "grad_norm": 0.2251998484134674, "learning_rate": 2.556804872335442e-05, "loss": 0.0088, "step": 16583 }, { "epoch": 11.65425158116655, "grad_norm": 0.11581951379776001, "learning_rate": 2.556758022956196e-05, "loss": 0.013, "step": 16584 }, { "epoch": 11.654954321855236, "grad_norm": 0.0777013748884201, "learning_rate": 2.5567111735769503e-05, "loss": 0.0147, "step": 16585 }, { "epoch": 11.655657062543922, "grad_norm": 0.1483154445886612, "learning_rate": 2.5566643241977046e-05, "loss": 0.028, "step": 16586 }, { "epoch": 11.656359803232608, "grad_norm": 0.09187323600053787, "learning_rate": 2.556617474818459e-05, "loss": 0.015, "step": 16587 }, { "epoch": 11.657062543921294, "grad_norm": 0.12146061658859253, "learning_rate": 2.556570625439213e-05, "loss": 0.0062, "step": 16588 }, { "epoch": 11.65776528460998, "grad_norm": 0.3412179946899414, "learning_rate": 2.556523776059967e-05, "loss": 0.0285, "step": 16589 }, { "epoch": 11.658468025298665, "grad_norm": 0.1918451339006424, "learning_rate": 2.5564769266807214e-05, "loss": 0.0243, "step": 16590 }, { "epoch": 11.659170765987351, "grad_norm": 0.19171486794948578, "learning_rate": 2.5564300773014758e-05, "loss": 0.0285, "step": 16591 }, { "epoch": 11.659873506676037, "grad_norm": 0.09364360570907593, "learning_rate": 2.55638322792223e-05, "loss": 0.0125, "step": 16592 }, { "epoch": 11.660576247364723, "grad_norm": 0.1624792069196701, "learning_rate": 2.5563363785429842e-05, "loss": 0.0323, "step": 16593 }, { "epoch": 11.66127898805341, "grad_norm": 0.4343149960041046, "learning_rate": 2.5562895291637386e-05, "loss": 0.0535, "step": 16594 }, { "epoch": 11.661981728742095, "grad_norm": 0.24013401567935944, "learning_rate": 2.556242679784493e-05, "loss": 0.0346, "step": 16595 }, { "epoch": 11.66268446943078, "grad_norm": 0.13137920200824738, "learning_rate": 2.5561958304052473e-05, "loss": 0.0255, "step": 16596 }, { "epoch": 11.663387210119465, "grad_norm": 0.6805097460746765, "learning_rate": 2.5561489810260014e-05, "loss": 0.0399, "step": 16597 }, { "epoch": 11.664089950808151, "grad_norm": 0.21633249521255493, "learning_rate": 2.5561021316467557e-05, "loss": 0.0441, "step": 16598 }, { "epoch": 11.664792691496837, "grad_norm": 0.21704596281051636, "learning_rate": 2.55605528226751e-05, "loss": 0.0688, "step": 16599 }, { "epoch": 11.665495432185523, "grad_norm": 0.3536650240421295, "learning_rate": 2.5560084328882645e-05, "loss": 0.1076, "step": 16600 }, { "epoch": 11.666198172874209, "grad_norm": 0.6645286083221436, "learning_rate": 2.5559615835090185e-05, "loss": 0.1072, "step": 16601 }, { "epoch": 11.666900913562895, "grad_norm": 0.6357182264328003, "learning_rate": 2.555914734129773e-05, "loss": 0.1753, "step": 16602 }, { "epoch": 11.66760365425158, "grad_norm": 1.7252395153045654, "learning_rate": 2.5558678847505273e-05, "loss": 0.1882, "step": 16603 }, { "epoch": 11.668306394940267, "grad_norm": 0.2396613508462906, "learning_rate": 2.5558210353712816e-05, "loss": 0.06, "step": 16604 }, { "epoch": 11.669009135628952, "grad_norm": 0.18199113011360168, "learning_rate": 2.5557741859920357e-05, "loss": 0.0296, "step": 16605 }, { "epoch": 11.669711876317638, "grad_norm": 0.14794236421585083, "learning_rate": 2.5557273366127897e-05, "loss": 0.0222, "step": 16606 }, { "epoch": 11.670414617006324, "grad_norm": 0.785233199596405, "learning_rate": 2.555680487233544e-05, "loss": 0.0359, "step": 16607 }, { "epoch": 11.67111735769501, "grad_norm": 0.17292505502700806, "learning_rate": 2.5556336378542985e-05, "loss": 0.0176, "step": 16608 }, { "epoch": 11.671820098383696, "grad_norm": 0.10755327343940735, "learning_rate": 2.555586788475053e-05, "loss": 0.0141, "step": 16609 }, { "epoch": 11.672522839072382, "grad_norm": 0.3736424148082733, "learning_rate": 2.555539939095807e-05, "loss": 0.0271, "step": 16610 }, { "epoch": 11.673225579761068, "grad_norm": 0.1571715921163559, "learning_rate": 2.5554930897165612e-05, "loss": 0.0295, "step": 16611 }, { "epoch": 11.673928320449754, "grad_norm": 0.13651549816131592, "learning_rate": 2.5554462403373156e-05, "loss": 0.0118, "step": 16612 }, { "epoch": 11.67463106113844, "grad_norm": 0.09288672357797623, "learning_rate": 2.55539939095807e-05, "loss": 0.0128, "step": 16613 }, { "epoch": 11.675333801827126, "grad_norm": 0.427103191614151, "learning_rate": 2.555352541578824e-05, "loss": 0.0223, "step": 16614 }, { "epoch": 11.676036542515812, "grad_norm": 0.19883884489536285, "learning_rate": 2.5553056921995784e-05, "loss": 0.0165, "step": 16615 }, { "epoch": 11.676739283204498, "grad_norm": 0.29805657267570496, "learning_rate": 2.5552588428203328e-05, "loss": 0.0275, "step": 16616 }, { "epoch": 11.677442023893184, "grad_norm": 0.11050590872764587, "learning_rate": 2.555211993441087e-05, "loss": 0.0147, "step": 16617 }, { "epoch": 11.67814476458187, "grad_norm": 0.22133949398994446, "learning_rate": 2.555165144061841e-05, "loss": 0.0351, "step": 16618 }, { "epoch": 11.678847505270555, "grad_norm": 0.1410233974456787, "learning_rate": 2.5551182946825955e-05, "loss": 0.0215, "step": 16619 }, { "epoch": 11.679550245959241, "grad_norm": 0.5907606482505798, "learning_rate": 2.55507144530335e-05, "loss": 0.0225, "step": 16620 }, { "epoch": 11.680252986647927, "grad_norm": 0.1492459625005722, "learning_rate": 2.5550245959241043e-05, "loss": 0.024, "step": 16621 }, { "epoch": 11.680955727336613, "grad_norm": 0.22048689424991608, "learning_rate": 2.5549777465448587e-05, "loss": 0.0358, "step": 16622 }, { "epoch": 11.681658468025299, "grad_norm": 0.2537413537502289, "learning_rate": 2.5549308971656124e-05, "loss": 0.0603, "step": 16623 }, { "epoch": 11.682361208713985, "grad_norm": 0.7074955701828003, "learning_rate": 2.5548840477863667e-05, "loss": 0.0666, "step": 16624 }, { "epoch": 11.683063949402671, "grad_norm": 0.9780179262161255, "learning_rate": 2.554837198407121e-05, "loss": 0.1282, "step": 16625 }, { "epoch": 11.683766690091357, "grad_norm": 0.4666306972503662, "learning_rate": 2.5547903490278755e-05, "loss": 0.1169, "step": 16626 }, { "epoch": 11.684469430780043, "grad_norm": 0.8696024417877197, "learning_rate": 2.5547434996486295e-05, "loss": 0.1866, "step": 16627 }, { "epoch": 11.685172171468729, "grad_norm": 0.9274701476097107, "learning_rate": 2.554696650269384e-05, "loss": 0.207, "step": 16628 }, { "epoch": 11.685874912157415, "grad_norm": 0.26229122281074524, "learning_rate": 2.5546498008901382e-05, "loss": 0.0663, "step": 16629 }, { "epoch": 11.6865776528461, "grad_norm": 0.13638822734355927, "learning_rate": 2.5546029515108926e-05, "loss": 0.0305, "step": 16630 }, { "epoch": 11.687280393534786, "grad_norm": 0.13288654386997223, "learning_rate": 2.5545561021316467e-05, "loss": 0.0195, "step": 16631 }, { "epoch": 11.687983134223472, "grad_norm": 0.1478557288646698, "learning_rate": 2.554509252752401e-05, "loss": 0.0237, "step": 16632 }, { "epoch": 11.688685874912156, "grad_norm": 0.07116306573152542, "learning_rate": 2.5544624033731554e-05, "loss": 0.0127, "step": 16633 }, { "epoch": 11.689388615600844, "grad_norm": 0.1043134406208992, "learning_rate": 2.5544155539939098e-05, "loss": 0.0099, "step": 16634 }, { "epoch": 11.690091356289528, "grad_norm": 0.07914597541093826, "learning_rate": 2.554368704614664e-05, "loss": 0.0161, "step": 16635 }, { "epoch": 11.690794096978214, "grad_norm": 0.11642919480800629, "learning_rate": 2.5543218552354182e-05, "loss": 0.0097, "step": 16636 }, { "epoch": 11.6914968376669, "grad_norm": 0.33205515146255493, "learning_rate": 2.5542750058561726e-05, "loss": 0.0182, "step": 16637 }, { "epoch": 11.692199578355586, "grad_norm": 0.10355202853679657, "learning_rate": 2.554228156476927e-05, "loss": 0.0127, "step": 16638 }, { "epoch": 11.692902319044272, "grad_norm": 0.8598862290382385, "learning_rate": 2.5541813070976813e-05, "loss": 0.0247, "step": 16639 }, { "epoch": 11.693605059732958, "grad_norm": 0.20833532512187958, "learning_rate": 2.5541344577184353e-05, "loss": 0.0208, "step": 16640 }, { "epoch": 11.694307800421644, "grad_norm": 0.19440700113773346, "learning_rate": 2.5540876083391894e-05, "loss": 0.023, "step": 16641 }, { "epoch": 11.69501054111033, "grad_norm": 0.12629878520965576, "learning_rate": 2.5540407589599437e-05, "loss": 0.0194, "step": 16642 }, { "epoch": 11.695713281799016, "grad_norm": 0.10526789724826813, "learning_rate": 2.553993909580698e-05, "loss": 0.0114, "step": 16643 }, { "epoch": 11.696416022487702, "grad_norm": 0.18084745109081268, "learning_rate": 2.553947060201452e-05, "loss": 0.0315, "step": 16644 }, { "epoch": 11.697118763176388, "grad_norm": 0.2577440142631531, "learning_rate": 2.5539002108222065e-05, "loss": 0.037, "step": 16645 }, { "epoch": 11.697821503865073, "grad_norm": 0.19849513471126556, "learning_rate": 2.553853361442961e-05, "loss": 0.028, "step": 16646 }, { "epoch": 11.69852424455376, "grad_norm": 0.32887697219848633, "learning_rate": 2.5538065120637153e-05, "loss": 0.0242, "step": 16647 }, { "epoch": 11.699226985242445, "grad_norm": 0.5057788491249084, "learning_rate": 2.5537596626844696e-05, "loss": 0.0683, "step": 16648 }, { "epoch": 11.699929725931131, "grad_norm": 0.3265496790409088, "learning_rate": 2.5537128133052237e-05, "loss": 0.0615, "step": 16649 }, { "epoch": 11.700632466619817, "grad_norm": 1.038973093032837, "learning_rate": 2.553665963925978e-05, "loss": 0.1015, "step": 16650 }, { "epoch": 11.701335207308503, "grad_norm": 0.6742702722549438, "learning_rate": 2.5536191145467324e-05, "loss": 0.1759, "step": 16651 }, { "epoch": 11.702037947997189, "grad_norm": 0.747299313545227, "learning_rate": 2.5535722651674868e-05, "loss": 0.1853, "step": 16652 }, { "epoch": 11.702740688685875, "grad_norm": 0.9425158500671387, "learning_rate": 2.5535254157882408e-05, "loss": 0.2298, "step": 16653 }, { "epoch": 11.70344342937456, "grad_norm": 0.15901298820972443, "learning_rate": 2.5534785664089952e-05, "loss": 0.06, "step": 16654 }, { "epoch": 11.704146170063247, "grad_norm": 0.08760005980730057, "learning_rate": 2.5534317170297496e-05, "loss": 0.0209, "step": 16655 }, { "epoch": 11.704848910751933, "grad_norm": 0.11815964430570602, "learning_rate": 2.553384867650504e-05, "loss": 0.0175, "step": 16656 }, { "epoch": 11.705551651440619, "grad_norm": 0.3479553461074829, "learning_rate": 2.553338018271258e-05, "loss": 0.0367, "step": 16657 }, { "epoch": 11.706254392129305, "grad_norm": 0.12358515709638596, "learning_rate": 2.553291168892012e-05, "loss": 0.0176, "step": 16658 }, { "epoch": 11.70695713281799, "grad_norm": 0.10126828402280807, "learning_rate": 2.5532443195127664e-05, "loss": 0.0098, "step": 16659 }, { "epoch": 11.707659873506676, "grad_norm": 0.1145879477262497, "learning_rate": 2.5531974701335207e-05, "loss": 0.0154, "step": 16660 }, { "epoch": 11.708362614195362, "grad_norm": 0.18824395537376404, "learning_rate": 2.553150620754275e-05, "loss": 0.0323, "step": 16661 }, { "epoch": 11.709065354884048, "grad_norm": 0.16361315548419952, "learning_rate": 2.553103771375029e-05, "loss": 0.0211, "step": 16662 }, { "epoch": 11.709768095572734, "grad_norm": 0.10521046817302704, "learning_rate": 2.5530569219957835e-05, "loss": 0.0144, "step": 16663 }, { "epoch": 11.71047083626142, "grad_norm": 0.14293193817138672, "learning_rate": 2.553010072616538e-05, "loss": 0.0262, "step": 16664 }, { "epoch": 11.711173576950106, "grad_norm": 0.1911434680223465, "learning_rate": 2.5529632232372923e-05, "loss": 0.0175, "step": 16665 }, { "epoch": 11.711876317638792, "grad_norm": 0.37459641695022583, "learning_rate": 2.5529163738580463e-05, "loss": 0.0365, "step": 16666 }, { "epoch": 11.712579058327478, "grad_norm": 0.23779194056987762, "learning_rate": 2.5528695244788007e-05, "loss": 0.0147, "step": 16667 }, { "epoch": 11.713281799016164, "grad_norm": 0.18997013568878174, "learning_rate": 2.552822675099555e-05, "loss": 0.0238, "step": 16668 }, { "epoch": 11.71398453970485, "grad_norm": 0.631173312664032, "learning_rate": 2.5527758257203094e-05, "loss": 0.0327, "step": 16669 }, { "epoch": 11.714687280393536, "grad_norm": 0.12770545482635498, "learning_rate": 2.5527289763410635e-05, "loss": 0.0141, "step": 16670 }, { "epoch": 11.715390021082221, "grad_norm": 0.2778374254703522, "learning_rate": 2.5526821269618178e-05, "loss": 0.038, "step": 16671 }, { "epoch": 11.716092761770906, "grad_norm": 0.30556416511535645, "learning_rate": 2.5526352775825722e-05, "loss": 0.0329, "step": 16672 }, { "epoch": 11.716795502459593, "grad_norm": 0.26571381092071533, "learning_rate": 2.5525884282033266e-05, "loss": 0.0397, "step": 16673 }, { "epoch": 11.717498243148277, "grad_norm": 0.4459531307220459, "learning_rate": 2.552541578824081e-05, "loss": 0.0666, "step": 16674 }, { "epoch": 11.718200983836963, "grad_norm": 0.58356112241745, "learning_rate": 2.552494729444835e-05, "loss": 0.0952, "step": 16675 }, { "epoch": 11.71890372452565, "grad_norm": 0.5791818499565125, "learning_rate": 2.552447880065589e-05, "loss": 0.1275, "step": 16676 }, { "epoch": 11.719606465214335, "grad_norm": 0.9414956569671631, "learning_rate": 2.5524010306863434e-05, "loss": 0.1591, "step": 16677 }, { "epoch": 11.720309205903021, "grad_norm": 1.0132181644439697, "learning_rate": 2.5523541813070978e-05, "loss": 0.2178, "step": 16678 }, { "epoch": 11.721011946591707, "grad_norm": 0.20740258693695068, "learning_rate": 2.5523073319278518e-05, "loss": 0.0547, "step": 16679 }, { "epoch": 11.721714687280393, "grad_norm": 0.13647732138633728, "learning_rate": 2.552260482548606e-05, "loss": 0.0257, "step": 16680 }, { "epoch": 11.722417427969079, "grad_norm": 0.15311922132968903, "learning_rate": 2.5522136331693605e-05, "loss": 0.0215, "step": 16681 }, { "epoch": 11.723120168657765, "grad_norm": 0.10678429156541824, "learning_rate": 2.552166783790115e-05, "loss": 0.0143, "step": 16682 }, { "epoch": 11.72382290934645, "grad_norm": 0.13662397861480713, "learning_rate": 2.552119934410869e-05, "loss": 0.0165, "step": 16683 }, { "epoch": 11.724525650035137, "grad_norm": 0.13219407200813293, "learning_rate": 2.5520730850316233e-05, "loss": 0.0114, "step": 16684 }, { "epoch": 11.725228390723823, "grad_norm": 0.105758436024189, "learning_rate": 2.5520262356523777e-05, "loss": 0.0149, "step": 16685 }, { "epoch": 11.725931131412509, "grad_norm": 0.14874570071697235, "learning_rate": 2.551979386273132e-05, "loss": 0.0164, "step": 16686 }, { "epoch": 11.726633872101194, "grad_norm": 0.24521033465862274, "learning_rate": 2.5519325368938864e-05, "loss": 0.018, "step": 16687 }, { "epoch": 11.72733661278988, "grad_norm": 0.1379794478416443, "learning_rate": 2.5518856875146405e-05, "loss": 0.0175, "step": 16688 }, { "epoch": 11.728039353478566, "grad_norm": 0.09377896040678024, "learning_rate": 2.551838838135395e-05, "loss": 0.018, "step": 16689 }, { "epoch": 11.728742094167252, "grad_norm": 0.1930302083492279, "learning_rate": 2.5517919887561492e-05, "loss": 0.0146, "step": 16690 }, { "epoch": 11.729444834855938, "grad_norm": 0.17936787009239197, "learning_rate": 2.5517451393769036e-05, "loss": 0.0381, "step": 16691 }, { "epoch": 11.730147575544624, "grad_norm": 0.11271140724420547, "learning_rate": 2.5516982899976576e-05, "loss": 0.0128, "step": 16692 }, { "epoch": 11.73085031623331, "grad_norm": 0.15976007282733917, "learning_rate": 2.5516514406184117e-05, "loss": 0.0314, "step": 16693 }, { "epoch": 11.731553056921996, "grad_norm": 0.4169343411922455, "learning_rate": 2.551604591239166e-05, "loss": 0.0332, "step": 16694 }, { "epoch": 11.732255797610682, "grad_norm": 0.11648917943239212, "learning_rate": 2.5515577418599204e-05, "loss": 0.0125, "step": 16695 }, { "epoch": 11.732958538299368, "grad_norm": 0.21101832389831543, "learning_rate": 2.5515108924806744e-05, "loss": 0.0475, "step": 16696 }, { "epoch": 11.733661278988054, "grad_norm": 0.18996135890483856, "learning_rate": 2.5514640431014288e-05, "loss": 0.0476, "step": 16697 }, { "epoch": 11.73436401967674, "grad_norm": 0.2832085192203522, "learning_rate": 2.5514171937221832e-05, "loss": 0.0481, "step": 16698 }, { "epoch": 11.735066760365426, "grad_norm": 0.26968297362327576, "learning_rate": 2.5513703443429375e-05, "loss": 0.064, "step": 16699 }, { "epoch": 11.735769501054111, "grad_norm": 0.2807481586933136, "learning_rate": 2.551323494963692e-05, "loss": 0.0842, "step": 16700 }, { "epoch": 11.736472241742797, "grad_norm": 1.0295958518981934, "learning_rate": 2.551276645584446e-05, "loss": 0.1292, "step": 16701 }, { "epoch": 11.737174982431483, "grad_norm": 0.7970324158668518, "learning_rate": 2.5512297962052003e-05, "loss": 0.1604, "step": 16702 }, { "epoch": 11.73787772312017, "grad_norm": 3.4591786861419678, "learning_rate": 2.5511829468259547e-05, "loss": 0.1752, "step": 16703 }, { "epoch": 11.738580463808855, "grad_norm": 0.17831382155418396, "learning_rate": 2.551136097446709e-05, "loss": 0.0508, "step": 16704 }, { "epoch": 11.739283204497541, "grad_norm": 0.3743182122707367, "learning_rate": 2.551089248067463e-05, "loss": 0.0209, "step": 16705 }, { "epoch": 11.739985945186227, "grad_norm": 0.2716158330440521, "learning_rate": 2.5510423986882175e-05, "loss": 0.0171, "step": 16706 }, { "epoch": 11.740688685874913, "grad_norm": 0.08664438128471375, "learning_rate": 2.550995549308972e-05, "loss": 0.0113, "step": 16707 }, { "epoch": 11.741391426563599, "grad_norm": 0.11234817653894424, "learning_rate": 2.5509486999297262e-05, "loss": 0.0123, "step": 16708 }, { "epoch": 11.742094167252285, "grad_norm": 0.12197640538215637, "learning_rate": 2.5509018505504803e-05, "loss": 0.016, "step": 16709 }, { "epoch": 11.74279690794097, "grad_norm": 0.09211434423923492, "learning_rate": 2.5508550011712343e-05, "loss": 0.0072, "step": 16710 }, { "epoch": 11.743499648629655, "grad_norm": 0.11952915042638779, "learning_rate": 2.5508081517919887e-05, "loss": 0.017, "step": 16711 }, { "epoch": 11.74420238931834, "grad_norm": 0.17465099692344666, "learning_rate": 2.550761302412743e-05, "loss": 0.0219, "step": 16712 }, { "epoch": 11.744905130007027, "grad_norm": 0.12804004549980164, "learning_rate": 2.5507144530334974e-05, "loss": 0.0115, "step": 16713 }, { "epoch": 11.745607870695713, "grad_norm": 0.2626453936100006, "learning_rate": 2.5506676036542514e-05, "loss": 0.0264, "step": 16714 }, { "epoch": 11.746310611384398, "grad_norm": 0.09037214517593384, "learning_rate": 2.5506207542750058e-05, "loss": 0.012, "step": 16715 }, { "epoch": 11.747013352073084, "grad_norm": 0.19992145895957947, "learning_rate": 2.5505739048957602e-05, "loss": 0.0234, "step": 16716 }, { "epoch": 11.74771609276177, "grad_norm": 0.19394677877426147, "learning_rate": 2.5505270555165146e-05, "loss": 0.0182, "step": 16717 }, { "epoch": 11.748418833450456, "grad_norm": 0.413448303937912, "learning_rate": 2.5504802061372686e-05, "loss": 0.018, "step": 16718 }, { "epoch": 11.749121574139142, "grad_norm": 0.233103945851326, "learning_rate": 2.550433356758023e-05, "loss": 0.03, "step": 16719 }, { "epoch": 11.749824314827828, "grad_norm": 0.12399274110794067, "learning_rate": 2.5503865073787773e-05, "loss": 0.0178, "step": 16720 }, { "epoch": 11.750527055516514, "grad_norm": 0.2787918448448181, "learning_rate": 2.5503396579995317e-05, "loss": 0.0316, "step": 16721 }, { "epoch": 11.7512297962052, "grad_norm": 0.2172432243824005, "learning_rate": 2.5502928086202857e-05, "loss": 0.0461, "step": 16722 }, { "epoch": 11.751932536893886, "grad_norm": 0.2636229991912842, "learning_rate": 2.55024595924104e-05, "loss": 0.0463, "step": 16723 }, { "epoch": 11.752635277582572, "grad_norm": 0.23304054141044617, "learning_rate": 2.5501991098617945e-05, "loss": 0.0575, "step": 16724 }, { "epoch": 11.753338018271258, "grad_norm": 2.219900369644165, "learning_rate": 2.550152260482549e-05, "loss": 0.0929, "step": 16725 }, { "epoch": 11.754040758959944, "grad_norm": 0.5137541890144348, "learning_rate": 2.5501054111033032e-05, "loss": 0.1426, "step": 16726 }, { "epoch": 11.75474349964863, "grad_norm": 0.7829116582870483, "learning_rate": 2.5500585617240573e-05, "loss": 0.1639, "step": 16727 }, { "epoch": 11.755446240337315, "grad_norm": 1.8146400451660156, "learning_rate": 2.5500117123448113e-05, "loss": 0.2262, "step": 16728 }, { "epoch": 11.756148981026001, "grad_norm": 0.21802546083927155, "learning_rate": 2.5499648629655657e-05, "loss": 0.0596, "step": 16729 }, { "epoch": 11.756851721714687, "grad_norm": 0.32598257064819336, "learning_rate": 2.54991801358632e-05, "loss": 0.0218, "step": 16730 }, { "epoch": 11.757554462403373, "grad_norm": 0.31696635484695435, "learning_rate": 2.549871164207074e-05, "loss": 0.0251, "step": 16731 }, { "epoch": 11.75825720309206, "grad_norm": 0.10001254081726074, "learning_rate": 2.5498243148278285e-05, "loss": 0.014, "step": 16732 }, { "epoch": 11.758959943780745, "grad_norm": 0.18877598643302917, "learning_rate": 2.5497774654485828e-05, "loss": 0.0164, "step": 16733 }, { "epoch": 11.759662684469431, "grad_norm": 0.27117717266082764, "learning_rate": 2.5497306160693372e-05, "loss": 0.0136, "step": 16734 }, { "epoch": 11.760365425158117, "grad_norm": 0.14145316183567047, "learning_rate": 2.5496837666900912e-05, "loss": 0.0193, "step": 16735 }, { "epoch": 11.761068165846803, "grad_norm": 0.17526397109031677, "learning_rate": 2.5496369173108456e-05, "loss": 0.0169, "step": 16736 }, { "epoch": 11.761770906535489, "grad_norm": 0.17544664442539215, "learning_rate": 2.5495900679316e-05, "loss": 0.0081, "step": 16737 }, { "epoch": 11.762473647224175, "grad_norm": 0.18959207832813263, "learning_rate": 2.5495432185523543e-05, "loss": 0.02, "step": 16738 }, { "epoch": 11.76317638791286, "grad_norm": 0.1256990283727646, "learning_rate": 2.5494963691731087e-05, "loss": 0.0221, "step": 16739 }, { "epoch": 11.763879128601546, "grad_norm": 0.1825517863035202, "learning_rate": 2.5494495197938628e-05, "loss": 0.0283, "step": 16740 }, { "epoch": 11.764581869290232, "grad_norm": 0.19260618090629578, "learning_rate": 2.549402670414617e-05, "loss": 0.0197, "step": 16741 }, { "epoch": 11.765284609978918, "grad_norm": 0.4609488844871521, "learning_rate": 2.5493558210353715e-05, "loss": 0.0167, "step": 16742 }, { "epoch": 11.765987350667604, "grad_norm": 0.2704358696937561, "learning_rate": 2.549308971656126e-05, "loss": 0.0145, "step": 16743 }, { "epoch": 11.76669009135629, "grad_norm": 0.1856154352426529, "learning_rate": 2.54926212227688e-05, "loss": 0.031, "step": 16744 }, { "epoch": 11.767392832044976, "grad_norm": 0.2688390016555786, "learning_rate": 2.549215272897634e-05, "loss": 0.0238, "step": 16745 }, { "epoch": 11.768095572733662, "grad_norm": 0.2253623604774475, "learning_rate": 2.5491684235183883e-05, "loss": 0.0291, "step": 16746 }, { "epoch": 11.768798313422348, "grad_norm": 0.2791750431060791, "learning_rate": 2.5491215741391427e-05, "loss": 0.0457, "step": 16747 }, { "epoch": 11.769501054111032, "grad_norm": 0.39772114157676697, "learning_rate": 2.549074724759897e-05, "loss": 0.0652, "step": 16748 }, { "epoch": 11.77020379479972, "grad_norm": 0.38846537470817566, "learning_rate": 2.549027875380651e-05, "loss": 0.0822, "step": 16749 }, { "epoch": 11.770906535488404, "grad_norm": 0.48103299736976624, "learning_rate": 2.5489810260014055e-05, "loss": 0.1064, "step": 16750 }, { "epoch": 11.77160927617709, "grad_norm": 0.6542701125144958, "learning_rate": 2.54893417662216e-05, "loss": 0.1431, "step": 16751 }, { "epoch": 11.772312016865776, "grad_norm": 0.8447095155715942, "learning_rate": 2.5488873272429142e-05, "loss": 0.1813, "step": 16752 }, { "epoch": 11.773014757554462, "grad_norm": 1.524601936340332, "learning_rate": 2.5488404778636682e-05, "loss": 0.2019, "step": 16753 }, { "epoch": 11.773717498243148, "grad_norm": 0.2714148759841919, "learning_rate": 2.5487936284844226e-05, "loss": 0.0553, "step": 16754 }, { "epoch": 11.774420238931834, "grad_norm": 0.3880786597728729, "learning_rate": 2.548746779105177e-05, "loss": 0.033, "step": 16755 }, { "epoch": 11.77512297962052, "grad_norm": 0.20828907191753387, "learning_rate": 2.5486999297259314e-05, "loss": 0.0218, "step": 16756 }, { "epoch": 11.775825720309205, "grad_norm": 0.08706528693437576, "learning_rate": 2.5486530803466854e-05, "loss": 0.0127, "step": 16757 }, { "epoch": 11.776528460997891, "grad_norm": 0.2697349786758423, "learning_rate": 2.5486062309674398e-05, "loss": 0.0245, "step": 16758 }, { "epoch": 11.777231201686577, "grad_norm": 0.0735410526394844, "learning_rate": 2.548559381588194e-05, "loss": 0.0065, "step": 16759 }, { "epoch": 11.777933942375263, "grad_norm": 0.11252603679895401, "learning_rate": 2.5485125322089485e-05, "loss": 0.0207, "step": 16760 }, { "epoch": 11.778636683063949, "grad_norm": 0.3095434904098511, "learning_rate": 2.5484656828297025e-05, "loss": 0.0203, "step": 16761 }, { "epoch": 11.779339423752635, "grad_norm": 0.12912940979003906, "learning_rate": 2.548418833450457e-05, "loss": 0.0191, "step": 16762 }, { "epoch": 11.780042164441321, "grad_norm": 0.09559831768274307, "learning_rate": 2.548371984071211e-05, "loss": 0.0118, "step": 16763 }, { "epoch": 11.780744905130007, "grad_norm": 0.1463533639907837, "learning_rate": 2.5483251346919653e-05, "loss": 0.023, "step": 16764 }, { "epoch": 11.781447645818693, "grad_norm": 0.08700130134820938, "learning_rate": 2.5482782853127197e-05, "loss": 0.0118, "step": 16765 }, { "epoch": 11.782150386507379, "grad_norm": 0.3886275291442871, "learning_rate": 2.5482314359334737e-05, "loss": 0.0263, "step": 16766 }, { "epoch": 11.782853127196065, "grad_norm": 0.09164790064096451, "learning_rate": 2.548184586554228e-05, "loss": 0.0116, "step": 16767 }, { "epoch": 11.78355586788475, "grad_norm": 0.3106815218925476, "learning_rate": 2.5481377371749825e-05, "loss": 0.0259, "step": 16768 }, { "epoch": 11.784258608573436, "grad_norm": 0.15115264058113098, "learning_rate": 2.548090887795737e-05, "loss": 0.0341, "step": 16769 }, { "epoch": 11.784961349262122, "grad_norm": 0.14586086571216583, "learning_rate": 2.548044038416491e-05, "loss": 0.0158, "step": 16770 }, { "epoch": 11.785664089950808, "grad_norm": 0.2752174437046051, "learning_rate": 2.5479971890372453e-05, "loss": 0.0387, "step": 16771 }, { "epoch": 11.786366830639494, "grad_norm": 0.17693254351615906, "learning_rate": 2.5479503396579996e-05, "loss": 0.037, "step": 16772 }, { "epoch": 11.78706957132818, "grad_norm": 0.2946535348892212, "learning_rate": 2.547903490278754e-05, "loss": 0.0567, "step": 16773 }, { "epoch": 11.787772312016866, "grad_norm": 0.5561230182647705, "learning_rate": 2.5478566408995084e-05, "loss": 0.0982, "step": 16774 }, { "epoch": 11.788475052705552, "grad_norm": 0.36487942934036255, "learning_rate": 2.5478097915202624e-05, "loss": 0.1154, "step": 16775 }, { "epoch": 11.789177793394238, "grad_norm": 1.1781561374664307, "learning_rate": 2.5477629421410168e-05, "loss": 0.1256, "step": 16776 }, { "epoch": 11.789880534082924, "grad_norm": 1.878255844116211, "learning_rate": 2.547716092761771e-05, "loss": 0.1479, "step": 16777 }, { "epoch": 11.79058327477161, "grad_norm": 1.2482554912567139, "learning_rate": 2.5476692433825255e-05, "loss": 0.1836, "step": 16778 }, { "epoch": 11.791286015460296, "grad_norm": 0.19231167435646057, "learning_rate": 2.5476223940032796e-05, "loss": 0.0793, "step": 16779 }, { "epoch": 11.791988756148982, "grad_norm": 0.5110605955123901, "learning_rate": 2.5475755446240336e-05, "loss": 0.0219, "step": 16780 }, { "epoch": 11.792691496837667, "grad_norm": 0.13344301283359528, "learning_rate": 2.547528695244788e-05, "loss": 0.0268, "step": 16781 }, { "epoch": 11.793394237526353, "grad_norm": 0.13396696746349335, "learning_rate": 2.5474818458655423e-05, "loss": 0.0198, "step": 16782 }, { "epoch": 11.79409697821504, "grad_norm": 0.08582428842782974, "learning_rate": 2.5474349964862964e-05, "loss": 0.0161, "step": 16783 }, { "epoch": 11.794799718903725, "grad_norm": 0.10623762011528015, "learning_rate": 2.5473881471070507e-05, "loss": 0.0155, "step": 16784 }, { "epoch": 11.795502459592411, "grad_norm": 0.10765279084444046, "learning_rate": 2.547341297727805e-05, "loss": 0.0194, "step": 16785 }, { "epoch": 11.796205200281097, "grad_norm": 0.09723082184791565, "learning_rate": 2.5472944483485595e-05, "loss": 0.0136, "step": 16786 }, { "epoch": 11.796907940969781, "grad_norm": 0.15927232801914215, "learning_rate": 2.547247598969314e-05, "loss": 0.0276, "step": 16787 }, { "epoch": 11.797610681658469, "grad_norm": 0.0625600665807724, "learning_rate": 2.547200749590068e-05, "loss": 0.0085, "step": 16788 }, { "epoch": 11.798313422347153, "grad_norm": 0.11263668537139893, "learning_rate": 2.5471539002108223e-05, "loss": 0.0234, "step": 16789 }, { "epoch": 11.799016163035839, "grad_norm": 0.2395634800195694, "learning_rate": 2.5471070508315766e-05, "loss": 0.0144, "step": 16790 }, { "epoch": 11.799718903724525, "grad_norm": 0.15303930640220642, "learning_rate": 2.547060201452331e-05, "loss": 0.0271, "step": 16791 }, { "epoch": 11.80042164441321, "grad_norm": 0.11738932877779007, "learning_rate": 2.547013352073085e-05, "loss": 0.0097, "step": 16792 }, { "epoch": 11.801124385101897, "grad_norm": 0.19592684507369995, "learning_rate": 2.5469665026938394e-05, "loss": 0.0204, "step": 16793 }, { "epoch": 11.801827125790583, "grad_norm": 0.22228367626667023, "learning_rate": 2.5469196533145938e-05, "loss": 0.0247, "step": 16794 }, { "epoch": 11.802529866479269, "grad_norm": 0.24840515851974487, "learning_rate": 2.546872803935348e-05, "loss": 0.0242, "step": 16795 }, { "epoch": 11.803232607167955, "grad_norm": 0.1400654911994934, "learning_rate": 2.5468259545561022e-05, "loss": 0.0207, "step": 16796 }, { "epoch": 11.80393534785664, "grad_norm": 0.21312539279460907, "learning_rate": 2.5467791051768566e-05, "loss": 0.047, "step": 16797 }, { "epoch": 11.804638088545326, "grad_norm": 0.5092722177505493, "learning_rate": 2.5467322557976106e-05, "loss": 0.0443, "step": 16798 }, { "epoch": 11.805340829234012, "grad_norm": 0.34618639945983887, "learning_rate": 2.546685406418365e-05, "loss": 0.0757, "step": 16799 }, { "epoch": 11.806043569922698, "grad_norm": 0.6720746755599976, "learning_rate": 2.5466385570391193e-05, "loss": 0.1229, "step": 16800 }, { "epoch": 11.806746310611384, "grad_norm": 0.4380231201648712, "learning_rate": 2.5465917076598734e-05, "loss": 0.1528, "step": 16801 }, { "epoch": 11.80744905130007, "grad_norm": 0.8164942860603333, "learning_rate": 2.5465448582806278e-05, "loss": 0.1452, "step": 16802 }, { "epoch": 11.808151791988756, "grad_norm": 1.0335530042648315, "learning_rate": 2.546498008901382e-05, "loss": 0.2051, "step": 16803 }, { "epoch": 11.808854532677442, "grad_norm": 0.18297624588012695, "learning_rate": 2.5464511595221365e-05, "loss": 0.0843, "step": 16804 }, { "epoch": 11.809557273366128, "grad_norm": 0.27591022849082947, "learning_rate": 2.5464043101428905e-05, "loss": 0.0177, "step": 16805 }, { "epoch": 11.810260014054814, "grad_norm": 0.09842165559530258, "learning_rate": 2.546357460763645e-05, "loss": 0.0183, "step": 16806 }, { "epoch": 11.8109627547435, "grad_norm": 0.3026145100593567, "learning_rate": 2.5463106113843993e-05, "loss": 0.0187, "step": 16807 }, { "epoch": 11.811665495432186, "grad_norm": 0.2856276333332062, "learning_rate": 2.5462637620051536e-05, "loss": 0.0105, "step": 16808 }, { "epoch": 11.812368236120872, "grad_norm": 0.2120019644498825, "learning_rate": 2.5462169126259077e-05, "loss": 0.0159, "step": 16809 }, { "epoch": 11.813070976809557, "grad_norm": 0.2829613983631134, "learning_rate": 2.546170063246662e-05, "loss": 0.0203, "step": 16810 }, { "epoch": 11.813773717498243, "grad_norm": 0.14168664813041687, "learning_rate": 2.5461232138674164e-05, "loss": 0.0208, "step": 16811 }, { "epoch": 11.81447645818693, "grad_norm": 0.13505813479423523, "learning_rate": 2.5460763644881708e-05, "loss": 0.0108, "step": 16812 }, { "epoch": 11.815179198875615, "grad_norm": 0.9603813290596008, "learning_rate": 2.5460295151089252e-05, "loss": 0.0331, "step": 16813 }, { "epoch": 11.815881939564301, "grad_norm": 0.0680113211274147, "learning_rate": 2.5459826657296792e-05, "loss": 0.008, "step": 16814 }, { "epoch": 11.816584680252987, "grad_norm": 0.20427820086479187, "learning_rate": 2.5459358163504332e-05, "loss": 0.0177, "step": 16815 }, { "epoch": 11.817287420941673, "grad_norm": 0.21295158565044403, "learning_rate": 2.5458889669711876e-05, "loss": 0.034, "step": 16816 }, { "epoch": 11.817990161630359, "grad_norm": 0.13086102902889252, "learning_rate": 2.545842117591942e-05, "loss": 0.0169, "step": 16817 }, { "epoch": 11.818692902319045, "grad_norm": 0.10417622327804565, "learning_rate": 2.545795268212696e-05, "loss": 0.0154, "step": 16818 }, { "epoch": 11.81939564300773, "grad_norm": 0.3933185338973999, "learning_rate": 2.5457484188334504e-05, "loss": 0.0254, "step": 16819 }, { "epoch": 11.820098383696417, "grad_norm": 0.14175572991371155, "learning_rate": 2.5457015694542048e-05, "loss": 0.0171, "step": 16820 }, { "epoch": 11.820801124385103, "grad_norm": 0.3371589183807373, "learning_rate": 2.545654720074959e-05, "loss": 0.0436, "step": 16821 }, { "epoch": 11.821503865073788, "grad_norm": 0.15079912543296814, "learning_rate": 2.5456078706957132e-05, "loss": 0.0272, "step": 16822 }, { "epoch": 11.822206605762474, "grad_norm": 0.1530991494655609, "learning_rate": 2.5455610213164675e-05, "loss": 0.0323, "step": 16823 }, { "epoch": 11.82290934645116, "grad_norm": 0.27608680725097656, "learning_rate": 2.545514171937222e-05, "loss": 0.0585, "step": 16824 }, { "epoch": 11.823612087139846, "grad_norm": 0.37110158801078796, "learning_rate": 2.5454673225579763e-05, "loss": 0.101, "step": 16825 }, { "epoch": 11.82431482782853, "grad_norm": 0.551073431968689, "learning_rate": 2.5454204731787307e-05, "loss": 0.1231, "step": 16826 }, { "epoch": 11.825017568517218, "grad_norm": 0.9376622438430786, "learning_rate": 2.5453736237994847e-05, "loss": 0.1685, "step": 16827 }, { "epoch": 11.825720309205902, "grad_norm": 1.2801889181137085, "learning_rate": 2.545326774420239e-05, "loss": 0.1785, "step": 16828 }, { "epoch": 11.826423049894588, "grad_norm": 0.21917657554149628, "learning_rate": 2.5452799250409934e-05, "loss": 0.0617, "step": 16829 }, { "epoch": 11.827125790583274, "grad_norm": 0.10967192053794861, "learning_rate": 2.5452330756617478e-05, "loss": 0.0249, "step": 16830 }, { "epoch": 11.82782853127196, "grad_norm": 0.5742688179016113, "learning_rate": 2.545186226282502e-05, "loss": 0.0237, "step": 16831 }, { "epoch": 11.828531271960646, "grad_norm": 0.10381476581096649, "learning_rate": 2.545139376903256e-05, "loss": 0.0136, "step": 16832 }, { "epoch": 11.829234012649332, "grad_norm": 0.07122374325990677, "learning_rate": 2.5450925275240103e-05, "loss": 0.0093, "step": 16833 }, { "epoch": 11.829936753338018, "grad_norm": 0.08984482288360596, "learning_rate": 2.5450456781447646e-05, "loss": 0.0068, "step": 16834 }, { "epoch": 11.830639494026704, "grad_norm": 0.17985309660434723, "learning_rate": 2.5449988287655187e-05, "loss": 0.0129, "step": 16835 }, { "epoch": 11.83134223471539, "grad_norm": 0.23349468410015106, "learning_rate": 2.544951979386273e-05, "loss": 0.0095, "step": 16836 }, { "epoch": 11.832044975404076, "grad_norm": 0.13173434138298035, "learning_rate": 2.5449051300070274e-05, "loss": 0.0251, "step": 16837 }, { "epoch": 11.832747716092761, "grad_norm": 0.1119716688990593, "learning_rate": 2.5448582806277818e-05, "loss": 0.0175, "step": 16838 }, { "epoch": 11.833450456781447, "grad_norm": 0.1792190968990326, "learning_rate": 2.544811431248536e-05, "loss": 0.022, "step": 16839 }, { "epoch": 11.834153197470133, "grad_norm": 0.1040559634566307, "learning_rate": 2.5447645818692902e-05, "loss": 0.015, "step": 16840 }, { "epoch": 11.83485593815882, "grad_norm": 0.246480792760849, "learning_rate": 2.5447177324900446e-05, "loss": 0.0175, "step": 16841 }, { "epoch": 11.835558678847505, "grad_norm": 0.08693663775920868, "learning_rate": 2.544670883110799e-05, "loss": 0.0145, "step": 16842 }, { "epoch": 11.836261419536191, "grad_norm": 0.13951823115348816, "learning_rate": 2.5446240337315533e-05, "loss": 0.0349, "step": 16843 }, { "epoch": 11.836964160224877, "grad_norm": 0.14233475923538208, "learning_rate": 2.5445771843523073e-05, "loss": 0.0199, "step": 16844 }, { "epoch": 11.837666900913563, "grad_norm": 0.17316490411758423, "learning_rate": 2.5445303349730617e-05, "loss": 0.0257, "step": 16845 }, { "epoch": 11.838369641602249, "grad_norm": 0.20891045033931732, "learning_rate": 2.544483485593816e-05, "loss": 0.0302, "step": 16846 }, { "epoch": 11.839072382290935, "grad_norm": 0.1949624866247177, "learning_rate": 2.5444366362145705e-05, "loss": 0.0187, "step": 16847 }, { "epoch": 11.83977512297962, "grad_norm": 0.27841123938560486, "learning_rate": 2.5443897868353245e-05, "loss": 0.0408, "step": 16848 }, { "epoch": 11.840477863668307, "grad_norm": 0.31398797035217285, "learning_rate": 2.544342937456079e-05, "loss": 0.0816, "step": 16849 }, { "epoch": 11.841180604356992, "grad_norm": 0.6378635168075562, "learning_rate": 2.544296088076833e-05, "loss": 0.1033, "step": 16850 }, { "epoch": 11.841883345045678, "grad_norm": 0.8630250692367554, "learning_rate": 2.5442492386975873e-05, "loss": 0.1485, "step": 16851 }, { "epoch": 11.842586085734364, "grad_norm": 0.8982089757919312, "learning_rate": 2.5442023893183416e-05, "loss": 0.1438, "step": 16852 }, { "epoch": 11.84328882642305, "grad_norm": 1.1280754804611206, "learning_rate": 2.5441555399390957e-05, "loss": 0.1875, "step": 16853 }, { "epoch": 11.843991567111736, "grad_norm": 0.2048802226781845, "learning_rate": 2.54410869055985e-05, "loss": 0.0637, "step": 16854 }, { "epoch": 11.844694307800422, "grad_norm": 0.17964385449886322, "learning_rate": 2.5440618411806044e-05, "loss": 0.0256, "step": 16855 }, { "epoch": 11.845397048489108, "grad_norm": 0.0889853984117508, "learning_rate": 2.5440149918013588e-05, "loss": 0.0152, "step": 16856 }, { "epoch": 11.846099789177794, "grad_norm": 0.25412923097610474, "learning_rate": 2.5439681424221128e-05, "loss": 0.0226, "step": 16857 }, { "epoch": 11.84680252986648, "grad_norm": 0.2770552635192871, "learning_rate": 2.5439212930428672e-05, "loss": 0.0127, "step": 16858 }, { "epoch": 11.847505270555166, "grad_norm": 0.1684289574623108, "learning_rate": 2.5438744436636216e-05, "loss": 0.0246, "step": 16859 }, { "epoch": 11.848208011243852, "grad_norm": 0.15216153860092163, "learning_rate": 2.543827594284376e-05, "loss": 0.0256, "step": 16860 }, { "epoch": 11.848910751932538, "grad_norm": 1.2065526247024536, "learning_rate": 2.54378074490513e-05, "loss": 0.0242, "step": 16861 }, { "epoch": 11.849613492621224, "grad_norm": 0.09607719630002975, "learning_rate": 2.5437338955258843e-05, "loss": 0.0134, "step": 16862 }, { "epoch": 11.85031623330991, "grad_norm": 0.07758746296167374, "learning_rate": 2.5436870461466387e-05, "loss": 0.0105, "step": 16863 }, { "epoch": 11.851018973998595, "grad_norm": 0.1850457787513733, "learning_rate": 2.543640196767393e-05, "loss": 0.0245, "step": 16864 }, { "epoch": 11.85172171468728, "grad_norm": 0.11705398559570312, "learning_rate": 2.5435933473881475e-05, "loss": 0.0211, "step": 16865 }, { "epoch": 11.852424455375965, "grad_norm": 0.15712107717990875, "learning_rate": 2.5435464980089015e-05, "loss": 0.0325, "step": 16866 }, { "epoch": 11.853127196064651, "grad_norm": 0.12891842424869537, "learning_rate": 2.5434996486296555e-05, "loss": 0.0161, "step": 16867 }, { "epoch": 11.853829936753337, "grad_norm": 0.5973770618438721, "learning_rate": 2.54345279925041e-05, "loss": 0.0357, "step": 16868 }, { "epoch": 11.854532677442023, "grad_norm": 0.4879818260669708, "learning_rate": 2.5434059498711643e-05, "loss": 0.032, "step": 16869 }, { "epoch": 11.85523541813071, "grad_norm": 0.1536475419998169, "learning_rate": 2.5433591004919183e-05, "loss": 0.0202, "step": 16870 }, { "epoch": 11.855938158819395, "grad_norm": 0.3193386197090149, "learning_rate": 2.5433122511126727e-05, "loss": 0.0336, "step": 16871 }, { "epoch": 11.856640899508081, "grad_norm": 0.1698041558265686, "learning_rate": 2.543265401733427e-05, "loss": 0.0325, "step": 16872 }, { "epoch": 11.857343640196767, "grad_norm": 0.2654719352722168, "learning_rate": 2.5432185523541814e-05, "loss": 0.0436, "step": 16873 }, { "epoch": 11.858046380885453, "grad_norm": 0.24095860123634338, "learning_rate": 2.5431717029749355e-05, "loss": 0.0518, "step": 16874 }, { "epoch": 11.858749121574139, "grad_norm": 0.3569384515285492, "learning_rate": 2.54312485359569e-05, "loss": 0.0996, "step": 16875 }, { "epoch": 11.859451862262825, "grad_norm": 0.4616755247116089, "learning_rate": 2.5430780042164442e-05, "loss": 0.1262, "step": 16876 }, { "epoch": 11.86015460295151, "grad_norm": 0.5498915910720825, "learning_rate": 2.5430311548371986e-05, "loss": 0.1676, "step": 16877 }, { "epoch": 11.860857343640197, "grad_norm": 0.8402672410011292, "learning_rate": 2.542984305457953e-05, "loss": 0.1904, "step": 16878 }, { "epoch": 11.861560084328882, "grad_norm": 0.2433534413576126, "learning_rate": 2.542937456078707e-05, "loss": 0.0735, "step": 16879 }, { "epoch": 11.862262825017568, "grad_norm": 0.15109819173812866, "learning_rate": 2.5428906066994614e-05, "loss": 0.0245, "step": 16880 }, { "epoch": 11.862965565706254, "grad_norm": 0.13450050354003906, "learning_rate": 2.5428437573202157e-05, "loss": 0.0314, "step": 16881 }, { "epoch": 11.86366830639494, "grad_norm": 0.16158868372440338, "learning_rate": 2.54279690794097e-05, "loss": 0.0185, "step": 16882 }, { "epoch": 11.864371047083626, "grad_norm": 0.15743908286094666, "learning_rate": 2.542750058561724e-05, "loss": 0.0201, "step": 16883 }, { "epoch": 11.865073787772312, "grad_norm": 0.053965479135513306, "learning_rate": 2.5427032091824785e-05, "loss": 0.0055, "step": 16884 }, { "epoch": 11.865776528460998, "grad_norm": 0.14952416718006134, "learning_rate": 2.5426563598032325e-05, "loss": 0.0128, "step": 16885 }, { "epoch": 11.866479269149684, "grad_norm": 0.07097707688808441, "learning_rate": 2.542609510423987e-05, "loss": 0.01, "step": 16886 }, { "epoch": 11.86718200983837, "grad_norm": 0.09990359097719193, "learning_rate": 2.542562661044741e-05, "loss": 0.0201, "step": 16887 }, { "epoch": 11.867884750527056, "grad_norm": 0.13023807108402252, "learning_rate": 2.5425158116654953e-05, "loss": 0.017, "step": 16888 }, { "epoch": 11.868587491215742, "grad_norm": 0.10841601341962814, "learning_rate": 2.5424689622862497e-05, "loss": 0.0204, "step": 16889 }, { "epoch": 11.869290231904428, "grad_norm": 0.13472606241703033, "learning_rate": 2.542422112907004e-05, "loss": 0.0099, "step": 16890 }, { "epoch": 11.869992972593113, "grad_norm": 0.22772525250911713, "learning_rate": 2.5423752635277584e-05, "loss": 0.0282, "step": 16891 }, { "epoch": 11.8706957132818, "grad_norm": 0.1278558373451233, "learning_rate": 2.5423284141485125e-05, "loss": 0.0238, "step": 16892 }, { "epoch": 11.871398453970485, "grad_norm": 0.07640431821346283, "learning_rate": 2.542281564769267e-05, "loss": 0.0132, "step": 16893 }, { "epoch": 11.872101194659171, "grad_norm": 0.2924892008304596, "learning_rate": 2.5422347153900212e-05, "loss": 0.0338, "step": 16894 }, { "epoch": 11.872803935347857, "grad_norm": 0.13960500061511993, "learning_rate": 2.5421878660107756e-05, "loss": 0.0363, "step": 16895 }, { "epoch": 11.873506676036543, "grad_norm": 0.2528298795223236, "learning_rate": 2.5421410166315296e-05, "loss": 0.0405, "step": 16896 }, { "epoch": 11.874209416725229, "grad_norm": 0.2006603181362152, "learning_rate": 2.542094167252284e-05, "loss": 0.0307, "step": 16897 }, { "epoch": 11.874912157413915, "grad_norm": 0.3005084693431854, "learning_rate": 2.5420473178730384e-05, "loss": 0.0494, "step": 16898 }, { "epoch": 11.8756148981026, "grad_norm": 0.46925896406173706, "learning_rate": 2.5420004684937927e-05, "loss": 0.043, "step": 16899 }, { "epoch": 11.876317638791287, "grad_norm": 0.8263867497444153, "learning_rate": 2.5419536191145468e-05, "loss": 0.098, "step": 16900 }, { "epoch": 11.877020379479973, "grad_norm": 0.44238725304603577, "learning_rate": 2.541906769735301e-05, "loss": 0.1568, "step": 16901 }, { "epoch": 11.877723120168657, "grad_norm": 1.8400908708572388, "learning_rate": 2.5418599203560552e-05, "loss": 0.1648, "step": 16902 }, { "epoch": 11.878425860857345, "grad_norm": 1.874725103378296, "learning_rate": 2.5418130709768096e-05, "loss": 0.2014, "step": 16903 }, { "epoch": 11.879128601546029, "grad_norm": 0.24192427098751068, "learning_rate": 2.541766221597564e-05, "loss": 0.0724, "step": 16904 }, { "epoch": 11.879831342234715, "grad_norm": 0.16180676221847534, "learning_rate": 2.541719372218318e-05, "loss": 0.0197, "step": 16905 }, { "epoch": 11.8805340829234, "grad_norm": 0.09778199344873428, "learning_rate": 2.5416725228390723e-05, "loss": 0.0155, "step": 16906 }, { "epoch": 11.881236823612086, "grad_norm": 0.1249760091304779, "learning_rate": 2.5416256734598267e-05, "loss": 0.018, "step": 16907 }, { "epoch": 11.881939564300772, "grad_norm": 0.14566127955913544, "learning_rate": 2.541578824080581e-05, "loss": 0.0149, "step": 16908 }, { "epoch": 11.882642304989458, "grad_norm": 0.0790444165468216, "learning_rate": 2.541531974701335e-05, "loss": 0.0063, "step": 16909 }, { "epoch": 11.883345045678144, "grad_norm": 0.21544942259788513, "learning_rate": 2.5414851253220895e-05, "loss": 0.0195, "step": 16910 }, { "epoch": 11.88404778636683, "grad_norm": 0.09885838627815247, "learning_rate": 2.541438275942844e-05, "loss": 0.0176, "step": 16911 }, { "epoch": 11.884750527055516, "grad_norm": 0.23108598589897156, "learning_rate": 2.5413914265635982e-05, "loss": 0.0368, "step": 16912 }, { "epoch": 11.885453267744202, "grad_norm": 0.08557355403900146, "learning_rate": 2.5413445771843523e-05, "loss": 0.0153, "step": 16913 }, { "epoch": 11.886156008432888, "grad_norm": 0.10592088848352432, "learning_rate": 2.5412977278051066e-05, "loss": 0.0306, "step": 16914 }, { "epoch": 11.886858749121574, "grad_norm": 0.09847532212734222, "learning_rate": 2.541250878425861e-05, "loss": 0.0098, "step": 16915 }, { "epoch": 11.88756148981026, "grad_norm": 0.1839814931154251, "learning_rate": 2.5412040290466154e-05, "loss": 0.0226, "step": 16916 }, { "epoch": 11.888264230498946, "grad_norm": 0.3624326288700104, "learning_rate": 2.5411571796673698e-05, "loss": 0.0306, "step": 16917 }, { "epoch": 11.888966971187632, "grad_norm": 0.09063968062400818, "learning_rate": 2.5411103302881238e-05, "loss": 0.0142, "step": 16918 }, { "epoch": 11.889669711876317, "grad_norm": 0.19640587270259857, "learning_rate": 2.5410634809088778e-05, "loss": 0.0319, "step": 16919 }, { "epoch": 11.890372452565003, "grad_norm": 0.16187672317028046, "learning_rate": 2.5410166315296322e-05, "loss": 0.0198, "step": 16920 }, { "epoch": 11.89107519325369, "grad_norm": 0.12472394853830338, "learning_rate": 2.5409697821503866e-05, "loss": 0.0226, "step": 16921 }, { "epoch": 11.891777933942375, "grad_norm": 0.19473019242286682, "learning_rate": 2.5409229327711406e-05, "loss": 0.0331, "step": 16922 }, { "epoch": 11.892480674631061, "grad_norm": 0.2515166699886322, "learning_rate": 2.540876083391895e-05, "loss": 0.0463, "step": 16923 }, { "epoch": 11.893183415319747, "grad_norm": 2.8646602630615234, "learning_rate": 2.5408292340126493e-05, "loss": 0.0587, "step": 16924 }, { "epoch": 11.893886156008433, "grad_norm": 0.44356048107147217, "learning_rate": 2.5407823846334037e-05, "loss": 0.1085, "step": 16925 }, { "epoch": 11.894588896697119, "grad_norm": 1.0078887939453125, "learning_rate": 2.5407355352541577e-05, "loss": 0.1497, "step": 16926 }, { "epoch": 11.895291637385805, "grad_norm": 1.1671428680419922, "learning_rate": 2.540688685874912e-05, "loss": 0.1716, "step": 16927 }, { "epoch": 11.89599437807449, "grad_norm": 1.1971027851104736, "learning_rate": 2.5406418364956665e-05, "loss": 0.1923, "step": 16928 }, { "epoch": 11.896697118763177, "grad_norm": 0.4485152065753937, "learning_rate": 2.540594987116421e-05, "loss": 0.068, "step": 16929 }, { "epoch": 11.897399859451863, "grad_norm": 0.1569640189409256, "learning_rate": 2.5405481377371752e-05, "loss": 0.0225, "step": 16930 }, { "epoch": 11.898102600140549, "grad_norm": 0.1145380288362503, "learning_rate": 2.5405012883579293e-05, "loss": 0.0185, "step": 16931 }, { "epoch": 11.898805340829234, "grad_norm": 0.1231112852692604, "learning_rate": 2.5404544389786836e-05, "loss": 0.0279, "step": 16932 }, { "epoch": 11.89950808151792, "grad_norm": 0.1121865063905716, "learning_rate": 2.540407589599438e-05, "loss": 0.016, "step": 16933 }, { "epoch": 11.900210822206606, "grad_norm": 0.0956224873661995, "learning_rate": 2.5403607402201924e-05, "loss": 0.0085, "step": 16934 }, { "epoch": 11.900913562895292, "grad_norm": 0.15504446625709534, "learning_rate": 2.5403138908409464e-05, "loss": 0.0172, "step": 16935 }, { "epoch": 11.901616303583978, "grad_norm": 0.15032336115837097, "learning_rate": 2.5402670414617008e-05, "loss": 0.0159, "step": 16936 }, { "epoch": 11.902319044272664, "grad_norm": 0.3021395206451416, "learning_rate": 2.5402201920824548e-05, "loss": 0.0243, "step": 16937 }, { "epoch": 11.90302178496135, "grad_norm": 0.13019564747810364, "learning_rate": 2.5401733427032092e-05, "loss": 0.022, "step": 16938 }, { "epoch": 11.903724525650036, "grad_norm": 0.6352089047431946, "learning_rate": 2.5401264933239636e-05, "loss": 0.0225, "step": 16939 }, { "epoch": 11.904427266338722, "grad_norm": 0.09534180909395218, "learning_rate": 2.5400796439447176e-05, "loss": 0.0102, "step": 16940 }, { "epoch": 11.905130007027406, "grad_norm": 0.11103449016809464, "learning_rate": 2.540032794565472e-05, "loss": 0.0207, "step": 16941 }, { "epoch": 11.905832747716094, "grad_norm": 0.06217774748802185, "learning_rate": 2.5399859451862264e-05, "loss": 0.011, "step": 16942 }, { "epoch": 11.906535488404778, "grad_norm": 0.137820303440094, "learning_rate": 2.5399390958069807e-05, "loss": 0.0297, "step": 16943 }, { "epoch": 11.907238229093464, "grad_norm": 0.242917999625206, "learning_rate": 2.5398922464277348e-05, "loss": 0.0259, "step": 16944 }, { "epoch": 11.90794096978215, "grad_norm": 0.48582544922828674, "learning_rate": 2.539845397048489e-05, "loss": 0.0126, "step": 16945 }, { "epoch": 11.908643710470836, "grad_norm": 0.22616054117679596, "learning_rate": 2.5397985476692435e-05, "loss": 0.0427, "step": 16946 }, { "epoch": 11.909346451159522, "grad_norm": 0.34094688296318054, "learning_rate": 2.539751698289998e-05, "loss": 0.0357, "step": 16947 }, { "epoch": 11.910049191848207, "grad_norm": 0.2716684639453888, "learning_rate": 2.539704848910752e-05, "loss": 0.0645, "step": 16948 }, { "epoch": 11.910751932536893, "grad_norm": 0.6645728349685669, "learning_rate": 2.5396579995315063e-05, "loss": 0.0756, "step": 16949 }, { "epoch": 11.91145467322558, "grad_norm": 0.5441806316375732, "learning_rate": 2.5396111501522607e-05, "loss": 0.0885, "step": 16950 }, { "epoch": 11.912157413914265, "grad_norm": 0.39100128412246704, "learning_rate": 2.539564300773015e-05, "loss": 0.13, "step": 16951 }, { "epoch": 11.912860154602951, "grad_norm": 0.8265925645828247, "learning_rate": 2.539517451393769e-05, "loss": 0.1757, "step": 16952 }, { "epoch": 11.913562895291637, "grad_norm": 0.9720310568809509, "learning_rate": 2.5394706020145234e-05, "loss": 0.1773, "step": 16953 }, { "epoch": 11.914265635980323, "grad_norm": 0.1695432811975479, "learning_rate": 2.5394237526352775e-05, "loss": 0.065, "step": 16954 }, { "epoch": 11.914968376669009, "grad_norm": 0.13938573002815247, "learning_rate": 2.539376903256032e-05, "loss": 0.0306, "step": 16955 }, { "epoch": 11.915671117357695, "grad_norm": 0.19407440721988678, "learning_rate": 2.5393300538767862e-05, "loss": 0.0155, "step": 16956 }, { "epoch": 11.91637385804638, "grad_norm": 0.09795796871185303, "learning_rate": 2.5392832044975402e-05, "loss": 0.0236, "step": 16957 }, { "epoch": 11.917076598735067, "grad_norm": 0.08384405076503754, "learning_rate": 2.5392363551182946e-05, "loss": 0.0126, "step": 16958 }, { "epoch": 11.917779339423753, "grad_norm": 0.12014775723218918, "learning_rate": 2.539189505739049e-05, "loss": 0.026, "step": 16959 }, { "epoch": 11.918482080112438, "grad_norm": 0.22150692343711853, "learning_rate": 2.5391426563598034e-05, "loss": 0.0176, "step": 16960 }, { "epoch": 11.919184820801124, "grad_norm": 0.16192235052585602, "learning_rate": 2.5390958069805574e-05, "loss": 0.0205, "step": 16961 }, { "epoch": 11.91988756148981, "grad_norm": 0.10625398904085159, "learning_rate": 2.5390489576013118e-05, "loss": 0.0147, "step": 16962 }, { "epoch": 11.920590302178496, "grad_norm": 0.11017915606498718, "learning_rate": 2.539002108222066e-05, "loss": 0.0099, "step": 16963 }, { "epoch": 11.921293042867182, "grad_norm": 0.2602790892124176, "learning_rate": 2.5389552588428205e-05, "loss": 0.0244, "step": 16964 }, { "epoch": 11.921995783555868, "grad_norm": 0.5281757116317749, "learning_rate": 2.538908409463575e-05, "loss": 0.0155, "step": 16965 }, { "epoch": 11.922698524244554, "grad_norm": 0.19015172123908997, "learning_rate": 2.538861560084329e-05, "loss": 0.0233, "step": 16966 }, { "epoch": 11.92340126493324, "grad_norm": 0.09974820166826248, "learning_rate": 2.5388147107050833e-05, "loss": 0.0065, "step": 16967 }, { "epoch": 11.924104005621926, "grad_norm": 0.17514559626579285, "learning_rate": 2.5387678613258377e-05, "loss": 0.0244, "step": 16968 }, { "epoch": 11.924806746310612, "grad_norm": 0.11749733984470367, "learning_rate": 2.538721011946592e-05, "loss": 0.0193, "step": 16969 }, { "epoch": 11.925509486999298, "grad_norm": 0.24492883682250977, "learning_rate": 2.538674162567346e-05, "loss": 0.0214, "step": 16970 }, { "epoch": 11.926212227687984, "grad_norm": 0.24852368235588074, "learning_rate": 2.5386273131881004e-05, "loss": 0.0326, "step": 16971 }, { "epoch": 11.92691496837667, "grad_norm": 0.2560189366340637, "learning_rate": 2.5385804638088545e-05, "loss": 0.0267, "step": 16972 }, { "epoch": 11.927617709065355, "grad_norm": 0.2858329713344574, "learning_rate": 2.538533614429609e-05, "loss": 0.0439, "step": 16973 }, { "epoch": 11.928320449754041, "grad_norm": 0.30599886178970337, "learning_rate": 2.538486765050363e-05, "loss": 0.0802, "step": 16974 }, { "epoch": 11.929023190442727, "grad_norm": 0.8377384543418884, "learning_rate": 2.5384399156711173e-05, "loss": 0.1069, "step": 16975 }, { "epoch": 11.929725931131413, "grad_norm": 0.42617443203926086, "learning_rate": 2.5383930662918716e-05, "loss": 0.1445, "step": 16976 }, { "epoch": 11.9304286718201, "grad_norm": 0.6888401508331299, "learning_rate": 2.538346216912626e-05, "loss": 0.177, "step": 16977 }, { "epoch": 11.931131412508785, "grad_norm": 1.0555082559585571, "learning_rate": 2.5382993675333804e-05, "loss": 0.194, "step": 16978 }, { "epoch": 11.931834153197471, "grad_norm": 0.4874127209186554, "learning_rate": 2.5382525181541344e-05, "loss": 0.0629, "step": 16979 }, { "epoch": 11.932536893886155, "grad_norm": 0.4126628637313843, "learning_rate": 2.5382056687748888e-05, "loss": 0.0226, "step": 16980 }, { "epoch": 11.933239634574843, "grad_norm": 0.1302282214164734, "learning_rate": 2.538158819395643e-05, "loss": 0.0177, "step": 16981 }, { "epoch": 11.933942375263527, "grad_norm": 0.09123246371746063, "learning_rate": 2.5381119700163975e-05, "loss": 0.0186, "step": 16982 }, { "epoch": 11.934645115952213, "grad_norm": 0.1145094558596611, "learning_rate": 2.5380651206371516e-05, "loss": 0.0281, "step": 16983 }, { "epoch": 11.935347856640899, "grad_norm": 0.15890975296497345, "learning_rate": 2.538018271257906e-05, "loss": 0.0228, "step": 16984 }, { "epoch": 11.936050597329585, "grad_norm": 0.16810861229896545, "learning_rate": 2.5379714218786603e-05, "loss": 0.011, "step": 16985 }, { "epoch": 11.93675333801827, "grad_norm": 0.2227092981338501, "learning_rate": 2.5379245724994147e-05, "loss": 0.0191, "step": 16986 }, { "epoch": 11.937456078706957, "grad_norm": 0.2701266407966614, "learning_rate": 2.5378777231201687e-05, "loss": 0.0324, "step": 16987 }, { "epoch": 11.938158819395642, "grad_norm": 0.11841876059770584, "learning_rate": 2.537830873740923e-05, "loss": 0.0084, "step": 16988 }, { "epoch": 11.938861560084328, "grad_norm": 0.17390255630016327, "learning_rate": 2.537784024361677e-05, "loss": 0.0287, "step": 16989 }, { "epoch": 11.939564300773014, "grad_norm": 0.1746884286403656, "learning_rate": 2.5377371749824315e-05, "loss": 0.013, "step": 16990 }, { "epoch": 11.9402670414617, "grad_norm": 0.2171437293291092, "learning_rate": 2.537690325603186e-05, "loss": 0.0312, "step": 16991 }, { "epoch": 11.940969782150386, "grad_norm": 0.09295309334993362, "learning_rate": 2.53764347622394e-05, "loss": 0.0178, "step": 16992 }, { "epoch": 11.941672522839072, "grad_norm": 0.17624536156654358, "learning_rate": 2.5375966268446943e-05, "loss": 0.0245, "step": 16993 }, { "epoch": 11.942375263527758, "grad_norm": 0.2985030710697174, "learning_rate": 2.5375497774654486e-05, "loss": 0.0455, "step": 16994 }, { "epoch": 11.943078004216444, "grad_norm": 0.1149185299873352, "learning_rate": 2.537502928086203e-05, "loss": 0.0175, "step": 16995 }, { "epoch": 11.94378074490513, "grad_norm": 0.6211010217666626, "learning_rate": 2.537456078706957e-05, "loss": 0.0426, "step": 16996 }, { "epoch": 11.944483485593816, "grad_norm": 0.21685871481895447, "learning_rate": 2.5374092293277114e-05, "loss": 0.05, "step": 16997 }, { "epoch": 11.945186226282502, "grad_norm": 0.39799806475639343, "learning_rate": 2.5373623799484658e-05, "loss": 0.0389, "step": 16998 }, { "epoch": 11.945888966971188, "grad_norm": 1.0030752420425415, "learning_rate": 2.53731553056922e-05, "loss": 0.0972, "step": 16999 }, { "epoch": 11.946591707659874, "grad_norm": 0.813326358795166, "learning_rate": 2.5372686811899742e-05, "loss": 0.1108, "step": 17000 }, { "epoch": 11.946591707659874, "eval_cer": 0.19344156765190446, "eval_loss": 0.2568434178829193, "eval_runtime": 18.6782, "eval_samples_per_second": 242.957, "eval_steps_per_second": 0.803, "eval_wer": 0.3444037876036982, "step": 17000 }, { "epoch": 11.94729444834856, "grad_norm": 0.9226441383361816, "learning_rate": 2.5372218318107286e-05, "loss": 0.1319, "step": 17001 }, { "epoch": 11.947997189037245, "grad_norm": 1.1209028959274292, "learning_rate": 2.537174982431483e-05, "loss": 0.183, "step": 17002 }, { "epoch": 11.948699929725931, "grad_norm": 0.974938154220581, "learning_rate": 2.5371281330522373e-05, "loss": 0.1813, "step": 17003 }, { "epoch": 11.949402670414617, "grad_norm": 0.2365744709968567, "learning_rate": 2.5370812836729917e-05, "loss": 0.0635, "step": 17004 }, { "epoch": 11.950105411103303, "grad_norm": 0.1516818255186081, "learning_rate": 2.5370344342937457e-05, "loss": 0.0174, "step": 17005 }, { "epoch": 11.950808151791989, "grad_norm": 0.30969852209091187, "learning_rate": 2.5369875849145e-05, "loss": 0.0165, "step": 17006 }, { "epoch": 11.951510892480675, "grad_norm": 0.09034895151853561, "learning_rate": 2.536940735535254e-05, "loss": 0.0144, "step": 17007 }, { "epoch": 11.952213633169361, "grad_norm": 0.16290949285030365, "learning_rate": 2.5368938861560085e-05, "loss": 0.0221, "step": 17008 }, { "epoch": 11.952916373858047, "grad_norm": 0.11266724020242691, "learning_rate": 2.5368470367767625e-05, "loss": 0.0124, "step": 17009 }, { "epoch": 11.953619114546733, "grad_norm": 0.16126155853271484, "learning_rate": 2.536800187397517e-05, "loss": 0.0081, "step": 17010 }, { "epoch": 11.954321855235419, "grad_norm": 0.1005125418305397, "learning_rate": 2.5367533380182713e-05, "loss": 0.0242, "step": 17011 }, { "epoch": 11.955024595924105, "grad_norm": 0.23411010205745697, "learning_rate": 2.5367064886390257e-05, "loss": 0.0308, "step": 17012 }, { "epoch": 11.95572733661279, "grad_norm": 0.12086363136768341, "learning_rate": 2.5366596392597797e-05, "loss": 0.022, "step": 17013 }, { "epoch": 11.956430077301476, "grad_norm": 0.14841899275779724, "learning_rate": 2.536612789880534e-05, "loss": 0.0222, "step": 17014 }, { "epoch": 11.957132817990162, "grad_norm": 0.12781471014022827, "learning_rate": 2.5365659405012884e-05, "loss": 0.0285, "step": 17015 }, { "epoch": 11.957835558678848, "grad_norm": 0.3631928563117981, "learning_rate": 2.5365190911220428e-05, "loss": 0.0229, "step": 17016 }, { "epoch": 11.958538299367534, "grad_norm": 0.5378190875053406, "learning_rate": 2.5364722417427972e-05, "loss": 0.0192, "step": 17017 }, { "epoch": 11.95924104005622, "grad_norm": 0.15211451053619385, "learning_rate": 2.5364253923635512e-05, "loss": 0.0138, "step": 17018 }, { "epoch": 11.959943780744904, "grad_norm": 0.18706445395946503, "learning_rate": 2.5363785429843056e-05, "loss": 0.026, "step": 17019 }, { "epoch": 11.96064652143359, "grad_norm": 0.16057619452476501, "learning_rate": 2.53633169360506e-05, "loss": 0.0213, "step": 17020 }, { "epoch": 11.961349262122276, "grad_norm": 0.1821940690279007, "learning_rate": 2.5362848442258143e-05, "loss": 0.0472, "step": 17021 }, { "epoch": 11.962052002810962, "grad_norm": 0.19180911779403687, "learning_rate": 2.5362379948465684e-05, "loss": 0.0443, "step": 17022 }, { "epoch": 11.962754743499648, "grad_norm": 0.23004968464374542, "learning_rate": 2.5361911454673227e-05, "loss": 0.0449, "step": 17023 }, { "epoch": 11.963457484188334, "grad_norm": 0.40378475189208984, "learning_rate": 2.5361442960880768e-05, "loss": 0.0734, "step": 17024 }, { "epoch": 11.96416022487702, "grad_norm": 0.45979607105255127, "learning_rate": 2.536097446708831e-05, "loss": 0.1213, "step": 17025 }, { "epoch": 11.964862965565706, "grad_norm": 0.860407829284668, "learning_rate": 2.5360505973295852e-05, "loss": 0.1284, "step": 17026 }, { "epoch": 11.965565706254392, "grad_norm": 1.2545465230941772, "learning_rate": 2.5360037479503395e-05, "loss": 0.1961, "step": 17027 }, { "epoch": 11.966268446943078, "grad_norm": 1.6692023277282715, "learning_rate": 2.535956898571094e-05, "loss": 0.2276, "step": 17028 }, { "epoch": 11.966971187631763, "grad_norm": 0.21718457341194153, "learning_rate": 2.5359100491918483e-05, "loss": 0.0642, "step": 17029 }, { "epoch": 11.96767392832045, "grad_norm": 0.12462334334850311, "learning_rate": 2.5358631998126027e-05, "loss": 0.0295, "step": 17030 }, { "epoch": 11.968376669009135, "grad_norm": 0.2369246780872345, "learning_rate": 2.5358163504333567e-05, "loss": 0.0258, "step": 17031 }, { "epoch": 11.969079409697821, "grad_norm": 0.35416126251220703, "learning_rate": 2.535769501054111e-05, "loss": 0.0171, "step": 17032 }, { "epoch": 11.969782150386507, "grad_norm": 0.48139140009880066, "learning_rate": 2.5357226516748654e-05, "loss": 0.0258, "step": 17033 }, { "epoch": 11.970484891075193, "grad_norm": 0.2580307722091675, "learning_rate": 2.5356758022956198e-05, "loss": 0.0135, "step": 17034 }, { "epoch": 11.971187631763879, "grad_norm": 0.1686822474002838, "learning_rate": 2.535628952916374e-05, "loss": 0.0135, "step": 17035 }, { "epoch": 11.971890372452565, "grad_norm": 0.09308740496635437, "learning_rate": 2.5355821035371282e-05, "loss": 0.0194, "step": 17036 }, { "epoch": 11.97259311314125, "grad_norm": 0.0995960533618927, "learning_rate": 2.5355352541578826e-05, "loss": 0.0131, "step": 17037 }, { "epoch": 11.973295853829937, "grad_norm": 0.09578315168619156, "learning_rate": 2.535488404778637e-05, "loss": 0.0106, "step": 17038 }, { "epoch": 11.973998594518623, "grad_norm": 0.19707341492176056, "learning_rate": 2.535441555399391e-05, "loss": 0.0198, "step": 17039 }, { "epoch": 11.974701335207309, "grad_norm": 0.2537609040737152, "learning_rate": 2.5353947060201454e-05, "loss": 0.0152, "step": 17040 }, { "epoch": 11.975404075895995, "grad_norm": 0.12615710496902466, "learning_rate": 2.5353478566408994e-05, "loss": 0.0235, "step": 17041 }, { "epoch": 11.97610681658468, "grad_norm": 0.17610952258110046, "learning_rate": 2.5353010072616538e-05, "loss": 0.021, "step": 17042 }, { "epoch": 11.976809557273366, "grad_norm": 0.36391347646713257, "learning_rate": 2.535254157882408e-05, "loss": 0.0345, "step": 17043 }, { "epoch": 11.977512297962052, "grad_norm": 0.257092148065567, "learning_rate": 2.5352073085031622e-05, "loss": 0.0262, "step": 17044 }, { "epoch": 11.978215038650738, "grad_norm": 0.18090881407260895, "learning_rate": 2.5351604591239166e-05, "loss": 0.0215, "step": 17045 }, { "epoch": 11.978917779339424, "grad_norm": 0.4947854280471802, "learning_rate": 2.535113609744671e-05, "loss": 0.0409, "step": 17046 }, { "epoch": 11.97962052002811, "grad_norm": 0.6372126340866089, "learning_rate": 2.5350667603654253e-05, "loss": 0.051, "step": 17047 }, { "epoch": 11.980323260716796, "grad_norm": 0.29413118958473206, "learning_rate": 2.5350199109861793e-05, "loss": 0.0418, "step": 17048 }, { "epoch": 11.981026001405482, "grad_norm": 0.39838966727256775, "learning_rate": 2.5349730616069337e-05, "loss": 0.0805, "step": 17049 }, { "epoch": 11.981728742094168, "grad_norm": 2.260450601577759, "learning_rate": 2.534926212227688e-05, "loss": 0.1002, "step": 17050 }, { "epoch": 11.982431482782854, "grad_norm": 0.681679904460907, "learning_rate": 2.5348793628484425e-05, "loss": 0.1196, "step": 17051 }, { "epoch": 11.98313422347154, "grad_norm": 1.490614891052246, "learning_rate": 2.5348325134691965e-05, "loss": 0.1959, "step": 17052 }, { "epoch": 11.983836964160226, "grad_norm": 1.3650236129760742, "learning_rate": 2.534785664089951e-05, "loss": 0.27, "step": 17053 }, { "epoch": 11.984539704848912, "grad_norm": 0.2887870669364929, "learning_rate": 2.5347388147107052e-05, "loss": 0.0611, "step": 17054 }, { "epoch": 11.985242445537597, "grad_norm": 0.07118618488311768, "learning_rate": 2.5346919653314596e-05, "loss": 0.0163, "step": 17055 }, { "epoch": 11.985945186226282, "grad_norm": 0.13195860385894775, "learning_rate": 2.534645115952214e-05, "loss": 0.0132, "step": 17056 }, { "epoch": 11.98664792691497, "grad_norm": 0.2480442374944687, "learning_rate": 2.534598266572968e-05, "loss": 0.0348, "step": 17057 }, { "epoch": 11.987350667603653, "grad_norm": 0.15550918877124786, "learning_rate": 2.5345514171937224e-05, "loss": 0.0057, "step": 17058 }, { "epoch": 11.98805340829234, "grad_norm": 0.12770743668079376, "learning_rate": 2.5345045678144764e-05, "loss": 0.0234, "step": 17059 }, { "epoch": 11.988756148981025, "grad_norm": 0.2042144536972046, "learning_rate": 2.5344577184352308e-05, "loss": 0.0238, "step": 17060 }, { "epoch": 11.989458889669711, "grad_norm": 0.20582956075668335, "learning_rate": 2.5344108690559848e-05, "loss": 0.0197, "step": 17061 }, { "epoch": 11.990161630358397, "grad_norm": 0.094959557056427, "learning_rate": 2.5343640196767392e-05, "loss": 0.0082, "step": 17062 }, { "epoch": 11.990864371047083, "grad_norm": 0.3050116002559662, "learning_rate": 2.5343171702974936e-05, "loss": 0.0173, "step": 17063 }, { "epoch": 11.991567111735769, "grad_norm": 0.10648827999830246, "learning_rate": 2.534270320918248e-05, "loss": 0.021, "step": 17064 }, { "epoch": 11.992269852424455, "grad_norm": 0.3664511740207672, "learning_rate": 2.534223471539002e-05, "loss": 0.0183, "step": 17065 }, { "epoch": 11.99297259311314, "grad_norm": 0.4137715697288513, "learning_rate": 2.5341766221597563e-05, "loss": 0.019, "step": 17066 }, { "epoch": 11.993675333801827, "grad_norm": 0.18360501527786255, "learning_rate": 2.5341297727805107e-05, "loss": 0.023, "step": 17067 }, { "epoch": 11.994378074490513, "grad_norm": 0.2830834984779358, "learning_rate": 2.534082923401265e-05, "loss": 0.0193, "step": 17068 }, { "epoch": 11.995080815179199, "grad_norm": 0.4545471668243408, "learning_rate": 2.5340360740220195e-05, "loss": 0.0247, "step": 17069 }, { "epoch": 11.995783555867884, "grad_norm": 0.21946527063846588, "learning_rate": 2.5339892246427735e-05, "loss": 0.0376, "step": 17070 }, { "epoch": 11.99648629655657, "grad_norm": 1.099852204322815, "learning_rate": 2.533942375263528e-05, "loss": 0.0598, "step": 17071 }, { "epoch": 11.997189037245256, "grad_norm": 1.4738523960113525, "learning_rate": 2.5338955258842822e-05, "loss": 0.0777, "step": 17072 }, { "epoch": 11.997891777933942, "grad_norm": 3.1807873249053955, "learning_rate": 2.5338486765050366e-05, "loss": 0.1066, "step": 17073 }, { "epoch": 11.998594518622628, "grad_norm": 0.6516090035438538, "learning_rate": 2.5338018271257907e-05, "loss": 0.1559, "step": 17074 }, { "epoch": 11.999297259311314, "grad_norm": 0.519162118434906, "learning_rate": 2.533754977746545e-05, "loss": 0.16, "step": 17075 }, { "epoch": 12.0, "grad_norm": 1.651314616203308, "learning_rate": 2.533708128367299e-05, "loss": 0.1366, "step": 17076 }, { "epoch": 12.000702740688686, "grad_norm": 0.527059018611908, "learning_rate": 2.5336612789880534e-05, "loss": 0.0584, "step": 17077 }, { "epoch": 12.001405481377372, "grad_norm": 0.1309351772069931, "learning_rate": 2.5336144296088075e-05, "loss": 0.019, "step": 17078 }, { "epoch": 12.002108222066058, "grad_norm": 0.18461857736110687, "learning_rate": 2.533567580229562e-05, "loss": 0.0354, "step": 17079 }, { "epoch": 12.002810962754744, "grad_norm": 0.0788215696811676, "learning_rate": 2.5335207308503162e-05, "loss": 0.0117, "step": 17080 }, { "epoch": 12.00351370344343, "grad_norm": 0.46820732951164246, "learning_rate": 2.5334738814710706e-05, "loss": 0.0161, "step": 17081 }, { "epoch": 12.004216444132116, "grad_norm": 0.08761020749807358, "learning_rate": 2.533427032091825e-05, "loss": 0.0113, "step": 17082 }, { "epoch": 12.004919184820801, "grad_norm": 0.12387120723724365, "learning_rate": 2.533380182712579e-05, "loss": 0.0205, "step": 17083 }, { "epoch": 12.005621925509487, "grad_norm": 0.09174603968858719, "learning_rate": 2.5333333333333334e-05, "loss": 0.0216, "step": 17084 }, { "epoch": 12.006324666198173, "grad_norm": 0.10747133195400238, "learning_rate": 2.5332864839540877e-05, "loss": 0.013, "step": 17085 }, { "epoch": 12.00702740688686, "grad_norm": 0.12458964437246323, "learning_rate": 2.533239634574842e-05, "loss": 0.0108, "step": 17086 }, { "epoch": 12.007730147575545, "grad_norm": 0.11706788092851639, "learning_rate": 2.533192785195596e-05, "loss": 0.0183, "step": 17087 }, { "epoch": 12.008432888264231, "grad_norm": 0.10424073040485382, "learning_rate": 2.5331459358163505e-05, "loss": 0.0127, "step": 17088 }, { "epoch": 12.009135628952917, "grad_norm": 0.23419998586177826, "learning_rate": 2.533099086437105e-05, "loss": 0.027, "step": 17089 }, { "epoch": 12.009838369641603, "grad_norm": 0.19101788103580475, "learning_rate": 2.5330522370578593e-05, "loss": 0.0173, "step": 17090 }, { "epoch": 12.010541110330289, "grad_norm": 0.15161892771720886, "learning_rate": 2.5330053876786133e-05, "loss": 0.0281, "step": 17091 }, { "epoch": 12.011243851018975, "grad_norm": 0.15569813549518585, "learning_rate": 2.5329585382993677e-05, "loss": 0.0341, "step": 17092 }, { "epoch": 12.01194659170766, "grad_norm": 0.20812474191188812, "learning_rate": 2.532911688920122e-05, "loss": 0.0238, "step": 17093 }, { "epoch": 12.012649332396347, "grad_norm": 0.3646785616874695, "learning_rate": 2.532864839540876e-05, "loss": 0.037, "step": 17094 }, { "epoch": 12.013352073085033, "grad_norm": 1.0656492710113525, "learning_rate": 2.5328179901616304e-05, "loss": 0.0254, "step": 17095 }, { "epoch": 12.014054813773717, "grad_norm": 0.25330471992492676, "learning_rate": 2.5327711407823845e-05, "loss": 0.0585, "step": 17096 }, { "epoch": 12.014757554462403, "grad_norm": 0.2909075915813446, "learning_rate": 2.532724291403139e-05, "loss": 0.0688, "step": 17097 }, { "epoch": 12.015460295151088, "grad_norm": 1.3688024282455444, "learning_rate": 2.5326774420238932e-05, "loss": 0.0861, "step": 17098 }, { "epoch": 12.016163035839774, "grad_norm": 0.6976176500320435, "learning_rate": 2.5326305926446476e-05, "loss": 0.1477, "step": 17099 }, { "epoch": 12.01686577652846, "grad_norm": 0.6226727962493896, "learning_rate": 2.5325837432654016e-05, "loss": 0.1552, "step": 17100 }, { "epoch": 12.017568517217146, "grad_norm": 1.1538052558898926, "learning_rate": 2.532536893886156e-05, "loss": 0.1993, "step": 17101 }, { "epoch": 12.018271257905832, "grad_norm": 0.19307343661785126, "learning_rate": 2.5324900445069104e-05, "loss": 0.0651, "step": 17102 }, { "epoch": 12.018973998594518, "grad_norm": 0.17894501984119415, "learning_rate": 2.5324431951276647e-05, "loss": 0.0445, "step": 17103 }, { "epoch": 12.019676739283204, "grad_norm": 0.13520562648773193, "learning_rate": 2.5323963457484188e-05, "loss": 0.0373, "step": 17104 }, { "epoch": 12.02037947997189, "grad_norm": 0.1130855455994606, "learning_rate": 2.532349496369173e-05, "loss": 0.0132, "step": 17105 }, { "epoch": 12.021082220660576, "grad_norm": 0.21108070015907288, "learning_rate": 2.5323026469899275e-05, "loss": 0.0176, "step": 17106 }, { "epoch": 12.021784961349262, "grad_norm": 0.16064634919166565, "learning_rate": 2.532255797610682e-05, "loss": 0.0129, "step": 17107 }, { "epoch": 12.022487702037948, "grad_norm": 0.13154567778110504, "learning_rate": 2.5322089482314363e-05, "loss": 0.0132, "step": 17108 }, { "epoch": 12.023190442726634, "grad_norm": 0.23282475769519806, "learning_rate": 2.5321620988521903e-05, "loss": 0.0294, "step": 17109 }, { "epoch": 12.02389318341532, "grad_norm": 0.15148605406284332, "learning_rate": 2.5321152494729447e-05, "loss": 0.0222, "step": 17110 }, { "epoch": 12.024595924104005, "grad_norm": 0.2350718379020691, "learning_rate": 2.5320684000936987e-05, "loss": 0.0192, "step": 17111 }, { "epoch": 12.025298664792691, "grad_norm": 0.1514674723148346, "learning_rate": 2.532021550714453e-05, "loss": 0.0173, "step": 17112 }, { "epoch": 12.026001405481377, "grad_norm": 0.14318592846393585, "learning_rate": 2.531974701335207e-05, "loss": 0.0207, "step": 17113 }, { "epoch": 12.026704146170063, "grad_norm": 0.1817423701286316, "learning_rate": 2.5319278519559615e-05, "loss": 0.0239, "step": 17114 }, { "epoch": 12.02740688685875, "grad_norm": 0.1138700470328331, "learning_rate": 2.531881002576716e-05, "loss": 0.0134, "step": 17115 }, { "epoch": 12.028109627547435, "grad_norm": 0.12689970433712006, "learning_rate": 2.5318341531974702e-05, "loss": 0.0233, "step": 17116 }, { "epoch": 12.028812368236121, "grad_norm": 0.3743530809879303, "learning_rate": 2.5317873038182243e-05, "loss": 0.0383, "step": 17117 }, { "epoch": 12.029515108924807, "grad_norm": 0.1590162068605423, "learning_rate": 2.5317404544389786e-05, "loss": 0.0121, "step": 17118 }, { "epoch": 12.030217849613493, "grad_norm": 0.21732981503009796, "learning_rate": 2.531693605059733e-05, "loss": 0.03, "step": 17119 }, { "epoch": 12.030920590302179, "grad_norm": 0.26262685656547546, "learning_rate": 2.5316467556804874e-05, "loss": 0.036, "step": 17120 }, { "epoch": 12.031623330990865, "grad_norm": 0.49242711067199707, "learning_rate": 2.5315999063012418e-05, "loss": 0.0585, "step": 17121 }, { "epoch": 12.03232607167955, "grad_norm": 0.27721381187438965, "learning_rate": 2.5315530569219958e-05, "loss": 0.0604, "step": 17122 }, { "epoch": 12.033028812368237, "grad_norm": 0.5348402261734009, "learning_rate": 2.53150620754275e-05, "loss": 0.1109, "step": 17123 }, { "epoch": 12.033731553056922, "grad_norm": 0.6155436038970947, "learning_rate": 2.5314593581635045e-05, "loss": 0.1405, "step": 17124 }, { "epoch": 12.034434293745608, "grad_norm": 0.8195634484291077, "learning_rate": 2.531412508784259e-05, "loss": 0.1866, "step": 17125 }, { "epoch": 12.035137034434294, "grad_norm": 5.392067909240723, "learning_rate": 2.531365659405013e-05, "loss": 0.1983, "step": 17126 }, { "epoch": 12.03583977512298, "grad_norm": 0.15787778794765472, "learning_rate": 2.5313188100257673e-05, "loss": 0.0489, "step": 17127 }, { "epoch": 12.036542515811666, "grad_norm": 0.2131349742412567, "learning_rate": 2.5312719606465213e-05, "loss": 0.033, "step": 17128 }, { "epoch": 12.037245256500352, "grad_norm": 0.06983647495508194, "learning_rate": 2.5312251112672757e-05, "loss": 0.0134, "step": 17129 }, { "epoch": 12.037947997189038, "grad_norm": 0.30098360776901245, "learning_rate": 2.5311782618880298e-05, "loss": 0.0191, "step": 17130 }, { "epoch": 12.038650737877724, "grad_norm": 0.18412184715270996, "learning_rate": 2.531131412508784e-05, "loss": 0.0169, "step": 17131 }, { "epoch": 12.03935347856641, "grad_norm": 0.09349480271339417, "learning_rate": 2.5310845631295385e-05, "loss": 0.0088, "step": 17132 }, { "epoch": 12.040056219255096, "grad_norm": 0.205882266163826, "learning_rate": 2.531037713750293e-05, "loss": 0.0177, "step": 17133 }, { "epoch": 12.04075895994378, "grad_norm": 0.3087867200374603, "learning_rate": 2.5309908643710472e-05, "loss": 0.0185, "step": 17134 }, { "epoch": 12.041461700632466, "grad_norm": 0.11242403090000153, "learning_rate": 2.5309440149918013e-05, "loss": 0.0154, "step": 17135 }, { "epoch": 12.042164441321152, "grad_norm": 0.07975919544696808, "learning_rate": 2.5308971656125556e-05, "loss": 0.0124, "step": 17136 }, { "epoch": 12.042867182009838, "grad_norm": 0.14937478303909302, "learning_rate": 2.53085031623331e-05, "loss": 0.0248, "step": 17137 }, { "epoch": 12.043569922698524, "grad_norm": 0.11953487992286682, "learning_rate": 2.5308034668540644e-05, "loss": 0.0171, "step": 17138 }, { "epoch": 12.04427266338721, "grad_norm": 0.4298396110534668, "learning_rate": 2.5307566174748184e-05, "loss": 0.0334, "step": 17139 }, { "epoch": 12.044975404075895, "grad_norm": 0.1649172008037567, "learning_rate": 2.5307097680955728e-05, "loss": 0.0103, "step": 17140 }, { "epoch": 12.045678144764581, "grad_norm": 0.13476645946502686, "learning_rate": 2.5306629187163272e-05, "loss": 0.0257, "step": 17141 }, { "epoch": 12.046380885453267, "grad_norm": 0.12203872203826904, "learning_rate": 2.5306160693370815e-05, "loss": 0.0293, "step": 17142 }, { "epoch": 12.047083626141953, "grad_norm": 0.18614737689495087, "learning_rate": 2.5305692199578356e-05, "loss": 0.0217, "step": 17143 }, { "epoch": 12.047786366830639, "grad_norm": 0.1934763640165329, "learning_rate": 2.53052237057859e-05, "loss": 0.0302, "step": 17144 }, { "epoch": 12.048489107519325, "grad_norm": 0.2752857804298401, "learning_rate": 2.5304755211993443e-05, "loss": 0.0339, "step": 17145 }, { "epoch": 12.049191848208011, "grad_norm": 0.25549083948135376, "learning_rate": 2.5304286718200984e-05, "loss": 0.0388, "step": 17146 }, { "epoch": 12.049894588896697, "grad_norm": 0.31559816002845764, "learning_rate": 2.5303818224408527e-05, "loss": 0.0737, "step": 17147 }, { "epoch": 12.050597329585383, "grad_norm": 0.36324596405029297, "learning_rate": 2.5303349730616068e-05, "loss": 0.0786, "step": 17148 }, { "epoch": 12.051300070274069, "grad_norm": 0.48957788944244385, "learning_rate": 2.530288123682361e-05, "loss": 0.1291, "step": 17149 }, { "epoch": 12.052002810962755, "grad_norm": 0.8007259368896484, "learning_rate": 2.5302412743031155e-05, "loss": 0.1715, "step": 17150 }, { "epoch": 12.05270555165144, "grad_norm": 1.7226510047912598, "learning_rate": 2.53019442492387e-05, "loss": 0.2338, "step": 17151 }, { "epoch": 12.053408292340126, "grad_norm": 0.35372817516326904, "learning_rate": 2.530147575544624e-05, "loss": 0.0518, "step": 17152 }, { "epoch": 12.054111033028812, "grad_norm": 0.10184646397829056, "learning_rate": 2.5301007261653783e-05, "loss": 0.024, "step": 17153 }, { "epoch": 12.054813773717498, "grad_norm": 0.15960152447223663, "learning_rate": 2.5300538767861327e-05, "loss": 0.0147, "step": 17154 }, { "epoch": 12.055516514406184, "grad_norm": 0.21251381933689117, "learning_rate": 2.530007027406887e-05, "loss": 0.0195, "step": 17155 }, { "epoch": 12.05621925509487, "grad_norm": 0.20211918652057648, "learning_rate": 2.529960178027641e-05, "loss": 0.0251, "step": 17156 }, { "epoch": 12.056921995783556, "grad_norm": 0.06387801468372345, "learning_rate": 2.5299133286483954e-05, "loss": 0.0061, "step": 17157 }, { "epoch": 12.057624736472242, "grad_norm": 0.12911204993724823, "learning_rate": 2.5298664792691498e-05, "loss": 0.0208, "step": 17158 }, { "epoch": 12.058327477160928, "grad_norm": 0.1279374659061432, "learning_rate": 2.5298196298899042e-05, "loss": 0.0161, "step": 17159 }, { "epoch": 12.059030217849614, "grad_norm": 0.12379968911409378, "learning_rate": 2.5297727805106586e-05, "loss": 0.0157, "step": 17160 }, { "epoch": 12.0597329585383, "grad_norm": 0.12915705144405365, "learning_rate": 2.5297259311314126e-05, "loss": 0.0156, "step": 17161 }, { "epoch": 12.060435699226986, "grad_norm": 0.22081761062145233, "learning_rate": 2.529679081752167e-05, "loss": 0.0253, "step": 17162 }, { "epoch": 12.061138439915672, "grad_norm": 0.08945410698652267, "learning_rate": 2.529632232372921e-05, "loss": 0.0094, "step": 17163 }, { "epoch": 12.061841180604358, "grad_norm": 0.12773825228214264, "learning_rate": 2.5295853829936754e-05, "loss": 0.0212, "step": 17164 }, { "epoch": 12.062543921293043, "grad_norm": 0.10673568397760391, "learning_rate": 2.5295385336144294e-05, "loss": 0.0089, "step": 17165 }, { "epoch": 12.06324666198173, "grad_norm": 0.31554022431373596, "learning_rate": 2.5294916842351838e-05, "loss": 0.0285, "step": 17166 }, { "epoch": 12.063949402670415, "grad_norm": 0.17582106590270996, "learning_rate": 2.529444834855938e-05, "loss": 0.0255, "step": 17167 }, { "epoch": 12.064652143359101, "grad_norm": 0.21842734515666962, "learning_rate": 2.5293979854766925e-05, "loss": 0.0318, "step": 17168 }, { "epoch": 12.065354884047787, "grad_norm": 0.10685747861862183, "learning_rate": 2.529351136097447e-05, "loss": 0.0177, "step": 17169 }, { "epoch": 12.066057624736473, "grad_norm": 0.22370512783527374, "learning_rate": 2.529304286718201e-05, "loss": 0.0538, "step": 17170 }, { "epoch": 12.066760365425159, "grad_norm": 0.2528049647808075, "learning_rate": 2.5292574373389553e-05, "loss": 0.0456, "step": 17171 }, { "epoch": 12.067463106113845, "grad_norm": 0.8020934462547302, "learning_rate": 2.5292105879597097e-05, "loss": 0.0918, "step": 17172 }, { "epoch": 12.068165846802529, "grad_norm": 0.4511389136314392, "learning_rate": 2.529163738580464e-05, "loss": 0.1259, "step": 17173 }, { "epoch": 12.068868587491215, "grad_norm": 0.5337373614311218, "learning_rate": 2.529116889201218e-05, "loss": 0.1455, "step": 17174 }, { "epoch": 12.0695713281799, "grad_norm": 0.8819316029548645, "learning_rate": 2.5290700398219724e-05, "loss": 0.149, "step": 17175 }, { "epoch": 12.070274068868587, "grad_norm": 0.6578412055969238, "learning_rate": 2.5290231904427268e-05, "loss": 0.1634, "step": 17176 }, { "epoch": 12.070976809557273, "grad_norm": 0.42074188590049744, "learning_rate": 2.5289763410634812e-05, "loss": 0.0767, "step": 17177 }, { "epoch": 12.071679550245959, "grad_norm": 0.14245331287384033, "learning_rate": 2.5289294916842352e-05, "loss": 0.0201, "step": 17178 }, { "epoch": 12.072382290934645, "grad_norm": 0.13536635041236877, "learning_rate": 2.5288826423049896e-05, "loss": 0.032, "step": 17179 }, { "epoch": 12.07308503162333, "grad_norm": 0.11697418987751007, "learning_rate": 2.528835792925744e-05, "loss": 0.0188, "step": 17180 }, { "epoch": 12.073787772312016, "grad_norm": 0.15354347229003906, "learning_rate": 2.528788943546498e-05, "loss": 0.022, "step": 17181 }, { "epoch": 12.074490513000702, "grad_norm": 0.15383461117744446, "learning_rate": 2.5287420941672524e-05, "loss": 0.0117, "step": 17182 }, { "epoch": 12.075193253689388, "grad_norm": 0.1118374839425087, "learning_rate": 2.5286952447880064e-05, "loss": 0.0087, "step": 17183 }, { "epoch": 12.075895994378074, "grad_norm": 0.1113860011100769, "learning_rate": 2.5286483954087608e-05, "loss": 0.024, "step": 17184 }, { "epoch": 12.07659873506676, "grad_norm": 0.21290932595729828, "learning_rate": 2.528601546029515e-05, "loss": 0.0203, "step": 17185 }, { "epoch": 12.077301475755446, "grad_norm": 0.22624897956848145, "learning_rate": 2.5285546966502695e-05, "loss": 0.0088, "step": 17186 }, { "epoch": 12.078004216444132, "grad_norm": 0.20860517024993896, "learning_rate": 2.5285078472710236e-05, "loss": 0.032, "step": 17187 }, { "epoch": 12.078706957132818, "grad_norm": 0.08168300241231918, "learning_rate": 2.528460997891778e-05, "loss": 0.0098, "step": 17188 }, { "epoch": 12.079409697821504, "grad_norm": 0.18274301290512085, "learning_rate": 2.5284141485125323e-05, "loss": 0.0233, "step": 17189 }, { "epoch": 12.08011243851019, "grad_norm": 0.1334647685289383, "learning_rate": 2.5283672991332867e-05, "loss": 0.021, "step": 17190 }, { "epoch": 12.080815179198876, "grad_norm": 0.35902297496795654, "learning_rate": 2.5283204497540407e-05, "loss": 0.0298, "step": 17191 }, { "epoch": 12.081517919887562, "grad_norm": 0.41341644525527954, "learning_rate": 2.528273600374795e-05, "loss": 0.0372, "step": 17192 }, { "epoch": 12.082220660576247, "grad_norm": 0.22141283750534058, "learning_rate": 2.5282267509955495e-05, "loss": 0.0268, "step": 17193 }, { "epoch": 12.082923401264933, "grad_norm": 0.3647365868091583, "learning_rate": 2.528179901616304e-05, "loss": 0.0304, "step": 17194 }, { "epoch": 12.08362614195362, "grad_norm": 0.2685339152812958, "learning_rate": 2.5281330522370582e-05, "loss": 0.0617, "step": 17195 }, { "epoch": 12.084328882642305, "grad_norm": 0.2905375063419342, "learning_rate": 2.5280862028578122e-05, "loss": 0.0312, "step": 17196 }, { "epoch": 12.085031623330991, "grad_norm": 0.19736520946025848, "learning_rate": 2.5280393534785666e-05, "loss": 0.0623, "step": 17197 }, { "epoch": 12.085734364019677, "grad_norm": 0.4724484384059906, "learning_rate": 2.5279925040993206e-05, "loss": 0.1244, "step": 17198 }, { "epoch": 12.086437104708363, "grad_norm": 1.2434800863265991, "learning_rate": 2.527945654720075e-05, "loss": 0.1104, "step": 17199 }, { "epoch": 12.087139845397049, "grad_norm": 1.0014866590499878, "learning_rate": 2.527898805340829e-05, "loss": 0.2062, "step": 17200 }, { "epoch": 12.087842586085735, "grad_norm": 1.3125629425048828, "learning_rate": 2.5278519559615834e-05, "loss": 0.1932, "step": 17201 }, { "epoch": 12.08854532677442, "grad_norm": 0.2823582887649536, "learning_rate": 2.5278051065823378e-05, "loss": 0.0519, "step": 17202 }, { "epoch": 12.089248067463107, "grad_norm": 0.11654414236545563, "learning_rate": 2.5277582572030922e-05, "loss": 0.0217, "step": 17203 }, { "epoch": 12.089950808151793, "grad_norm": 0.10684049129486084, "learning_rate": 2.5277114078238462e-05, "loss": 0.0159, "step": 17204 }, { "epoch": 12.090653548840478, "grad_norm": 0.10133939236402512, "learning_rate": 2.5276645584446006e-05, "loss": 0.0152, "step": 17205 }, { "epoch": 12.091356289529164, "grad_norm": 0.06881746649742126, "learning_rate": 2.527617709065355e-05, "loss": 0.0098, "step": 17206 }, { "epoch": 12.09205903021785, "grad_norm": 0.08412002772092819, "learning_rate": 2.5275708596861093e-05, "loss": 0.0119, "step": 17207 }, { "epoch": 12.092761770906536, "grad_norm": 0.12016284465789795, "learning_rate": 2.5275240103068637e-05, "loss": 0.0248, "step": 17208 }, { "epoch": 12.093464511595222, "grad_norm": 0.09848343580961227, "learning_rate": 2.5274771609276177e-05, "loss": 0.0194, "step": 17209 }, { "epoch": 12.094167252283908, "grad_norm": 0.19333316385746002, "learning_rate": 2.527430311548372e-05, "loss": 0.0186, "step": 17210 }, { "epoch": 12.094869992972592, "grad_norm": 0.11909200251102448, "learning_rate": 2.5273834621691265e-05, "loss": 0.021, "step": 17211 }, { "epoch": 12.095572733661278, "grad_norm": 0.23731541633605957, "learning_rate": 2.527336612789881e-05, "loss": 0.0205, "step": 17212 }, { "epoch": 12.096275474349964, "grad_norm": 0.1254202276468277, "learning_rate": 2.527289763410635e-05, "loss": 0.0141, "step": 17213 }, { "epoch": 12.09697821503865, "grad_norm": 0.17125940322875977, "learning_rate": 2.5272429140313893e-05, "loss": 0.0237, "step": 17214 }, { "epoch": 12.097680955727336, "grad_norm": 0.08260706812143326, "learning_rate": 2.5271960646521433e-05, "loss": 0.0115, "step": 17215 }, { "epoch": 12.098383696416022, "grad_norm": 0.15123595297336578, "learning_rate": 2.5271492152728977e-05, "loss": 0.0359, "step": 17216 }, { "epoch": 12.099086437104708, "grad_norm": 0.13873380422592163, "learning_rate": 2.5271023658936517e-05, "loss": 0.031, "step": 17217 }, { "epoch": 12.099789177793394, "grad_norm": 0.19841265678405762, "learning_rate": 2.527055516514406e-05, "loss": 0.0178, "step": 17218 }, { "epoch": 12.10049191848208, "grad_norm": 0.14078335464000702, "learning_rate": 2.5270086671351604e-05, "loss": 0.027, "step": 17219 }, { "epoch": 12.101194659170766, "grad_norm": 0.3792499899864197, "learning_rate": 2.5269618177559148e-05, "loss": 0.0334, "step": 17220 }, { "epoch": 12.101897399859451, "grad_norm": 0.24893689155578613, "learning_rate": 2.5269149683766692e-05, "loss": 0.0414, "step": 17221 }, { "epoch": 12.102600140548137, "grad_norm": 0.3059646785259247, "learning_rate": 2.5268681189974232e-05, "loss": 0.0498, "step": 17222 }, { "epoch": 12.103302881236823, "grad_norm": 0.804859459400177, "learning_rate": 2.5268212696181776e-05, "loss": 0.1101, "step": 17223 }, { "epoch": 12.10400562192551, "grad_norm": 1.4036844968795776, "learning_rate": 2.526774420238932e-05, "loss": 0.1496, "step": 17224 }, { "epoch": 12.104708362614195, "grad_norm": 1.1066569089889526, "learning_rate": 2.5267275708596863e-05, "loss": 0.2091, "step": 17225 }, { "epoch": 12.105411103302881, "grad_norm": 1.1165776252746582, "learning_rate": 2.5266807214804404e-05, "loss": 0.2076, "step": 17226 }, { "epoch": 12.106113843991567, "grad_norm": 0.24511931836605072, "learning_rate": 2.5266338721011947e-05, "loss": 0.0587, "step": 17227 }, { "epoch": 12.106816584680253, "grad_norm": 0.20655246078968048, "learning_rate": 2.526587022721949e-05, "loss": 0.0326, "step": 17228 }, { "epoch": 12.107519325368939, "grad_norm": 0.16757109761238098, "learning_rate": 2.5265401733427035e-05, "loss": 0.0225, "step": 17229 }, { "epoch": 12.108222066057625, "grad_norm": 0.7886034846305847, "learning_rate": 2.5264933239634575e-05, "loss": 0.0145, "step": 17230 }, { "epoch": 12.10892480674631, "grad_norm": 0.17674250900745392, "learning_rate": 2.526446474584212e-05, "loss": 0.0319, "step": 17231 }, { "epoch": 12.109627547434997, "grad_norm": 0.2083306610584259, "learning_rate": 2.5263996252049663e-05, "loss": 0.0062, "step": 17232 }, { "epoch": 12.110330288123683, "grad_norm": 0.30741646885871887, "learning_rate": 2.5263527758257203e-05, "loss": 0.0213, "step": 17233 }, { "epoch": 12.111033028812368, "grad_norm": 0.11316266655921936, "learning_rate": 2.5263059264464747e-05, "loss": 0.0197, "step": 17234 }, { "epoch": 12.111735769501054, "grad_norm": 0.15217070281505585, "learning_rate": 2.5262590770672287e-05, "loss": 0.018, "step": 17235 }, { "epoch": 12.11243851018974, "grad_norm": 0.053157053887844086, "learning_rate": 2.526212227687983e-05, "loss": 0.0121, "step": 17236 }, { "epoch": 12.113141250878426, "grad_norm": 0.10689634084701538, "learning_rate": 2.5261653783087374e-05, "loss": 0.0129, "step": 17237 }, { "epoch": 12.113843991567112, "grad_norm": 0.15216949582099915, "learning_rate": 2.5261185289294918e-05, "loss": 0.0083, "step": 17238 }, { "epoch": 12.114546732255798, "grad_norm": 0.22044701874256134, "learning_rate": 2.526071679550246e-05, "loss": 0.03, "step": 17239 }, { "epoch": 12.115249472944484, "grad_norm": 0.11364519596099854, "learning_rate": 2.5260248301710002e-05, "loss": 0.0188, "step": 17240 }, { "epoch": 12.11595221363317, "grad_norm": 0.28085723519325256, "learning_rate": 2.5259779807917546e-05, "loss": 0.0348, "step": 17241 }, { "epoch": 12.116654954321856, "grad_norm": 0.2665574550628662, "learning_rate": 2.525931131412509e-05, "loss": 0.0435, "step": 17242 }, { "epoch": 12.117357695010542, "grad_norm": 0.22985278069972992, "learning_rate": 2.525884282033263e-05, "loss": 0.0229, "step": 17243 }, { "epoch": 12.118060435699228, "grad_norm": 0.4678786098957062, "learning_rate": 2.5258374326540174e-05, "loss": 0.0229, "step": 17244 }, { "epoch": 12.118763176387914, "grad_norm": 0.16865938901901245, "learning_rate": 2.5257905832747718e-05, "loss": 0.0282, "step": 17245 }, { "epoch": 12.1194659170766, "grad_norm": 0.353384405374527, "learning_rate": 2.525743733895526e-05, "loss": 0.0646, "step": 17246 }, { "epoch": 12.120168657765285, "grad_norm": 1.0466973781585693, "learning_rate": 2.5256968845162805e-05, "loss": 0.1017, "step": 17247 }, { "epoch": 12.120871398453971, "grad_norm": 0.48900264501571655, "learning_rate": 2.5256500351370345e-05, "loss": 0.0989, "step": 17248 }, { "epoch": 12.121574139142655, "grad_norm": 0.6583696007728577, "learning_rate": 2.525603185757789e-05, "loss": 0.1477, "step": 17249 }, { "epoch": 12.122276879831341, "grad_norm": 0.8312402963638306, "learning_rate": 2.525556336378543e-05, "loss": 0.1949, "step": 17250 }, { "epoch": 12.122979620520027, "grad_norm": 3.5036752223968506, "learning_rate": 2.5255094869992973e-05, "loss": 0.2171, "step": 17251 }, { "epoch": 12.123682361208713, "grad_norm": 0.49023914337158203, "learning_rate": 2.5254626376200513e-05, "loss": 0.0712, "step": 17252 }, { "epoch": 12.1243851018974, "grad_norm": 0.37402674555778503, "learning_rate": 2.5254157882408057e-05, "loss": 0.0188, "step": 17253 }, { "epoch": 12.125087842586085, "grad_norm": 0.21457603573799133, "learning_rate": 2.52536893886156e-05, "loss": 0.0408, "step": 17254 }, { "epoch": 12.125790583274771, "grad_norm": 0.30035558342933655, "learning_rate": 2.5253220894823145e-05, "loss": 0.0253, "step": 17255 }, { "epoch": 12.126493323963457, "grad_norm": 0.3847390115261078, "learning_rate": 2.5252752401030685e-05, "loss": 0.0301, "step": 17256 }, { "epoch": 12.127196064652143, "grad_norm": 0.12268763035535812, "learning_rate": 2.525228390723823e-05, "loss": 0.022, "step": 17257 }, { "epoch": 12.127898805340829, "grad_norm": 0.2739257216453552, "learning_rate": 2.5251815413445772e-05, "loss": 0.018, "step": 17258 }, { "epoch": 12.128601546029515, "grad_norm": 0.3114722967147827, "learning_rate": 2.5251346919653316e-05, "loss": 0.0261, "step": 17259 }, { "epoch": 12.1293042867182, "grad_norm": 0.15375177562236786, "learning_rate": 2.525087842586086e-05, "loss": 0.0158, "step": 17260 }, { "epoch": 12.130007027406887, "grad_norm": 0.12461009621620178, "learning_rate": 2.52504099320684e-05, "loss": 0.0162, "step": 17261 }, { "epoch": 12.130709768095572, "grad_norm": 0.11170870810747147, "learning_rate": 2.5249941438275944e-05, "loss": 0.0194, "step": 17262 }, { "epoch": 12.131412508784258, "grad_norm": 0.1132269874215126, "learning_rate": 2.5249472944483488e-05, "loss": 0.0146, "step": 17263 }, { "epoch": 12.132115249472944, "grad_norm": 0.15839867293834686, "learning_rate": 2.524900445069103e-05, "loss": 0.0346, "step": 17264 }, { "epoch": 12.13281799016163, "grad_norm": 0.1502683460712433, "learning_rate": 2.524853595689857e-05, "loss": 0.0133, "step": 17265 }, { "epoch": 12.133520730850316, "grad_norm": 0.18729637563228607, "learning_rate": 2.5248067463106115e-05, "loss": 0.0325, "step": 17266 }, { "epoch": 12.134223471539002, "grad_norm": 0.1919219046831131, "learning_rate": 2.524759896931366e-05, "loss": 0.03, "step": 17267 }, { "epoch": 12.134926212227688, "grad_norm": 0.12229171395301819, "learning_rate": 2.52471304755212e-05, "loss": 0.0129, "step": 17268 }, { "epoch": 12.135628952916374, "grad_norm": 0.21975331008434296, "learning_rate": 2.524666198172874e-05, "loss": 0.0315, "step": 17269 }, { "epoch": 12.13633169360506, "grad_norm": 0.47196078300476074, "learning_rate": 2.5246193487936284e-05, "loss": 0.0274, "step": 17270 }, { "epoch": 12.137034434293746, "grad_norm": 0.28273841738700867, "learning_rate": 2.5245724994143827e-05, "loss": 0.052, "step": 17271 }, { "epoch": 12.137737174982432, "grad_norm": 0.8059970736503601, "learning_rate": 2.524525650035137e-05, "loss": 0.0685, "step": 17272 }, { "epoch": 12.138439915671118, "grad_norm": 0.5536470413208008, "learning_rate": 2.5244788006558915e-05, "loss": 0.0975, "step": 17273 }, { "epoch": 12.139142656359803, "grad_norm": 1.7759125232696533, "learning_rate": 2.5244319512766455e-05, "loss": 0.1558, "step": 17274 }, { "epoch": 12.13984539704849, "grad_norm": 0.5611337423324585, "learning_rate": 2.5243851018974e-05, "loss": 0.1692, "step": 17275 }, { "epoch": 12.140548137737175, "grad_norm": 0.9325129985809326, "learning_rate": 2.5243382525181542e-05, "loss": 0.1603, "step": 17276 }, { "epoch": 12.141250878425861, "grad_norm": 0.3094492554664612, "learning_rate": 2.5242914031389086e-05, "loss": 0.0628, "step": 17277 }, { "epoch": 12.141953619114547, "grad_norm": 0.09379373490810394, "learning_rate": 2.5242445537596627e-05, "loss": 0.0158, "step": 17278 }, { "epoch": 12.142656359803233, "grad_norm": 0.15364734828472137, "learning_rate": 2.524197704380417e-05, "loss": 0.0225, "step": 17279 }, { "epoch": 12.143359100491919, "grad_norm": 0.1563972532749176, "learning_rate": 2.5241508550011714e-05, "loss": 0.0152, "step": 17280 }, { "epoch": 12.144061841180605, "grad_norm": 0.08130397647619247, "learning_rate": 2.5241040056219258e-05, "loss": 0.0112, "step": 17281 }, { "epoch": 12.14476458186929, "grad_norm": 0.072791188955307, "learning_rate": 2.5240571562426798e-05, "loss": 0.0207, "step": 17282 }, { "epoch": 12.145467322557977, "grad_norm": 0.15403850376605988, "learning_rate": 2.5240103068634342e-05, "loss": 0.0202, "step": 17283 }, { "epoch": 12.146170063246663, "grad_norm": 0.4282413721084595, "learning_rate": 2.5239634574841886e-05, "loss": 0.0173, "step": 17284 }, { "epoch": 12.146872803935349, "grad_norm": 0.17211489379405975, "learning_rate": 2.5239166081049426e-05, "loss": 0.0265, "step": 17285 }, { "epoch": 12.147575544624035, "grad_norm": 0.3840314447879791, "learning_rate": 2.523869758725697e-05, "loss": 0.0089, "step": 17286 }, { "epoch": 12.14827828531272, "grad_norm": 0.13315054774284363, "learning_rate": 2.523822909346451e-05, "loss": 0.0272, "step": 17287 }, { "epoch": 12.148981026001405, "grad_norm": 0.18861305713653564, "learning_rate": 2.5237760599672054e-05, "loss": 0.0105, "step": 17288 }, { "epoch": 12.14968376669009, "grad_norm": 0.3094063699245453, "learning_rate": 2.5237292105879597e-05, "loss": 0.0282, "step": 17289 }, { "epoch": 12.150386507378776, "grad_norm": 0.14378590881824493, "learning_rate": 2.523682361208714e-05, "loss": 0.0297, "step": 17290 }, { "epoch": 12.151089248067462, "grad_norm": 0.24324026703834534, "learning_rate": 2.523635511829468e-05, "loss": 0.0251, "step": 17291 }, { "epoch": 12.151791988756148, "grad_norm": 0.12312789261341095, "learning_rate": 2.5235886624502225e-05, "loss": 0.0266, "step": 17292 }, { "epoch": 12.152494729444834, "grad_norm": 0.1892370581626892, "learning_rate": 2.523541813070977e-05, "loss": 0.0282, "step": 17293 }, { "epoch": 12.15319747013352, "grad_norm": 0.20467329025268555, "learning_rate": 2.5234949636917313e-05, "loss": 0.0575, "step": 17294 }, { "epoch": 12.153900210822206, "grad_norm": 0.7502613663673401, "learning_rate": 2.5234481143124853e-05, "loss": 0.0468, "step": 17295 }, { "epoch": 12.154602951510892, "grad_norm": 0.2804122269153595, "learning_rate": 2.5234012649332397e-05, "loss": 0.0441, "step": 17296 }, { "epoch": 12.155305692199578, "grad_norm": 0.6846539974212646, "learning_rate": 2.523354415553994e-05, "loss": 0.0893, "step": 17297 }, { "epoch": 12.156008432888264, "grad_norm": 0.3402632772922516, "learning_rate": 2.5233075661747484e-05, "loss": 0.0929, "step": 17298 }, { "epoch": 12.15671117357695, "grad_norm": 0.5726075768470764, "learning_rate": 2.5232607167955028e-05, "loss": 0.1469, "step": 17299 }, { "epoch": 12.157413914265636, "grad_norm": 0.8055588006973267, "learning_rate": 2.5232138674162568e-05, "loss": 0.1487, "step": 17300 }, { "epoch": 12.158116654954322, "grad_norm": 1.376402735710144, "learning_rate": 2.5231670180370112e-05, "loss": 0.1864, "step": 17301 }, { "epoch": 12.158819395643008, "grad_norm": 0.20696698129177094, "learning_rate": 2.5231201686577656e-05, "loss": 0.0624, "step": 17302 }, { "epoch": 12.159522136331693, "grad_norm": 0.3944851756095886, "learning_rate": 2.5230733192785196e-05, "loss": 0.0189, "step": 17303 }, { "epoch": 12.16022487702038, "grad_norm": 0.15268965065479279, "learning_rate": 2.5230264698992736e-05, "loss": 0.0209, "step": 17304 }, { "epoch": 12.160927617709065, "grad_norm": 0.4709901809692383, "learning_rate": 2.522979620520028e-05, "loss": 0.0139, "step": 17305 }, { "epoch": 12.161630358397751, "grad_norm": 0.09189803153276443, "learning_rate": 2.5229327711407824e-05, "loss": 0.0164, "step": 17306 }, { "epoch": 12.162333099086437, "grad_norm": 0.1290646642446518, "learning_rate": 2.5228859217615367e-05, "loss": 0.0142, "step": 17307 }, { "epoch": 12.163035839775123, "grad_norm": 0.23744510114192963, "learning_rate": 2.5228390723822908e-05, "loss": 0.0115, "step": 17308 }, { "epoch": 12.163738580463809, "grad_norm": 0.15608680248260498, "learning_rate": 2.522792223003045e-05, "loss": 0.014, "step": 17309 }, { "epoch": 12.164441321152495, "grad_norm": 0.23121827840805054, "learning_rate": 2.5227453736237995e-05, "loss": 0.0217, "step": 17310 }, { "epoch": 12.16514406184118, "grad_norm": 0.10942675173282623, "learning_rate": 2.522698524244554e-05, "loss": 0.0091, "step": 17311 }, { "epoch": 12.165846802529867, "grad_norm": 0.13115333020687103, "learning_rate": 2.5226516748653083e-05, "loss": 0.0149, "step": 17312 }, { "epoch": 12.166549543218553, "grad_norm": 0.09517110139131546, "learning_rate": 2.5226048254860623e-05, "loss": 0.0128, "step": 17313 }, { "epoch": 12.167252283907239, "grad_norm": 0.15481169521808624, "learning_rate": 2.5225579761068167e-05, "loss": 0.0293, "step": 17314 }, { "epoch": 12.167955024595924, "grad_norm": 0.1516045778989792, "learning_rate": 2.522511126727571e-05, "loss": 0.0245, "step": 17315 }, { "epoch": 12.16865776528461, "grad_norm": 0.19441218674182892, "learning_rate": 2.5224642773483254e-05, "loss": 0.0278, "step": 17316 }, { "epoch": 12.169360505973296, "grad_norm": 0.22131101787090302, "learning_rate": 2.5224174279690795e-05, "loss": 0.0271, "step": 17317 }, { "epoch": 12.170063246661982, "grad_norm": 0.17202213406562805, "learning_rate": 2.5223705785898338e-05, "loss": 0.0148, "step": 17318 }, { "epoch": 12.170765987350668, "grad_norm": 0.24872824549674988, "learning_rate": 2.5223237292105882e-05, "loss": 0.0371, "step": 17319 }, { "epoch": 12.171468728039354, "grad_norm": 0.142433300614357, "learning_rate": 2.5222768798313422e-05, "loss": 0.0271, "step": 17320 }, { "epoch": 12.17217146872804, "grad_norm": 0.6047397255897522, "learning_rate": 2.5222300304520963e-05, "loss": 0.0874, "step": 17321 }, { "epoch": 12.172874209416726, "grad_norm": 0.24229112267494202, "learning_rate": 2.5221831810728506e-05, "loss": 0.0737, "step": 17322 }, { "epoch": 12.173576950105412, "grad_norm": 0.39167261123657227, "learning_rate": 2.522136331693605e-05, "loss": 0.1182, "step": 17323 }, { "epoch": 12.174279690794098, "grad_norm": 0.9868462681770325, "learning_rate": 2.5220894823143594e-05, "loss": 0.1257, "step": 17324 }, { "epoch": 12.174982431482784, "grad_norm": 1.0219918489456177, "learning_rate": 2.5220426329351138e-05, "loss": 0.1608, "step": 17325 }, { "epoch": 12.17568517217147, "grad_norm": 0.7595489025115967, "learning_rate": 2.5219957835558678e-05, "loss": 0.1673, "step": 17326 }, { "epoch": 12.176387912860154, "grad_norm": 0.268610417842865, "learning_rate": 2.521948934176622e-05, "loss": 0.0725, "step": 17327 }, { "epoch": 12.17709065354884, "grad_norm": 0.35740017890930176, "learning_rate": 2.5219020847973765e-05, "loss": 0.0466, "step": 17328 }, { "epoch": 12.177793394237526, "grad_norm": 0.15107594430446625, "learning_rate": 2.521855235418131e-05, "loss": 0.0208, "step": 17329 }, { "epoch": 12.178496134926212, "grad_norm": 0.14444120228290558, "learning_rate": 2.521808386038885e-05, "loss": 0.0251, "step": 17330 }, { "epoch": 12.179198875614897, "grad_norm": 0.09601577371358871, "learning_rate": 2.5217615366596393e-05, "loss": 0.0074, "step": 17331 }, { "epoch": 12.179901616303583, "grad_norm": 0.22825665771961212, "learning_rate": 2.5217146872803937e-05, "loss": 0.024, "step": 17332 }, { "epoch": 12.18060435699227, "grad_norm": 0.10524643212556839, "learning_rate": 2.521667837901148e-05, "loss": 0.009, "step": 17333 }, { "epoch": 12.181307097680955, "grad_norm": 0.07486487179994583, "learning_rate": 2.521620988521902e-05, "loss": 0.006, "step": 17334 }, { "epoch": 12.182009838369641, "grad_norm": 0.3585636019706726, "learning_rate": 2.5215741391426565e-05, "loss": 0.0237, "step": 17335 }, { "epoch": 12.182712579058327, "grad_norm": 0.13269948959350586, "learning_rate": 2.521527289763411e-05, "loss": 0.0143, "step": 17336 }, { "epoch": 12.183415319747013, "grad_norm": 0.12826666235923767, "learning_rate": 2.521480440384165e-05, "loss": 0.0098, "step": 17337 }, { "epoch": 12.184118060435699, "grad_norm": 0.11868906021118164, "learning_rate": 2.5214335910049192e-05, "loss": 0.02, "step": 17338 }, { "epoch": 12.184820801124385, "grad_norm": 0.5064955353736877, "learning_rate": 2.5213867416256733e-05, "loss": 0.0189, "step": 17339 }, { "epoch": 12.18552354181307, "grad_norm": 0.40718576312065125, "learning_rate": 2.5213398922464277e-05, "loss": 0.0263, "step": 17340 }, { "epoch": 12.186226282501757, "grad_norm": 0.11469347029924393, "learning_rate": 2.521293042867182e-05, "loss": 0.0128, "step": 17341 }, { "epoch": 12.186929023190443, "grad_norm": 0.17349764704704285, "learning_rate": 2.5212461934879364e-05, "loss": 0.0217, "step": 17342 }, { "epoch": 12.187631763879128, "grad_norm": 0.29894766211509705, "learning_rate": 2.5211993441086904e-05, "loss": 0.0203, "step": 17343 }, { "epoch": 12.188334504567814, "grad_norm": 0.21103979647159576, "learning_rate": 2.5211524947294448e-05, "loss": 0.0281, "step": 17344 }, { "epoch": 12.1890372452565, "grad_norm": 0.33133479952812195, "learning_rate": 2.5211056453501992e-05, "loss": 0.0551, "step": 17345 }, { "epoch": 12.189739985945186, "grad_norm": 0.2369145005941391, "learning_rate": 2.5210587959709535e-05, "loss": 0.0433, "step": 17346 }, { "epoch": 12.190442726633872, "grad_norm": 0.32875218987464905, "learning_rate": 2.5210119465917076e-05, "loss": 0.0668, "step": 17347 }, { "epoch": 12.191145467322558, "grad_norm": 0.6673150062561035, "learning_rate": 2.520965097212462e-05, "loss": 0.0795, "step": 17348 }, { "epoch": 12.191848208011244, "grad_norm": 0.6243224143981934, "learning_rate": 2.5209182478332163e-05, "loss": 0.13, "step": 17349 }, { "epoch": 12.19255094869993, "grad_norm": 1.0897912979125977, "learning_rate": 2.5208713984539707e-05, "loss": 0.1523, "step": 17350 }, { "epoch": 12.193253689388616, "grad_norm": 1.4332603216171265, "learning_rate": 2.520824549074725e-05, "loss": 0.1814, "step": 17351 }, { "epoch": 12.193956430077302, "grad_norm": 0.47494813799858093, "learning_rate": 2.520777699695479e-05, "loss": 0.0677, "step": 17352 }, { "epoch": 12.194659170765988, "grad_norm": 0.11468932777643204, "learning_rate": 2.5207308503162335e-05, "loss": 0.0266, "step": 17353 }, { "epoch": 12.195361911454674, "grad_norm": 0.14986436069011688, "learning_rate": 2.520684000936988e-05, "loss": 0.0176, "step": 17354 }, { "epoch": 12.19606465214336, "grad_norm": 0.12577641010284424, "learning_rate": 2.520637151557742e-05, "loss": 0.0248, "step": 17355 }, { "epoch": 12.196767392832045, "grad_norm": 0.1100577637553215, "learning_rate": 2.520590302178496e-05, "loss": 0.014, "step": 17356 }, { "epoch": 12.197470133520731, "grad_norm": 0.2309659868478775, "learning_rate": 2.5205434527992503e-05, "loss": 0.0104, "step": 17357 }, { "epoch": 12.198172874209417, "grad_norm": 0.10053470730781555, "learning_rate": 2.5204966034200047e-05, "loss": 0.0206, "step": 17358 }, { "epoch": 12.198875614898103, "grad_norm": 0.20533950626850128, "learning_rate": 2.520449754040759e-05, "loss": 0.0297, "step": 17359 }, { "epoch": 12.19957835558679, "grad_norm": 0.07668321579694748, "learning_rate": 2.5204029046615134e-05, "loss": 0.0147, "step": 17360 }, { "epoch": 12.200281096275475, "grad_norm": 0.09303544461727142, "learning_rate": 2.5203560552822674e-05, "loss": 0.0081, "step": 17361 }, { "epoch": 12.200983836964161, "grad_norm": 0.1192338615655899, "learning_rate": 2.5203092059030218e-05, "loss": 0.0242, "step": 17362 }, { "epoch": 12.201686577652847, "grad_norm": 0.1741200089454651, "learning_rate": 2.5202623565237762e-05, "loss": 0.0239, "step": 17363 }, { "epoch": 12.202389318341533, "grad_norm": 0.19982820749282837, "learning_rate": 2.5202155071445306e-05, "loss": 0.0223, "step": 17364 }, { "epoch": 12.203092059030217, "grad_norm": 0.10805684328079224, "learning_rate": 2.5201686577652846e-05, "loss": 0.0123, "step": 17365 }, { "epoch": 12.203794799718903, "grad_norm": 0.2982516586780548, "learning_rate": 2.520121808386039e-05, "loss": 0.0317, "step": 17366 }, { "epoch": 12.204497540407589, "grad_norm": 0.2290612906217575, "learning_rate": 2.5200749590067933e-05, "loss": 0.0406, "step": 17367 }, { "epoch": 12.205200281096275, "grad_norm": 0.11044656485319138, "learning_rate": 2.5200281096275477e-05, "loss": 0.0175, "step": 17368 }, { "epoch": 12.20590302178496, "grad_norm": 0.2501337230205536, "learning_rate": 2.5199812602483017e-05, "loss": 0.0468, "step": 17369 }, { "epoch": 12.206605762473647, "grad_norm": 0.16683761775493622, "learning_rate": 2.519934410869056e-05, "loss": 0.0218, "step": 17370 }, { "epoch": 12.207308503162333, "grad_norm": 0.2864370048046112, "learning_rate": 2.5198875614898105e-05, "loss": 0.045, "step": 17371 }, { "epoch": 12.208011243851018, "grad_norm": 0.26222655177116394, "learning_rate": 2.5198407121105645e-05, "loss": 0.0535, "step": 17372 }, { "epoch": 12.208713984539704, "grad_norm": 0.9495143890380859, "learning_rate": 2.519793862731319e-05, "loss": 0.1011, "step": 17373 }, { "epoch": 12.20941672522839, "grad_norm": 0.4247438907623291, "learning_rate": 2.519747013352073e-05, "loss": 0.1384, "step": 17374 }, { "epoch": 12.210119465917076, "grad_norm": 1.150169014930725, "learning_rate": 2.5197001639728273e-05, "loss": 0.1514, "step": 17375 }, { "epoch": 12.210822206605762, "grad_norm": 1.6373587846755981, "learning_rate": 2.5196533145935817e-05, "loss": 0.1873, "step": 17376 }, { "epoch": 12.211524947294448, "grad_norm": 0.5113210678100586, "learning_rate": 2.519606465214336e-05, "loss": 0.0837, "step": 17377 }, { "epoch": 12.212227687983134, "grad_norm": 0.30473026633262634, "learning_rate": 2.51955961583509e-05, "loss": 0.0571, "step": 17378 }, { "epoch": 12.21293042867182, "grad_norm": 0.21805404126644135, "learning_rate": 2.5195127664558445e-05, "loss": 0.0241, "step": 17379 }, { "epoch": 12.213633169360506, "grad_norm": 0.09602918475866318, "learning_rate": 2.5194659170765988e-05, "loss": 0.0113, "step": 17380 }, { "epoch": 12.214335910049192, "grad_norm": 0.07990198582410812, "learning_rate": 2.5194190676973532e-05, "loss": 0.0115, "step": 17381 }, { "epoch": 12.215038650737878, "grad_norm": 0.08392294496297836, "learning_rate": 2.5193722183181072e-05, "loss": 0.0096, "step": 17382 }, { "epoch": 12.215741391426564, "grad_norm": 0.20793400704860687, "learning_rate": 2.5193253689388616e-05, "loss": 0.0296, "step": 17383 }, { "epoch": 12.21644413211525, "grad_norm": 0.2470574975013733, "learning_rate": 2.519278519559616e-05, "loss": 0.0123, "step": 17384 }, { "epoch": 12.217146872803935, "grad_norm": 0.19416609406471252, "learning_rate": 2.5192316701803704e-05, "loss": 0.0165, "step": 17385 }, { "epoch": 12.217849613492621, "grad_norm": 0.10245171189308167, "learning_rate": 2.5191848208011247e-05, "loss": 0.0129, "step": 17386 }, { "epoch": 12.218552354181307, "grad_norm": 0.25189393758773804, "learning_rate": 2.5191379714218788e-05, "loss": 0.0177, "step": 17387 }, { "epoch": 12.219255094869993, "grad_norm": 0.17486050724983215, "learning_rate": 2.519091122042633e-05, "loss": 0.0104, "step": 17388 }, { "epoch": 12.219957835558679, "grad_norm": 0.264785498380661, "learning_rate": 2.5190442726633875e-05, "loss": 0.0289, "step": 17389 }, { "epoch": 12.220660576247365, "grad_norm": 0.43149086833000183, "learning_rate": 2.5189974232841415e-05, "loss": 0.0241, "step": 17390 }, { "epoch": 12.221363316936051, "grad_norm": 0.11043938249349594, "learning_rate": 2.5189505739048956e-05, "loss": 0.0143, "step": 17391 }, { "epoch": 12.222066057624737, "grad_norm": 0.36255404353141785, "learning_rate": 2.51890372452565e-05, "loss": 0.0411, "step": 17392 }, { "epoch": 12.222768798313423, "grad_norm": 0.17665117979049683, "learning_rate": 2.5188568751464043e-05, "loss": 0.0323, "step": 17393 }, { "epoch": 12.223471539002109, "grad_norm": 0.37941062450408936, "learning_rate": 2.5188100257671587e-05, "loss": 0.0309, "step": 17394 }, { "epoch": 12.224174279690795, "grad_norm": 0.24181689321994781, "learning_rate": 2.5187631763879127e-05, "loss": 0.0554, "step": 17395 }, { "epoch": 12.22487702037948, "grad_norm": 0.3264326751232147, "learning_rate": 2.518716327008667e-05, "loss": 0.0389, "step": 17396 }, { "epoch": 12.225579761068166, "grad_norm": 0.5153281092643738, "learning_rate": 2.5186694776294215e-05, "loss": 0.0914, "step": 17397 }, { "epoch": 12.226282501756852, "grad_norm": 0.3555139899253845, "learning_rate": 2.518622628250176e-05, "loss": 0.0796, "step": 17398 }, { "epoch": 12.226985242445538, "grad_norm": 0.6807113289833069, "learning_rate": 2.5185757788709302e-05, "loss": 0.1164, "step": 17399 }, { "epoch": 12.227687983134224, "grad_norm": 0.911214292049408, "learning_rate": 2.5185289294916842e-05, "loss": 0.193, "step": 17400 }, { "epoch": 12.22839072382291, "grad_norm": 1.4335848093032837, "learning_rate": 2.5184820801124386e-05, "loss": 0.2027, "step": 17401 }, { "epoch": 12.229093464511596, "grad_norm": 0.698336124420166, "learning_rate": 2.518435230733193e-05, "loss": 0.0708, "step": 17402 }, { "epoch": 12.22979620520028, "grad_norm": 0.1511356681585312, "learning_rate": 2.5183883813539474e-05, "loss": 0.0262, "step": 17403 }, { "epoch": 12.230498945888966, "grad_norm": 0.1553908884525299, "learning_rate": 2.5183415319747014e-05, "loss": 0.0319, "step": 17404 }, { "epoch": 12.231201686577652, "grad_norm": 0.17986883223056793, "learning_rate": 2.5182946825954558e-05, "loss": 0.0142, "step": 17405 }, { "epoch": 12.231904427266338, "grad_norm": 0.08730291575193405, "learning_rate": 2.51824783321621e-05, "loss": 0.0164, "step": 17406 }, { "epoch": 12.232607167955024, "grad_norm": 0.06855738162994385, "learning_rate": 2.5182009838369642e-05, "loss": 0.01, "step": 17407 }, { "epoch": 12.23330990864371, "grad_norm": 0.10249730944633484, "learning_rate": 2.5181541344577182e-05, "loss": 0.0122, "step": 17408 }, { "epoch": 12.234012649332396, "grad_norm": 0.07898280769586563, "learning_rate": 2.5181072850784726e-05, "loss": 0.0146, "step": 17409 }, { "epoch": 12.234715390021082, "grad_norm": 0.1416245698928833, "learning_rate": 2.518060435699227e-05, "loss": 0.0145, "step": 17410 }, { "epoch": 12.235418130709768, "grad_norm": 0.08197522908449173, "learning_rate": 2.5180135863199813e-05, "loss": 0.0165, "step": 17411 }, { "epoch": 12.236120871398454, "grad_norm": 0.13516592979431152, "learning_rate": 2.5179667369407357e-05, "loss": 0.0101, "step": 17412 }, { "epoch": 12.23682361208714, "grad_norm": 0.11586224287748337, "learning_rate": 2.5179198875614897e-05, "loss": 0.0115, "step": 17413 }, { "epoch": 12.237526352775825, "grad_norm": 0.20749516785144806, "learning_rate": 2.517873038182244e-05, "loss": 0.0474, "step": 17414 }, { "epoch": 12.238229093464511, "grad_norm": 0.09464436769485474, "learning_rate": 2.5178261888029985e-05, "loss": 0.0109, "step": 17415 }, { "epoch": 12.238931834153197, "grad_norm": 0.1801188588142395, "learning_rate": 2.517779339423753e-05, "loss": 0.0254, "step": 17416 }, { "epoch": 12.239634574841883, "grad_norm": 0.1944698691368103, "learning_rate": 2.517732490044507e-05, "loss": 0.0434, "step": 17417 }, { "epoch": 12.240337315530569, "grad_norm": 0.11431904882192612, "learning_rate": 2.5176856406652613e-05, "loss": 0.0179, "step": 17418 }, { "epoch": 12.241040056219255, "grad_norm": 0.17799411714076996, "learning_rate": 2.5176387912860156e-05, "loss": 0.0408, "step": 17419 }, { "epoch": 12.24174279690794, "grad_norm": 0.2801159620285034, "learning_rate": 2.51759194190677e-05, "loss": 0.0528, "step": 17420 }, { "epoch": 12.242445537596627, "grad_norm": 0.7366117238998413, "learning_rate": 2.517545092527524e-05, "loss": 0.0419, "step": 17421 }, { "epoch": 12.243148278285313, "grad_norm": 0.30375993251800537, "learning_rate": 2.5174982431482784e-05, "loss": 0.0658, "step": 17422 }, { "epoch": 12.243851018973999, "grad_norm": 0.5718865990638733, "learning_rate": 2.5174513937690328e-05, "loss": 0.1063, "step": 17423 }, { "epoch": 12.244553759662685, "grad_norm": 0.44820642471313477, "learning_rate": 2.5174045443897868e-05, "loss": 0.1475, "step": 17424 }, { "epoch": 12.24525650035137, "grad_norm": 0.8336661458015442, "learning_rate": 2.5173576950105412e-05, "loss": 0.168, "step": 17425 }, { "epoch": 12.245959241040056, "grad_norm": 0.7536218762397766, "learning_rate": 2.5173108456312952e-05, "loss": 0.1727, "step": 17426 }, { "epoch": 12.246661981728742, "grad_norm": 0.20636923611164093, "learning_rate": 2.5172639962520496e-05, "loss": 0.0591, "step": 17427 }, { "epoch": 12.247364722417428, "grad_norm": 0.10498528927564621, "learning_rate": 2.517217146872804e-05, "loss": 0.0205, "step": 17428 }, { "epoch": 12.248067463106114, "grad_norm": 0.09056118875741959, "learning_rate": 2.5171702974935583e-05, "loss": 0.0182, "step": 17429 }, { "epoch": 12.2487702037948, "grad_norm": 0.1775544136762619, "learning_rate": 2.5171234481143124e-05, "loss": 0.0129, "step": 17430 }, { "epoch": 12.249472944483486, "grad_norm": 0.07991055399179459, "learning_rate": 2.5170765987350667e-05, "loss": 0.0094, "step": 17431 }, { "epoch": 12.250175685172172, "grad_norm": 0.08322419226169586, "learning_rate": 2.517029749355821e-05, "loss": 0.0133, "step": 17432 }, { "epoch": 12.250878425860858, "grad_norm": 0.09829235076904297, "learning_rate": 2.5169828999765755e-05, "loss": 0.0171, "step": 17433 }, { "epoch": 12.251581166549544, "grad_norm": 0.11896228045225143, "learning_rate": 2.5169360505973295e-05, "loss": 0.0192, "step": 17434 }, { "epoch": 12.25228390723823, "grad_norm": 0.24483606219291687, "learning_rate": 2.516889201218084e-05, "loss": 0.0221, "step": 17435 }, { "epoch": 12.252986647926916, "grad_norm": 0.12823304533958435, "learning_rate": 2.5168423518388383e-05, "loss": 0.0092, "step": 17436 }, { "epoch": 12.253689388615602, "grad_norm": 0.4001700282096863, "learning_rate": 2.5167955024595926e-05, "loss": 0.022, "step": 17437 }, { "epoch": 12.254392129304287, "grad_norm": 0.08425269275903702, "learning_rate": 2.516748653080347e-05, "loss": 0.0128, "step": 17438 }, { "epoch": 12.255094869992973, "grad_norm": 0.32363006472587585, "learning_rate": 2.516701803701101e-05, "loss": 0.0236, "step": 17439 }, { "epoch": 12.25579761068166, "grad_norm": 0.11425954103469849, "learning_rate": 2.5166549543218554e-05, "loss": 0.0174, "step": 17440 }, { "epoch": 12.256500351370345, "grad_norm": 0.13551391661167145, "learning_rate": 2.5166081049426098e-05, "loss": 0.0224, "step": 17441 }, { "epoch": 12.25720309205903, "grad_norm": 0.2799399793148041, "learning_rate": 2.5165612555633638e-05, "loss": 0.0337, "step": 17442 }, { "epoch": 12.257905832747715, "grad_norm": 0.12896300852298737, "learning_rate": 2.516514406184118e-05, "loss": 0.0207, "step": 17443 }, { "epoch": 12.258608573436401, "grad_norm": 0.25503936409950256, "learning_rate": 2.5164675568048722e-05, "loss": 0.0331, "step": 17444 }, { "epoch": 12.259311314125087, "grad_norm": 0.22804959118366241, "learning_rate": 2.5164207074256266e-05, "loss": 0.0488, "step": 17445 }, { "epoch": 12.260014054813773, "grad_norm": 0.29124197363853455, "learning_rate": 2.516373858046381e-05, "loss": 0.0481, "step": 17446 }, { "epoch": 12.260716795502459, "grad_norm": 0.5404981374740601, "learning_rate": 2.516327008667135e-05, "loss": 0.0537, "step": 17447 }, { "epoch": 12.261419536191145, "grad_norm": 0.3127504885196686, "learning_rate": 2.5162801592878894e-05, "loss": 0.0963, "step": 17448 }, { "epoch": 12.26212227687983, "grad_norm": 0.28484493494033813, "learning_rate": 2.5162333099086438e-05, "loss": 0.1092, "step": 17449 }, { "epoch": 12.262825017568517, "grad_norm": 0.6643813848495483, "learning_rate": 2.516186460529398e-05, "loss": 0.1615, "step": 17450 }, { "epoch": 12.263527758257203, "grad_norm": 0.8434536457061768, "learning_rate": 2.5161396111501525e-05, "loss": 0.1969, "step": 17451 }, { "epoch": 12.264230498945889, "grad_norm": 0.2613718509674072, "learning_rate": 2.5160927617709065e-05, "loss": 0.0617, "step": 17452 }, { "epoch": 12.264933239634574, "grad_norm": 1.0646299123764038, "learning_rate": 2.516045912391661e-05, "loss": 0.0293, "step": 17453 }, { "epoch": 12.26563598032326, "grad_norm": 0.09971752762794495, "learning_rate": 2.5159990630124153e-05, "loss": 0.0166, "step": 17454 }, { "epoch": 12.266338721011946, "grad_norm": 0.10468588024377823, "learning_rate": 2.5159522136331697e-05, "loss": 0.0192, "step": 17455 }, { "epoch": 12.267041461700632, "grad_norm": 0.09568632394075394, "learning_rate": 2.5159053642539237e-05, "loss": 0.014, "step": 17456 }, { "epoch": 12.267744202389318, "grad_norm": 0.09673486649990082, "learning_rate": 2.515858514874678e-05, "loss": 0.0093, "step": 17457 }, { "epoch": 12.268446943078004, "grad_norm": 0.12464891374111176, "learning_rate": 2.5158116654954324e-05, "loss": 0.0165, "step": 17458 }, { "epoch": 12.26914968376669, "grad_norm": 0.08222194761037827, "learning_rate": 2.5157648161161865e-05, "loss": 0.0174, "step": 17459 }, { "epoch": 12.269852424455376, "grad_norm": 0.07998686283826828, "learning_rate": 2.5157179667369405e-05, "loss": 0.0081, "step": 17460 }, { "epoch": 12.270555165144062, "grad_norm": 0.0935538187623024, "learning_rate": 2.515671117357695e-05, "loss": 0.0091, "step": 17461 }, { "epoch": 12.271257905832748, "grad_norm": 0.22928179800510406, "learning_rate": 2.5156242679784492e-05, "loss": 0.0236, "step": 17462 }, { "epoch": 12.271960646521434, "grad_norm": 0.1113334596157074, "learning_rate": 2.5155774185992036e-05, "loss": 0.0147, "step": 17463 }, { "epoch": 12.27266338721012, "grad_norm": 0.13679173588752747, "learning_rate": 2.515530569219958e-05, "loss": 0.0266, "step": 17464 }, { "epoch": 12.273366127898806, "grad_norm": 0.1689434051513672, "learning_rate": 2.515483719840712e-05, "loss": 0.0121, "step": 17465 }, { "epoch": 12.274068868587491, "grad_norm": 0.18678772449493408, "learning_rate": 2.5154368704614664e-05, "loss": 0.017, "step": 17466 }, { "epoch": 12.274771609276177, "grad_norm": 0.19845391809940338, "learning_rate": 2.5153900210822208e-05, "loss": 0.0389, "step": 17467 }, { "epoch": 12.275474349964863, "grad_norm": 0.2225046157836914, "learning_rate": 2.515343171702975e-05, "loss": 0.0177, "step": 17468 }, { "epoch": 12.27617709065355, "grad_norm": 0.2541176974773407, "learning_rate": 2.5152963223237292e-05, "loss": 0.0381, "step": 17469 }, { "epoch": 12.276879831342235, "grad_norm": 0.42868348956108093, "learning_rate": 2.5152494729444835e-05, "loss": 0.0462, "step": 17470 }, { "epoch": 12.277582572030921, "grad_norm": 0.2835710644721985, "learning_rate": 2.515202623565238e-05, "loss": 0.0336, "step": 17471 }, { "epoch": 12.278285312719607, "grad_norm": 0.2536314129829407, "learning_rate": 2.5151557741859923e-05, "loss": 0.051, "step": 17472 }, { "epoch": 12.278988053408293, "grad_norm": 0.5741526484489441, "learning_rate": 2.5151089248067463e-05, "loss": 0.0909, "step": 17473 }, { "epoch": 12.279690794096979, "grad_norm": 1.4133254289627075, "learning_rate": 2.5150620754275007e-05, "loss": 0.115, "step": 17474 }, { "epoch": 12.280393534785665, "grad_norm": 0.7232682108879089, "learning_rate": 2.515015226048255e-05, "loss": 0.1476, "step": 17475 }, { "epoch": 12.28109627547435, "grad_norm": 1.321476936340332, "learning_rate": 2.5149683766690094e-05, "loss": 0.2068, "step": 17476 }, { "epoch": 12.281799016163037, "grad_norm": 0.2608757019042969, "learning_rate": 2.5149215272897635e-05, "loss": 0.0821, "step": 17477 }, { "epoch": 12.282501756851723, "grad_norm": 0.14404414594173431, "learning_rate": 2.5148746779105175e-05, "loss": 0.0193, "step": 17478 }, { "epoch": 12.283204497540408, "grad_norm": 0.177231103181839, "learning_rate": 2.514827828531272e-05, "loss": 0.0214, "step": 17479 }, { "epoch": 12.283907238229094, "grad_norm": 0.1906968057155609, "learning_rate": 2.5147809791520263e-05, "loss": 0.0366, "step": 17480 }, { "epoch": 12.284609978917779, "grad_norm": 0.17634527385234833, "learning_rate": 2.5147341297727806e-05, "loss": 0.0154, "step": 17481 }, { "epoch": 12.285312719606464, "grad_norm": 0.4469943046569824, "learning_rate": 2.5146872803935347e-05, "loss": 0.0176, "step": 17482 }, { "epoch": 12.28601546029515, "grad_norm": 0.10201293975114822, "learning_rate": 2.514640431014289e-05, "loss": 0.0131, "step": 17483 }, { "epoch": 12.286718200983836, "grad_norm": 0.15216688811779022, "learning_rate": 2.5145935816350434e-05, "loss": 0.0224, "step": 17484 }, { "epoch": 12.287420941672522, "grad_norm": 0.2531445026397705, "learning_rate": 2.5145467322557978e-05, "loss": 0.0148, "step": 17485 }, { "epoch": 12.288123682361208, "grad_norm": 0.1417347639799118, "learning_rate": 2.5144998828765518e-05, "loss": 0.0079, "step": 17486 }, { "epoch": 12.288826423049894, "grad_norm": 0.3266652524471283, "learning_rate": 2.5144530334973062e-05, "loss": 0.0248, "step": 17487 }, { "epoch": 12.28952916373858, "grad_norm": 0.610281229019165, "learning_rate": 2.5144061841180606e-05, "loss": 0.0232, "step": 17488 }, { "epoch": 12.290231904427266, "grad_norm": 0.20096446573734283, "learning_rate": 2.514359334738815e-05, "loss": 0.0291, "step": 17489 }, { "epoch": 12.290934645115952, "grad_norm": 0.4157260060310364, "learning_rate": 2.5143124853595693e-05, "loss": 0.0188, "step": 17490 }, { "epoch": 12.291637385804638, "grad_norm": 0.28733253479003906, "learning_rate": 2.5142656359803233e-05, "loss": 0.0178, "step": 17491 }, { "epoch": 12.292340126493324, "grad_norm": 0.2521912455558777, "learning_rate": 2.5142187866010777e-05, "loss": 0.0229, "step": 17492 }, { "epoch": 12.29304286718201, "grad_norm": 0.16952431201934814, "learning_rate": 2.514171937221832e-05, "loss": 0.0253, "step": 17493 }, { "epoch": 12.293745607870695, "grad_norm": 0.35322651267051697, "learning_rate": 2.514125087842586e-05, "loss": 0.0336, "step": 17494 }, { "epoch": 12.294448348559381, "grad_norm": 0.20391932129859924, "learning_rate": 2.51407823846334e-05, "loss": 0.0252, "step": 17495 }, { "epoch": 12.295151089248067, "grad_norm": 1.6295337677001953, "learning_rate": 2.5140313890840945e-05, "loss": 0.0418, "step": 17496 }, { "epoch": 12.295853829936753, "grad_norm": 0.3642842471599579, "learning_rate": 2.513984539704849e-05, "loss": 0.0552, "step": 17497 }, { "epoch": 12.29655657062544, "grad_norm": 0.4560343027114868, "learning_rate": 2.5139376903256033e-05, "loss": 0.0854, "step": 17498 }, { "epoch": 12.297259311314125, "grad_norm": 1.2937356233596802, "learning_rate": 2.5138908409463573e-05, "loss": 0.1297, "step": 17499 }, { "epoch": 12.297962052002811, "grad_norm": 0.7560743093490601, "learning_rate": 2.5138439915671117e-05, "loss": 0.1694, "step": 17500 }, { "epoch": 12.298664792691497, "grad_norm": 0.8201974630355835, "learning_rate": 2.513797142187866e-05, "loss": 0.213, "step": 17501 }, { "epoch": 12.299367533380183, "grad_norm": 0.22563323378562927, "learning_rate": 2.5137502928086204e-05, "loss": 0.0566, "step": 17502 }, { "epoch": 12.300070274068869, "grad_norm": 0.1475132405757904, "learning_rate": 2.5137034434293748e-05, "loss": 0.0245, "step": 17503 }, { "epoch": 12.300773014757555, "grad_norm": 0.10215780884027481, "learning_rate": 2.5136565940501288e-05, "loss": 0.0173, "step": 17504 }, { "epoch": 12.30147575544624, "grad_norm": 0.8322394490242004, "learning_rate": 2.5136097446708832e-05, "loss": 0.0238, "step": 17505 }, { "epoch": 12.302178496134927, "grad_norm": 0.17430327832698822, "learning_rate": 2.5135628952916376e-05, "loss": 0.0278, "step": 17506 }, { "epoch": 12.302881236823612, "grad_norm": 0.09181272983551025, "learning_rate": 2.513516045912392e-05, "loss": 0.0047, "step": 17507 }, { "epoch": 12.303583977512298, "grad_norm": 0.12018250674009323, "learning_rate": 2.513469196533146e-05, "loss": 0.0172, "step": 17508 }, { "epoch": 12.304286718200984, "grad_norm": 0.08575356006622314, "learning_rate": 2.5134223471539003e-05, "loss": 0.0197, "step": 17509 }, { "epoch": 12.30498945888967, "grad_norm": 0.3325352668762207, "learning_rate": 2.5133754977746547e-05, "loss": 0.02, "step": 17510 }, { "epoch": 12.305692199578356, "grad_norm": 0.0753302052617073, "learning_rate": 2.513328648395409e-05, "loss": 0.0066, "step": 17511 }, { "epoch": 12.306394940267042, "grad_norm": 0.11363550275564194, "learning_rate": 2.5132817990161628e-05, "loss": 0.0219, "step": 17512 }, { "epoch": 12.307097680955728, "grad_norm": 0.11417989432811737, "learning_rate": 2.513234949636917e-05, "loss": 0.0123, "step": 17513 }, { "epoch": 12.307800421644414, "grad_norm": 0.19690510630607605, "learning_rate": 2.5131881002576715e-05, "loss": 0.0159, "step": 17514 }, { "epoch": 12.3085031623331, "grad_norm": 0.12292148917913437, "learning_rate": 2.513141250878426e-05, "loss": 0.0165, "step": 17515 }, { "epoch": 12.309205903021786, "grad_norm": 0.40723833441734314, "learning_rate": 2.5130944014991803e-05, "loss": 0.0228, "step": 17516 }, { "epoch": 12.309908643710472, "grad_norm": 0.28155285120010376, "learning_rate": 2.5130475521199343e-05, "loss": 0.0333, "step": 17517 }, { "epoch": 12.310611384399156, "grad_norm": 0.13137467205524445, "learning_rate": 2.5130007027406887e-05, "loss": 0.0096, "step": 17518 }, { "epoch": 12.311314125087842, "grad_norm": 0.33971938490867615, "learning_rate": 2.512953853361443e-05, "loss": 0.0291, "step": 17519 }, { "epoch": 12.312016865776528, "grad_norm": 0.47474154829978943, "learning_rate": 2.5129070039821974e-05, "loss": 0.0331, "step": 17520 }, { "epoch": 12.312719606465214, "grad_norm": 0.386156290769577, "learning_rate": 2.5128601546029515e-05, "loss": 0.0636, "step": 17521 }, { "epoch": 12.3134223471539, "grad_norm": 0.347369521856308, "learning_rate": 2.512813305223706e-05, "loss": 0.0748, "step": 17522 }, { "epoch": 12.314125087842585, "grad_norm": 0.3692971467971802, "learning_rate": 2.5127664558444602e-05, "loss": 0.1026, "step": 17523 }, { "epoch": 12.314827828531271, "grad_norm": 0.42679867148399353, "learning_rate": 2.5127196064652146e-05, "loss": 0.1301, "step": 17524 }, { "epoch": 12.315530569219957, "grad_norm": 3.3017735481262207, "learning_rate": 2.5126727570859686e-05, "loss": 0.1516, "step": 17525 }, { "epoch": 12.316233309908643, "grad_norm": 2.4265246391296387, "learning_rate": 2.512625907706723e-05, "loss": 0.1789, "step": 17526 }, { "epoch": 12.316936050597329, "grad_norm": 0.43294021487236023, "learning_rate": 2.5125790583274774e-05, "loss": 0.0609, "step": 17527 }, { "epoch": 12.317638791286015, "grad_norm": 0.1338386833667755, "learning_rate": 2.5125322089482317e-05, "loss": 0.0201, "step": 17528 }, { "epoch": 12.318341531974701, "grad_norm": 0.13523279130458832, "learning_rate": 2.5124853595689858e-05, "loss": 0.0187, "step": 17529 }, { "epoch": 12.319044272663387, "grad_norm": 0.48914748430252075, "learning_rate": 2.5124385101897398e-05, "loss": 0.0132, "step": 17530 }, { "epoch": 12.319747013352073, "grad_norm": 0.0953521579504013, "learning_rate": 2.512391660810494e-05, "loss": 0.0172, "step": 17531 }, { "epoch": 12.320449754040759, "grad_norm": 0.10495804995298386, "learning_rate": 2.5123448114312485e-05, "loss": 0.0125, "step": 17532 }, { "epoch": 12.321152494729445, "grad_norm": 0.146479532122612, "learning_rate": 2.512297962052003e-05, "loss": 0.0113, "step": 17533 }, { "epoch": 12.32185523541813, "grad_norm": 0.17326907813549042, "learning_rate": 2.512251112672757e-05, "loss": 0.0182, "step": 17534 }, { "epoch": 12.322557976106816, "grad_norm": 0.2351868897676468, "learning_rate": 2.5122042632935113e-05, "loss": 0.0228, "step": 17535 }, { "epoch": 12.323260716795502, "grad_norm": 0.08519141376018524, "learning_rate": 2.5121574139142657e-05, "loss": 0.0143, "step": 17536 }, { "epoch": 12.323963457484188, "grad_norm": 0.15744903683662415, "learning_rate": 2.51211056453502e-05, "loss": 0.0262, "step": 17537 }, { "epoch": 12.324666198172874, "grad_norm": 0.09742938727140427, "learning_rate": 2.512063715155774e-05, "loss": 0.0145, "step": 17538 }, { "epoch": 12.32536893886156, "grad_norm": 0.11672244966030121, "learning_rate": 2.5120168657765285e-05, "loss": 0.0195, "step": 17539 }, { "epoch": 12.326071679550246, "grad_norm": 0.08909130096435547, "learning_rate": 2.511970016397283e-05, "loss": 0.012, "step": 17540 }, { "epoch": 12.326774420238932, "grad_norm": 0.23639729619026184, "learning_rate": 2.5119231670180372e-05, "loss": 0.0282, "step": 17541 }, { "epoch": 12.327477160927618, "grad_norm": 0.20012620091438293, "learning_rate": 2.5118763176387916e-05, "loss": 0.0265, "step": 17542 }, { "epoch": 12.328179901616304, "grad_norm": 0.16785109043121338, "learning_rate": 2.5118294682595456e-05, "loss": 0.0199, "step": 17543 }, { "epoch": 12.32888264230499, "grad_norm": 0.17039768397808075, "learning_rate": 2.5117826188803e-05, "loss": 0.0257, "step": 17544 }, { "epoch": 12.329585382993676, "grad_norm": 0.20343999564647675, "learning_rate": 2.5117357695010544e-05, "loss": 0.055, "step": 17545 }, { "epoch": 12.330288123682362, "grad_norm": 0.2882664203643799, "learning_rate": 2.5116889201218084e-05, "loss": 0.0552, "step": 17546 }, { "epoch": 12.330990864371048, "grad_norm": 0.34058502316474915, "learning_rate": 2.5116420707425624e-05, "loss": 0.0619, "step": 17547 }, { "epoch": 12.331693605059733, "grad_norm": 0.46160566806793213, "learning_rate": 2.5115952213633168e-05, "loss": 0.1113, "step": 17548 }, { "epoch": 12.33239634574842, "grad_norm": 0.924674391746521, "learning_rate": 2.5115483719840712e-05, "loss": 0.1394, "step": 17549 }, { "epoch": 12.333099086437105, "grad_norm": 0.7359849810600281, "learning_rate": 2.5115015226048256e-05, "loss": 0.1566, "step": 17550 }, { "epoch": 12.333801827125791, "grad_norm": 0.7639681100845337, "learning_rate": 2.51145467322558e-05, "loss": 0.2017, "step": 17551 }, { "epoch": 12.334504567814477, "grad_norm": 0.25958654284477234, "learning_rate": 2.511407823846334e-05, "loss": 0.06, "step": 17552 }, { "epoch": 12.335207308503163, "grad_norm": 0.1653335690498352, "learning_rate": 2.5113609744670883e-05, "loss": 0.0284, "step": 17553 }, { "epoch": 12.335910049191849, "grad_norm": 0.12890101969242096, "learning_rate": 2.5113141250878427e-05, "loss": 0.0276, "step": 17554 }, { "epoch": 12.336612789880535, "grad_norm": 0.0817977637052536, "learning_rate": 2.511267275708597e-05, "loss": 0.0117, "step": 17555 }, { "epoch": 12.33731553056922, "grad_norm": 0.2458314150571823, "learning_rate": 2.511220426329351e-05, "loss": 0.0134, "step": 17556 }, { "epoch": 12.338018271257905, "grad_norm": 0.2543620765209198, "learning_rate": 2.5111735769501055e-05, "loss": 0.007, "step": 17557 }, { "epoch": 12.33872101194659, "grad_norm": 1.0711166858673096, "learning_rate": 2.51112672757086e-05, "loss": 0.0167, "step": 17558 }, { "epoch": 12.339423752635277, "grad_norm": 0.22479034960269928, "learning_rate": 2.5110798781916142e-05, "loss": 0.0173, "step": 17559 }, { "epoch": 12.340126493323963, "grad_norm": 0.1117686927318573, "learning_rate": 2.5110330288123683e-05, "loss": 0.0124, "step": 17560 }, { "epoch": 12.340829234012649, "grad_norm": 0.08649961650371552, "learning_rate": 2.5109861794331226e-05, "loss": 0.0124, "step": 17561 }, { "epoch": 12.341531974701335, "grad_norm": 0.3421894907951355, "learning_rate": 2.510939330053877e-05, "loss": 0.0265, "step": 17562 }, { "epoch": 12.34223471539002, "grad_norm": 0.0708751529455185, "learning_rate": 2.5108924806746314e-05, "loss": 0.0055, "step": 17563 }, { "epoch": 12.342937456078706, "grad_norm": 0.20988258719444275, "learning_rate": 2.5108456312953854e-05, "loss": 0.021, "step": 17564 }, { "epoch": 12.343640196767392, "grad_norm": 0.12280458956956863, "learning_rate": 2.5107987819161394e-05, "loss": 0.0298, "step": 17565 }, { "epoch": 12.344342937456078, "grad_norm": 0.10648494958877563, "learning_rate": 2.5107519325368938e-05, "loss": 0.0116, "step": 17566 }, { "epoch": 12.345045678144764, "grad_norm": 0.0794953927397728, "learning_rate": 2.5107050831576482e-05, "loss": 0.0179, "step": 17567 }, { "epoch": 12.34574841883345, "grad_norm": 0.22892674803733826, "learning_rate": 2.5106582337784026e-05, "loss": 0.0378, "step": 17568 }, { "epoch": 12.346451159522136, "grad_norm": 0.23629985749721527, "learning_rate": 2.5106113843991566e-05, "loss": 0.0338, "step": 17569 }, { "epoch": 12.347153900210822, "grad_norm": 0.28627288341522217, "learning_rate": 2.510564535019911e-05, "loss": 0.0439, "step": 17570 }, { "epoch": 12.347856640899508, "grad_norm": 0.22563494741916656, "learning_rate": 2.5105176856406653e-05, "loss": 0.0413, "step": 17571 }, { "epoch": 12.348559381588194, "grad_norm": 0.28222334384918213, "learning_rate": 2.5104708362614197e-05, "loss": 0.0727, "step": 17572 }, { "epoch": 12.34926212227688, "grad_norm": 0.48036959767341614, "learning_rate": 2.5104239868821737e-05, "loss": 0.1007, "step": 17573 }, { "epoch": 12.349964862965566, "grad_norm": 0.7289139032363892, "learning_rate": 2.510377137502928e-05, "loss": 0.1416, "step": 17574 }, { "epoch": 12.350667603654252, "grad_norm": 0.8155667781829834, "learning_rate": 2.5103302881236825e-05, "loss": 0.1627, "step": 17575 }, { "epoch": 12.351370344342937, "grad_norm": 1.3969602584838867, "learning_rate": 2.510283438744437e-05, "loss": 0.2019, "step": 17576 }, { "epoch": 12.352073085031623, "grad_norm": 0.24338498711585999, "learning_rate": 2.5102365893651912e-05, "loss": 0.0885, "step": 17577 }, { "epoch": 12.35277582572031, "grad_norm": 0.1902095526456833, "learning_rate": 2.5101897399859453e-05, "loss": 0.0252, "step": 17578 }, { "epoch": 12.353478566408995, "grad_norm": 0.1574499011039734, "learning_rate": 2.5101428906066996e-05, "loss": 0.0383, "step": 17579 }, { "epoch": 12.354181307097681, "grad_norm": 0.12964412569999695, "learning_rate": 2.510096041227454e-05, "loss": 0.0185, "step": 17580 }, { "epoch": 12.354884047786367, "grad_norm": 0.10272374749183655, "learning_rate": 2.510049191848208e-05, "loss": 0.0165, "step": 17581 }, { "epoch": 12.355586788475053, "grad_norm": 0.10591025650501251, "learning_rate": 2.510002342468962e-05, "loss": 0.0107, "step": 17582 }, { "epoch": 12.356289529163739, "grad_norm": 0.1048780083656311, "learning_rate": 2.5099554930897165e-05, "loss": 0.0137, "step": 17583 }, { "epoch": 12.356992269852425, "grad_norm": 0.4466235339641571, "learning_rate": 2.5099086437104708e-05, "loss": 0.0138, "step": 17584 }, { "epoch": 12.35769501054111, "grad_norm": 0.2012730836868286, "learning_rate": 2.5098617943312252e-05, "loss": 0.0376, "step": 17585 }, { "epoch": 12.358397751229797, "grad_norm": 0.3915250301361084, "learning_rate": 2.5098149449519792e-05, "loss": 0.022, "step": 17586 }, { "epoch": 12.359100491918483, "grad_norm": 0.2568013370037079, "learning_rate": 2.5097680955727336e-05, "loss": 0.0439, "step": 17587 }, { "epoch": 12.359803232607169, "grad_norm": 0.11473729461431503, "learning_rate": 2.509721246193488e-05, "loss": 0.0117, "step": 17588 }, { "epoch": 12.360505973295854, "grad_norm": 0.1855042278766632, "learning_rate": 2.5096743968142424e-05, "loss": 0.0224, "step": 17589 }, { "epoch": 12.36120871398454, "grad_norm": 0.29153111577033997, "learning_rate": 2.5096275474349967e-05, "loss": 0.0189, "step": 17590 }, { "epoch": 12.361911454673226, "grad_norm": 0.27839553356170654, "learning_rate": 2.5095806980557508e-05, "loss": 0.0199, "step": 17591 }, { "epoch": 12.362614195361912, "grad_norm": 0.22896677255630493, "learning_rate": 2.509533848676505e-05, "loss": 0.0385, "step": 17592 }, { "epoch": 12.363316936050598, "grad_norm": 0.15561996400356293, "learning_rate": 2.5094869992972595e-05, "loss": 0.0151, "step": 17593 }, { "epoch": 12.364019676739284, "grad_norm": 0.18490853905677795, "learning_rate": 2.509440149918014e-05, "loss": 0.0358, "step": 17594 }, { "epoch": 12.36472241742797, "grad_norm": 0.24373582005500793, "learning_rate": 2.509393300538768e-05, "loss": 0.0434, "step": 17595 }, { "epoch": 12.365425158116654, "grad_norm": 0.2556886374950409, "learning_rate": 2.5093464511595223e-05, "loss": 0.0632, "step": 17596 }, { "epoch": 12.36612789880534, "grad_norm": 0.29223763942718506, "learning_rate": 2.5092996017802767e-05, "loss": 0.062, "step": 17597 }, { "epoch": 12.366830639494026, "grad_norm": 2.0497865676879883, "learning_rate": 2.509252752401031e-05, "loss": 0.1089, "step": 17598 }, { "epoch": 12.367533380182712, "grad_norm": 1.3295515775680542, "learning_rate": 2.5092059030217847e-05, "loss": 0.1187, "step": 17599 }, { "epoch": 12.368236120871398, "grad_norm": 0.551310122013092, "learning_rate": 2.509159053642539e-05, "loss": 0.1628, "step": 17600 }, { "epoch": 12.368938861560084, "grad_norm": 0.718397319316864, "learning_rate": 2.5091122042632935e-05, "loss": 0.2093, "step": 17601 }, { "epoch": 12.36964160224877, "grad_norm": 0.31853190064430237, "learning_rate": 2.509065354884048e-05, "loss": 0.058, "step": 17602 }, { "epoch": 12.370344342937456, "grad_norm": 0.2273302525281906, "learning_rate": 2.5090185055048022e-05, "loss": 0.0205, "step": 17603 }, { "epoch": 12.371047083626141, "grad_norm": 0.21199806034564972, "learning_rate": 2.5089716561255562e-05, "loss": 0.0265, "step": 17604 }, { "epoch": 12.371749824314827, "grad_norm": 0.1098494902253151, "learning_rate": 2.5089248067463106e-05, "loss": 0.0182, "step": 17605 }, { "epoch": 12.372452565003513, "grad_norm": 0.12435369938611984, "learning_rate": 2.508877957367065e-05, "loss": 0.0187, "step": 17606 }, { "epoch": 12.3731553056922, "grad_norm": 0.08844909816980362, "learning_rate": 2.5088311079878194e-05, "loss": 0.0086, "step": 17607 }, { "epoch": 12.373858046380885, "grad_norm": 0.07798044383525848, "learning_rate": 2.5087842586085734e-05, "loss": 0.0142, "step": 17608 }, { "epoch": 12.374560787069571, "grad_norm": 0.13041359186172485, "learning_rate": 2.5087374092293278e-05, "loss": 0.0233, "step": 17609 }, { "epoch": 12.375263527758257, "grad_norm": 0.16173885762691498, "learning_rate": 2.508690559850082e-05, "loss": 0.0217, "step": 17610 }, { "epoch": 12.375966268446943, "grad_norm": 0.12224218994379044, "learning_rate": 2.5086437104708365e-05, "loss": 0.0157, "step": 17611 }, { "epoch": 12.376669009135629, "grad_norm": 0.26948070526123047, "learning_rate": 2.5085968610915906e-05, "loss": 0.0176, "step": 17612 }, { "epoch": 12.377371749824315, "grad_norm": 0.15197928249835968, "learning_rate": 2.508550011712345e-05, "loss": 0.0102, "step": 17613 }, { "epoch": 12.378074490513, "grad_norm": 0.15893374383449554, "learning_rate": 2.5085031623330993e-05, "loss": 0.0183, "step": 17614 }, { "epoch": 12.378777231201687, "grad_norm": 0.3399137556552887, "learning_rate": 2.5084563129538537e-05, "loss": 0.0147, "step": 17615 }, { "epoch": 12.379479971890373, "grad_norm": 0.09483283013105392, "learning_rate": 2.5084094635746077e-05, "loss": 0.0163, "step": 17616 }, { "epoch": 12.380182712579058, "grad_norm": 0.2673202455043793, "learning_rate": 2.5083626141953617e-05, "loss": 0.0357, "step": 17617 }, { "epoch": 12.380885453267744, "grad_norm": 0.16074669361114502, "learning_rate": 2.508315764816116e-05, "loss": 0.0153, "step": 17618 }, { "epoch": 12.38158819395643, "grad_norm": 0.28406447172164917, "learning_rate": 2.5082689154368705e-05, "loss": 0.0373, "step": 17619 }, { "epoch": 12.382290934645116, "grad_norm": 0.1704649031162262, "learning_rate": 2.508222066057625e-05, "loss": 0.0376, "step": 17620 }, { "epoch": 12.382993675333802, "grad_norm": 0.2697002589702606, "learning_rate": 2.508175216678379e-05, "loss": 0.0567, "step": 17621 }, { "epoch": 12.383696416022488, "grad_norm": 0.7105581164360046, "learning_rate": 2.5081283672991333e-05, "loss": 0.0684, "step": 17622 }, { "epoch": 12.384399156711174, "grad_norm": 0.5621397495269775, "learning_rate": 2.5080815179198876e-05, "loss": 0.0894, "step": 17623 }, { "epoch": 12.38510189739986, "grad_norm": 0.5416638255119324, "learning_rate": 2.508034668540642e-05, "loss": 0.1106, "step": 17624 }, { "epoch": 12.385804638088546, "grad_norm": 0.9908743500709534, "learning_rate": 2.507987819161396e-05, "loss": 0.153, "step": 17625 }, { "epoch": 12.386507378777232, "grad_norm": 0.771547257900238, "learning_rate": 2.5079409697821504e-05, "loss": 0.1744, "step": 17626 }, { "epoch": 12.387210119465918, "grad_norm": 0.18693962693214417, "learning_rate": 2.5078941204029048e-05, "loss": 0.0537, "step": 17627 }, { "epoch": 12.387912860154604, "grad_norm": 0.19454768300056458, "learning_rate": 2.507847271023659e-05, "loss": 0.0302, "step": 17628 }, { "epoch": 12.38861560084329, "grad_norm": 0.09475011378526688, "learning_rate": 2.5078004216444135e-05, "loss": 0.0168, "step": 17629 }, { "epoch": 12.389318341531975, "grad_norm": 0.10152055323123932, "learning_rate": 2.5077535722651676e-05, "loss": 0.0133, "step": 17630 }, { "epoch": 12.390021082220661, "grad_norm": 0.07150039076805115, "learning_rate": 2.507706722885922e-05, "loss": 0.01, "step": 17631 }, { "epoch": 12.390723822909347, "grad_norm": 0.09996328502893448, "learning_rate": 2.5076598735066763e-05, "loss": 0.0141, "step": 17632 }, { "epoch": 12.391426563598033, "grad_norm": 0.06991909444332123, "learning_rate": 2.5076130241274303e-05, "loss": 0.007, "step": 17633 }, { "epoch": 12.392129304286719, "grad_norm": 0.11662372201681137, "learning_rate": 2.5075661747481844e-05, "loss": 0.024, "step": 17634 }, { "epoch": 12.392832044975403, "grad_norm": 0.1668948233127594, "learning_rate": 2.5075193253689387e-05, "loss": 0.0172, "step": 17635 }, { "epoch": 12.39353478566409, "grad_norm": 0.14004407823085785, "learning_rate": 2.507472475989693e-05, "loss": 0.0103, "step": 17636 }, { "epoch": 12.394237526352775, "grad_norm": 0.08601437509059906, "learning_rate": 2.5074256266104475e-05, "loss": 0.0156, "step": 17637 }, { "epoch": 12.394940267041461, "grad_norm": 0.10448506474494934, "learning_rate": 2.5073787772312015e-05, "loss": 0.01, "step": 17638 }, { "epoch": 12.395643007730147, "grad_norm": 0.1161288395524025, "learning_rate": 2.507331927851956e-05, "loss": 0.0236, "step": 17639 }, { "epoch": 12.396345748418833, "grad_norm": 0.18976087868213654, "learning_rate": 2.5072850784727103e-05, "loss": 0.0159, "step": 17640 }, { "epoch": 12.397048489107519, "grad_norm": 0.1789071261882782, "learning_rate": 2.5072382290934646e-05, "loss": 0.0227, "step": 17641 }, { "epoch": 12.397751229796205, "grad_norm": 0.2881259620189667, "learning_rate": 2.507191379714219e-05, "loss": 0.0334, "step": 17642 }, { "epoch": 12.39845397048489, "grad_norm": 0.18194356560707092, "learning_rate": 2.507144530334973e-05, "loss": 0.0157, "step": 17643 }, { "epoch": 12.399156711173577, "grad_norm": 0.18607571721076965, "learning_rate": 2.5070976809557274e-05, "loss": 0.0389, "step": 17644 }, { "epoch": 12.399859451862262, "grad_norm": 0.2995443344116211, "learning_rate": 2.5070508315764818e-05, "loss": 0.0658, "step": 17645 }, { "epoch": 12.400562192550948, "grad_norm": 0.2618078887462616, "learning_rate": 2.507003982197236e-05, "loss": 0.051, "step": 17646 }, { "epoch": 12.401264933239634, "grad_norm": 0.3462502956390381, "learning_rate": 2.5069571328179902e-05, "loss": 0.0596, "step": 17647 }, { "epoch": 12.40196767392832, "grad_norm": 0.4590485095977783, "learning_rate": 2.5069102834387446e-05, "loss": 0.1136, "step": 17648 }, { "epoch": 12.402670414617006, "grad_norm": 0.7079879641532898, "learning_rate": 2.506863434059499e-05, "loss": 0.1753, "step": 17649 }, { "epoch": 12.403373155305692, "grad_norm": 3.8351666927337646, "learning_rate": 2.5068165846802533e-05, "loss": 0.1532, "step": 17650 }, { "epoch": 12.404075895994378, "grad_norm": 1.4024771451950073, "learning_rate": 2.506769735301007e-05, "loss": 0.1652, "step": 17651 }, { "epoch": 12.404778636683064, "grad_norm": 0.22596867382526398, "learning_rate": 2.5067228859217614e-05, "loss": 0.07, "step": 17652 }, { "epoch": 12.40548137737175, "grad_norm": 0.11663524806499481, "learning_rate": 2.5066760365425158e-05, "loss": 0.0201, "step": 17653 }, { "epoch": 12.406184118060436, "grad_norm": 0.12330009788274765, "learning_rate": 2.50662918716327e-05, "loss": 0.0239, "step": 17654 }, { "epoch": 12.406886858749122, "grad_norm": 0.09159864485263824, "learning_rate": 2.5065823377840245e-05, "loss": 0.0161, "step": 17655 }, { "epoch": 12.407589599437808, "grad_norm": 0.13453568518161774, "learning_rate": 2.5065354884047785e-05, "loss": 0.0211, "step": 17656 }, { "epoch": 12.408292340126494, "grad_norm": 0.1792091280221939, "learning_rate": 2.506488639025533e-05, "loss": 0.0156, "step": 17657 }, { "epoch": 12.40899508081518, "grad_norm": 0.14522868394851685, "learning_rate": 2.5064417896462873e-05, "loss": 0.0138, "step": 17658 }, { "epoch": 12.409697821503865, "grad_norm": 0.11145489662885666, "learning_rate": 2.5063949402670417e-05, "loss": 0.0258, "step": 17659 }, { "epoch": 12.410400562192551, "grad_norm": 0.742091953754425, "learning_rate": 2.5063480908877957e-05, "loss": 0.0266, "step": 17660 }, { "epoch": 12.411103302881237, "grad_norm": 0.1142115518450737, "learning_rate": 2.50630124150855e-05, "loss": 0.0122, "step": 17661 }, { "epoch": 12.411806043569923, "grad_norm": 0.11718379706144333, "learning_rate": 2.5062543921293044e-05, "loss": 0.0167, "step": 17662 }, { "epoch": 12.412508784258609, "grad_norm": 0.0811251774430275, "learning_rate": 2.5062075427500588e-05, "loss": 0.0166, "step": 17663 }, { "epoch": 12.413211524947295, "grad_norm": 0.20312513411045074, "learning_rate": 2.506160693370813e-05, "loss": 0.0228, "step": 17664 }, { "epoch": 12.41391426563598, "grad_norm": 0.14383138716220856, "learning_rate": 2.5061138439915672e-05, "loss": 0.0161, "step": 17665 }, { "epoch": 12.414617006324667, "grad_norm": 0.23440420627593994, "learning_rate": 2.5060669946123216e-05, "loss": 0.0247, "step": 17666 }, { "epoch": 12.415319747013353, "grad_norm": 0.1430150419473648, "learning_rate": 2.506020145233076e-05, "loss": 0.0241, "step": 17667 }, { "epoch": 12.416022487702039, "grad_norm": 0.14514315128326416, "learning_rate": 2.50597329585383e-05, "loss": 0.0376, "step": 17668 }, { "epoch": 12.416725228390725, "grad_norm": 0.29440632462501526, "learning_rate": 2.505926446474584e-05, "loss": 0.0274, "step": 17669 }, { "epoch": 12.41742796907941, "grad_norm": 0.2513270676136017, "learning_rate": 2.5058795970953384e-05, "loss": 0.0555, "step": 17670 }, { "epoch": 12.418130709768096, "grad_norm": 0.3645046055316925, "learning_rate": 2.5058327477160928e-05, "loss": 0.0387, "step": 17671 }, { "epoch": 12.41883345045678, "grad_norm": 0.25082582235336304, "learning_rate": 2.505785898336847e-05, "loss": 0.0645, "step": 17672 }, { "epoch": 12.419536191145466, "grad_norm": 1.3201872110366821, "learning_rate": 2.5057390489576012e-05, "loss": 0.091, "step": 17673 }, { "epoch": 12.420238931834152, "grad_norm": 0.5596537590026855, "learning_rate": 2.5056921995783555e-05, "loss": 0.103, "step": 17674 }, { "epoch": 12.420941672522838, "grad_norm": 0.7897912859916687, "learning_rate": 2.50564535019911e-05, "loss": 0.1442, "step": 17675 }, { "epoch": 12.421644413211524, "grad_norm": 0.8183876872062683, "learning_rate": 2.5055985008198643e-05, "loss": 0.1895, "step": 17676 }, { "epoch": 12.42234715390021, "grad_norm": 0.19688084721565247, "learning_rate": 2.5055516514406183e-05, "loss": 0.0717, "step": 17677 }, { "epoch": 12.423049894588896, "grad_norm": 0.1426258534193039, "learning_rate": 2.5055048020613727e-05, "loss": 0.017, "step": 17678 }, { "epoch": 12.423752635277582, "grad_norm": 0.2711925506591797, "learning_rate": 2.505457952682127e-05, "loss": 0.0318, "step": 17679 }, { "epoch": 12.424455375966268, "grad_norm": 0.09099645167589188, "learning_rate": 2.5054111033028814e-05, "loss": 0.0138, "step": 17680 }, { "epoch": 12.425158116654954, "grad_norm": 0.12908388674259186, "learning_rate": 2.5053642539236358e-05, "loss": 0.0174, "step": 17681 }, { "epoch": 12.42586085734364, "grad_norm": 0.20045824348926544, "learning_rate": 2.50531740454439e-05, "loss": 0.0178, "step": 17682 }, { "epoch": 12.426563598032326, "grad_norm": 0.12930235266685486, "learning_rate": 2.5052705551651442e-05, "loss": 0.0117, "step": 17683 }, { "epoch": 12.427266338721012, "grad_norm": 0.12404520064592361, "learning_rate": 2.5052237057858986e-05, "loss": 0.0133, "step": 17684 }, { "epoch": 12.427969079409698, "grad_norm": 0.6362631916999817, "learning_rate": 2.505176856406653e-05, "loss": 0.0285, "step": 17685 }, { "epoch": 12.428671820098383, "grad_norm": 0.15882690250873566, "learning_rate": 2.5051300070274067e-05, "loss": 0.0126, "step": 17686 }, { "epoch": 12.42937456078707, "grad_norm": 0.6551979184150696, "learning_rate": 2.505083157648161e-05, "loss": 0.0488, "step": 17687 }, { "epoch": 12.430077301475755, "grad_norm": 0.120657779276371, "learning_rate": 2.5050363082689154e-05, "loss": 0.0137, "step": 17688 }, { "epoch": 12.430780042164441, "grad_norm": 0.18490083515644073, "learning_rate": 2.5049894588896698e-05, "loss": 0.0227, "step": 17689 }, { "epoch": 12.431482782853127, "grad_norm": 0.27715420722961426, "learning_rate": 2.5049426095104238e-05, "loss": 0.0431, "step": 17690 }, { "epoch": 12.432185523541813, "grad_norm": 0.12177397310733795, "learning_rate": 2.5048957601311782e-05, "loss": 0.0148, "step": 17691 }, { "epoch": 12.432888264230499, "grad_norm": 0.1673388034105301, "learning_rate": 2.5048489107519326e-05, "loss": 0.0223, "step": 17692 }, { "epoch": 12.433591004919185, "grad_norm": 0.36262086033821106, "learning_rate": 2.504802061372687e-05, "loss": 0.0287, "step": 17693 }, { "epoch": 12.43429374560787, "grad_norm": 0.16649846732616425, "learning_rate": 2.5047552119934413e-05, "loss": 0.0261, "step": 17694 }, { "epoch": 12.434996486296557, "grad_norm": 0.3565656244754791, "learning_rate": 2.5047083626141953e-05, "loss": 0.0415, "step": 17695 }, { "epoch": 12.435699226985243, "grad_norm": 0.18733997642993927, "learning_rate": 2.5046615132349497e-05, "loss": 0.0256, "step": 17696 }, { "epoch": 12.436401967673929, "grad_norm": 0.33895421028137207, "learning_rate": 2.504614663855704e-05, "loss": 0.0893, "step": 17697 }, { "epoch": 12.437104708362615, "grad_norm": 0.2970855236053467, "learning_rate": 2.5045678144764585e-05, "loss": 0.0874, "step": 17698 }, { "epoch": 12.4378074490513, "grad_norm": 0.41788625717163086, "learning_rate": 2.5045209650972125e-05, "loss": 0.1225, "step": 17699 }, { "epoch": 12.438510189739986, "grad_norm": 0.6374841928482056, "learning_rate": 2.504474115717967e-05, "loss": 0.1587, "step": 17700 }, { "epoch": 12.439212930428672, "grad_norm": 6.361976146697998, "learning_rate": 2.5044272663387212e-05, "loss": 0.1973, "step": 17701 }, { "epoch": 12.439915671117358, "grad_norm": 0.23799222707748413, "learning_rate": 2.5043804169594756e-05, "loss": 0.0645, "step": 17702 }, { "epoch": 12.440618411806044, "grad_norm": 0.4451868236064911, "learning_rate": 2.5043335675802293e-05, "loss": 0.0244, "step": 17703 }, { "epoch": 12.44132115249473, "grad_norm": 0.1098121851682663, "learning_rate": 2.5042867182009837e-05, "loss": 0.0167, "step": 17704 }, { "epoch": 12.442023893183416, "grad_norm": 0.07384580373764038, "learning_rate": 2.504239868821738e-05, "loss": 0.0123, "step": 17705 }, { "epoch": 12.442726633872102, "grad_norm": 0.10810230672359467, "learning_rate": 2.5041930194424924e-05, "loss": 0.0154, "step": 17706 }, { "epoch": 12.443429374560788, "grad_norm": 0.1677965670824051, "learning_rate": 2.5041461700632468e-05, "loss": 0.0104, "step": 17707 }, { "epoch": 12.444132115249474, "grad_norm": 0.14793765544891357, "learning_rate": 2.5040993206840008e-05, "loss": 0.0173, "step": 17708 }, { "epoch": 12.44483485593816, "grad_norm": 0.1959816962480545, "learning_rate": 2.5040524713047552e-05, "loss": 0.0198, "step": 17709 }, { "epoch": 12.445537596626846, "grad_norm": 0.15803489089012146, "learning_rate": 2.5040056219255096e-05, "loss": 0.0246, "step": 17710 }, { "epoch": 12.44624033731553, "grad_norm": 0.5921167731285095, "learning_rate": 2.503958772546264e-05, "loss": 0.0153, "step": 17711 }, { "epoch": 12.446943078004216, "grad_norm": 0.11312820017337799, "learning_rate": 2.503911923167018e-05, "loss": 0.0191, "step": 17712 }, { "epoch": 12.447645818692902, "grad_norm": 0.07721252739429474, "learning_rate": 2.5038650737877723e-05, "loss": 0.009, "step": 17713 }, { "epoch": 12.448348559381587, "grad_norm": 0.11872278898954391, "learning_rate": 2.5038182244085267e-05, "loss": 0.0124, "step": 17714 }, { "epoch": 12.449051300070273, "grad_norm": 0.17094749212265015, "learning_rate": 2.503771375029281e-05, "loss": 0.0144, "step": 17715 }, { "epoch": 12.44975404075896, "grad_norm": 0.12570379674434662, "learning_rate": 2.503724525650035e-05, "loss": 0.0335, "step": 17716 }, { "epoch": 12.450456781447645, "grad_norm": 0.17491690814495087, "learning_rate": 2.5036776762707895e-05, "loss": 0.028, "step": 17717 }, { "epoch": 12.451159522136331, "grad_norm": 0.20426243543624878, "learning_rate": 2.503630826891544e-05, "loss": 0.0219, "step": 17718 }, { "epoch": 12.451862262825017, "grad_norm": 0.18633192777633667, "learning_rate": 2.5035839775122982e-05, "loss": 0.0258, "step": 17719 }, { "epoch": 12.452565003513703, "grad_norm": 0.2354319542646408, "learning_rate": 2.5035371281330526e-05, "loss": 0.029, "step": 17720 }, { "epoch": 12.453267744202389, "grad_norm": 0.3194800317287445, "learning_rate": 2.5034902787538063e-05, "loss": 0.07, "step": 17721 }, { "epoch": 12.453970484891075, "grad_norm": 0.31583863496780396, "learning_rate": 2.5034434293745607e-05, "loss": 0.0501, "step": 17722 }, { "epoch": 12.45467322557976, "grad_norm": 0.4980621039867401, "learning_rate": 2.503396579995315e-05, "loss": 0.0912, "step": 17723 }, { "epoch": 12.455375966268447, "grad_norm": 0.5228587985038757, "learning_rate": 2.5033497306160694e-05, "loss": 0.122, "step": 17724 }, { "epoch": 12.456078706957133, "grad_norm": 0.8725787401199341, "learning_rate": 2.5033028812368235e-05, "loss": 0.156, "step": 17725 }, { "epoch": 12.456781447645819, "grad_norm": 0.9237148761749268, "learning_rate": 2.503256031857578e-05, "loss": 0.1854, "step": 17726 }, { "epoch": 12.457484188334504, "grad_norm": 0.1639995276927948, "learning_rate": 2.5032091824783322e-05, "loss": 0.0594, "step": 17727 }, { "epoch": 12.45818692902319, "grad_norm": 0.145196795463562, "learning_rate": 2.5031623330990866e-05, "loss": 0.0196, "step": 17728 }, { "epoch": 12.458889669711876, "grad_norm": 0.12763206660747528, "learning_rate": 2.5031154837198406e-05, "loss": 0.0264, "step": 17729 }, { "epoch": 12.459592410400562, "grad_norm": 0.08403603732585907, "learning_rate": 2.503068634340595e-05, "loss": 0.0193, "step": 17730 }, { "epoch": 12.460295151089248, "grad_norm": 0.16463328897953033, "learning_rate": 2.5030217849613494e-05, "loss": 0.0154, "step": 17731 }, { "epoch": 12.460997891777934, "grad_norm": 0.21028631925582886, "learning_rate": 2.5029749355821037e-05, "loss": 0.0216, "step": 17732 }, { "epoch": 12.46170063246662, "grad_norm": 0.2519943118095398, "learning_rate": 2.502928086202858e-05, "loss": 0.0235, "step": 17733 }, { "epoch": 12.462403373155306, "grad_norm": 0.11903855204582214, "learning_rate": 2.502881236823612e-05, "loss": 0.0353, "step": 17734 }, { "epoch": 12.463106113843992, "grad_norm": 0.23710544407367706, "learning_rate": 2.5028343874443665e-05, "loss": 0.0276, "step": 17735 }, { "epoch": 12.463808854532678, "grad_norm": 0.13764849305152893, "learning_rate": 2.502787538065121e-05, "loss": 0.0103, "step": 17736 }, { "epoch": 12.464511595221364, "grad_norm": 0.09768187999725342, "learning_rate": 2.5027406886858753e-05, "loss": 0.0113, "step": 17737 }, { "epoch": 12.46521433591005, "grad_norm": 0.286088228225708, "learning_rate": 2.502693839306629e-05, "loss": 0.0195, "step": 17738 }, { "epoch": 12.465917076598735, "grad_norm": 0.10411256551742554, "learning_rate": 2.5026469899273833e-05, "loss": 0.0181, "step": 17739 }, { "epoch": 12.466619817287421, "grad_norm": 0.07627470046281815, "learning_rate": 2.5026001405481377e-05, "loss": 0.013, "step": 17740 }, { "epoch": 12.467322557976107, "grad_norm": 0.15830297768115997, "learning_rate": 2.502553291168892e-05, "loss": 0.0283, "step": 17741 }, { "epoch": 12.468025298664793, "grad_norm": 0.18944652378559113, "learning_rate": 2.502506441789646e-05, "loss": 0.0295, "step": 17742 }, { "epoch": 12.46872803935348, "grad_norm": 0.2633588910102844, "learning_rate": 2.5024595924104005e-05, "loss": 0.023, "step": 17743 }, { "epoch": 12.469430780042165, "grad_norm": 0.5716639161109924, "learning_rate": 2.502412743031155e-05, "loss": 0.034, "step": 17744 }, { "epoch": 12.470133520730851, "grad_norm": 0.13972270488739014, "learning_rate": 2.5023658936519092e-05, "loss": 0.0249, "step": 17745 }, { "epoch": 12.470836261419537, "grad_norm": 0.21437489986419678, "learning_rate": 2.5023190442726636e-05, "loss": 0.0289, "step": 17746 }, { "epoch": 12.471539002108223, "grad_norm": 1.2417229413986206, "learning_rate": 2.5022721948934176e-05, "loss": 0.0783, "step": 17747 }, { "epoch": 12.472241742796909, "grad_norm": 0.3683626353740692, "learning_rate": 2.502225345514172e-05, "loss": 0.1146, "step": 17748 }, { "epoch": 12.472944483485595, "grad_norm": 0.8261152505874634, "learning_rate": 2.5021784961349264e-05, "loss": 0.1345, "step": 17749 }, { "epoch": 12.473647224174279, "grad_norm": 0.8236269354820251, "learning_rate": 2.5021316467556807e-05, "loss": 0.1679, "step": 17750 }, { "epoch": 12.474349964862965, "grad_norm": 4.695771217346191, "learning_rate": 2.5020847973764348e-05, "loss": 0.2095, "step": 17751 }, { "epoch": 12.47505270555165, "grad_norm": 0.17260190844535828, "learning_rate": 2.502037947997189e-05, "loss": 0.0486, "step": 17752 }, { "epoch": 12.475755446240337, "grad_norm": 0.48173394799232483, "learning_rate": 2.5019910986179435e-05, "loss": 0.0207, "step": 17753 }, { "epoch": 12.476458186929023, "grad_norm": 0.16840405762195587, "learning_rate": 2.501944249238698e-05, "loss": 0.0295, "step": 17754 }, { "epoch": 12.477160927617708, "grad_norm": 0.07449311763048172, "learning_rate": 2.501897399859452e-05, "loss": 0.0111, "step": 17755 }, { "epoch": 12.477863668306394, "grad_norm": 0.09372365474700928, "learning_rate": 2.501850550480206e-05, "loss": 0.0209, "step": 17756 }, { "epoch": 12.47856640899508, "grad_norm": 0.48556816577911377, "learning_rate": 2.5018037011009603e-05, "loss": 0.011, "step": 17757 }, { "epoch": 12.479269149683766, "grad_norm": 0.1664675623178482, "learning_rate": 2.5017568517217147e-05, "loss": 0.0133, "step": 17758 }, { "epoch": 12.479971890372452, "grad_norm": 0.12795570492744446, "learning_rate": 2.501710002342469e-05, "loss": 0.0152, "step": 17759 }, { "epoch": 12.480674631061138, "grad_norm": 0.3722449541091919, "learning_rate": 2.501663152963223e-05, "loss": 0.0265, "step": 17760 }, { "epoch": 12.481377371749824, "grad_norm": 0.4171329736709595, "learning_rate": 2.5016163035839775e-05, "loss": 0.0138, "step": 17761 }, { "epoch": 12.48208011243851, "grad_norm": 0.3304683566093445, "learning_rate": 2.501569454204732e-05, "loss": 0.0275, "step": 17762 }, { "epoch": 12.482782853127196, "grad_norm": 0.2570972740650177, "learning_rate": 2.5015226048254862e-05, "loss": 0.0216, "step": 17763 }, { "epoch": 12.483485593815882, "grad_norm": 0.2131836861371994, "learning_rate": 2.5014757554462403e-05, "loss": 0.0181, "step": 17764 }, { "epoch": 12.484188334504568, "grad_norm": 0.2650841176509857, "learning_rate": 2.5014289060669946e-05, "loss": 0.027, "step": 17765 }, { "epoch": 12.484891075193254, "grad_norm": 0.11776352673768997, "learning_rate": 2.501382056687749e-05, "loss": 0.0256, "step": 17766 }, { "epoch": 12.48559381588194, "grad_norm": 0.19163987040519714, "learning_rate": 2.5013352073085034e-05, "loss": 0.0349, "step": 17767 }, { "epoch": 12.486296556570625, "grad_norm": 0.15931406617164612, "learning_rate": 2.5012883579292574e-05, "loss": 0.0249, "step": 17768 }, { "epoch": 12.486999297259311, "grad_norm": 0.22938287258148193, "learning_rate": 2.5012415085500118e-05, "loss": 0.0451, "step": 17769 }, { "epoch": 12.487702037947997, "grad_norm": 0.33050936460494995, "learning_rate": 2.501194659170766e-05, "loss": 0.0301, "step": 17770 }, { "epoch": 12.488404778636683, "grad_norm": 1.1449306011199951, "learning_rate": 2.5011478097915205e-05, "loss": 0.0559, "step": 17771 }, { "epoch": 12.489107519325369, "grad_norm": 0.46403369307518005, "learning_rate": 2.501100960412275e-05, "loss": 0.0593, "step": 17772 }, { "epoch": 12.489810260014055, "grad_norm": 0.9433634877204895, "learning_rate": 2.5010541110330286e-05, "loss": 0.0817, "step": 17773 }, { "epoch": 12.490513000702741, "grad_norm": 0.7223585247993469, "learning_rate": 2.501007261653783e-05, "loss": 0.1464, "step": 17774 }, { "epoch": 12.491215741391427, "grad_norm": 0.6452442407608032, "learning_rate": 2.5009604122745373e-05, "loss": 0.1601, "step": 17775 }, { "epoch": 12.491918482080113, "grad_norm": 1.227434515953064, "learning_rate": 2.5009135628952917e-05, "loss": 0.1661, "step": 17776 }, { "epoch": 12.492621222768799, "grad_norm": 0.34154385328292847, "learning_rate": 2.5008667135160458e-05, "loss": 0.081, "step": 17777 }, { "epoch": 12.493323963457485, "grad_norm": 0.13490380346775055, "learning_rate": 2.5008198641368e-05, "loss": 0.0178, "step": 17778 }, { "epoch": 12.49402670414617, "grad_norm": 0.19492383301258087, "learning_rate": 2.5007730147575545e-05, "loss": 0.0221, "step": 17779 }, { "epoch": 12.494729444834856, "grad_norm": 0.13195830583572388, "learning_rate": 2.500726165378309e-05, "loss": 0.0177, "step": 17780 }, { "epoch": 12.495432185523542, "grad_norm": 0.110221728682518, "learning_rate": 2.5006793159990632e-05, "loss": 0.0152, "step": 17781 }, { "epoch": 12.496134926212228, "grad_norm": 0.10837051272392273, "learning_rate": 2.5006324666198173e-05, "loss": 0.0117, "step": 17782 }, { "epoch": 12.496837666900914, "grad_norm": 0.1300802081823349, "learning_rate": 2.5005856172405716e-05, "loss": 0.0242, "step": 17783 }, { "epoch": 12.4975404075896, "grad_norm": 0.43539685010910034, "learning_rate": 2.500538767861326e-05, "loss": 0.0232, "step": 17784 }, { "epoch": 12.498243148278286, "grad_norm": 0.11228159815073013, "learning_rate": 2.5004919184820804e-05, "loss": 0.0159, "step": 17785 }, { "epoch": 12.498945888966972, "grad_norm": 0.12690408527851105, "learning_rate": 2.5004450691028344e-05, "loss": 0.0139, "step": 17786 }, { "epoch": 12.499648629655656, "grad_norm": 0.1192822977900505, "learning_rate": 2.5003982197235888e-05, "loss": 0.0222, "step": 17787 }, { "epoch": 12.500351370344344, "grad_norm": 0.15463009476661682, "learning_rate": 2.5003513703443432e-05, "loss": 0.0098, "step": 17788 }, { "epoch": 12.501054111033028, "grad_norm": 0.6285418272018433, "learning_rate": 2.5003045209650975e-05, "loss": 0.0254, "step": 17789 }, { "epoch": 12.501756851721714, "grad_norm": 0.16032688319683075, "learning_rate": 2.5002576715858512e-05, "loss": 0.0136, "step": 17790 }, { "epoch": 12.5024595924104, "grad_norm": 0.10945563018321991, "learning_rate": 2.5002108222066056e-05, "loss": 0.0264, "step": 17791 }, { "epoch": 12.503162333099086, "grad_norm": 1.9840288162231445, "learning_rate": 2.50016397282736e-05, "loss": 0.0185, "step": 17792 }, { "epoch": 12.503865073787772, "grad_norm": 0.18050521612167358, "learning_rate": 2.5001171234481144e-05, "loss": 0.0166, "step": 17793 }, { "epoch": 12.504567814476458, "grad_norm": 0.18491972982883453, "learning_rate": 2.5000702740688687e-05, "loss": 0.0346, "step": 17794 }, { "epoch": 12.505270555165144, "grad_norm": 0.5194740295410156, "learning_rate": 2.5000234246896228e-05, "loss": 0.0312, "step": 17795 }, { "epoch": 12.50597329585383, "grad_norm": 0.3112214207649231, "learning_rate": 2.499976575310377e-05, "loss": 0.0429, "step": 17796 }, { "epoch": 12.506676036542515, "grad_norm": 0.41705748438835144, "learning_rate": 2.4999297259311315e-05, "loss": 0.0456, "step": 17797 }, { "epoch": 12.507378777231201, "grad_norm": 0.8350920081138611, "learning_rate": 2.499882876551886e-05, "loss": 0.1395, "step": 17798 }, { "epoch": 12.508081517919887, "grad_norm": 1.1158641576766968, "learning_rate": 2.49983602717264e-05, "loss": 0.1313, "step": 17799 }, { "epoch": 12.508784258608573, "grad_norm": 2.826781749725342, "learning_rate": 2.4997891777933943e-05, "loss": 0.1679, "step": 17800 }, { "epoch": 12.509486999297259, "grad_norm": 1.3073205947875977, "learning_rate": 2.4997423284141487e-05, "loss": 0.2168, "step": 17801 }, { "epoch": 12.510189739985945, "grad_norm": 0.23377393186092377, "learning_rate": 2.499695479034903e-05, "loss": 0.0737, "step": 17802 }, { "epoch": 12.510892480674631, "grad_norm": 0.1559968739748001, "learning_rate": 2.499648629655657e-05, "loss": 0.0279, "step": 17803 }, { "epoch": 12.511595221363317, "grad_norm": 0.23125632107257843, "learning_rate": 2.4996017802764114e-05, "loss": 0.0213, "step": 17804 }, { "epoch": 12.512297962052003, "grad_norm": 0.11613134294748306, "learning_rate": 2.4995549308971658e-05, "loss": 0.0134, "step": 17805 }, { "epoch": 12.513000702740689, "grad_norm": 0.14081329107284546, "learning_rate": 2.4995080815179202e-05, "loss": 0.018, "step": 17806 }, { "epoch": 12.513703443429375, "grad_norm": 0.06617619842290878, "learning_rate": 2.4994612321386746e-05, "loss": 0.0096, "step": 17807 }, { "epoch": 12.51440618411806, "grad_norm": 0.10751168429851532, "learning_rate": 2.4994143827594283e-05, "loss": 0.0127, "step": 17808 }, { "epoch": 12.515108924806746, "grad_norm": 0.1353665441274643, "learning_rate": 2.4993675333801826e-05, "loss": 0.0198, "step": 17809 }, { "epoch": 12.515811665495432, "grad_norm": 0.5883471369743347, "learning_rate": 2.499320684000937e-05, "loss": 0.0219, "step": 17810 }, { "epoch": 12.516514406184118, "grad_norm": 0.2637944519519806, "learning_rate": 2.4992738346216914e-05, "loss": 0.0086, "step": 17811 }, { "epoch": 12.517217146872804, "grad_norm": 0.3331305682659149, "learning_rate": 2.4992269852424454e-05, "loss": 0.0207, "step": 17812 }, { "epoch": 12.51791988756149, "grad_norm": 0.14192254841327667, "learning_rate": 2.4991801358631998e-05, "loss": 0.0122, "step": 17813 }, { "epoch": 12.518622628250176, "grad_norm": 0.25451692938804626, "learning_rate": 2.499133286483954e-05, "loss": 0.02, "step": 17814 }, { "epoch": 12.519325368938862, "grad_norm": 0.15665194392204285, "learning_rate": 2.4990864371047085e-05, "loss": 0.029, "step": 17815 }, { "epoch": 12.520028109627548, "grad_norm": 0.509064257144928, "learning_rate": 2.4990395877254626e-05, "loss": 0.0387, "step": 17816 }, { "epoch": 12.520730850316234, "grad_norm": 0.21455487608909607, "learning_rate": 2.498992738346217e-05, "loss": 0.0218, "step": 17817 }, { "epoch": 12.52143359100492, "grad_norm": 0.19272390007972717, "learning_rate": 2.4989458889669713e-05, "loss": 0.0189, "step": 17818 }, { "epoch": 12.522136331693606, "grad_norm": 0.24202868342399597, "learning_rate": 2.4988990395877257e-05, "loss": 0.0383, "step": 17819 }, { "epoch": 12.522839072382292, "grad_norm": 0.26912158727645874, "learning_rate": 2.49885219020848e-05, "loss": 0.0279, "step": 17820 }, { "epoch": 12.523541813070977, "grad_norm": 1.078781008720398, "learning_rate": 2.498805340829234e-05, "loss": 0.0471, "step": 17821 }, { "epoch": 12.524244553759663, "grad_norm": 0.7076261639595032, "learning_rate": 2.4987584914499885e-05, "loss": 0.0631, "step": 17822 }, { "epoch": 12.52494729444835, "grad_norm": 0.5254914164543152, "learning_rate": 2.4987116420707428e-05, "loss": 0.1291, "step": 17823 }, { "epoch": 12.525650035137035, "grad_norm": 0.6717126965522766, "learning_rate": 2.4986647926914972e-05, "loss": 0.1268, "step": 17824 }, { "epoch": 12.526352775825721, "grad_norm": 0.839453399181366, "learning_rate": 2.498617943312251e-05, "loss": 0.1718, "step": 17825 }, { "epoch": 12.527055516514405, "grad_norm": 1.2330455780029297, "learning_rate": 2.4985710939330053e-05, "loss": 0.1642, "step": 17826 }, { "epoch": 12.527758257203093, "grad_norm": 0.1736983358860016, "learning_rate": 2.4985242445537596e-05, "loss": 0.0529, "step": 17827 }, { "epoch": 12.528460997891777, "grad_norm": 0.1463164985179901, "learning_rate": 2.498477395174514e-05, "loss": 0.0284, "step": 17828 }, { "epoch": 12.529163738580463, "grad_norm": 0.10251811891794205, "learning_rate": 2.498430545795268e-05, "loss": 0.015, "step": 17829 }, { "epoch": 12.529866479269149, "grad_norm": 0.12551763653755188, "learning_rate": 2.4983836964160224e-05, "loss": 0.0081, "step": 17830 }, { "epoch": 12.530569219957835, "grad_norm": 0.20131328701972961, "learning_rate": 2.4983368470367768e-05, "loss": 0.0206, "step": 17831 }, { "epoch": 12.53127196064652, "grad_norm": 0.26451122760772705, "learning_rate": 2.498289997657531e-05, "loss": 0.0179, "step": 17832 }, { "epoch": 12.531974701335207, "grad_norm": 0.22018800675868988, "learning_rate": 2.4982431482782855e-05, "loss": 0.0185, "step": 17833 }, { "epoch": 12.532677442023893, "grad_norm": 0.060952991247177124, "learning_rate": 2.4981962988990396e-05, "loss": 0.0058, "step": 17834 }, { "epoch": 12.533380182712579, "grad_norm": 0.09504242986440659, "learning_rate": 2.498149449519794e-05, "loss": 0.0188, "step": 17835 }, { "epoch": 12.534082923401265, "grad_norm": 0.1037026196718216, "learning_rate": 2.4981026001405483e-05, "loss": 0.0107, "step": 17836 }, { "epoch": 12.53478566408995, "grad_norm": 0.19542881846427917, "learning_rate": 2.4980557507613027e-05, "loss": 0.0249, "step": 17837 }, { "epoch": 12.535488404778636, "grad_norm": 0.22397419810295105, "learning_rate": 2.4980089013820567e-05, "loss": 0.0118, "step": 17838 }, { "epoch": 12.536191145467322, "grad_norm": 0.08371249586343765, "learning_rate": 2.497962052002811e-05, "loss": 0.0119, "step": 17839 }, { "epoch": 12.536893886156008, "grad_norm": 0.32718348503112793, "learning_rate": 2.4979152026235655e-05, "loss": 0.0351, "step": 17840 }, { "epoch": 12.537596626844694, "grad_norm": 0.12914450466632843, "learning_rate": 2.49786835324432e-05, "loss": 0.0247, "step": 17841 }, { "epoch": 12.53829936753338, "grad_norm": 0.3107295632362366, "learning_rate": 2.4978215038650735e-05, "loss": 0.0306, "step": 17842 }, { "epoch": 12.539002108222066, "grad_norm": 0.10933071374893188, "learning_rate": 2.497774654485828e-05, "loss": 0.0115, "step": 17843 }, { "epoch": 12.539704848910752, "grad_norm": 0.17560674250125885, "learning_rate": 2.4977278051065823e-05, "loss": 0.0404, "step": 17844 }, { "epoch": 12.540407589599438, "grad_norm": 0.4472585916519165, "learning_rate": 2.4976809557273366e-05, "loss": 0.0252, "step": 17845 }, { "epoch": 12.541110330288124, "grad_norm": 0.4121381640434265, "learning_rate": 2.497634106348091e-05, "loss": 0.0528, "step": 17846 }, { "epoch": 12.54181307097681, "grad_norm": 0.339534193277359, "learning_rate": 2.497587256968845e-05, "loss": 0.0573, "step": 17847 }, { "epoch": 12.542515811665496, "grad_norm": 0.8756337761878967, "learning_rate": 2.4975404075895994e-05, "loss": 0.0932, "step": 17848 }, { "epoch": 12.543218552354181, "grad_norm": 0.4650201201438904, "learning_rate": 2.4974935582103538e-05, "loss": 0.1467, "step": 17849 }, { "epoch": 12.543921293042867, "grad_norm": 0.8360643982887268, "learning_rate": 2.4974467088311082e-05, "loss": 0.1799, "step": 17850 }, { "epoch": 12.544624033731553, "grad_norm": 1.5040223598480225, "learning_rate": 2.4973998594518622e-05, "loss": 0.2016, "step": 17851 }, { "epoch": 12.54532677442024, "grad_norm": 0.1766781210899353, "learning_rate": 2.4973530100726166e-05, "loss": 0.0688, "step": 17852 }, { "epoch": 12.546029515108925, "grad_norm": 0.1525607854127884, "learning_rate": 2.497306160693371e-05, "loss": 0.0173, "step": 17853 }, { "epoch": 12.546732255797611, "grad_norm": 0.14899127185344696, "learning_rate": 2.4972593113141253e-05, "loss": 0.0191, "step": 17854 }, { "epoch": 12.547434996486297, "grad_norm": 0.13949857652187347, "learning_rate": 2.4972124619348794e-05, "loss": 0.022, "step": 17855 }, { "epoch": 12.548137737174983, "grad_norm": 0.10442400723695755, "learning_rate": 2.4971656125556337e-05, "loss": 0.0154, "step": 17856 }, { "epoch": 12.548840477863669, "grad_norm": 0.15226975083351135, "learning_rate": 2.497118763176388e-05, "loss": 0.0137, "step": 17857 }, { "epoch": 12.549543218552355, "grad_norm": 0.10315598547458649, "learning_rate": 2.4970719137971425e-05, "loss": 0.0109, "step": 17858 }, { "epoch": 12.55024595924104, "grad_norm": 0.08893214166164398, "learning_rate": 2.497025064417897e-05, "loss": 0.0157, "step": 17859 }, { "epoch": 12.550948699929727, "grad_norm": 0.20507369935512543, "learning_rate": 2.4969782150386505e-05, "loss": 0.0226, "step": 17860 }, { "epoch": 12.551651440618413, "grad_norm": 0.16273438930511475, "learning_rate": 2.496931365659405e-05, "loss": 0.0188, "step": 17861 }, { "epoch": 12.552354181307098, "grad_norm": 0.1741442084312439, "learning_rate": 2.4968845162801593e-05, "loss": 0.0287, "step": 17862 }, { "epoch": 12.553056921995784, "grad_norm": 0.33950814604759216, "learning_rate": 2.4968376669009137e-05, "loss": 0.0102, "step": 17863 }, { "epoch": 12.55375966268447, "grad_norm": 0.21467521786689758, "learning_rate": 2.4967908175216677e-05, "loss": 0.022, "step": 17864 }, { "epoch": 12.554462403373154, "grad_norm": 0.23169980943202972, "learning_rate": 2.496743968142422e-05, "loss": 0.0121, "step": 17865 }, { "epoch": 12.55516514406184, "grad_norm": 0.20773158967494965, "learning_rate": 2.4966971187631764e-05, "loss": 0.017, "step": 17866 }, { "epoch": 12.555867884750526, "grad_norm": 0.31467655301094055, "learning_rate": 2.4966502693839308e-05, "loss": 0.0408, "step": 17867 }, { "epoch": 12.556570625439212, "grad_norm": 0.16224151849746704, "learning_rate": 2.496603420004685e-05, "loss": 0.015, "step": 17868 }, { "epoch": 12.557273366127898, "grad_norm": 0.27356573939323425, "learning_rate": 2.4965565706254392e-05, "loss": 0.0266, "step": 17869 }, { "epoch": 12.557976106816584, "grad_norm": 0.2192254364490509, "learning_rate": 2.4965097212461936e-05, "loss": 0.0473, "step": 17870 }, { "epoch": 12.55867884750527, "grad_norm": 0.23231208324432373, "learning_rate": 2.496462871866948e-05, "loss": 0.0527, "step": 17871 }, { "epoch": 12.559381588193956, "grad_norm": 0.3772641718387604, "learning_rate": 2.4964160224877023e-05, "loss": 0.0633, "step": 17872 }, { "epoch": 12.560084328882642, "grad_norm": 0.3999934494495392, "learning_rate": 2.4963691731084564e-05, "loss": 0.0784, "step": 17873 }, { "epoch": 12.560787069571328, "grad_norm": 3.2434303760528564, "learning_rate": 2.4963223237292107e-05, "loss": 0.1362, "step": 17874 }, { "epoch": 12.561489810260014, "grad_norm": 0.5994413495063782, "learning_rate": 2.496275474349965e-05, "loss": 0.1489, "step": 17875 }, { "epoch": 12.5621925509487, "grad_norm": 1.2989124059677124, "learning_rate": 2.4962286249707195e-05, "loss": 0.2362, "step": 17876 }, { "epoch": 12.562895291637385, "grad_norm": 0.29141849279403687, "learning_rate": 2.4961817755914732e-05, "loss": 0.0757, "step": 17877 }, { "epoch": 12.563598032326071, "grad_norm": 0.3267176151275635, "learning_rate": 2.4961349262122276e-05, "loss": 0.023, "step": 17878 }, { "epoch": 12.564300773014757, "grad_norm": 0.12624330818653107, "learning_rate": 2.496088076832982e-05, "loss": 0.0174, "step": 17879 }, { "epoch": 12.565003513703443, "grad_norm": 0.11212077736854553, "learning_rate": 2.4960412274537363e-05, "loss": 0.0164, "step": 17880 }, { "epoch": 12.56570625439213, "grad_norm": 0.08360809832811356, "learning_rate": 2.4959943780744903e-05, "loss": 0.0208, "step": 17881 }, { "epoch": 12.566408995080815, "grad_norm": 0.1797352284193039, "learning_rate": 2.4959475286952447e-05, "loss": 0.0147, "step": 17882 }, { "epoch": 12.567111735769501, "grad_norm": 0.09032529592514038, "learning_rate": 2.495900679315999e-05, "loss": 0.0116, "step": 17883 }, { "epoch": 12.567814476458187, "grad_norm": 0.08724325895309448, "learning_rate": 2.4958538299367534e-05, "loss": 0.0198, "step": 17884 }, { "epoch": 12.568517217146873, "grad_norm": 0.25262686610221863, "learning_rate": 2.4958069805575078e-05, "loss": 0.0189, "step": 17885 }, { "epoch": 12.569219957835559, "grad_norm": 0.13968448340892792, "learning_rate": 2.495760131178262e-05, "loss": 0.0234, "step": 17886 }, { "epoch": 12.569922698524245, "grad_norm": 0.10657142102718353, "learning_rate": 2.4957132817990162e-05, "loss": 0.024, "step": 17887 }, { "epoch": 12.57062543921293, "grad_norm": 0.1246768981218338, "learning_rate": 2.4956664324197706e-05, "loss": 0.0168, "step": 17888 }, { "epoch": 12.571328179901617, "grad_norm": 0.174928680062294, "learning_rate": 2.495619583040525e-05, "loss": 0.0309, "step": 17889 }, { "epoch": 12.572030920590302, "grad_norm": 0.10855799913406372, "learning_rate": 2.495572733661279e-05, "loss": 0.0103, "step": 17890 }, { "epoch": 12.572733661278988, "grad_norm": 0.125235915184021, "learning_rate": 2.4955258842820334e-05, "loss": 0.0181, "step": 17891 }, { "epoch": 12.573436401967674, "grad_norm": 0.13330334424972534, "learning_rate": 2.4954790349027878e-05, "loss": 0.0271, "step": 17892 }, { "epoch": 12.57413914265636, "grad_norm": 0.13728448748588562, "learning_rate": 2.495432185523542e-05, "loss": 0.0221, "step": 17893 }, { "epoch": 12.574841883345046, "grad_norm": 0.18730323016643524, "learning_rate": 2.495385336144296e-05, "loss": 0.0514, "step": 17894 }, { "epoch": 12.575544624033732, "grad_norm": 0.23283924162387848, "learning_rate": 2.4953384867650502e-05, "loss": 0.0345, "step": 17895 }, { "epoch": 12.576247364722418, "grad_norm": 0.46302589774131775, "learning_rate": 2.4952916373858046e-05, "loss": 0.0592, "step": 17896 }, { "epoch": 12.576950105411104, "grad_norm": 0.19106963276863098, "learning_rate": 2.495244788006559e-05, "loss": 0.0546, "step": 17897 }, { "epoch": 12.57765284609979, "grad_norm": 0.3994178771972656, "learning_rate": 2.4951979386273133e-05, "loss": 0.0974, "step": 17898 }, { "epoch": 12.578355586788476, "grad_norm": 0.45775580406188965, "learning_rate": 2.4951510892480673e-05, "loss": 0.1549, "step": 17899 }, { "epoch": 12.579058327477162, "grad_norm": 1.1866530179977417, "learning_rate": 2.4951042398688217e-05, "loss": 0.1726, "step": 17900 }, { "epoch": 12.579761068165848, "grad_norm": 1.2994675636291504, "learning_rate": 2.495057390489576e-05, "loss": 0.1711, "step": 17901 }, { "epoch": 12.580463808854532, "grad_norm": 0.25782352685928345, "learning_rate": 2.4950105411103305e-05, "loss": 0.0656, "step": 17902 }, { "epoch": 12.58116654954322, "grad_norm": 0.10636984556913376, "learning_rate": 2.4949636917310845e-05, "loss": 0.0191, "step": 17903 }, { "epoch": 12.581869290231904, "grad_norm": 0.10661318153142929, "learning_rate": 2.494916842351839e-05, "loss": 0.0277, "step": 17904 }, { "epoch": 12.58257203092059, "grad_norm": 0.19232240319252014, "learning_rate": 2.4948699929725932e-05, "loss": 0.0209, "step": 17905 }, { "epoch": 12.583274771609275, "grad_norm": 0.09897097945213318, "learning_rate": 2.4948231435933476e-05, "loss": 0.0136, "step": 17906 }, { "epoch": 12.583977512297961, "grad_norm": 0.09782106429338455, "learning_rate": 2.4947762942141016e-05, "loss": 0.0147, "step": 17907 }, { "epoch": 12.584680252986647, "grad_norm": 0.11410778015851974, "learning_rate": 2.494729444834856e-05, "loss": 0.0133, "step": 17908 }, { "epoch": 12.585382993675333, "grad_norm": 0.15027591586112976, "learning_rate": 2.4946825954556104e-05, "loss": 0.0344, "step": 17909 }, { "epoch": 12.58608573436402, "grad_norm": 0.14076711237430573, "learning_rate": 2.4946357460763648e-05, "loss": 0.0294, "step": 17910 }, { "epoch": 12.586788475052705, "grad_norm": 0.05813632905483246, "learning_rate": 2.494588896697119e-05, "loss": 0.0085, "step": 17911 }, { "epoch": 12.587491215741391, "grad_norm": 0.22758394479751587, "learning_rate": 2.4945420473178728e-05, "loss": 0.022, "step": 17912 }, { "epoch": 12.588193956430077, "grad_norm": 0.13726264238357544, "learning_rate": 2.4944951979386272e-05, "loss": 0.0122, "step": 17913 }, { "epoch": 12.588896697118763, "grad_norm": 0.10444221645593643, "learning_rate": 2.4944483485593816e-05, "loss": 0.0182, "step": 17914 }, { "epoch": 12.589599437807449, "grad_norm": 0.4764136075973511, "learning_rate": 2.494401499180136e-05, "loss": 0.0153, "step": 17915 }, { "epoch": 12.590302178496135, "grad_norm": 0.1734360158443451, "learning_rate": 2.49435464980089e-05, "loss": 0.03, "step": 17916 }, { "epoch": 12.59100491918482, "grad_norm": 0.17173674702644348, "learning_rate": 2.4943078004216444e-05, "loss": 0.0365, "step": 17917 }, { "epoch": 12.591707659873506, "grad_norm": 0.13947820663452148, "learning_rate": 2.4942609510423987e-05, "loss": 0.0158, "step": 17918 }, { "epoch": 12.592410400562192, "grad_norm": 0.16175076365470886, "learning_rate": 2.494214101663153e-05, "loss": 0.0347, "step": 17919 }, { "epoch": 12.593113141250878, "grad_norm": 0.27630814909935, "learning_rate": 2.494167252283907e-05, "loss": 0.0625, "step": 17920 }, { "epoch": 12.593815881939564, "grad_norm": 0.3035903573036194, "learning_rate": 2.4941204029046615e-05, "loss": 0.038, "step": 17921 }, { "epoch": 12.59451862262825, "grad_norm": 0.7424221038818359, "learning_rate": 2.494073553525416e-05, "loss": 0.0714, "step": 17922 }, { "epoch": 12.595221363316936, "grad_norm": 0.6466213464736938, "learning_rate": 2.4940267041461702e-05, "loss": 0.0932, "step": 17923 }, { "epoch": 12.595924104005622, "grad_norm": 1.0775381326675415, "learning_rate": 2.4939798547669246e-05, "loss": 0.1286, "step": 17924 }, { "epoch": 12.596626844694308, "grad_norm": 0.9584313035011292, "learning_rate": 2.4939330053876787e-05, "loss": 0.1718, "step": 17925 }, { "epoch": 12.597329585382994, "grad_norm": 1.7538354396820068, "learning_rate": 2.493886156008433e-05, "loss": 0.1986, "step": 17926 }, { "epoch": 12.59803232607168, "grad_norm": 0.16307543218135834, "learning_rate": 2.4938393066291874e-05, "loss": 0.0625, "step": 17927 }, { "epoch": 12.598735066760366, "grad_norm": 0.11159342527389526, "learning_rate": 2.4937924572499418e-05, "loss": 0.0172, "step": 17928 }, { "epoch": 12.599437807449052, "grad_norm": 0.17595702409744263, "learning_rate": 2.4937456078706958e-05, "loss": 0.0133, "step": 17929 }, { "epoch": 12.600140548137738, "grad_norm": 0.06483189016580582, "learning_rate": 2.49369875849145e-05, "loss": 0.0093, "step": 17930 }, { "epoch": 12.600843288826423, "grad_norm": 0.10223883390426636, "learning_rate": 2.4936519091122042e-05, "loss": 0.019, "step": 17931 }, { "epoch": 12.60154602951511, "grad_norm": 0.07138120383024216, "learning_rate": 2.4936050597329586e-05, "loss": 0.009, "step": 17932 }, { "epoch": 12.602248770203795, "grad_norm": 0.12723861634731293, "learning_rate": 2.4935582103537126e-05, "loss": 0.0143, "step": 17933 }, { "epoch": 12.602951510892481, "grad_norm": 0.1736944317817688, "learning_rate": 2.493511360974467e-05, "loss": 0.016, "step": 17934 }, { "epoch": 12.603654251581167, "grad_norm": 0.2377680540084839, "learning_rate": 2.4934645115952214e-05, "loss": 0.027, "step": 17935 }, { "epoch": 12.604356992269853, "grad_norm": 0.12351670861244202, "learning_rate": 2.4934176622159757e-05, "loss": 0.012, "step": 17936 }, { "epoch": 12.605059732958539, "grad_norm": 0.5112298727035522, "learning_rate": 2.49337081283673e-05, "loss": 0.0299, "step": 17937 }, { "epoch": 12.605762473647225, "grad_norm": 0.11840961873531342, "learning_rate": 2.493323963457484e-05, "loss": 0.0122, "step": 17938 }, { "epoch": 12.60646521433591, "grad_norm": 0.19946406781673431, "learning_rate": 2.4932771140782385e-05, "loss": 0.0295, "step": 17939 }, { "epoch": 12.607167955024597, "grad_norm": 0.1487312614917755, "learning_rate": 2.493230264698993e-05, "loss": 0.0223, "step": 17940 }, { "epoch": 12.607870695713281, "grad_norm": 0.11222396790981293, "learning_rate": 2.4931834153197473e-05, "loss": 0.0241, "step": 17941 }, { "epoch": 12.608573436401969, "grad_norm": 0.20705612003803253, "learning_rate": 2.4931365659405013e-05, "loss": 0.0398, "step": 17942 }, { "epoch": 12.609276177090653, "grad_norm": 0.13629893958568573, "learning_rate": 2.4930897165612557e-05, "loss": 0.0206, "step": 17943 }, { "epoch": 12.609978917779339, "grad_norm": 0.901459813117981, "learning_rate": 2.49304286718201e-05, "loss": 0.032, "step": 17944 }, { "epoch": 12.610681658468025, "grad_norm": 0.2289549857378006, "learning_rate": 2.4929960178027644e-05, "loss": 0.0299, "step": 17945 }, { "epoch": 12.61138439915671, "grad_norm": 0.2460455447435379, "learning_rate": 2.4929491684235184e-05, "loss": 0.0439, "step": 17946 }, { "epoch": 12.612087139845396, "grad_norm": 0.36400020122528076, "learning_rate": 2.4929023190442725e-05, "loss": 0.0565, "step": 17947 }, { "epoch": 12.612789880534082, "grad_norm": 0.8371872901916504, "learning_rate": 2.492855469665027e-05, "loss": 0.1139, "step": 17948 }, { "epoch": 12.613492621222768, "grad_norm": 1.6465826034545898, "learning_rate": 2.4928086202857812e-05, "loss": 0.1369, "step": 17949 }, { "epoch": 12.614195361911454, "grad_norm": 1.6583545207977295, "learning_rate": 2.4927617709065356e-05, "loss": 0.1885, "step": 17950 }, { "epoch": 12.61489810260014, "grad_norm": 0.9725984334945679, "learning_rate": 2.4927149215272896e-05, "loss": 0.185, "step": 17951 }, { "epoch": 12.615600843288826, "grad_norm": 0.2241353839635849, "learning_rate": 2.492668072148044e-05, "loss": 0.0549, "step": 17952 }, { "epoch": 12.616303583977512, "grad_norm": 0.09501868486404419, "learning_rate": 2.4926212227687984e-05, "loss": 0.0186, "step": 17953 }, { "epoch": 12.617006324666198, "grad_norm": 0.0865088477730751, "learning_rate": 2.4925743733895527e-05, "loss": 0.0194, "step": 17954 }, { "epoch": 12.617709065354884, "grad_norm": 0.14443229138851166, "learning_rate": 2.4925275240103068e-05, "loss": 0.0214, "step": 17955 }, { "epoch": 12.61841180604357, "grad_norm": 0.5202565789222717, "learning_rate": 2.492480674631061e-05, "loss": 0.0103, "step": 17956 }, { "epoch": 12.619114546732256, "grad_norm": 0.11539514362812042, "learning_rate": 2.4924338252518155e-05, "loss": 0.0102, "step": 17957 }, { "epoch": 12.619817287420942, "grad_norm": 0.09027117490768433, "learning_rate": 2.49238697587257e-05, "loss": 0.0134, "step": 17958 }, { "epoch": 12.620520028109627, "grad_norm": 0.17386461794376373, "learning_rate": 2.492340126493324e-05, "loss": 0.0218, "step": 17959 }, { "epoch": 12.621222768798313, "grad_norm": 0.10422319173812866, "learning_rate": 2.4922932771140783e-05, "loss": 0.0128, "step": 17960 }, { "epoch": 12.621925509487, "grad_norm": 0.18732582032680511, "learning_rate": 2.4922464277348327e-05, "loss": 0.0208, "step": 17961 }, { "epoch": 12.622628250175685, "grad_norm": 0.12044090032577515, "learning_rate": 2.492199578355587e-05, "loss": 0.0173, "step": 17962 }, { "epoch": 12.623330990864371, "grad_norm": 0.09777134656906128, "learning_rate": 2.4921527289763414e-05, "loss": 0.0153, "step": 17963 }, { "epoch": 12.624033731553057, "grad_norm": 0.103151336312294, "learning_rate": 2.492105879597095e-05, "loss": 0.0281, "step": 17964 }, { "epoch": 12.624736472241743, "grad_norm": 0.06942661851644516, "learning_rate": 2.4920590302178495e-05, "loss": 0.0127, "step": 17965 }, { "epoch": 12.625439212930429, "grad_norm": 0.2149573266506195, "learning_rate": 2.492012180838604e-05, "loss": 0.0518, "step": 17966 }, { "epoch": 12.626141953619115, "grad_norm": 0.13016724586486816, "learning_rate": 2.4919653314593582e-05, "loss": 0.0319, "step": 17967 }, { "epoch": 12.6268446943078, "grad_norm": 0.40546295046806335, "learning_rate": 2.4919184820801123e-05, "loss": 0.0211, "step": 17968 }, { "epoch": 12.627547434996487, "grad_norm": 0.802260160446167, "learning_rate": 2.4918716327008666e-05, "loss": 0.0379, "step": 17969 }, { "epoch": 12.628250175685173, "grad_norm": 0.14843535423278809, "learning_rate": 2.491824783321621e-05, "loss": 0.0204, "step": 17970 }, { "epoch": 12.628952916373859, "grad_norm": 0.20426496863365173, "learning_rate": 2.4917779339423754e-05, "loss": 0.0486, "step": 17971 }, { "epoch": 12.629655657062544, "grad_norm": 0.3342609703540802, "learning_rate": 2.4917310845631298e-05, "loss": 0.0726, "step": 17972 }, { "epoch": 12.63035839775123, "grad_norm": 0.5705327391624451, "learning_rate": 2.4916842351838838e-05, "loss": 0.087, "step": 17973 }, { "epoch": 12.631061138439916, "grad_norm": 0.4367873966693878, "learning_rate": 2.491637385804638e-05, "loss": 0.1143, "step": 17974 }, { "epoch": 12.631763879128602, "grad_norm": 0.7896393537521362, "learning_rate": 2.4915905364253925e-05, "loss": 0.1481, "step": 17975 }, { "epoch": 12.632466619817288, "grad_norm": 0.6370651721954346, "learning_rate": 2.491543687046147e-05, "loss": 0.1821, "step": 17976 }, { "epoch": 12.633169360505974, "grad_norm": 0.21888141334056854, "learning_rate": 2.491496837666901e-05, "loss": 0.058, "step": 17977 }, { "epoch": 12.63387210119466, "grad_norm": 0.13973645865917206, "learning_rate": 2.4914499882876553e-05, "loss": 0.0201, "step": 17978 }, { "epoch": 12.634574841883346, "grad_norm": 0.12661008536815643, "learning_rate": 2.4914031389084097e-05, "loss": 0.0187, "step": 17979 }, { "epoch": 12.63527758257203, "grad_norm": 0.15759365260601044, "learning_rate": 2.491356289529164e-05, "loss": 0.018, "step": 17980 }, { "epoch": 12.635980323260716, "grad_norm": 0.509740948677063, "learning_rate": 2.491309440149918e-05, "loss": 0.0297, "step": 17981 }, { "epoch": 12.636683063949402, "grad_norm": 0.07121780514717102, "learning_rate": 2.491262590770672e-05, "loss": 0.006, "step": 17982 }, { "epoch": 12.637385804638088, "grad_norm": 0.151448056101799, "learning_rate": 2.4912157413914265e-05, "loss": 0.0179, "step": 17983 }, { "epoch": 12.638088545326774, "grad_norm": 0.19736234843730927, "learning_rate": 2.491168892012181e-05, "loss": 0.017, "step": 17984 }, { "epoch": 12.63879128601546, "grad_norm": 0.15415821969509125, "learning_rate": 2.4911220426329352e-05, "loss": 0.0169, "step": 17985 }, { "epoch": 12.639494026704146, "grad_norm": 0.043119367212057114, "learning_rate": 2.4910751932536893e-05, "loss": 0.0069, "step": 17986 }, { "epoch": 12.640196767392831, "grad_norm": 0.07282719016075134, "learning_rate": 2.4910283438744437e-05, "loss": 0.0081, "step": 17987 }, { "epoch": 12.640899508081517, "grad_norm": 0.3527582585811615, "learning_rate": 2.490981494495198e-05, "loss": 0.0219, "step": 17988 }, { "epoch": 12.641602248770203, "grad_norm": 0.12979324162006378, "learning_rate": 2.4909346451159524e-05, "loss": 0.0175, "step": 17989 }, { "epoch": 12.64230498945889, "grad_norm": 0.15899622440338135, "learning_rate": 2.4908877957367064e-05, "loss": 0.0252, "step": 17990 }, { "epoch": 12.643007730147575, "grad_norm": 0.24339249730110168, "learning_rate": 2.4908409463574608e-05, "loss": 0.0322, "step": 17991 }, { "epoch": 12.643710470836261, "grad_norm": 1.2418116331100464, "learning_rate": 2.4907940969782152e-05, "loss": 0.0298, "step": 17992 }, { "epoch": 12.644413211524947, "grad_norm": 0.09746818244457245, "learning_rate": 2.4907472475989695e-05, "loss": 0.0115, "step": 17993 }, { "epoch": 12.645115952213633, "grad_norm": 0.25167492032051086, "learning_rate": 2.4907003982197236e-05, "loss": 0.0307, "step": 17994 }, { "epoch": 12.645818692902319, "grad_norm": 0.16116555035114288, "learning_rate": 2.490653548840478e-05, "loss": 0.0205, "step": 17995 }, { "epoch": 12.646521433591005, "grad_norm": 0.18744874000549316, "learning_rate": 2.4906066994612323e-05, "loss": 0.0437, "step": 17996 }, { "epoch": 12.64722417427969, "grad_norm": 0.28368818759918213, "learning_rate": 2.4905598500819867e-05, "loss": 0.0579, "step": 17997 }, { "epoch": 12.647926914968377, "grad_norm": 0.9048852920532227, "learning_rate": 2.490513000702741e-05, "loss": 0.1062, "step": 17998 }, { "epoch": 12.648629655657063, "grad_norm": 0.5003566145896912, "learning_rate": 2.4904661513234948e-05, "loss": 0.1178, "step": 17999 }, { "epoch": 12.649332396345748, "grad_norm": 0.8997273445129395, "learning_rate": 2.490419301944249e-05, "loss": 0.176, "step": 18000 }, { "epoch": 12.649332396345748, "eval_cer": 0.19223047799853796, "eval_loss": 0.2594671845436096, "eval_runtime": 18.1162, "eval_samples_per_second": 250.495, "eval_steps_per_second": 0.828, "eval_wer": 0.3395156559872629, "step": 18000 }, { "epoch": 12.650035137034434, "grad_norm": 0.8432363271713257, "learning_rate": 2.4903724525650035e-05, "loss": 0.1954, "step": 18001 }, { "epoch": 12.65073787772312, "grad_norm": 0.39742782711982727, "learning_rate": 2.490325603185758e-05, "loss": 0.0711, "step": 18002 }, { "epoch": 12.651440618411806, "grad_norm": 0.09113101661205292, "learning_rate": 2.490278753806512e-05, "loss": 0.0233, "step": 18003 }, { "epoch": 12.652143359100492, "grad_norm": 0.09394107013940811, "learning_rate": 2.4902319044272663e-05, "loss": 0.0204, "step": 18004 }, { "epoch": 12.652846099789178, "grad_norm": 0.07945311814546585, "learning_rate": 2.4901850550480207e-05, "loss": 0.0131, "step": 18005 }, { "epoch": 12.653548840477864, "grad_norm": 0.11681398004293442, "learning_rate": 2.490138205668775e-05, "loss": 0.0174, "step": 18006 }, { "epoch": 12.65425158116655, "grad_norm": 0.2697887718677521, "learning_rate": 2.490091356289529e-05, "loss": 0.0249, "step": 18007 }, { "epoch": 12.654954321855236, "grad_norm": 0.2254757285118103, "learning_rate": 2.4900445069102834e-05, "loss": 0.0171, "step": 18008 }, { "epoch": 12.655657062543922, "grad_norm": 0.08955363929271698, "learning_rate": 2.4899976575310378e-05, "loss": 0.0183, "step": 18009 }, { "epoch": 12.656359803232608, "grad_norm": 0.1272512674331665, "learning_rate": 2.4899508081517922e-05, "loss": 0.0196, "step": 18010 }, { "epoch": 12.657062543921294, "grad_norm": 0.06781185418367386, "learning_rate": 2.4899039587725466e-05, "loss": 0.0057, "step": 18011 }, { "epoch": 12.65776528460998, "grad_norm": 0.10703978687524796, "learning_rate": 2.4898571093933006e-05, "loss": 0.0199, "step": 18012 }, { "epoch": 12.658468025298665, "grad_norm": 0.43573594093322754, "learning_rate": 2.489810260014055e-05, "loss": 0.0231, "step": 18013 }, { "epoch": 12.659170765987351, "grad_norm": 0.16162721812725067, "learning_rate": 2.4897634106348093e-05, "loss": 0.019, "step": 18014 }, { "epoch": 12.659873506676037, "grad_norm": 0.2776155173778534, "learning_rate": 2.4897165612555637e-05, "loss": 0.0167, "step": 18015 }, { "epoch": 12.660576247364723, "grad_norm": 0.56288081407547, "learning_rate": 2.4896697118763177e-05, "loss": 0.0442, "step": 18016 }, { "epoch": 12.66127898805341, "grad_norm": 0.5702728629112244, "learning_rate": 2.4896228624970718e-05, "loss": 0.0324, "step": 18017 }, { "epoch": 12.661981728742095, "grad_norm": 0.12717880308628082, "learning_rate": 2.489576013117826e-05, "loss": 0.0101, "step": 18018 }, { "epoch": 12.66268446943078, "grad_norm": 0.3253785967826843, "learning_rate": 2.4895291637385805e-05, "loss": 0.029, "step": 18019 }, { "epoch": 12.663387210119465, "grad_norm": 0.6965747475624084, "learning_rate": 2.4894823143593346e-05, "loss": 0.027, "step": 18020 }, { "epoch": 12.664089950808151, "grad_norm": 0.3577766418457031, "learning_rate": 2.489435464980089e-05, "loss": 0.053, "step": 18021 }, { "epoch": 12.664792691496837, "grad_norm": 0.23127874732017517, "learning_rate": 2.4893886156008433e-05, "loss": 0.0647, "step": 18022 }, { "epoch": 12.665495432185523, "grad_norm": 0.4755116105079651, "learning_rate": 2.4893417662215977e-05, "loss": 0.0999, "step": 18023 }, { "epoch": 12.666198172874209, "grad_norm": 0.8427507877349854, "learning_rate": 2.489294916842352e-05, "loss": 0.1444, "step": 18024 }, { "epoch": 12.666900913562895, "grad_norm": 1.0595566034317017, "learning_rate": 2.489248067463106e-05, "loss": 0.1926, "step": 18025 }, { "epoch": 12.66760365425158, "grad_norm": 0.58275306224823, "learning_rate": 2.4892012180838605e-05, "loss": 0.1664, "step": 18026 }, { "epoch": 12.668306394940267, "grad_norm": 0.8847208023071289, "learning_rate": 2.4891543687046148e-05, "loss": 0.0959, "step": 18027 }, { "epoch": 12.669009135628952, "grad_norm": 0.1195911094546318, "learning_rate": 2.4891075193253692e-05, "loss": 0.0245, "step": 18028 }, { "epoch": 12.669711876317638, "grad_norm": 0.23560121655464172, "learning_rate": 2.4890606699461232e-05, "loss": 0.0285, "step": 18029 }, { "epoch": 12.670414617006324, "grad_norm": 0.09337843209505081, "learning_rate": 2.4890138205668776e-05, "loss": 0.0152, "step": 18030 }, { "epoch": 12.67111735769501, "grad_norm": 0.08819843828678131, "learning_rate": 2.488966971187632e-05, "loss": 0.0179, "step": 18031 }, { "epoch": 12.671820098383696, "grad_norm": 0.10023728758096695, "learning_rate": 2.4889201218083864e-05, "loss": 0.0099, "step": 18032 }, { "epoch": 12.672522839072382, "grad_norm": 0.2447744756937027, "learning_rate": 2.4888732724291404e-05, "loss": 0.0235, "step": 18033 }, { "epoch": 12.673225579761068, "grad_norm": 0.2570244073867798, "learning_rate": 2.4888264230498944e-05, "loss": 0.0376, "step": 18034 }, { "epoch": 12.673928320449754, "grad_norm": 0.15993811190128326, "learning_rate": 2.4887795736706488e-05, "loss": 0.025, "step": 18035 }, { "epoch": 12.67463106113844, "grad_norm": 0.08392554521560669, "learning_rate": 2.488732724291403e-05, "loss": 0.0133, "step": 18036 }, { "epoch": 12.675333801827126, "grad_norm": 0.17926296591758728, "learning_rate": 2.4886858749121575e-05, "loss": 0.0237, "step": 18037 }, { "epoch": 12.676036542515812, "grad_norm": 0.17388597130775452, "learning_rate": 2.4886390255329116e-05, "loss": 0.0131, "step": 18038 }, { "epoch": 12.676739283204498, "grad_norm": 0.20038175582885742, "learning_rate": 2.488592176153666e-05, "loss": 0.0165, "step": 18039 }, { "epoch": 12.677442023893184, "grad_norm": 0.09407300502061844, "learning_rate": 2.4885453267744203e-05, "loss": 0.0097, "step": 18040 }, { "epoch": 12.67814476458187, "grad_norm": 0.20389407873153687, "learning_rate": 2.4884984773951747e-05, "loss": 0.0485, "step": 18041 }, { "epoch": 12.678847505270555, "grad_norm": 0.5067119002342224, "learning_rate": 2.4884516280159287e-05, "loss": 0.0325, "step": 18042 }, { "epoch": 12.679550245959241, "grad_norm": 0.13467521965503693, "learning_rate": 2.488404778636683e-05, "loss": 0.0217, "step": 18043 }, { "epoch": 12.680252986647927, "grad_norm": 0.1501537412405014, "learning_rate": 2.4883579292574375e-05, "loss": 0.0217, "step": 18044 }, { "epoch": 12.680955727336613, "grad_norm": 0.22485092282295227, "learning_rate": 2.488311079878192e-05, "loss": 0.0417, "step": 18045 }, { "epoch": 12.681658468025299, "grad_norm": 0.22876422107219696, "learning_rate": 2.488264230498946e-05, "loss": 0.0429, "step": 18046 }, { "epoch": 12.682361208713985, "grad_norm": 0.4925469756126404, "learning_rate": 2.4882173811197002e-05, "loss": 0.0709, "step": 18047 }, { "epoch": 12.683063949402671, "grad_norm": 1.0270915031433105, "learning_rate": 2.4881705317404546e-05, "loss": 0.1002, "step": 18048 }, { "epoch": 12.683766690091357, "grad_norm": 1.1143887042999268, "learning_rate": 2.488123682361209e-05, "loss": 0.148, "step": 18049 }, { "epoch": 12.684469430780043, "grad_norm": 1.0603160858154297, "learning_rate": 2.4880768329819634e-05, "loss": 0.1372, "step": 18050 }, { "epoch": 12.685172171468729, "grad_norm": 1.7095441818237305, "learning_rate": 2.488029983602717e-05, "loss": 0.184, "step": 18051 }, { "epoch": 12.685874912157415, "grad_norm": 0.22907385230064392, "learning_rate": 2.4879831342234714e-05, "loss": 0.0613, "step": 18052 }, { "epoch": 12.6865776528461, "grad_norm": 0.24973998963832855, "learning_rate": 2.4879362848442258e-05, "loss": 0.0166, "step": 18053 }, { "epoch": 12.687280393534786, "grad_norm": 0.155496746301651, "learning_rate": 2.4878894354649802e-05, "loss": 0.0163, "step": 18054 }, { "epoch": 12.687983134223472, "grad_norm": 0.10903023928403854, "learning_rate": 2.4878425860857342e-05, "loss": 0.0141, "step": 18055 }, { "epoch": 12.688685874912156, "grad_norm": 0.13403470814228058, "learning_rate": 2.4877957367064886e-05, "loss": 0.0165, "step": 18056 }, { "epoch": 12.689388615600844, "grad_norm": 0.2379140853881836, "learning_rate": 2.487748887327243e-05, "loss": 0.0108, "step": 18057 }, { "epoch": 12.690091356289528, "grad_norm": 0.17885980010032654, "learning_rate": 2.4877020379479973e-05, "loss": 0.0185, "step": 18058 }, { "epoch": 12.690794096978214, "grad_norm": 0.10189413279294968, "learning_rate": 2.4876551885687514e-05, "loss": 0.0118, "step": 18059 }, { "epoch": 12.6914968376669, "grad_norm": 0.24454160034656525, "learning_rate": 2.4876083391895057e-05, "loss": 0.0305, "step": 18060 }, { "epoch": 12.692199578355586, "grad_norm": 0.12482577562332153, "learning_rate": 2.48756148981026e-05, "loss": 0.0148, "step": 18061 }, { "epoch": 12.692902319044272, "grad_norm": 0.15514323115348816, "learning_rate": 2.4875146404310145e-05, "loss": 0.0243, "step": 18062 }, { "epoch": 12.693605059732958, "grad_norm": 0.08841359615325928, "learning_rate": 2.487467791051769e-05, "loss": 0.0076, "step": 18063 }, { "epoch": 12.694307800421644, "grad_norm": 0.14433225989341736, "learning_rate": 2.487420941672523e-05, "loss": 0.0199, "step": 18064 }, { "epoch": 12.69501054111033, "grad_norm": 0.17544078826904297, "learning_rate": 2.4873740922932773e-05, "loss": 0.0114, "step": 18065 }, { "epoch": 12.695713281799016, "grad_norm": 0.3129976689815521, "learning_rate": 2.4873272429140316e-05, "loss": 0.0347, "step": 18066 }, { "epoch": 12.696416022487702, "grad_norm": 0.22885175049304962, "learning_rate": 2.487280393534786e-05, "loss": 0.0278, "step": 18067 }, { "epoch": 12.697118763176388, "grad_norm": 0.6019541621208191, "learning_rate": 2.48723354415554e-05, "loss": 0.0151, "step": 18068 }, { "epoch": 12.697821503865073, "grad_norm": 0.29912835359573364, "learning_rate": 2.487186694776294e-05, "loss": 0.0633, "step": 18069 }, { "epoch": 12.69852424455376, "grad_norm": 1.1055364608764648, "learning_rate": 2.4871398453970484e-05, "loss": 0.0269, "step": 18070 }, { "epoch": 12.699226985242445, "grad_norm": 0.463472843170166, "learning_rate": 2.4870929960178028e-05, "loss": 0.0525, "step": 18071 }, { "epoch": 12.699929725931131, "grad_norm": 0.3748433589935303, "learning_rate": 2.487046146638557e-05, "loss": 0.0528, "step": 18072 }, { "epoch": 12.700632466619817, "grad_norm": 0.4733225703239441, "learning_rate": 2.4869992972593112e-05, "loss": 0.1008, "step": 18073 }, { "epoch": 12.701335207308503, "grad_norm": 0.5822198390960693, "learning_rate": 2.4869524478800656e-05, "loss": 0.1557, "step": 18074 }, { "epoch": 12.702037947997189, "grad_norm": 2.90738582611084, "learning_rate": 2.48690559850082e-05, "loss": 0.1508, "step": 18075 }, { "epoch": 12.702740688685875, "grad_norm": 1.577455997467041, "learning_rate": 2.4868587491215743e-05, "loss": 0.1907, "step": 18076 }, { "epoch": 12.70344342937456, "grad_norm": 0.562537431716919, "learning_rate": 2.4868118997423284e-05, "loss": 0.0638, "step": 18077 }, { "epoch": 12.704146170063247, "grad_norm": 0.07446050643920898, "learning_rate": 2.4867650503630827e-05, "loss": 0.019, "step": 18078 }, { "epoch": 12.704848910751933, "grad_norm": 0.11314809322357178, "learning_rate": 2.486718200983837e-05, "loss": 0.0152, "step": 18079 }, { "epoch": 12.705551651440619, "grad_norm": 0.20604437589645386, "learning_rate": 2.4866713516045915e-05, "loss": 0.0361, "step": 18080 }, { "epoch": 12.706254392129305, "grad_norm": 0.10894114524126053, "learning_rate": 2.4866245022253455e-05, "loss": 0.0186, "step": 18081 }, { "epoch": 12.70695713281799, "grad_norm": 0.2539278566837311, "learning_rate": 2.4865776528461e-05, "loss": 0.0064, "step": 18082 }, { "epoch": 12.707659873506676, "grad_norm": 0.11835112422704697, "learning_rate": 2.4865308034668543e-05, "loss": 0.0143, "step": 18083 }, { "epoch": 12.708362614195362, "grad_norm": 0.17052388191223145, "learning_rate": 2.4864839540876086e-05, "loss": 0.0194, "step": 18084 }, { "epoch": 12.709065354884048, "grad_norm": 0.2674165666103363, "learning_rate": 2.4864371047083627e-05, "loss": 0.0252, "step": 18085 }, { "epoch": 12.709768095572734, "grad_norm": 0.11732344329357147, "learning_rate": 2.4863902553291167e-05, "loss": 0.0152, "step": 18086 }, { "epoch": 12.71047083626142, "grad_norm": 0.2605810761451721, "learning_rate": 2.486343405949871e-05, "loss": 0.0354, "step": 18087 }, { "epoch": 12.711173576950106, "grad_norm": 0.22304365038871765, "learning_rate": 2.4862965565706255e-05, "loss": 0.0138, "step": 18088 }, { "epoch": 12.711876317638792, "grad_norm": 0.21511255204677582, "learning_rate": 2.4862497071913798e-05, "loss": 0.025, "step": 18089 }, { "epoch": 12.712579058327478, "grad_norm": 0.06972914934158325, "learning_rate": 2.486202857812134e-05, "loss": 0.0101, "step": 18090 }, { "epoch": 12.713281799016164, "grad_norm": 0.19097687304019928, "learning_rate": 2.4861560084328882e-05, "loss": 0.0307, "step": 18091 }, { "epoch": 12.71398453970485, "grad_norm": 0.2813788056373596, "learning_rate": 2.4861091590536426e-05, "loss": 0.0255, "step": 18092 }, { "epoch": 12.714687280393536, "grad_norm": 0.280853271484375, "learning_rate": 2.486062309674397e-05, "loss": 0.0262, "step": 18093 }, { "epoch": 12.715390021082221, "grad_norm": 0.2283957451581955, "learning_rate": 2.486015460295151e-05, "loss": 0.0319, "step": 18094 }, { "epoch": 12.716092761770906, "grad_norm": 1.0894136428833008, "learning_rate": 2.4859686109159054e-05, "loss": 0.0516, "step": 18095 }, { "epoch": 12.716795502459593, "grad_norm": 0.4124087393283844, "learning_rate": 2.4859217615366598e-05, "loss": 0.0502, "step": 18096 }, { "epoch": 12.717498243148277, "grad_norm": 0.4704740643501282, "learning_rate": 2.485874912157414e-05, "loss": 0.0665, "step": 18097 }, { "epoch": 12.718200983836963, "grad_norm": 1.4009977579116821, "learning_rate": 2.485828062778168e-05, "loss": 0.137, "step": 18098 }, { "epoch": 12.71890372452565, "grad_norm": 0.6530400514602661, "learning_rate": 2.4857812133989225e-05, "loss": 0.1198, "step": 18099 }, { "epoch": 12.719606465214335, "grad_norm": 0.7962614297866821, "learning_rate": 2.485734364019677e-05, "loss": 0.1745, "step": 18100 }, { "epoch": 12.720309205903021, "grad_norm": 3.0489447116851807, "learning_rate": 2.4856875146404313e-05, "loss": 0.1892, "step": 18101 }, { "epoch": 12.721011946591707, "grad_norm": 0.3153766989707947, "learning_rate": 2.4856406652611857e-05, "loss": 0.0595, "step": 18102 }, { "epoch": 12.721714687280393, "grad_norm": 0.10168817639350891, "learning_rate": 2.4855938158819397e-05, "loss": 0.023, "step": 18103 }, { "epoch": 12.722417427969079, "grad_norm": 0.10364098101854324, "learning_rate": 2.4855469665026937e-05, "loss": 0.0241, "step": 18104 }, { "epoch": 12.723120168657765, "grad_norm": 0.10058176517486572, "learning_rate": 2.485500117123448e-05, "loss": 0.0149, "step": 18105 }, { "epoch": 12.72382290934645, "grad_norm": 0.05045224726200104, "learning_rate": 2.4854532677442025e-05, "loss": 0.0049, "step": 18106 }, { "epoch": 12.724525650035137, "grad_norm": 0.3516433835029602, "learning_rate": 2.4854064183649565e-05, "loss": 0.0121, "step": 18107 }, { "epoch": 12.725228390723823, "grad_norm": 0.10526417940855026, "learning_rate": 2.485359568985711e-05, "loss": 0.0178, "step": 18108 }, { "epoch": 12.725931131412509, "grad_norm": 0.2563922703266144, "learning_rate": 2.4853127196064652e-05, "loss": 0.0239, "step": 18109 }, { "epoch": 12.726633872101194, "grad_norm": 0.4987080991268158, "learning_rate": 2.4852658702272196e-05, "loss": 0.0195, "step": 18110 }, { "epoch": 12.72733661278988, "grad_norm": 0.09121762961149216, "learning_rate": 2.4852190208479736e-05, "loss": 0.0152, "step": 18111 }, { "epoch": 12.728039353478566, "grad_norm": 0.21625138819217682, "learning_rate": 2.485172171468728e-05, "loss": 0.029, "step": 18112 }, { "epoch": 12.728742094167252, "grad_norm": 0.14345808327198029, "learning_rate": 2.4851253220894824e-05, "loss": 0.0158, "step": 18113 }, { "epoch": 12.729444834855938, "grad_norm": 0.6932727694511414, "learning_rate": 2.4850784727102368e-05, "loss": 0.0475, "step": 18114 }, { "epoch": 12.730147575544624, "grad_norm": 0.44764837622642517, "learning_rate": 2.485031623330991e-05, "loss": 0.0179, "step": 18115 }, { "epoch": 12.73085031623331, "grad_norm": 0.17258554697036743, "learning_rate": 2.4849847739517452e-05, "loss": 0.0329, "step": 18116 }, { "epoch": 12.731553056921996, "grad_norm": 0.17921532690525055, "learning_rate": 2.4849379245724995e-05, "loss": 0.037, "step": 18117 }, { "epoch": 12.732255797610682, "grad_norm": 0.1296689659357071, "learning_rate": 2.484891075193254e-05, "loss": 0.018, "step": 18118 }, { "epoch": 12.732958538299368, "grad_norm": 0.24722734093666077, "learning_rate": 2.4848442258140083e-05, "loss": 0.0524, "step": 18119 }, { "epoch": 12.733661278988054, "grad_norm": 0.2774752378463745, "learning_rate": 2.4847973764347623e-05, "loss": 0.0351, "step": 18120 }, { "epoch": 12.73436401967674, "grad_norm": 0.3088684678077698, "learning_rate": 2.4847505270555164e-05, "loss": 0.0374, "step": 18121 }, { "epoch": 12.735066760365426, "grad_norm": 0.33795464038848877, "learning_rate": 2.4847036776762707e-05, "loss": 0.0957, "step": 18122 }, { "epoch": 12.735769501054111, "grad_norm": 0.33809199929237366, "learning_rate": 2.484656828297025e-05, "loss": 0.1063, "step": 18123 }, { "epoch": 12.736472241742797, "grad_norm": 1.9543321132659912, "learning_rate": 2.484609978917779e-05, "loss": 0.1302, "step": 18124 }, { "epoch": 12.737174982431483, "grad_norm": 0.6660453677177429, "learning_rate": 2.4845631295385335e-05, "loss": 0.1826, "step": 18125 }, { "epoch": 12.73787772312017, "grad_norm": 1.7491519451141357, "learning_rate": 2.484516280159288e-05, "loss": 0.2147, "step": 18126 }, { "epoch": 12.738580463808855, "grad_norm": 0.23842334747314453, "learning_rate": 2.4844694307800423e-05, "loss": 0.0635, "step": 18127 }, { "epoch": 12.739283204497541, "grad_norm": 0.17336739599704742, "learning_rate": 2.4844225814007966e-05, "loss": 0.0244, "step": 18128 }, { "epoch": 12.739985945186227, "grad_norm": 0.09423927962779999, "learning_rate": 2.4843757320215507e-05, "loss": 0.0211, "step": 18129 }, { "epoch": 12.740688685874913, "grad_norm": 0.17115166783332825, "learning_rate": 2.484328882642305e-05, "loss": 0.0226, "step": 18130 }, { "epoch": 12.741391426563599, "grad_norm": 0.20937371253967285, "learning_rate": 2.4842820332630594e-05, "loss": 0.0229, "step": 18131 }, { "epoch": 12.742094167252285, "grad_norm": 0.21604835987091064, "learning_rate": 2.4842351838838138e-05, "loss": 0.0164, "step": 18132 }, { "epoch": 12.74279690794097, "grad_norm": 0.15765628218650818, "learning_rate": 2.4841883345045678e-05, "loss": 0.0115, "step": 18133 }, { "epoch": 12.743499648629655, "grad_norm": 0.20342108607292175, "learning_rate": 2.4841414851253222e-05, "loss": 0.018, "step": 18134 }, { "epoch": 12.74420238931834, "grad_norm": 0.11464844644069672, "learning_rate": 2.4840946357460766e-05, "loss": 0.0167, "step": 18135 }, { "epoch": 12.744905130007027, "grad_norm": 0.10259183496236801, "learning_rate": 2.484047786366831e-05, "loss": 0.0172, "step": 18136 }, { "epoch": 12.745607870695713, "grad_norm": 0.22528673708438873, "learning_rate": 2.484000936987585e-05, "loss": 0.0309, "step": 18137 }, { "epoch": 12.746310611384398, "grad_norm": 0.10314788669347763, "learning_rate": 2.483954087608339e-05, "loss": 0.0131, "step": 18138 }, { "epoch": 12.747013352073084, "grad_norm": 0.1584046632051468, "learning_rate": 2.4839072382290934e-05, "loss": 0.03, "step": 18139 }, { "epoch": 12.74771609276177, "grad_norm": 0.1427580863237381, "learning_rate": 2.4838603888498477e-05, "loss": 0.0146, "step": 18140 }, { "epoch": 12.748418833450456, "grad_norm": 0.17691288888454437, "learning_rate": 2.483813539470602e-05, "loss": 0.0408, "step": 18141 }, { "epoch": 12.749121574139142, "grad_norm": 0.0999113991856575, "learning_rate": 2.483766690091356e-05, "loss": 0.0227, "step": 18142 }, { "epoch": 12.749824314827828, "grad_norm": 0.16498419642448425, "learning_rate": 2.4837198407121105e-05, "loss": 0.019, "step": 18143 }, { "epoch": 12.750527055516514, "grad_norm": 0.3036607801914215, "learning_rate": 2.483672991332865e-05, "loss": 0.0448, "step": 18144 }, { "epoch": 12.7512297962052, "grad_norm": 0.23582138121128082, "learning_rate": 2.4836261419536193e-05, "loss": 0.063, "step": 18145 }, { "epoch": 12.751932536893886, "grad_norm": 0.3525506854057312, "learning_rate": 2.4835792925743733e-05, "loss": 0.0459, "step": 18146 }, { "epoch": 12.752635277582572, "grad_norm": 0.36046063899993896, "learning_rate": 2.4835324431951277e-05, "loss": 0.0759, "step": 18147 }, { "epoch": 12.753338018271258, "grad_norm": 0.3367122411727905, "learning_rate": 2.483485593815882e-05, "loss": 0.0981, "step": 18148 }, { "epoch": 12.754040758959944, "grad_norm": 0.6757754683494568, "learning_rate": 2.4834387444366364e-05, "loss": 0.1406, "step": 18149 }, { "epoch": 12.75474349964863, "grad_norm": 0.5835115313529968, "learning_rate": 2.4833918950573904e-05, "loss": 0.1884, "step": 18150 }, { "epoch": 12.755446240337315, "grad_norm": 0.9360138177871704, "learning_rate": 2.4833450456781448e-05, "loss": 0.179, "step": 18151 }, { "epoch": 12.756148981026001, "grad_norm": 0.21037274599075317, "learning_rate": 2.4832981962988992e-05, "loss": 0.0607, "step": 18152 }, { "epoch": 12.756851721714687, "grad_norm": 0.17765939235687256, "learning_rate": 2.4832513469196536e-05, "loss": 0.0234, "step": 18153 }, { "epoch": 12.757554462403373, "grad_norm": 0.09408574551343918, "learning_rate": 2.483204497540408e-05, "loss": 0.0206, "step": 18154 }, { "epoch": 12.75825720309206, "grad_norm": 0.10661623626947403, "learning_rate": 2.483157648161162e-05, "loss": 0.0168, "step": 18155 }, { "epoch": 12.758959943780745, "grad_norm": 0.34145107865333557, "learning_rate": 2.483110798781916e-05, "loss": 0.0263, "step": 18156 }, { "epoch": 12.759662684469431, "grad_norm": 0.10714651644229889, "learning_rate": 2.4830639494026704e-05, "loss": 0.0133, "step": 18157 }, { "epoch": 12.760365425158117, "grad_norm": 0.13255855441093445, "learning_rate": 2.4830171000234248e-05, "loss": 0.0275, "step": 18158 }, { "epoch": 12.761068165846803, "grad_norm": 0.11121782660484314, "learning_rate": 2.4829702506441788e-05, "loss": 0.0184, "step": 18159 }, { "epoch": 12.761770906535489, "grad_norm": 0.17582935094833374, "learning_rate": 2.482923401264933e-05, "loss": 0.0214, "step": 18160 }, { "epoch": 12.762473647224175, "grad_norm": 0.15336471796035767, "learning_rate": 2.4828765518856875e-05, "loss": 0.0088, "step": 18161 }, { "epoch": 12.76317638791286, "grad_norm": 0.1357845813035965, "learning_rate": 2.482829702506442e-05, "loss": 0.0283, "step": 18162 }, { "epoch": 12.763879128601546, "grad_norm": 0.3096931278705597, "learning_rate": 2.482782853127196e-05, "loss": 0.0102, "step": 18163 }, { "epoch": 12.764581869290232, "grad_norm": 0.33017081022262573, "learning_rate": 2.4827360037479503e-05, "loss": 0.021, "step": 18164 }, { "epoch": 12.765284609978918, "grad_norm": 0.3553949296474457, "learning_rate": 2.4826891543687047e-05, "loss": 0.0149, "step": 18165 }, { "epoch": 12.765987350667604, "grad_norm": 0.18161644041538239, "learning_rate": 2.482642304989459e-05, "loss": 0.0356, "step": 18166 }, { "epoch": 12.76669009135629, "grad_norm": 0.13016267120838165, "learning_rate": 2.4825954556102134e-05, "loss": 0.0214, "step": 18167 }, { "epoch": 12.767392832044976, "grad_norm": 0.21149471402168274, "learning_rate": 2.4825486062309675e-05, "loss": 0.0209, "step": 18168 }, { "epoch": 12.768095572733662, "grad_norm": 0.23368264734745026, "learning_rate": 2.482501756851722e-05, "loss": 0.0417, "step": 18169 }, { "epoch": 12.768798313422348, "grad_norm": 0.20144639909267426, "learning_rate": 2.4824549074724762e-05, "loss": 0.0339, "step": 18170 }, { "epoch": 12.769501054111032, "grad_norm": 0.20295119285583496, "learning_rate": 2.4824080580932306e-05, "loss": 0.0388, "step": 18171 }, { "epoch": 12.77020379479972, "grad_norm": 0.3239545524120331, "learning_rate": 2.4823612087139846e-05, "loss": 0.0535, "step": 18172 }, { "epoch": 12.770906535488404, "grad_norm": 0.6103905439376831, "learning_rate": 2.4823143593347386e-05, "loss": 0.1002, "step": 18173 }, { "epoch": 12.77160927617709, "grad_norm": 1.2308720350265503, "learning_rate": 2.482267509955493e-05, "loss": 0.1423, "step": 18174 }, { "epoch": 12.772312016865776, "grad_norm": 0.6920691132545471, "learning_rate": 2.4822206605762474e-05, "loss": 0.1775, "step": 18175 }, { "epoch": 12.773014757554462, "grad_norm": 0.9356858134269714, "learning_rate": 2.4821738111970018e-05, "loss": 0.1598, "step": 18176 }, { "epoch": 12.773717498243148, "grad_norm": 0.20996437966823578, "learning_rate": 2.4821269618177558e-05, "loss": 0.056, "step": 18177 }, { "epoch": 12.774420238931834, "grad_norm": 0.10175113379955292, "learning_rate": 2.4820801124385102e-05, "loss": 0.0217, "step": 18178 }, { "epoch": 12.77512297962052, "grad_norm": 0.0935303196310997, "learning_rate": 2.4820332630592645e-05, "loss": 0.0187, "step": 18179 }, { "epoch": 12.775825720309205, "grad_norm": 0.07234872132539749, "learning_rate": 2.481986413680019e-05, "loss": 0.0117, "step": 18180 }, { "epoch": 12.776528460997891, "grad_norm": 0.22557750344276428, "learning_rate": 2.481939564300773e-05, "loss": 0.0193, "step": 18181 }, { "epoch": 12.777231201686577, "grad_norm": 0.16583071649074554, "learning_rate": 2.4818927149215273e-05, "loss": 0.0112, "step": 18182 }, { "epoch": 12.777933942375263, "grad_norm": 0.062225136905908585, "learning_rate": 2.4818458655422817e-05, "loss": 0.0108, "step": 18183 }, { "epoch": 12.778636683063949, "grad_norm": 0.16730016469955444, "learning_rate": 2.481799016163036e-05, "loss": 0.0291, "step": 18184 }, { "epoch": 12.779339423752635, "grad_norm": 0.08751232177019119, "learning_rate": 2.48175216678379e-05, "loss": 0.0202, "step": 18185 }, { "epoch": 12.780042164441321, "grad_norm": 0.1770651936531067, "learning_rate": 2.4817053174045445e-05, "loss": 0.0208, "step": 18186 }, { "epoch": 12.780744905130007, "grad_norm": 0.10403399169445038, "learning_rate": 2.481658468025299e-05, "loss": 0.0256, "step": 18187 }, { "epoch": 12.781447645818693, "grad_norm": 0.22307349741458893, "learning_rate": 2.4816116186460532e-05, "loss": 0.0137, "step": 18188 }, { "epoch": 12.782150386507379, "grad_norm": 0.15365223586559296, "learning_rate": 2.4815647692668076e-05, "loss": 0.0358, "step": 18189 }, { "epoch": 12.782853127196065, "grad_norm": 0.0934280976653099, "learning_rate": 2.4815179198875616e-05, "loss": 0.0132, "step": 18190 }, { "epoch": 12.78355586788475, "grad_norm": 0.15098372101783752, "learning_rate": 2.4814710705083157e-05, "loss": 0.0289, "step": 18191 }, { "epoch": 12.784258608573436, "grad_norm": 1.2304747104644775, "learning_rate": 2.48142422112907e-05, "loss": 0.0346, "step": 18192 }, { "epoch": 12.784961349262122, "grad_norm": 0.1523517519235611, "learning_rate": 2.4813773717498244e-05, "loss": 0.0252, "step": 18193 }, { "epoch": 12.785664089950808, "grad_norm": 0.10508359223604202, "learning_rate": 2.4813305223705784e-05, "loss": 0.0194, "step": 18194 }, { "epoch": 12.786366830639494, "grad_norm": 0.17655950784683228, "learning_rate": 2.4812836729913328e-05, "loss": 0.0638, "step": 18195 }, { "epoch": 12.78706957132818, "grad_norm": 0.8358632922172546, "learning_rate": 2.4812368236120872e-05, "loss": 0.0437, "step": 18196 }, { "epoch": 12.787772312016866, "grad_norm": 0.43834614753723145, "learning_rate": 2.4811899742328416e-05, "loss": 0.0978, "step": 18197 }, { "epoch": 12.788475052705552, "grad_norm": 0.6173011064529419, "learning_rate": 2.4811431248535956e-05, "loss": 0.0963, "step": 18198 }, { "epoch": 12.789177793394238, "grad_norm": 0.38154080510139465, "learning_rate": 2.48109627547435e-05, "loss": 0.129, "step": 18199 }, { "epoch": 12.789880534082924, "grad_norm": 0.657703161239624, "learning_rate": 2.4810494260951043e-05, "loss": 0.1635, "step": 18200 }, { "epoch": 12.79058327477161, "grad_norm": 1.5342798233032227, "learning_rate": 2.4810025767158587e-05, "loss": 0.2228, "step": 18201 }, { "epoch": 12.791286015460296, "grad_norm": 0.24896547198295593, "learning_rate": 2.480955727336613e-05, "loss": 0.0693, "step": 18202 }, { "epoch": 12.791988756148982, "grad_norm": 0.11846821010112762, "learning_rate": 2.480908877957367e-05, "loss": 0.0247, "step": 18203 }, { "epoch": 12.792691496837667, "grad_norm": 0.2595027685165405, "learning_rate": 2.4808620285781215e-05, "loss": 0.0234, "step": 18204 }, { "epoch": 12.793394237526353, "grad_norm": 0.1412077397108078, "learning_rate": 2.480815179198876e-05, "loss": 0.0195, "step": 18205 }, { "epoch": 12.79409697821504, "grad_norm": 0.07688204944133759, "learning_rate": 2.4807683298196302e-05, "loss": 0.0183, "step": 18206 }, { "epoch": 12.794799718903725, "grad_norm": 0.08741477131843567, "learning_rate": 2.4807214804403843e-05, "loss": 0.0088, "step": 18207 }, { "epoch": 12.795502459592411, "grad_norm": 0.1455973982810974, "learning_rate": 2.4806746310611383e-05, "loss": 0.0179, "step": 18208 }, { "epoch": 12.796205200281097, "grad_norm": 0.10963565111160278, "learning_rate": 2.4806277816818927e-05, "loss": 0.0225, "step": 18209 }, { "epoch": 12.796907940969781, "grad_norm": 0.08216733485460281, "learning_rate": 2.480580932302647e-05, "loss": 0.0102, "step": 18210 }, { "epoch": 12.797610681658469, "grad_norm": 0.0781768411397934, "learning_rate": 2.480534082923401e-05, "loss": 0.0088, "step": 18211 }, { "epoch": 12.798313422347153, "grad_norm": 0.09697957336902618, "learning_rate": 2.4804872335441554e-05, "loss": 0.0175, "step": 18212 }, { "epoch": 12.799016163035839, "grad_norm": 0.18010316789150238, "learning_rate": 2.4804403841649098e-05, "loss": 0.0135, "step": 18213 }, { "epoch": 12.799718903724525, "grad_norm": 0.3372942805290222, "learning_rate": 2.4803935347856642e-05, "loss": 0.0265, "step": 18214 }, { "epoch": 12.80042164441321, "grad_norm": 0.11334694176912308, "learning_rate": 2.4803466854064186e-05, "loss": 0.0134, "step": 18215 }, { "epoch": 12.801124385101897, "grad_norm": 1.182091474533081, "learning_rate": 2.4802998360271726e-05, "loss": 0.0212, "step": 18216 }, { "epoch": 12.801827125790583, "grad_norm": 0.1696598380804062, "learning_rate": 2.480252986647927e-05, "loss": 0.0348, "step": 18217 }, { "epoch": 12.802529866479269, "grad_norm": 0.1992063671350479, "learning_rate": 2.4802061372686813e-05, "loss": 0.0143, "step": 18218 }, { "epoch": 12.803232607167955, "grad_norm": 0.31217825412750244, "learning_rate": 2.4801592878894357e-05, "loss": 0.0335, "step": 18219 }, { "epoch": 12.80393534785664, "grad_norm": 0.23023778200149536, "learning_rate": 2.4801124385101897e-05, "loss": 0.0252, "step": 18220 }, { "epoch": 12.804638088545326, "grad_norm": 0.32161569595336914, "learning_rate": 2.480065589130944e-05, "loss": 0.0799, "step": 18221 }, { "epoch": 12.805340829234012, "grad_norm": 0.22662915289402008, "learning_rate": 2.4800187397516985e-05, "loss": 0.052, "step": 18222 }, { "epoch": 12.806043569922698, "grad_norm": 0.36377280950546265, "learning_rate": 2.479971890372453e-05, "loss": 0.094, "step": 18223 }, { "epoch": 12.806746310611384, "grad_norm": 3.6974828243255615, "learning_rate": 2.479925040993207e-05, "loss": 0.1336, "step": 18224 }, { "epoch": 12.80744905130007, "grad_norm": 0.5083566308021545, "learning_rate": 2.4798781916139613e-05, "loss": 0.1683, "step": 18225 }, { "epoch": 12.808151791988756, "grad_norm": 0.803160548210144, "learning_rate": 2.4798313422347153e-05, "loss": 0.1941, "step": 18226 }, { "epoch": 12.808854532677442, "grad_norm": 0.18740876019001007, "learning_rate": 2.4797844928554697e-05, "loss": 0.0758, "step": 18227 }, { "epoch": 12.809557273366128, "grad_norm": 0.11564568430185318, "learning_rate": 2.479737643476224e-05, "loss": 0.0266, "step": 18228 }, { "epoch": 12.810260014054814, "grad_norm": 0.17197385430335999, "learning_rate": 2.479690794096978e-05, "loss": 0.0203, "step": 18229 }, { "epoch": 12.8109627547435, "grad_norm": 0.1440933793783188, "learning_rate": 2.4796439447177325e-05, "loss": 0.0138, "step": 18230 }, { "epoch": 12.811665495432186, "grad_norm": 0.13537967205047607, "learning_rate": 2.479597095338487e-05, "loss": 0.0184, "step": 18231 }, { "epoch": 12.812368236120872, "grad_norm": 0.06508415192365646, "learning_rate": 2.4795502459592412e-05, "loss": 0.0113, "step": 18232 }, { "epoch": 12.813070976809557, "grad_norm": 0.17631548643112183, "learning_rate": 2.4795033965799952e-05, "loss": 0.021, "step": 18233 }, { "epoch": 12.813773717498243, "grad_norm": 0.13484133780002594, "learning_rate": 2.4794565472007496e-05, "loss": 0.0225, "step": 18234 }, { "epoch": 12.81447645818693, "grad_norm": 0.21587592363357544, "learning_rate": 2.479409697821504e-05, "loss": 0.0257, "step": 18235 }, { "epoch": 12.815179198875615, "grad_norm": 0.09474843740463257, "learning_rate": 2.4793628484422584e-05, "loss": 0.0083, "step": 18236 }, { "epoch": 12.815881939564301, "grad_norm": 0.138018399477005, "learning_rate": 2.4793159990630124e-05, "loss": 0.0175, "step": 18237 }, { "epoch": 12.816584680252987, "grad_norm": 0.5437408089637756, "learning_rate": 2.4792691496837668e-05, "loss": 0.0199, "step": 18238 }, { "epoch": 12.817287420941673, "grad_norm": 0.11250875890254974, "learning_rate": 2.479222300304521e-05, "loss": 0.0212, "step": 18239 }, { "epoch": 12.817990161630359, "grad_norm": 0.14410953223705292, "learning_rate": 2.4791754509252755e-05, "loss": 0.0118, "step": 18240 }, { "epoch": 12.818692902319045, "grad_norm": 0.6223078966140747, "learning_rate": 2.47912860154603e-05, "loss": 0.0158, "step": 18241 }, { "epoch": 12.81939564300773, "grad_norm": 5.946137428283691, "learning_rate": 2.479081752166784e-05, "loss": 0.0195, "step": 18242 }, { "epoch": 12.820098383696417, "grad_norm": 0.15842685103416443, "learning_rate": 2.479034902787538e-05, "loss": 0.0169, "step": 18243 }, { "epoch": 12.820801124385103, "grad_norm": 0.3569159209728241, "learning_rate": 2.4789880534082923e-05, "loss": 0.0354, "step": 18244 }, { "epoch": 12.821503865073788, "grad_norm": 0.24102114140987396, "learning_rate": 2.4789412040290467e-05, "loss": 0.0501, "step": 18245 }, { "epoch": 12.822206605762474, "grad_norm": 0.31990593671798706, "learning_rate": 2.4788943546498007e-05, "loss": 0.0456, "step": 18246 }, { "epoch": 12.82290934645116, "grad_norm": 0.1858617216348648, "learning_rate": 2.478847505270555e-05, "loss": 0.0637, "step": 18247 }, { "epoch": 12.823612087139846, "grad_norm": 0.26984167098999023, "learning_rate": 2.4788006558913095e-05, "loss": 0.1092, "step": 18248 }, { "epoch": 12.82431482782853, "grad_norm": 0.4522288739681244, "learning_rate": 2.478753806512064e-05, "loss": 0.1548, "step": 18249 }, { "epoch": 12.825017568517218, "grad_norm": 0.7477418780326843, "learning_rate": 2.478706957132818e-05, "loss": 0.1506, "step": 18250 }, { "epoch": 12.825720309205902, "grad_norm": 1.2639703750610352, "learning_rate": 2.4786601077535722e-05, "loss": 0.173, "step": 18251 }, { "epoch": 12.826423049894588, "grad_norm": 0.18201206624507904, "learning_rate": 2.4786132583743266e-05, "loss": 0.0487, "step": 18252 }, { "epoch": 12.827125790583274, "grad_norm": 0.13285386562347412, "learning_rate": 2.478566408995081e-05, "loss": 0.0287, "step": 18253 }, { "epoch": 12.82782853127196, "grad_norm": 0.12356598675251007, "learning_rate": 2.4785195596158354e-05, "loss": 0.021, "step": 18254 }, { "epoch": 12.828531271960646, "grad_norm": 0.1426941305398941, "learning_rate": 2.4784727102365894e-05, "loss": 0.0222, "step": 18255 }, { "epoch": 12.829234012649332, "grad_norm": 0.292305052280426, "learning_rate": 2.4784258608573438e-05, "loss": 0.0263, "step": 18256 }, { "epoch": 12.829936753338018, "grad_norm": 0.20542114973068237, "learning_rate": 2.478379011478098e-05, "loss": 0.0113, "step": 18257 }, { "epoch": 12.830639494026704, "grad_norm": 0.269899845123291, "learning_rate": 2.4783321620988525e-05, "loss": 0.022, "step": 18258 }, { "epoch": 12.83134223471539, "grad_norm": 0.15717267990112305, "learning_rate": 2.4782853127196066e-05, "loss": 0.0219, "step": 18259 }, { "epoch": 12.832044975404076, "grad_norm": 0.1271020621061325, "learning_rate": 2.4782384633403606e-05, "loss": 0.0229, "step": 18260 }, { "epoch": 12.832747716092761, "grad_norm": 0.07078973203897476, "learning_rate": 2.478191613961115e-05, "loss": 0.0093, "step": 18261 }, { "epoch": 12.833450456781447, "grad_norm": 0.14031893014907837, "learning_rate": 2.4781447645818693e-05, "loss": 0.0231, "step": 18262 }, { "epoch": 12.834153197470133, "grad_norm": 0.10564789175987244, "learning_rate": 2.4780979152026234e-05, "loss": 0.0141, "step": 18263 }, { "epoch": 12.83485593815882, "grad_norm": 0.283038854598999, "learning_rate": 2.4780510658233777e-05, "loss": 0.0237, "step": 18264 }, { "epoch": 12.835558678847505, "grad_norm": 0.155440554022789, "learning_rate": 2.478004216444132e-05, "loss": 0.0194, "step": 18265 }, { "epoch": 12.836261419536191, "grad_norm": 0.13302083313465118, "learning_rate": 2.4779573670648865e-05, "loss": 0.0284, "step": 18266 }, { "epoch": 12.836964160224877, "grad_norm": 0.1601664125919342, "learning_rate": 2.477910517685641e-05, "loss": 0.0199, "step": 18267 }, { "epoch": 12.837666900913563, "grad_norm": 0.28866884112358093, "learning_rate": 2.477863668306395e-05, "loss": 0.0274, "step": 18268 }, { "epoch": 12.838369641602249, "grad_norm": 0.19413086771965027, "learning_rate": 2.4778168189271493e-05, "loss": 0.0149, "step": 18269 }, { "epoch": 12.839072382290935, "grad_norm": 0.9314768314361572, "learning_rate": 2.4777699695479036e-05, "loss": 0.0617, "step": 18270 }, { "epoch": 12.83977512297962, "grad_norm": 0.34254974126815796, "learning_rate": 2.477723120168658e-05, "loss": 0.0341, "step": 18271 }, { "epoch": 12.840477863668307, "grad_norm": 0.8949034214019775, "learning_rate": 2.477676270789412e-05, "loss": 0.0896, "step": 18272 }, { "epoch": 12.841180604356992, "grad_norm": 0.4063762128353119, "learning_rate": 2.4776294214101664e-05, "loss": 0.0913, "step": 18273 }, { "epoch": 12.841883345045678, "grad_norm": 0.6878186464309692, "learning_rate": 2.4775825720309208e-05, "loss": 0.1346, "step": 18274 }, { "epoch": 12.842586085734364, "grad_norm": 0.736648440361023, "learning_rate": 2.477535722651675e-05, "loss": 0.1667, "step": 18275 }, { "epoch": 12.84328882642305, "grad_norm": 4.4168195724487305, "learning_rate": 2.4774888732724292e-05, "loss": 0.1648, "step": 18276 }, { "epoch": 12.843991567111736, "grad_norm": 0.28384891152381897, "learning_rate": 2.4774420238931836e-05, "loss": 0.0531, "step": 18277 }, { "epoch": 12.844694307800422, "grad_norm": 0.14415571093559265, "learning_rate": 2.4773951745139376e-05, "loss": 0.0256, "step": 18278 }, { "epoch": 12.845397048489108, "grad_norm": 0.12314032018184662, "learning_rate": 2.477348325134692e-05, "loss": 0.0169, "step": 18279 }, { "epoch": 12.846099789177794, "grad_norm": 0.11256052553653717, "learning_rate": 2.4773014757554463e-05, "loss": 0.0219, "step": 18280 }, { "epoch": 12.84680252986648, "grad_norm": 0.1466144174337387, "learning_rate": 2.4772546263762004e-05, "loss": 0.0177, "step": 18281 }, { "epoch": 12.847505270555166, "grad_norm": 0.08223260194063187, "learning_rate": 2.4772077769969547e-05, "loss": 0.0127, "step": 18282 }, { "epoch": 12.848208011243852, "grad_norm": 0.13425874710083008, "learning_rate": 2.477160927617709e-05, "loss": 0.0172, "step": 18283 }, { "epoch": 12.848910751932538, "grad_norm": 0.0759212076663971, "learning_rate": 2.4771140782384635e-05, "loss": 0.0088, "step": 18284 }, { "epoch": 12.849613492621224, "grad_norm": 0.14351674914360046, "learning_rate": 2.4770672288592175e-05, "loss": 0.0382, "step": 18285 }, { "epoch": 12.85031623330991, "grad_norm": 0.1047251969575882, "learning_rate": 2.477020379479972e-05, "loss": 0.0158, "step": 18286 }, { "epoch": 12.851018973998595, "grad_norm": 0.8397150635719299, "learning_rate": 2.4769735301007263e-05, "loss": 0.0285, "step": 18287 }, { "epoch": 12.85172171468728, "grad_norm": 0.12292760610580444, "learning_rate": 2.4769266807214806e-05, "loss": 0.012, "step": 18288 }, { "epoch": 12.852424455375965, "grad_norm": 0.21563951671123505, "learning_rate": 2.4768798313422347e-05, "loss": 0.0217, "step": 18289 }, { "epoch": 12.853127196064651, "grad_norm": 0.18365363776683807, "learning_rate": 2.476832981962989e-05, "loss": 0.0277, "step": 18290 }, { "epoch": 12.853829936753337, "grad_norm": 0.15948602557182312, "learning_rate": 2.4767861325837434e-05, "loss": 0.0268, "step": 18291 }, { "epoch": 12.854532677442023, "grad_norm": 0.10433785617351532, "learning_rate": 2.4767392832044978e-05, "loss": 0.0216, "step": 18292 }, { "epoch": 12.85523541813071, "grad_norm": 0.2533809244632721, "learning_rate": 2.476692433825252e-05, "loss": 0.0218, "step": 18293 }, { "epoch": 12.855938158819395, "grad_norm": 0.18075449764728546, "learning_rate": 2.4766455844460062e-05, "loss": 0.026, "step": 18294 }, { "epoch": 12.856640899508081, "grad_norm": 0.33184877038002014, "learning_rate": 2.4765987350667602e-05, "loss": 0.0419, "step": 18295 }, { "epoch": 12.857343640196767, "grad_norm": 0.364889532327652, "learning_rate": 2.4765518856875146e-05, "loss": 0.0825, "step": 18296 }, { "epoch": 12.858046380885453, "grad_norm": 0.6978863477706909, "learning_rate": 2.476505036308269e-05, "loss": 0.0866, "step": 18297 }, { "epoch": 12.858749121574139, "grad_norm": 0.2785596251487732, "learning_rate": 2.476458186929023e-05, "loss": 0.095, "step": 18298 }, { "epoch": 12.859451862262825, "grad_norm": 0.38022857904434204, "learning_rate": 2.4764113375497774e-05, "loss": 0.1385, "step": 18299 }, { "epoch": 12.86015460295151, "grad_norm": 0.4395565688610077, "learning_rate": 2.4763644881705318e-05, "loss": 0.1285, "step": 18300 }, { "epoch": 12.860857343640197, "grad_norm": 1.26347017288208, "learning_rate": 2.476317638791286e-05, "loss": 0.1739, "step": 18301 }, { "epoch": 12.861560084328882, "grad_norm": 0.1665511131286621, "learning_rate": 2.47627078941204e-05, "loss": 0.0597, "step": 18302 }, { "epoch": 12.862262825017568, "grad_norm": 0.1546773463487625, "learning_rate": 2.4762239400327945e-05, "loss": 0.0249, "step": 18303 }, { "epoch": 12.862965565706254, "grad_norm": 0.15394751727581024, "learning_rate": 2.476177090653549e-05, "loss": 0.0227, "step": 18304 }, { "epoch": 12.86366830639494, "grad_norm": 0.14745798707008362, "learning_rate": 2.4761302412743033e-05, "loss": 0.0262, "step": 18305 }, { "epoch": 12.864371047083626, "grad_norm": 0.125745490193367, "learning_rate": 2.4760833918950577e-05, "loss": 0.0144, "step": 18306 }, { "epoch": 12.865073787772312, "grad_norm": 0.12829522788524628, "learning_rate": 2.4760365425158117e-05, "loss": 0.0144, "step": 18307 }, { "epoch": 12.865776528460998, "grad_norm": 0.11108724772930145, "learning_rate": 2.475989693136566e-05, "loss": 0.0166, "step": 18308 }, { "epoch": 12.866479269149684, "grad_norm": 0.1615164428949356, "learning_rate": 2.4759428437573204e-05, "loss": 0.0164, "step": 18309 }, { "epoch": 12.86718200983837, "grad_norm": 0.17184990644454956, "learning_rate": 2.4758959943780748e-05, "loss": 0.0154, "step": 18310 }, { "epoch": 12.867884750527056, "grad_norm": 0.10900871455669403, "learning_rate": 2.475849144998829e-05, "loss": 0.0121, "step": 18311 }, { "epoch": 12.868587491215742, "grad_norm": 0.09698324650526047, "learning_rate": 2.4758022956195832e-05, "loss": 0.0184, "step": 18312 }, { "epoch": 12.869290231904428, "grad_norm": 0.11517680436372757, "learning_rate": 2.4757554462403372e-05, "loss": 0.0155, "step": 18313 }, { "epoch": 12.869992972593113, "grad_norm": 0.12722721695899963, "learning_rate": 2.4757085968610916e-05, "loss": 0.0168, "step": 18314 }, { "epoch": 12.8706957132818, "grad_norm": 0.14447270333766937, "learning_rate": 2.4756617474818457e-05, "loss": 0.018, "step": 18315 }, { "epoch": 12.871398453970485, "grad_norm": 0.2679043710231781, "learning_rate": 2.4756148981026e-05, "loss": 0.0225, "step": 18316 }, { "epoch": 12.872101194659171, "grad_norm": 0.51043301820755, "learning_rate": 2.4755680487233544e-05, "loss": 0.0283, "step": 18317 }, { "epoch": 12.872803935347857, "grad_norm": 0.17086121439933777, "learning_rate": 2.4755211993441088e-05, "loss": 0.0366, "step": 18318 }, { "epoch": 12.873506676036543, "grad_norm": 0.21195738017559052, "learning_rate": 2.475474349964863e-05, "loss": 0.0324, "step": 18319 }, { "epoch": 12.874209416725229, "grad_norm": 0.20640821754932404, "learning_rate": 2.4754275005856172e-05, "loss": 0.0297, "step": 18320 }, { "epoch": 12.874912157413915, "grad_norm": 0.15988244116306305, "learning_rate": 2.4753806512063715e-05, "loss": 0.038, "step": 18321 }, { "epoch": 12.8756148981026, "grad_norm": 0.28152894973754883, "learning_rate": 2.475333801827126e-05, "loss": 0.0673, "step": 18322 }, { "epoch": 12.876317638791287, "grad_norm": 0.34594371914863586, "learning_rate": 2.4752869524478803e-05, "loss": 0.1064, "step": 18323 }, { "epoch": 12.877020379479973, "grad_norm": 0.6549326181411743, "learning_rate": 2.4752401030686343e-05, "loss": 0.1383, "step": 18324 }, { "epoch": 12.877723120168657, "grad_norm": 0.5379546284675598, "learning_rate": 2.4751932536893887e-05, "loss": 0.1857, "step": 18325 }, { "epoch": 12.878425860857345, "grad_norm": 0.8581241965293884, "learning_rate": 2.475146404310143e-05, "loss": 0.1632, "step": 18326 }, { "epoch": 12.879128601546029, "grad_norm": 0.18845610320568085, "learning_rate": 2.4750995549308974e-05, "loss": 0.0616, "step": 18327 }, { "epoch": 12.879831342234715, "grad_norm": 0.23134325444698334, "learning_rate": 2.4750527055516515e-05, "loss": 0.0291, "step": 18328 }, { "epoch": 12.8805340829234, "grad_norm": 0.1100551038980484, "learning_rate": 2.475005856172406e-05, "loss": 0.0258, "step": 18329 }, { "epoch": 12.881236823612086, "grad_norm": 0.1926976889371872, "learning_rate": 2.47495900679316e-05, "loss": 0.0204, "step": 18330 }, { "epoch": 12.881939564300772, "grad_norm": 0.22294363379478455, "learning_rate": 2.4749121574139143e-05, "loss": 0.0226, "step": 18331 }, { "epoch": 12.882642304989458, "grad_norm": 0.0850028246641159, "learning_rate": 2.4748653080346686e-05, "loss": 0.0148, "step": 18332 }, { "epoch": 12.883345045678144, "grad_norm": 0.14314474165439606, "learning_rate": 2.4748184586554227e-05, "loss": 0.0132, "step": 18333 }, { "epoch": 12.88404778636683, "grad_norm": 0.08697199821472168, "learning_rate": 2.474771609276177e-05, "loss": 0.0133, "step": 18334 }, { "epoch": 12.884750527055516, "grad_norm": 0.10677383840084076, "learning_rate": 2.4747247598969314e-05, "loss": 0.0166, "step": 18335 }, { "epoch": 12.885453267744202, "grad_norm": 0.08139988034963608, "learning_rate": 2.4746779105176858e-05, "loss": 0.0206, "step": 18336 }, { "epoch": 12.886156008432888, "grad_norm": 0.13232861459255219, "learning_rate": 2.4746310611384398e-05, "loss": 0.0159, "step": 18337 }, { "epoch": 12.886858749121574, "grad_norm": 0.14210231602191925, "learning_rate": 2.4745842117591942e-05, "loss": 0.0086, "step": 18338 }, { "epoch": 12.88756148981026, "grad_norm": 0.24013108015060425, "learning_rate": 2.4745373623799486e-05, "loss": 0.0237, "step": 18339 }, { "epoch": 12.888264230498946, "grad_norm": 0.18526175618171692, "learning_rate": 2.474490513000703e-05, "loss": 0.0177, "step": 18340 }, { "epoch": 12.888966971187632, "grad_norm": 0.21658748388290405, "learning_rate": 2.474443663621457e-05, "loss": 0.0191, "step": 18341 }, { "epoch": 12.889669711876317, "grad_norm": 0.195146843791008, "learning_rate": 2.4743968142422113e-05, "loss": 0.0292, "step": 18342 }, { "epoch": 12.890372452565003, "grad_norm": 0.13610537350177765, "learning_rate": 2.4743499648629657e-05, "loss": 0.0287, "step": 18343 }, { "epoch": 12.89107519325369, "grad_norm": 0.137898787856102, "learning_rate": 2.47430311548372e-05, "loss": 0.0272, "step": 18344 }, { "epoch": 12.891777933942375, "grad_norm": 0.275777131319046, "learning_rate": 2.4742562661044745e-05, "loss": 0.0388, "step": 18345 }, { "epoch": 12.892480674631061, "grad_norm": 0.8706154823303223, "learning_rate": 2.4742094167252285e-05, "loss": 0.0601, "step": 18346 }, { "epoch": 12.893183415319747, "grad_norm": 0.23933398723602295, "learning_rate": 2.4741625673459825e-05, "loss": 0.069, "step": 18347 }, { "epoch": 12.893886156008433, "grad_norm": 0.4292737543582916, "learning_rate": 2.474115717966737e-05, "loss": 0.0953, "step": 18348 }, { "epoch": 12.894588896697119, "grad_norm": 0.5811780095100403, "learning_rate": 2.4740688685874913e-05, "loss": 0.1347, "step": 18349 }, { "epoch": 12.895291637385805, "grad_norm": 0.8842989802360535, "learning_rate": 2.4740220192082453e-05, "loss": 0.158, "step": 18350 }, { "epoch": 12.89599437807449, "grad_norm": 2.4819397926330566, "learning_rate": 2.4739751698289997e-05, "loss": 0.187, "step": 18351 }, { "epoch": 12.896697118763177, "grad_norm": 0.23696525394916534, "learning_rate": 2.473928320449754e-05, "loss": 0.0664, "step": 18352 }, { "epoch": 12.897399859451863, "grad_norm": 0.1412971317768097, "learning_rate": 2.4738814710705084e-05, "loss": 0.0291, "step": 18353 }, { "epoch": 12.898102600140549, "grad_norm": 0.22729620337486267, "learning_rate": 2.4738346216912625e-05, "loss": 0.0192, "step": 18354 }, { "epoch": 12.898805340829234, "grad_norm": 0.11181014031171799, "learning_rate": 2.4737877723120168e-05, "loss": 0.0163, "step": 18355 }, { "epoch": 12.89950808151792, "grad_norm": 0.12248391658067703, "learning_rate": 2.4737409229327712e-05, "loss": 0.0224, "step": 18356 }, { "epoch": 12.900210822206606, "grad_norm": 0.08454791456460953, "learning_rate": 2.4736940735535256e-05, "loss": 0.0135, "step": 18357 }, { "epoch": 12.900913562895292, "grad_norm": 0.2575766444206238, "learning_rate": 2.47364722417428e-05, "loss": 0.0128, "step": 18358 }, { "epoch": 12.901616303583978, "grad_norm": 0.48277246952056885, "learning_rate": 2.473600374795034e-05, "loss": 0.0346, "step": 18359 }, { "epoch": 12.902319044272664, "grad_norm": 0.09898988157510757, "learning_rate": 2.4735535254157883e-05, "loss": 0.0239, "step": 18360 }, { "epoch": 12.90302178496135, "grad_norm": 0.09181375801563263, "learning_rate": 2.4735066760365427e-05, "loss": 0.0105, "step": 18361 }, { "epoch": 12.903724525650036, "grad_norm": 0.18145740032196045, "learning_rate": 2.473459826657297e-05, "loss": 0.0244, "step": 18362 }, { "epoch": 12.904427266338722, "grad_norm": 0.15807078778743744, "learning_rate": 2.473412977278051e-05, "loss": 0.0182, "step": 18363 }, { "epoch": 12.905130007027406, "grad_norm": 1.1228748559951782, "learning_rate": 2.4733661278988055e-05, "loss": 0.0294, "step": 18364 }, { "epoch": 12.905832747716094, "grad_norm": 0.0901975929737091, "learning_rate": 2.4733192785195595e-05, "loss": 0.0153, "step": 18365 }, { "epoch": 12.906535488404778, "grad_norm": 0.14209522306919098, "learning_rate": 2.473272429140314e-05, "loss": 0.0248, "step": 18366 }, { "epoch": 12.907238229093464, "grad_norm": 0.22027072310447693, "learning_rate": 2.4732255797610683e-05, "loss": 0.0256, "step": 18367 }, { "epoch": 12.90794096978215, "grad_norm": 0.12994880974292755, "learning_rate": 2.4731787303818223e-05, "loss": 0.0219, "step": 18368 }, { "epoch": 12.908643710470836, "grad_norm": 0.1573476791381836, "learning_rate": 2.4731318810025767e-05, "loss": 0.0281, "step": 18369 }, { "epoch": 12.909346451159522, "grad_norm": 0.261483371257782, "learning_rate": 2.473085031623331e-05, "loss": 0.0404, "step": 18370 }, { "epoch": 12.910049191848207, "grad_norm": 0.21203802525997162, "learning_rate": 2.4730381822440854e-05, "loss": 0.0503, "step": 18371 }, { "epoch": 12.910751932536893, "grad_norm": 0.33638888597488403, "learning_rate": 2.4729913328648395e-05, "loss": 0.078, "step": 18372 }, { "epoch": 12.91145467322558, "grad_norm": 0.4863472580909729, "learning_rate": 2.472944483485594e-05, "loss": 0.0783, "step": 18373 }, { "epoch": 12.912157413914265, "grad_norm": 0.5927860736846924, "learning_rate": 2.4728976341063482e-05, "loss": 0.1284, "step": 18374 }, { "epoch": 12.912860154602951, "grad_norm": 1.219070553779602, "learning_rate": 2.4728507847271026e-05, "loss": 0.1501, "step": 18375 }, { "epoch": 12.913562895291637, "grad_norm": 1.9948588609695435, "learning_rate": 2.4728039353478566e-05, "loss": 0.1996, "step": 18376 }, { "epoch": 12.914265635980323, "grad_norm": 0.1714964509010315, "learning_rate": 2.472757085968611e-05, "loss": 0.0676, "step": 18377 }, { "epoch": 12.914968376669009, "grad_norm": 0.20825302600860596, "learning_rate": 2.4727102365893654e-05, "loss": 0.0366, "step": 18378 }, { "epoch": 12.915671117357695, "grad_norm": 0.10946976393461227, "learning_rate": 2.4726633872101197e-05, "loss": 0.0148, "step": 18379 }, { "epoch": 12.91637385804638, "grad_norm": 0.09525566548109055, "learning_rate": 2.4726165378308738e-05, "loss": 0.0156, "step": 18380 }, { "epoch": 12.917076598735067, "grad_norm": 0.06902667880058289, "learning_rate": 2.472569688451628e-05, "loss": 0.0101, "step": 18381 }, { "epoch": 12.917779339423753, "grad_norm": 0.07544300705194473, "learning_rate": 2.4725228390723822e-05, "loss": 0.0108, "step": 18382 }, { "epoch": 12.918482080112438, "grad_norm": 0.07686825841665268, "learning_rate": 2.4724759896931365e-05, "loss": 0.0125, "step": 18383 }, { "epoch": 12.919184820801124, "grad_norm": 0.11658834666013718, "learning_rate": 2.472429140313891e-05, "loss": 0.016, "step": 18384 }, { "epoch": 12.91988756148981, "grad_norm": 0.1024327352643013, "learning_rate": 2.472382290934645e-05, "loss": 0.0196, "step": 18385 }, { "epoch": 12.920590302178496, "grad_norm": 0.17011792957782745, "learning_rate": 2.4723354415553993e-05, "loss": 0.0175, "step": 18386 }, { "epoch": 12.921293042867182, "grad_norm": 0.11000148206949234, "learning_rate": 2.4722885921761537e-05, "loss": 0.0187, "step": 18387 }, { "epoch": 12.921995783555868, "grad_norm": 0.10272105038166046, "learning_rate": 2.472241742796908e-05, "loss": 0.0114, "step": 18388 }, { "epoch": 12.922698524244554, "grad_norm": 0.1254201978445053, "learning_rate": 2.472194893417662e-05, "loss": 0.0255, "step": 18389 }, { "epoch": 12.92340126493324, "grad_norm": 0.1043698713183403, "learning_rate": 2.4721480440384165e-05, "loss": 0.0174, "step": 18390 }, { "epoch": 12.924104005621926, "grad_norm": 0.1988629698753357, "learning_rate": 2.472101194659171e-05, "loss": 0.0355, "step": 18391 }, { "epoch": 12.924806746310612, "grad_norm": 0.2773939371109009, "learning_rate": 2.4720543452799252e-05, "loss": 0.0354, "step": 18392 }, { "epoch": 12.925509486999298, "grad_norm": 0.10777490586042404, "learning_rate": 2.4720074959006796e-05, "loss": 0.017, "step": 18393 }, { "epoch": 12.926212227687984, "grad_norm": 0.2089804857969284, "learning_rate": 2.4719606465214336e-05, "loss": 0.0225, "step": 18394 }, { "epoch": 12.92691496837667, "grad_norm": 0.15856994688510895, "learning_rate": 2.471913797142188e-05, "loss": 0.032, "step": 18395 }, { "epoch": 12.927617709065355, "grad_norm": 0.7787186503410339, "learning_rate": 2.4718669477629424e-05, "loss": 0.0407, "step": 18396 }, { "epoch": 12.928320449754041, "grad_norm": 2.4186947345733643, "learning_rate": 2.4718200983836967e-05, "loss": 0.0437, "step": 18397 }, { "epoch": 12.929023190442727, "grad_norm": 0.4225822389125824, "learning_rate": 2.4717732490044508e-05, "loss": 0.0928, "step": 18398 }, { "epoch": 12.929725931131413, "grad_norm": 0.607822597026825, "learning_rate": 2.471726399625205e-05, "loss": 0.1329, "step": 18399 }, { "epoch": 12.9304286718201, "grad_norm": 0.7432097792625427, "learning_rate": 2.4716795502459592e-05, "loss": 0.1592, "step": 18400 }, { "epoch": 12.931131412508785, "grad_norm": 1.1157805919647217, "learning_rate": 2.4716327008667136e-05, "loss": 0.2183, "step": 18401 }, { "epoch": 12.931834153197471, "grad_norm": 1.0842279195785522, "learning_rate": 2.4715858514874676e-05, "loss": 0.0549, "step": 18402 }, { "epoch": 12.932536893886155, "grad_norm": 0.10928057134151459, "learning_rate": 2.471539002108222e-05, "loss": 0.0165, "step": 18403 }, { "epoch": 12.933239634574843, "grad_norm": 0.09425061196088791, "learning_rate": 2.4714921527289763e-05, "loss": 0.0237, "step": 18404 }, { "epoch": 12.933942375263527, "grad_norm": 0.12897741794586182, "learning_rate": 2.4714453033497307e-05, "loss": 0.0264, "step": 18405 }, { "epoch": 12.934645115952213, "grad_norm": 0.16456395387649536, "learning_rate": 2.471398453970485e-05, "loss": 0.0097, "step": 18406 }, { "epoch": 12.935347856640899, "grad_norm": 0.17674964666366577, "learning_rate": 2.471351604591239e-05, "loss": 0.019, "step": 18407 }, { "epoch": 12.936050597329585, "grad_norm": 0.38451114296913147, "learning_rate": 2.4713047552119935e-05, "loss": 0.0061, "step": 18408 }, { "epoch": 12.93675333801827, "grad_norm": 0.12302364408969879, "learning_rate": 2.471257905832748e-05, "loss": 0.0214, "step": 18409 }, { "epoch": 12.937456078706957, "grad_norm": 0.21328331530094147, "learning_rate": 2.4712110564535022e-05, "loss": 0.0144, "step": 18410 }, { "epoch": 12.938158819395642, "grad_norm": 0.2637327015399933, "learning_rate": 2.4711642070742563e-05, "loss": 0.0512, "step": 18411 }, { "epoch": 12.938861560084328, "grad_norm": 0.0806613638997078, "learning_rate": 2.4711173576950106e-05, "loss": 0.0095, "step": 18412 }, { "epoch": 12.939564300773014, "grad_norm": 0.08403951674699783, "learning_rate": 2.471070508315765e-05, "loss": 0.0132, "step": 18413 }, { "epoch": 12.9402670414617, "grad_norm": 0.138656347990036, "learning_rate": 2.4710236589365194e-05, "loss": 0.0228, "step": 18414 }, { "epoch": 12.940969782150386, "grad_norm": 0.21302853524684906, "learning_rate": 2.4709768095572734e-05, "loss": 0.0237, "step": 18415 }, { "epoch": 12.941672522839072, "grad_norm": 0.3154173195362091, "learning_rate": 2.4709299601780278e-05, "loss": 0.0237, "step": 18416 }, { "epoch": 12.942375263527758, "grad_norm": 0.17548871040344238, "learning_rate": 2.4708831107987818e-05, "loss": 0.0489, "step": 18417 }, { "epoch": 12.943078004216444, "grad_norm": 0.2729601263999939, "learning_rate": 2.4708362614195362e-05, "loss": 0.0424, "step": 18418 }, { "epoch": 12.94378074490513, "grad_norm": 0.13813017308712006, "learning_rate": 2.4707894120402906e-05, "loss": 0.0215, "step": 18419 }, { "epoch": 12.944483485593816, "grad_norm": 0.4029909372329712, "learning_rate": 2.4707425626610446e-05, "loss": 0.0609, "step": 18420 }, { "epoch": 12.945186226282502, "grad_norm": 0.36586976051330566, "learning_rate": 2.470695713281799e-05, "loss": 0.0555, "step": 18421 }, { "epoch": 12.945888966971188, "grad_norm": 0.684110701084137, "learning_rate": 2.4706488639025533e-05, "loss": 0.063, "step": 18422 }, { "epoch": 12.946591707659874, "grad_norm": 0.9568367600440979, "learning_rate": 2.4706020145233077e-05, "loss": 0.1124, "step": 18423 }, { "epoch": 12.94729444834856, "grad_norm": 0.9688555002212524, "learning_rate": 2.4705551651440618e-05, "loss": 0.119, "step": 18424 }, { "epoch": 12.947997189037245, "grad_norm": 1.4107670783996582, "learning_rate": 2.470508315764816e-05, "loss": 0.1584, "step": 18425 }, { "epoch": 12.948699929725931, "grad_norm": 1.5312873125076294, "learning_rate": 2.4704614663855705e-05, "loss": 0.1828, "step": 18426 }, { "epoch": 12.949402670414617, "grad_norm": 0.3476126492023468, "learning_rate": 2.470414617006325e-05, "loss": 0.0784, "step": 18427 }, { "epoch": 12.950105411103303, "grad_norm": 0.12631221115589142, "learning_rate": 2.470367767627079e-05, "loss": 0.0325, "step": 18428 }, { "epoch": 12.950808151791989, "grad_norm": 0.09873290359973907, "learning_rate": 2.4703209182478333e-05, "loss": 0.0199, "step": 18429 }, { "epoch": 12.951510892480675, "grad_norm": 0.1298520714044571, "learning_rate": 2.4702740688685876e-05, "loss": 0.0169, "step": 18430 }, { "epoch": 12.952213633169361, "grad_norm": 0.2435922920703888, "learning_rate": 2.470227219489342e-05, "loss": 0.0234, "step": 18431 }, { "epoch": 12.952916373858047, "grad_norm": 0.0952220931649208, "learning_rate": 2.4701803701100964e-05, "loss": 0.0063, "step": 18432 }, { "epoch": 12.953619114546733, "grad_norm": 0.08530265837907791, "learning_rate": 2.4701335207308504e-05, "loss": 0.0143, "step": 18433 }, { "epoch": 12.954321855235419, "grad_norm": 0.2372998297214508, "learning_rate": 2.4700866713516048e-05, "loss": 0.0083, "step": 18434 }, { "epoch": 12.955024595924105, "grad_norm": 0.20771099627017975, "learning_rate": 2.470039821972359e-05, "loss": 0.0158, "step": 18435 }, { "epoch": 12.95572733661279, "grad_norm": 0.10103125870227814, "learning_rate": 2.4699929725931132e-05, "loss": 0.0181, "step": 18436 }, { "epoch": 12.956430077301476, "grad_norm": 0.27543210983276367, "learning_rate": 2.4699461232138672e-05, "loss": 0.0281, "step": 18437 }, { "epoch": 12.957132817990162, "grad_norm": 0.42437395453453064, "learning_rate": 2.4698992738346216e-05, "loss": 0.0361, "step": 18438 }, { "epoch": 12.957835558678848, "grad_norm": 0.19563813507556915, "learning_rate": 2.469852424455376e-05, "loss": 0.03, "step": 18439 }, { "epoch": 12.958538299367534, "grad_norm": 0.14152565598487854, "learning_rate": 2.4698055750761304e-05, "loss": 0.0124, "step": 18440 }, { "epoch": 12.95924104005622, "grad_norm": 0.16743354499340057, "learning_rate": 2.4697587256968844e-05, "loss": 0.0298, "step": 18441 }, { "epoch": 12.959943780744904, "grad_norm": 0.15355831384658813, "learning_rate": 2.4697118763176388e-05, "loss": 0.0284, "step": 18442 }, { "epoch": 12.96064652143359, "grad_norm": 0.10880179703235626, "learning_rate": 2.469665026938393e-05, "loss": 0.0176, "step": 18443 }, { "epoch": 12.961349262122276, "grad_norm": 0.16529427468776703, "learning_rate": 2.4696181775591475e-05, "loss": 0.0298, "step": 18444 }, { "epoch": 12.962052002810962, "grad_norm": 0.15442821383476257, "learning_rate": 2.469571328179902e-05, "loss": 0.0257, "step": 18445 }, { "epoch": 12.962754743499648, "grad_norm": 0.5596612095832825, "learning_rate": 2.469524478800656e-05, "loss": 0.0591, "step": 18446 }, { "epoch": 12.963457484188334, "grad_norm": 0.2807126045227051, "learning_rate": 2.4694776294214103e-05, "loss": 0.0732, "step": 18447 }, { "epoch": 12.96416022487702, "grad_norm": 0.2966386377811432, "learning_rate": 2.4694307800421647e-05, "loss": 0.0933, "step": 18448 }, { "epoch": 12.964862965565706, "grad_norm": 0.9394522905349731, "learning_rate": 2.469383930662919e-05, "loss": 0.1347, "step": 18449 }, { "epoch": 12.965565706254392, "grad_norm": 0.8665157556533813, "learning_rate": 2.469337081283673e-05, "loss": 0.1635, "step": 18450 }, { "epoch": 12.966268446943078, "grad_norm": 3.91904878616333, "learning_rate": 2.4692902319044274e-05, "loss": 0.1987, "step": 18451 }, { "epoch": 12.966971187631763, "grad_norm": 0.2757722735404968, "learning_rate": 2.4692433825251815e-05, "loss": 0.0671, "step": 18452 }, { "epoch": 12.96767392832045, "grad_norm": 0.18353113532066345, "learning_rate": 2.469196533145936e-05, "loss": 0.0192, "step": 18453 }, { "epoch": 12.968376669009135, "grad_norm": 0.16747912764549255, "learning_rate": 2.46914968376669e-05, "loss": 0.0202, "step": 18454 }, { "epoch": 12.969079409697821, "grad_norm": 0.08874363452196121, "learning_rate": 2.4691028343874443e-05, "loss": 0.0188, "step": 18455 }, { "epoch": 12.969782150386507, "grad_norm": 0.11927612125873566, "learning_rate": 2.4690559850081986e-05, "loss": 0.0165, "step": 18456 }, { "epoch": 12.970484891075193, "grad_norm": 0.16334040462970734, "learning_rate": 2.469009135628953e-05, "loss": 0.0204, "step": 18457 }, { "epoch": 12.971187631763879, "grad_norm": 0.18274977803230286, "learning_rate": 2.4689622862497074e-05, "loss": 0.0106, "step": 18458 }, { "epoch": 12.971890372452565, "grad_norm": 0.1400052160024643, "learning_rate": 2.4689154368704614e-05, "loss": 0.0082, "step": 18459 }, { "epoch": 12.97259311314125, "grad_norm": 0.7512627840042114, "learning_rate": 2.4688685874912158e-05, "loss": 0.0239, "step": 18460 }, { "epoch": 12.973295853829937, "grad_norm": 0.21707434952259064, "learning_rate": 2.46882173811197e-05, "loss": 0.012, "step": 18461 }, { "epoch": 12.973998594518623, "grad_norm": 0.33141177892684937, "learning_rate": 2.4687748887327245e-05, "loss": 0.0279, "step": 18462 }, { "epoch": 12.974701335207309, "grad_norm": 0.133554145693779, "learning_rate": 2.4687280393534786e-05, "loss": 0.0181, "step": 18463 }, { "epoch": 12.975404075895995, "grad_norm": 0.14333590865135193, "learning_rate": 2.468681189974233e-05, "loss": 0.0177, "step": 18464 }, { "epoch": 12.97610681658468, "grad_norm": 0.15841613709926605, "learning_rate": 2.4686343405949873e-05, "loss": 0.018, "step": 18465 }, { "epoch": 12.976809557273366, "grad_norm": 0.1913401186466217, "learning_rate": 2.4685874912157417e-05, "loss": 0.0289, "step": 18466 }, { "epoch": 12.977512297962052, "grad_norm": 0.2575125992298126, "learning_rate": 2.4685406418364957e-05, "loss": 0.0262, "step": 18467 }, { "epoch": 12.978215038650738, "grad_norm": 0.14600034058094025, "learning_rate": 2.46849379245725e-05, "loss": 0.0188, "step": 18468 }, { "epoch": 12.978917779339424, "grad_norm": 0.27054017782211304, "learning_rate": 2.468446943078004e-05, "loss": 0.0316, "step": 18469 }, { "epoch": 12.97962052002811, "grad_norm": 0.18664897978305817, "learning_rate": 2.4684000936987585e-05, "loss": 0.0342, "step": 18470 }, { "epoch": 12.980323260716796, "grad_norm": 0.3692598044872284, "learning_rate": 2.468353244319513e-05, "loss": 0.0502, "step": 18471 }, { "epoch": 12.981026001405482, "grad_norm": 0.3828796446323395, "learning_rate": 2.468306394940267e-05, "loss": 0.0596, "step": 18472 }, { "epoch": 12.981728742094168, "grad_norm": 0.30779537558555603, "learning_rate": 2.4682595455610213e-05, "loss": 0.0905, "step": 18473 }, { "epoch": 12.982431482782854, "grad_norm": 0.5738338232040405, "learning_rate": 2.4682126961817756e-05, "loss": 0.1713, "step": 18474 }, { "epoch": 12.98313422347154, "grad_norm": 0.7854738235473633, "learning_rate": 2.46816584680253e-05, "loss": 0.1518, "step": 18475 }, { "epoch": 12.983836964160226, "grad_norm": 1.0201458930969238, "learning_rate": 2.468118997423284e-05, "loss": 0.2094, "step": 18476 }, { "epoch": 12.984539704848912, "grad_norm": 0.16271652281284332, "learning_rate": 2.4680721480440384e-05, "loss": 0.0601, "step": 18477 }, { "epoch": 12.985242445537597, "grad_norm": 0.14751070737838745, "learning_rate": 2.4680252986647928e-05, "loss": 0.0185, "step": 18478 }, { "epoch": 12.985945186226282, "grad_norm": 0.12345702946186066, "learning_rate": 2.467978449285547e-05, "loss": 0.0118, "step": 18479 }, { "epoch": 12.98664792691497, "grad_norm": 0.1835738569498062, "learning_rate": 2.4679315999063012e-05, "loss": 0.0173, "step": 18480 }, { "epoch": 12.987350667603653, "grad_norm": 0.19051265716552734, "learning_rate": 2.4678847505270556e-05, "loss": 0.0115, "step": 18481 }, { "epoch": 12.98805340829234, "grad_norm": 0.25769156217575073, "learning_rate": 2.46783790114781e-05, "loss": 0.0216, "step": 18482 }, { "epoch": 12.988756148981025, "grad_norm": 0.09505195170640945, "learning_rate": 2.4677910517685643e-05, "loss": 0.0118, "step": 18483 }, { "epoch": 12.989458889669711, "grad_norm": 0.128792867064476, "learning_rate": 2.4677442023893187e-05, "loss": 0.0139, "step": 18484 }, { "epoch": 12.990161630358397, "grad_norm": 0.10467003285884857, "learning_rate": 2.4676973530100727e-05, "loss": 0.015, "step": 18485 }, { "epoch": 12.990864371047083, "grad_norm": 0.18369421362876892, "learning_rate": 2.467650503630827e-05, "loss": 0.0313, "step": 18486 }, { "epoch": 12.991567111735769, "grad_norm": 0.142600417137146, "learning_rate": 2.467603654251581e-05, "loss": 0.0319, "step": 18487 }, { "epoch": 12.992269852424455, "grad_norm": 0.21765944361686707, "learning_rate": 2.4675568048723355e-05, "loss": 0.0272, "step": 18488 }, { "epoch": 12.99297259311314, "grad_norm": 0.10590828955173492, "learning_rate": 2.4675099554930895e-05, "loss": 0.0144, "step": 18489 }, { "epoch": 12.993675333801827, "grad_norm": 0.2827707827091217, "learning_rate": 2.467463106113844e-05, "loss": 0.0431, "step": 18490 }, { "epoch": 12.994378074490513, "grad_norm": 0.29167914390563965, "learning_rate": 2.4674162567345983e-05, "loss": 0.0255, "step": 18491 }, { "epoch": 12.995080815179199, "grad_norm": 0.24518261849880219, "learning_rate": 2.4673694073553526e-05, "loss": 0.033, "step": 18492 }, { "epoch": 12.995783555867884, "grad_norm": 0.2212599664926529, "learning_rate": 2.4673225579761067e-05, "loss": 0.0488, "step": 18493 }, { "epoch": 12.99648629655657, "grad_norm": 0.4144318103790283, "learning_rate": 2.467275708596861e-05, "loss": 0.034, "step": 18494 }, { "epoch": 12.997189037245256, "grad_norm": 0.29024064540863037, "learning_rate": 2.4672288592176154e-05, "loss": 0.0842, "step": 18495 }, { "epoch": 12.997891777933942, "grad_norm": 0.4860488474369049, "learning_rate": 2.4671820098383698e-05, "loss": 0.0887, "step": 18496 }, { "epoch": 12.998594518622628, "grad_norm": 0.8123059272766113, "learning_rate": 2.4671351604591242e-05, "loss": 0.1649, "step": 18497 }, { "epoch": 12.999297259311314, "grad_norm": 0.5810554027557373, "learning_rate": 2.4670883110798782e-05, "loss": 0.1301, "step": 18498 }, { "epoch": 13.0, "grad_norm": 0.7503758668899536, "learning_rate": 2.4670414617006326e-05, "loss": 0.1121, "step": 18499 }, { "epoch": 13.000702740688686, "grad_norm": 0.168060302734375, "learning_rate": 2.466994612321387e-05, "loss": 0.0547, "step": 18500 }, { "epoch": 13.001405481377372, "grad_norm": 0.06649822741746902, "learning_rate": 2.4669477629421413e-05, "loss": 0.0125, "step": 18501 }, { "epoch": 13.002108222066058, "grad_norm": 0.11594687402248383, "learning_rate": 2.4669009135628954e-05, "loss": 0.0202, "step": 18502 }, { "epoch": 13.002810962754744, "grad_norm": 0.26654481887817383, "learning_rate": 2.4668540641836497e-05, "loss": 0.0344, "step": 18503 }, { "epoch": 13.00351370344343, "grad_norm": 0.13080233335494995, "learning_rate": 2.4668072148044038e-05, "loss": 0.0173, "step": 18504 }, { "epoch": 13.004216444132116, "grad_norm": 0.1035563200712204, "learning_rate": 2.466760365425158e-05, "loss": 0.0212, "step": 18505 }, { "epoch": 13.004919184820801, "grad_norm": 0.11312659084796906, "learning_rate": 2.466713516045912e-05, "loss": 0.0229, "step": 18506 }, { "epoch": 13.005621925509487, "grad_norm": 0.09812384098768234, "learning_rate": 2.4666666666666665e-05, "loss": 0.0141, "step": 18507 }, { "epoch": 13.006324666198173, "grad_norm": 0.09690176695585251, "learning_rate": 2.466619817287421e-05, "loss": 0.0135, "step": 18508 }, { "epoch": 13.00702740688686, "grad_norm": 0.2184789925813675, "learning_rate": 2.4665729679081753e-05, "loss": 0.0302, "step": 18509 }, { "epoch": 13.007730147575545, "grad_norm": 0.13525304198265076, "learning_rate": 2.4665261185289297e-05, "loss": 0.0166, "step": 18510 }, { "epoch": 13.008432888264231, "grad_norm": 0.19814832508563995, "learning_rate": 2.4664792691496837e-05, "loss": 0.0166, "step": 18511 }, { "epoch": 13.009135628952917, "grad_norm": 0.1190832182765007, "learning_rate": 2.466432419770438e-05, "loss": 0.0207, "step": 18512 }, { "epoch": 13.009838369641603, "grad_norm": 0.12326256185770035, "learning_rate": 2.4663855703911924e-05, "loss": 0.0201, "step": 18513 }, { "epoch": 13.010541110330289, "grad_norm": 0.19295985996723175, "learning_rate": 2.4663387210119468e-05, "loss": 0.0335, "step": 18514 }, { "epoch": 13.011243851018975, "grad_norm": 0.28618597984313965, "learning_rate": 2.466291871632701e-05, "loss": 0.0384, "step": 18515 }, { "epoch": 13.01194659170766, "grad_norm": 0.25275176763534546, "learning_rate": 2.4662450222534552e-05, "loss": 0.0544, "step": 18516 }, { "epoch": 13.012649332396347, "grad_norm": 0.17657126486301422, "learning_rate": 2.4661981728742096e-05, "loss": 0.0162, "step": 18517 }, { "epoch": 13.013352073085033, "grad_norm": 0.12447217851877213, "learning_rate": 2.466151323494964e-05, "loss": 0.0178, "step": 18518 }, { "epoch": 13.014054813773717, "grad_norm": 0.25328919291496277, "learning_rate": 2.466104474115718e-05, "loss": 0.0538, "step": 18519 }, { "epoch": 13.014757554462403, "grad_norm": 0.2453470230102539, "learning_rate": 2.4660576247364724e-05, "loss": 0.0668, "step": 18520 }, { "epoch": 13.015460295151088, "grad_norm": 0.33894863724708557, "learning_rate": 2.4660107753572267e-05, "loss": 0.108, "step": 18521 }, { "epoch": 13.016163035839774, "grad_norm": 0.5655122995376587, "learning_rate": 2.4659639259779808e-05, "loss": 0.1313, "step": 18522 }, { "epoch": 13.01686577652846, "grad_norm": 0.6342031359672546, "learning_rate": 2.465917076598735e-05, "loss": 0.1414, "step": 18523 }, { "epoch": 13.017568517217146, "grad_norm": 1.6340038776397705, "learning_rate": 2.4658702272194892e-05, "loss": 0.1741, "step": 18524 }, { "epoch": 13.018271257905832, "grad_norm": 0.15882478654384613, "learning_rate": 2.4658233778402436e-05, "loss": 0.0647, "step": 18525 }, { "epoch": 13.018973998594518, "grad_norm": 0.12372621148824692, "learning_rate": 2.465776528460998e-05, "loss": 0.0261, "step": 18526 }, { "epoch": 13.019676739283204, "grad_norm": 0.11921662092208862, "learning_rate": 2.4657296790817523e-05, "loss": 0.0123, "step": 18527 }, { "epoch": 13.02037947997189, "grad_norm": 0.23629894852638245, "learning_rate": 2.4656828297025063e-05, "loss": 0.0118, "step": 18528 }, { "epoch": 13.021082220660576, "grad_norm": 0.12628266215324402, "learning_rate": 2.4656359803232607e-05, "loss": 0.0192, "step": 18529 }, { "epoch": 13.021784961349262, "grad_norm": 0.06747916340827942, "learning_rate": 2.465589130944015e-05, "loss": 0.008, "step": 18530 }, { "epoch": 13.022487702037948, "grad_norm": 0.08280869573354721, "learning_rate": 2.4655422815647694e-05, "loss": 0.0065, "step": 18531 }, { "epoch": 13.023190442726634, "grad_norm": 0.7480136752128601, "learning_rate": 2.4654954321855235e-05, "loss": 0.0477, "step": 18532 }, { "epoch": 13.02389318341532, "grad_norm": 0.08525025844573975, "learning_rate": 2.465448582806278e-05, "loss": 0.0174, "step": 18533 }, { "epoch": 13.024595924104005, "grad_norm": 0.07752720266580582, "learning_rate": 2.4654017334270322e-05, "loss": 0.0067, "step": 18534 }, { "epoch": 13.025298664792691, "grad_norm": 0.13394111394882202, "learning_rate": 2.4653548840477866e-05, "loss": 0.0141, "step": 18535 }, { "epoch": 13.026001405481377, "grad_norm": 0.09269136190414429, "learning_rate": 2.465308034668541e-05, "loss": 0.0223, "step": 18536 }, { "epoch": 13.026704146170063, "grad_norm": 0.18186631798744202, "learning_rate": 2.465261185289295e-05, "loss": 0.0124, "step": 18537 }, { "epoch": 13.02740688685875, "grad_norm": 0.1242147907614708, "learning_rate": 2.4652143359100494e-05, "loss": 0.0152, "step": 18538 }, { "epoch": 13.028109627547435, "grad_norm": 0.2334948033094406, "learning_rate": 2.4651674865308034e-05, "loss": 0.0341, "step": 18539 }, { "epoch": 13.028812368236121, "grad_norm": 0.22023259103298187, "learning_rate": 2.4651206371515578e-05, "loss": 0.0313, "step": 18540 }, { "epoch": 13.029515108924807, "grad_norm": 0.1795574277639389, "learning_rate": 2.4650737877723118e-05, "loss": 0.0199, "step": 18541 }, { "epoch": 13.030217849613493, "grad_norm": 0.22308248281478882, "learning_rate": 2.4650269383930662e-05, "loss": 0.0382, "step": 18542 }, { "epoch": 13.030920590302179, "grad_norm": 0.6981537938117981, "learning_rate": 2.4649800890138206e-05, "loss": 0.0224, "step": 18543 }, { "epoch": 13.031623330990865, "grad_norm": 0.33592069149017334, "learning_rate": 2.464933239634575e-05, "loss": 0.0419, "step": 18544 }, { "epoch": 13.03232607167955, "grad_norm": 0.4708542227745056, "learning_rate": 2.464886390255329e-05, "loss": 0.0565, "step": 18545 }, { "epoch": 13.033028812368237, "grad_norm": 0.413415789604187, "learning_rate": 2.4648395408760833e-05, "loss": 0.0837, "step": 18546 }, { "epoch": 13.033731553056922, "grad_norm": 0.44954776763916016, "learning_rate": 2.4647926914968377e-05, "loss": 0.1446, "step": 18547 }, { "epoch": 13.034434293745608, "grad_norm": 0.7414532899856567, "learning_rate": 2.464745842117592e-05, "loss": 0.1555, "step": 18548 }, { "epoch": 13.035137034434294, "grad_norm": 0.9729117155075073, "learning_rate": 2.4646989927383465e-05, "loss": 0.1855, "step": 18549 }, { "epoch": 13.03583977512298, "grad_norm": 0.4808861315250397, "learning_rate": 2.4646521433591005e-05, "loss": 0.0767, "step": 18550 }, { "epoch": 13.036542515811666, "grad_norm": 0.12099452316761017, "learning_rate": 2.464605293979855e-05, "loss": 0.0188, "step": 18551 }, { "epoch": 13.037245256500352, "grad_norm": 0.08428915590047836, "learning_rate": 2.4645584446006092e-05, "loss": 0.014, "step": 18552 }, { "epoch": 13.037947997189038, "grad_norm": 0.1634232997894287, "learning_rate": 2.4645115952213636e-05, "loss": 0.0116, "step": 18553 }, { "epoch": 13.038650737877724, "grad_norm": 0.11853662878274918, "learning_rate": 2.4644647458421176e-05, "loss": 0.0247, "step": 18554 }, { "epoch": 13.03935347856641, "grad_norm": 0.0944572314620018, "learning_rate": 2.464417896462872e-05, "loss": 0.0106, "step": 18555 }, { "epoch": 13.040056219255096, "grad_norm": 0.07382137328386307, "learning_rate": 2.464371047083626e-05, "loss": 0.0104, "step": 18556 }, { "epoch": 13.04075895994378, "grad_norm": 0.19164294004440308, "learning_rate": 2.4643241977043804e-05, "loss": 0.0121, "step": 18557 }, { "epoch": 13.041461700632466, "grad_norm": 0.11037357151508331, "learning_rate": 2.4642773483251348e-05, "loss": 0.0168, "step": 18558 }, { "epoch": 13.042164441321152, "grad_norm": 0.07882235199213028, "learning_rate": 2.4642304989458888e-05, "loss": 0.0068, "step": 18559 }, { "epoch": 13.042867182009838, "grad_norm": 0.1923792064189911, "learning_rate": 2.4641836495666432e-05, "loss": 0.0421, "step": 18560 }, { "epoch": 13.043569922698524, "grad_norm": 0.11677702516317368, "learning_rate": 2.4641368001873976e-05, "loss": 0.0141, "step": 18561 }, { "epoch": 13.04427266338721, "grad_norm": 0.1622200757265091, "learning_rate": 2.464089950808152e-05, "loss": 0.0304, "step": 18562 }, { "epoch": 13.044975404075895, "grad_norm": 0.15874435007572174, "learning_rate": 2.464043101428906e-05, "loss": 0.0066, "step": 18563 }, { "epoch": 13.045678144764581, "grad_norm": 0.6321475505828857, "learning_rate": 2.4639962520496604e-05, "loss": 0.0539, "step": 18564 }, { "epoch": 13.046380885453267, "grad_norm": 0.196133092045784, "learning_rate": 2.4639494026704147e-05, "loss": 0.0296, "step": 18565 }, { "epoch": 13.047083626141953, "grad_norm": 0.21184587478637695, "learning_rate": 2.463902553291169e-05, "loss": 0.0245, "step": 18566 }, { "epoch": 13.047786366830639, "grad_norm": 0.15182144939899445, "learning_rate": 2.463855703911923e-05, "loss": 0.028, "step": 18567 }, { "epoch": 13.048489107519325, "grad_norm": 0.16640950739383698, "learning_rate": 2.4638088545326775e-05, "loss": 0.035, "step": 18568 }, { "epoch": 13.049191848208011, "grad_norm": 0.17596882581710815, "learning_rate": 2.463762005153432e-05, "loss": 0.0299, "step": 18569 }, { "epoch": 13.049894588896697, "grad_norm": 0.434390127658844, "learning_rate": 2.4637151557741862e-05, "loss": 0.06, "step": 18570 }, { "epoch": 13.050597329585383, "grad_norm": 0.5160647630691528, "learning_rate": 2.4636683063949403e-05, "loss": 0.1078, "step": 18571 }, { "epoch": 13.051300070274069, "grad_norm": 0.661831796169281, "learning_rate": 2.4636214570156947e-05, "loss": 0.1393, "step": 18572 }, { "epoch": 13.052002810962755, "grad_norm": 1.3882418870925903, "learning_rate": 2.463574607636449e-05, "loss": 0.1444, "step": 18573 }, { "epoch": 13.05270555165144, "grad_norm": 1.4505043029785156, "learning_rate": 2.463527758257203e-05, "loss": 0.191, "step": 18574 }, { "epoch": 13.053408292340126, "grad_norm": 0.29374077916145325, "learning_rate": 2.4634809088779574e-05, "loss": 0.0831, "step": 18575 }, { "epoch": 13.054111033028812, "grad_norm": 0.17182524502277374, "learning_rate": 2.4634340594987115e-05, "loss": 0.0342, "step": 18576 }, { "epoch": 13.054813773717498, "grad_norm": 0.09584947675466537, "learning_rate": 2.463387210119466e-05, "loss": 0.0085, "step": 18577 }, { "epoch": 13.055516514406184, "grad_norm": 0.4153992533683777, "learning_rate": 2.4633403607402202e-05, "loss": 0.0212, "step": 18578 }, { "epoch": 13.05621925509487, "grad_norm": 0.10845930874347687, "learning_rate": 2.4632935113609746e-05, "loss": 0.0149, "step": 18579 }, { "epoch": 13.056921995783556, "grad_norm": 0.28872376680374146, "learning_rate": 2.4632466619817286e-05, "loss": 0.0186, "step": 18580 }, { "epoch": 13.057624736472242, "grad_norm": 0.08046048134565353, "learning_rate": 2.463199812602483e-05, "loss": 0.0196, "step": 18581 }, { "epoch": 13.058327477160928, "grad_norm": 0.19065573811531067, "learning_rate": 2.4631529632232374e-05, "loss": 0.0137, "step": 18582 }, { "epoch": 13.059030217849614, "grad_norm": 0.10929995030164719, "learning_rate": 2.4631061138439917e-05, "loss": 0.0272, "step": 18583 }, { "epoch": 13.0597329585383, "grad_norm": 0.13938286900520325, "learning_rate": 2.463059264464746e-05, "loss": 0.0221, "step": 18584 }, { "epoch": 13.060435699226986, "grad_norm": 0.178181990981102, "learning_rate": 2.4630124150855e-05, "loss": 0.0195, "step": 18585 }, { "epoch": 13.061138439915672, "grad_norm": 0.11640509217977524, "learning_rate": 2.4629655657062545e-05, "loss": 0.0079, "step": 18586 }, { "epoch": 13.061841180604358, "grad_norm": 0.16038818657398224, "learning_rate": 2.462918716327009e-05, "loss": 0.0194, "step": 18587 }, { "epoch": 13.062543921293043, "grad_norm": 0.22954687476158142, "learning_rate": 2.4628718669477633e-05, "loss": 0.0124, "step": 18588 }, { "epoch": 13.06324666198173, "grad_norm": 0.1420469582080841, "learning_rate": 2.4628250175685173e-05, "loss": 0.014, "step": 18589 }, { "epoch": 13.063949402670415, "grad_norm": 0.2634880244731903, "learning_rate": 2.4627781681892717e-05, "loss": 0.0355, "step": 18590 }, { "epoch": 13.064652143359101, "grad_norm": 0.22510932385921478, "learning_rate": 2.4627313188100257e-05, "loss": 0.0117, "step": 18591 }, { "epoch": 13.065354884047787, "grad_norm": 0.10772275179624557, "learning_rate": 2.46268446943078e-05, "loss": 0.0277, "step": 18592 }, { "epoch": 13.066057624736473, "grad_norm": 0.2580512762069702, "learning_rate": 2.462637620051534e-05, "loss": 0.0288, "step": 18593 }, { "epoch": 13.066760365425159, "grad_norm": 0.21408745646476746, "learning_rate": 2.4625907706722885e-05, "loss": 0.0418, "step": 18594 }, { "epoch": 13.067463106113845, "grad_norm": 0.6531588435173035, "learning_rate": 2.462543921293043e-05, "loss": 0.0531, "step": 18595 }, { "epoch": 13.068165846802529, "grad_norm": 0.7640916109085083, "learning_rate": 2.4624970719137972e-05, "loss": 0.0728, "step": 18596 }, { "epoch": 13.068868587491215, "grad_norm": 1.0346428155899048, "learning_rate": 2.4624502225345516e-05, "loss": 0.1577, "step": 18597 }, { "epoch": 13.0695713281799, "grad_norm": 0.8989861011505127, "learning_rate": 2.4624033731553056e-05, "loss": 0.1401, "step": 18598 }, { "epoch": 13.070274068868587, "grad_norm": 1.008493423461914, "learning_rate": 2.46235652377606e-05, "loss": 0.174, "step": 18599 }, { "epoch": 13.070976809557273, "grad_norm": 0.3124697804450989, "learning_rate": 2.4623096743968144e-05, "loss": 0.0478, "step": 18600 }, { "epoch": 13.071679550245959, "grad_norm": 0.2847104072570801, "learning_rate": 2.4622628250175687e-05, "loss": 0.0269, "step": 18601 }, { "epoch": 13.072382290934645, "grad_norm": 0.1235419511795044, "learning_rate": 2.4622159756383228e-05, "loss": 0.0249, "step": 18602 }, { "epoch": 13.07308503162333, "grad_norm": 0.10361078381538391, "learning_rate": 2.462169126259077e-05, "loss": 0.013, "step": 18603 }, { "epoch": 13.073787772312016, "grad_norm": 0.10881944745779037, "learning_rate": 2.4621222768798315e-05, "loss": 0.0087, "step": 18604 }, { "epoch": 13.074490513000702, "grad_norm": 0.10082919895648956, "learning_rate": 2.462075427500586e-05, "loss": 0.0127, "step": 18605 }, { "epoch": 13.075193253689388, "grad_norm": 0.1424817591905594, "learning_rate": 2.46202857812134e-05, "loss": 0.0145, "step": 18606 }, { "epoch": 13.075895994378074, "grad_norm": 0.10792720317840576, "learning_rate": 2.4619817287420943e-05, "loss": 0.0167, "step": 18607 }, { "epoch": 13.07659873506676, "grad_norm": 0.42416146397590637, "learning_rate": 2.4619348793628487e-05, "loss": 0.0175, "step": 18608 }, { "epoch": 13.077301475755446, "grad_norm": 0.06898590177297592, "learning_rate": 2.4618880299836027e-05, "loss": 0.0065, "step": 18609 }, { "epoch": 13.078004216444132, "grad_norm": 0.2205992490053177, "learning_rate": 2.461841180604357e-05, "loss": 0.0281, "step": 18610 }, { "epoch": 13.078706957132818, "grad_norm": 0.13458378612995148, "learning_rate": 2.461794331225111e-05, "loss": 0.0235, "step": 18611 }, { "epoch": 13.079409697821504, "grad_norm": 0.3432783782482147, "learning_rate": 2.4617474818458655e-05, "loss": 0.0212, "step": 18612 }, { "epoch": 13.08011243851019, "grad_norm": 0.1825275868177414, "learning_rate": 2.46170063246662e-05, "loss": 0.025, "step": 18613 }, { "epoch": 13.080815179198876, "grad_norm": 0.17827914655208588, "learning_rate": 2.4616537830873742e-05, "loss": 0.0291, "step": 18614 }, { "epoch": 13.081517919887562, "grad_norm": 0.49953219294548035, "learning_rate": 2.4616069337081283e-05, "loss": 0.0434, "step": 18615 }, { "epoch": 13.082220660576247, "grad_norm": 0.1376103311777115, "learning_rate": 2.4615600843288826e-05, "loss": 0.0176, "step": 18616 }, { "epoch": 13.082923401264933, "grad_norm": 0.3533742427825928, "learning_rate": 2.461513234949637e-05, "loss": 0.0439, "step": 18617 }, { "epoch": 13.08362614195362, "grad_norm": 0.23741407692432404, "learning_rate": 2.4614663855703914e-05, "loss": 0.0248, "step": 18618 }, { "epoch": 13.084328882642305, "grad_norm": 0.8486049771308899, "learning_rate": 2.4614195361911454e-05, "loss": 0.0469, "step": 18619 }, { "epoch": 13.085031623330991, "grad_norm": 0.2770600914955139, "learning_rate": 2.4613726868118998e-05, "loss": 0.0665, "step": 18620 }, { "epoch": 13.085734364019677, "grad_norm": 1.1300615072250366, "learning_rate": 2.461325837432654e-05, "loss": 0.1124, "step": 18621 }, { "epoch": 13.086437104708363, "grad_norm": 1.1472982168197632, "learning_rate": 2.4612789880534085e-05, "loss": 0.1247, "step": 18622 }, { "epoch": 13.087139845397049, "grad_norm": 0.9803206920623779, "learning_rate": 2.461232138674163e-05, "loss": 0.1602, "step": 18623 }, { "epoch": 13.087842586085735, "grad_norm": 1.4734498262405396, "learning_rate": 2.461185289294917e-05, "loss": 0.1812, "step": 18624 }, { "epoch": 13.08854532677442, "grad_norm": 0.17861908674240112, "learning_rate": 2.4611384399156713e-05, "loss": 0.062, "step": 18625 }, { "epoch": 13.089248067463107, "grad_norm": 0.1433819979429245, "learning_rate": 2.4610915905364254e-05, "loss": 0.0171, "step": 18626 }, { "epoch": 13.089950808151793, "grad_norm": 0.10611774027347565, "learning_rate": 2.4610447411571797e-05, "loss": 0.0217, "step": 18627 }, { "epoch": 13.090653548840478, "grad_norm": 0.11962906271219254, "learning_rate": 2.4609978917779338e-05, "loss": 0.0213, "step": 18628 }, { "epoch": 13.091356289529164, "grad_norm": 0.23700641095638275, "learning_rate": 2.460951042398688e-05, "loss": 0.0394, "step": 18629 }, { "epoch": 13.09205903021785, "grad_norm": 0.07112816721200943, "learning_rate": 2.4609041930194425e-05, "loss": 0.0062, "step": 18630 }, { "epoch": 13.092761770906536, "grad_norm": 0.0844600722193718, "learning_rate": 2.460857343640197e-05, "loss": 0.0082, "step": 18631 }, { "epoch": 13.093464511595222, "grad_norm": 0.200514018535614, "learning_rate": 2.460810494260951e-05, "loss": 0.0174, "step": 18632 }, { "epoch": 13.094167252283908, "grad_norm": 0.14914929866790771, "learning_rate": 2.4607636448817053e-05, "loss": 0.0225, "step": 18633 }, { "epoch": 13.094869992972592, "grad_norm": 0.3022434115409851, "learning_rate": 2.4607167955024597e-05, "loss": 0.0165, "step": 18634 }, { "epoch": 13.095572733661278, "grad_norm": 0.25525161623954773, "learning_rate": 2.460669946123214e-05, "loss": 0.0111, "step": 18635 }, { "epoch": 13.096275474349964, "grad_norm": 0.10273291915655136, "learning_rate": 2.4606230967439684e-05, "loss": 0.0178, "step": 18636 }, { "epoch": 13.09697821503865, "grad_norm": 0.11245965957641602, "learning_rate": 2.4605762473647224e-05, "loss": 0.0112, "step": 18637 }, { "epoch": 13.097680955727336, "grad_norm": 0.10985011607408524, "learning_rate": 2.4605293979854768e-05, "loss": 0.0243, "step": 18638 }, { "epoch": 13.098383696416022, "grad_norm": 0.17426125705242157, "learning_rate": 2.4604825486062312e-05, "loss": 0.0257, "step": 18639 }, { "epoch": 13.099086437104708, "grad_norm": 0.16200612485408783, "learning_rate": 2.4604356992269855e-05, "loss": 0.0292, "step": 18640 }, { "epoch": 13.099789177793394, "grad_norm": 0.22787122428417206, "learning_rate": 2.4603888498477396e-05, "loss": 0.0361, "step": 18641 }, { "epoch": 13.10049191848208, "grad_norm": 0.28818991780281067, "learning_rate": 2.460342000468494e-05, "loss": 0.0109, "step": 18642 }, { "epoch": 13.101194659170766, "grad_norm": 0.39411109685897827, "learning_rate": 2.4602951510892483e-05, "loss": 0.0326, "step": 18643 }, { "epoch": 13.101897399859451, "grad_norm": 0.21267899870872498, "learning_rate": 2.4602483017100024e-05, "loss": 0.0294, "step": 18644 }, { "epoch": 13.102600140548137, "grad_norm": 0.3787429928779602, "learning_rate": 2.4602014523307564e-05, "loss": 0.0929, "step": 18645 }, { "epoch": 13.103302881236823, "grad_norm": 0.5927630662918091, "learning_rate": 2.4601546029515108e-05, "loss": 0.0904, "step": 18646 }, { "epoch": 13.10400562192551, "grad_norm": 0.379700243473053, "learning_rate": 2.460107753572265e-05, "loss": 0.1286, "step": 18647 }, { "epoch": 13.104708362614195, "grad_norm": 0.7867081761360168, "learning_rate": 2.4600609041930195e-05, "loss": 0.1736, "step": 18648 }, { "epoch": 13.105411103302881, "grad_norm": 0.9323194026947021, "learning_rate": 2.460014054813774e-05, "loss": 0.1639, "step": 18649 }, { "epoch": 13.106113843991567, "grad_norm": 0.27301278710365295, "learning_rate": 2.459967205434528e-05, "loss": 0.0524, "step": 18650 }, { "epoch": 13.106816584680253, "grad_norm": 0.07664067298173904, "learning_rate": 2.4599203560552823e-05, "loss": 0.0165, "step": 18651 }, { "epoch": 13.107519325368939, "grad_norm": 0.11759579926729202, "learning_rate": 2.4598735066760367e-05, "loss": 0.0193, "step": 18652 }, { "epoch": 13.108222066057625, "grad_norm": 0.14342741668224335, "learning_rate": 2.459826657296791e-05, "loss": 0.0228, "step": 18653 }, { "epoch": 13.10892480674631, "grad_norm": 0.18089112639427185, "learning_rate": 2.459779807917545e-05, "loss": 0.0296, "step": 18654 }, { "epoch": 13.109627547434997, "grad_norm": 0.2002035677433014, "learning_rate": 2.4597329585382994e-05, "loss": 0.0192, "step": 18655 }, { "epoch": 13.110330288123683, "grad_norm": 0.11688072234392166, "learning_rate": 2.4596861091590538e-05, "loss": 0.0163, "step": 18656 }, { "epoch": 13.111033028812368, "grad_norm": 0.19388683140277863, "learning_rate": 2.4596392597798082e-05, "loss": 0.0226, "step": 18657 }, { "epoch": 13.111735769501054, "grad_norm": 0.10524438321590424, "learning_rate": 2.4595924104005622e-05, "loss": 0.0155, "step": 18658 }, { "epoch": 13.11243851018974, "grad_norm": 0.271589994430542, "learning_rate": 2.4595455610213166e-05, "loss": 0.0207, "step": 18659 }, { "epoch": 13.113141250878426, "grad_norm": 0.11120559275150299, "learning_rate": 2.459498711642071e-05, "loss": 0.0149, "step": 18660 }, { "epoch": 13.113843991567112, "grad_norm": 0.06665433198213577, "learning_rate": 2.459451862262825e-05, "loss": 0.0082, "step": 18661 }, { "epoch": 13.114546732255798, "grad_norm": 0.151397243142128, "learning_rate": 2.4594050128835794e-05, "loss": 0.0216, "step": 18662 }, { "epoch": 13.115249472944484, "grad_norm": 0.12187936156988144, "learning_rate": 2.4593581635043334e-05, "loss": 0.0141, "step": 18663 }, { "epoch": 13.11595221363317, "grad_norm": 0.20008927583694458, "learning_rate": 2.4593113141250878e-05, "loss": 0.0421, "step": 18664 }, { "epoch": 13.116654954321856, "grad_norm": 0.14970193803310394, "learning_rate": 2.459264464745842e-05, "loss": 0.0319, "step": 18665 }, { "epoch": 13.117357695010542, "grad_norm": 0.3479498028755188, "learning_rate": 2.4592176153665965e-05, "loss": 0.015, "step": 18666 }, { "epoch": 13.118060435699228, "grad_norm": 0.13271909952163696, "learning_rate": 2.4591707659873506e-05, "loss": 0.0332, "step": 18667 }, { "epoch": 13.118763176387914, "grad_norm": 0.6495242118835449, "learning_rate": 2.459123916608105e-05, "loss": 0.0528, "step": 18668 }, { "epoch": 13.1194659170766, "grad_norm": 0.2107241153717041, "learning_rate": 2.4590770672288593e-05, "loss": 0.0348, "step": 18669 }, { "epoch": 13.120168657765285, "grad_norm": 0.3219349980354309, "learning_rate": 2.4590302178496137e-05, "loss": 0.0846, "step": 18670 }, { "epoch": 13.120871398453971, "grad_norm": 0.3405572772026062, "learning_rate": 2.4589833684703677e-05, "loss": 0.1064, "step": 18671 }, { "epoch": 13.121574139142655, "grad_norm": 1.259527564048767, "learning_rate": 2.458936519091122e-05, "loss": 0.1131, "step": 18672 }, { "epoch": 13.122276879831341, "grad_norm": 0.8369753360748291, "learning_rate": 2.4588896697118765e-05, "loss": 0.1572, "step": 18673 }, { "epoch": 13.122979620520027, "grad_norm": 0.6910134553909302, "learning_rate": 2.4588428203326308e-05, "loss": 0.1903, "step": 18674 }, { "epoch": 13.123682361208713, "grad_norm": 0.19179530441761017, "learning_rate": 2.4587959709533852e-05, "loss": 0.0498, "step": 18675 }, { "epoch": 13.1243851018974, "grad_norm": 0.08831533044576645, "learning_rate": 2.4587491215741392e-05, "loss": 0.02, "step": 18676 }, { "epoch": 13.125087842586085, "grad_norm": 0.20794019103050232, "learning_rate": 2.4587022721948936e-05, "loss": 0.0137, "step": 18677 }, { "epoch": 13.125790583274771, "grad_norm": 0.06276681274175644, "learning_rate": 2.4586554228156476e-05, "loss": 0.0095, "step": 18678 }, { "epoch": 13.126493323963457, "grad_norm": 0.11138172447681427, "learning_rate": 2.458608573436402e-05, "loss": 0.0151, "step": 18679 }, { "epoch": 13.127196064652143, "grad_norm": 0.17023511230945587, "learning_rate": 2.458561724057156e-05, "loss": 0.0171, "step": 18680 }, { "epoch": 13.127898805340829, "grad_norm": 0.10463574528694153, "learning_rate": 2.4585148746779104e-05, "loss": 0.0111, "step": 18681 }, { "epoch": 13.128601546029515, "grad_norm": 0.0652313157916069, "learning_rate": 2.4584680252986648e-05, "loss": 0.0134, "step": 18682 }, { "epoch": 13.1293042867182, "grad_norm": 0.08909182995557785, "learning_rate": 2.458421175919419e-05, "loss": 0.0204, "step": 18683 }, { "epoch": 13.130007027406887, "grad_norm": 0.08428553491830826, "learning_rate": 2.4583743265401732e-05, "loss": 0.0109, "step": 18684 }, { "epoch": 13.130709768095572, "grad_norm": 0.10639864206314087, "learning_rate": 2.4583274771609276e-05, "loss": 0.0146, "step": 18685 }, { "epoch": 13.131412508784258, "grad_norm": 0.18432995676994324, "learning_rate": 2.458280627781682e-05, "loss": 0.0142, "step": 18686 }, { "epoch": 13.132115249472944, "grad_norm": 0.1413840502500534, "learning_rate": 2.4582337784024363e-05, "loss": 0.0265, "step": 18687 }, { "epoch": 13.13281799016163, "grad_norm": 0.13511279225349426, "learning_rate": 2.4581869290231907e-05, "loss": 0.0141, "step": 18688 }, { "epoch": 13.133520730850316, "grad_norm": 0.20341119170188904, "learning_rate": 2.4581400796439447e-05, "loss": 0.031, "step": 18689 }, { "epoch": 13.134223471539002, "grad_norm": 0.19900797307491302, "learning_rate": 2.458093230264699e-05, "loss": 0.0268, "step": 18690 }, { "epoch": 13.134926212227688, "grad_norm": 0.1738770306110382, "learning_rate": 2.4580463808854535e-05, "loss": 0.0252, "step": 18691 }, { "epoch": 13.135628952916374, "grad_norm": 0.12680014967918396, "learning_rate": 2.457999531506208e-05, "loss": 0.0377, "step": 18692 }, { "epoch": 13.13633169360506, "grad_norm": 0.29206374287605286, "learning_rate": 2.457952682126962e-05, "loss": 0.0559, "step": 18693 }, { "epoch": 13.137034434293746, "grad_norm": 0.2640987038612366, "learning_rate": 2.4579058327477162e-05, "loss": 0.0428, "step": 18694 }, { "epoch": 13.137737174982432, "grad_norm": 0.29073306918144226, "learning_rate": 2.4578589833684706e-05, "loss": 0.0719, "step": 18695 }, { "epoch": 13.138439915671118, "grad_norm": 0.38469430804252625, "learning_rate": 2.4578121339892247e-05, "loss": 0.0883, "step": 18696 }, { "epoch": 13.139142656359803, "grad_norm": 0.6947897672653198, "learning_rate": 2.4577652846099787e-05, "loss": 0.1353, "step": 18697 }, { "epoch": 13.13984539704849, "grad_norm": 0.8098961710929871, "learning_rate": 2.457718435230733e-05, "loss": 0.1776, "step": 18698 }, { "epoch": 13.140548137737175, "grad_norm": 1.3150242567062378, "learning_rate": 2.4576715858514874e-05, "loss": 0.1707, "step": 18699 }, { "epoch": 13.141250878425861, "grad_norm": 0.24579647183418274, "learning_rate": 2.4576247364722418e-05, "loss": 0.0615, "step": 18700 }, { "epoch": 13.141953619114547, "grad_norm": 0.10368344932794571, "learning_rate": 2.4575778870929962e-05, "loss": 0.0217, "step": 18701 }, { "epoch": 13.142656359803233, "grad_norm": 0.11902104318141937, "learning_rate": 2.4575310377137502e-05, "loss": 0.0173, "step": 18702 }, { "epoch": 13.143359100491919, "grad_norm": 0.09565030038356781, "learning_rate": 2.4574841883345046e-05, "loss": 0.0138, "step": 18703 }, { "epoch": 13.144061841180605, "grad_norm": 0.10937809199094772, "learning_rate": 2.457437338955259e-05, "loss": 0.0107, "step": 18704 }, { "epoch": 13.14476458186929, "grad_norm": 0.07908185571432114, "learning_rate": 2.4573904895760133e-05, "loss": 0.009, "step": 18705 }, { "epoch": 13.145467322557977, "grad_norm": 0.06186619773507118, "learning_rate": 2.4573436401967674e-05, "loss": 0.0104, "step": 18706 }, { "epoch": 13.146170063246663, "grad_norm": 0.07982789725065231, "learning_rate": 2.4572967908175217e-05, "loss": 0.0103, "step": 18707 }, { "epoch": 13.146872803935349, "grad_norm": 0.15402719378471375, "learning_rate": 2.457249941438276e-05, "loss": 0.0089, "step": 18708 }, { "epoch": 13.147575544624035, "grad_norm": 0.20577141642570496, "learning_rate": 2.4572030920590305e-05, "loss": 0.0253, "step": 18709 }, { "epoch": 13.14827828531272, "grad_norm": 0.12410350143909454, "learning_rate": 2.4571562426797845e-05, "loss": 0.0145, "step": 18710 }, { "epoch": 13.148981026001405, "grad_norm": 0.15314260125160217, "learning_rate": 2.457109393300539e-05, "loss": 0.0243, "step": 18711 }, { "epoch": 13.14968376669009, "grad_norm": 0.20485274493694305, "learning_rate": 2.4570625439212933e-05, "loss": 0.0252, "step": 18712 }, { "epoch": 13.150386507378776, "grad_norm": 0.6756203174591064, "learning_rate": 2.4570156945420473e-05, "loss": 0.0124, "step": 18713 }, { "epoch": 13.151089248067462, "grad_norm": 0.2840125858783722, "learning_rate": 2.4569688451628017e-05, "loss": 0.022, "step": 18714 }, { "epoch": 13.151791988756148, "grad_norm": 0.16943730413913727, "learning_rate": 2.4569219957835557e-05, "loss": 0.0235, "step": 18715 }, { "epoch": 13.152494729444834, "grad_norm": 0.20157243311405182, "learning_rate": 2.45687514640431e-05, "loss": 0.0136, "step": 18716 }, { "epoch": 13.15319747013352, "grad_norm": 0.33288484811782837, "learning_rate": 2.4568282970250644e-05, "loss": 0.0302, "step": 18717 }, { "epoch": 13.153900210822206, "grad_norm": 0.36992526054382324, "learning_rate": 2.4567814476458188e-05, "loss": 0.0554, "step": 18718 }, { "epoch": 13.154602951510892, "grad_norm": 0.3193305432796478, "learning_rate": 2.456734598266573e-05, "loss": 0.0465, "step": 18719 }, { "epoch": 13.155305692199578, "grad_norm": 0.3015175759792328, "learning_rate": 2.4566877488873272e-05, "loss": 0.0835, "step": 18720 }, { "epoch": 13.156008432888264, "grad_norm": 0.3172239065170288, "learning_rate": 2.4566408995080816e-05, "loss": 0.0753, "step": 18721 }, { "epoch": 13.15671117357695, "grad_norm": 0.6089237928390503, "learning_rate": 2.456594050128836e-05, "loss": 0.1208, "step": 18722 }, { "epoch": 13.157413914265636, "grad_norm": 0.5417016744613647, "learning_rate": 2.45654720074959e-05, "loss": 0.1399, "step": 18723 }, { "epoch": 13.158116654954322, "grad_norm": 1.5492459535598755, "learning_rate": 2.4565003513703444e-05, "loss": 0.1961, "step": 18724 }, { "epoch": 13.158819395643008, "grad_norm": 0.21952831745147705, "learning_rate": 2.4564535019910987e-05, "loss": 0.0503, "step": 18725 }, { "epoch": 13.159522136331693, "grad_norm": 0.28058478236198425, "learning_rate": 2.456406652611853e-05, "loss": 0.0375, "step": 18726 }, { "epoch": 13.16022487702038, "grad_norm": 0.16837242245674133, "learning_rate": 2.4563598032326075e-05, "loss": 0.035, "step": 18727 }, { "epoch": 13.160927617709065, "grad_norm": 0.1556439995765686, "learning_rate": 2.4563129538533615e-05, "loss": 0.0115, "step": 18728 }, { "epoch": 13.161630358397751, "grad_norm": 0.06466522812843323, "learning_rate": 2.456266104474116e-05, "loss": 0.0079, "step": 18729 }, { "epoch": 13.162333099086437, "grad_norm": 0.0764058455824852, "learning_rate": 2.4562192550948703e-05, "loss": 0.0106, "step": 18730 }, { "epoch": 13.163035839775123, "grad_norm": 0.07922681421041489, "learning_rate": 2.4561724057156243e-05, "loss": 0.0096, "step": 18731 }, { "epoch": 13.163738580463809, "grad_norm": 0.09082353115081787, "learning_rate": 2.4561255563363783e-05, "loss": 0.019, "step": 18732 }, { "epoch": 13.164441321152495, "grad_norm": 0.1695619523525238, "learning_rate": 2.4560787069571327e-05, "loss": 0.0106, "step": 18733 }, { "epoch": 13.16514406184118, "grad_norm": 0.07802603393793106, "learning_rate": 2.456031857577887e-05, "loss": 0.0092, "step": 18734 }, { "epoch": 13.165846802529867, "grad_norm": 0.1301029771566391, "learning_rate": 2.4559850081986415e-05, "loss": 0.0168, "step": 18735 }, { "epoch": 13.166549543218553, "grad_norm": 0.24384813010692596, "learning_rate": 2.4559381588193955e-05, "loss": 0.0113, "step": 18736 }, { "epoch": 13.167252283907239, "grad_norm": 0.09155186265707016, "learning_rate": 2.45589130944015e-05, "loss": 0.0171, "step": 18737 }, { "epoch": 13.167955024595924, "grad_norm": 0.09639396518468857, "learning_rate": 2.4558444600609042e-05, "loss": 0.0103, "step": 18738 }, { "epoch": 13.16865776528461, "grad_norm": 0.1098441630601883, "learning_rate": 2.4557976106816586e-05, "loss": 0.0213, "step": 18739 }, { "epoch": 13.169360505973296, "grad_norm": 0.18762144446372986, "learning_rate": 2.455750761302413e-05, "loss": 0.041, "step": 18740 }, { "epoch": 13.170063246661982, "grad_norm": 0.14766491949558258, "learning_rate": 2.455703911923167e-05, "loss": 0.0202, "step": 18741 }, { "epoch": 13.170765987350668, "grad_norm": 0.16379690170288086, "learning_rate": 2.4556570625439214e-05, "loss": 0.0258, "step": 18742 }, { "epoch": 13.171468728039354, "grad_norm": 0.5486988425254822, "learning_rate": 2.4556102131646758e-05, "loss": 0.0264, "step": 18743 }, { "epoch": 13.17217146872804, "grad_norm": 0.59217768907547, "learning_rate": 2.45556336378543e-05, "loss": 0.0724, "step": 18744 }, { "epoch": 13.172874209416726, "grad_norm": 0.21607759594917297, "learning_rate": 2.455516514406184e-05, "loss": 0.045, "step": 18745 }, { "epoch": 13.173576950105412, "grad_norm": 0.29896804690361023, "learning_rate": 2.4554696650269385e-05, "loss": 0.0777, "step": 18746 }, { "epoch": 13.174279690794098, "grad_norm": 0.8135021924972534, "learning_rate": 2.455422815647693e-05, "loss": 0.1601, "step": 18747 }, { "epoch": 13.174982431482784, "grad_norm": 0.7978629469871521, "learning_rate": 2.455375966268447e-05, "loss": 0.1838, "step": 18748 }, { "epoch": 13.17568517217147, "grad_norm": 0.7196577191352844, "learning_rate": 2.455329116889201e-05, "loss": 0.184, "step": 18749 }, { "epoch": 13.176387912860154, "grad_norm": 0.27029862999916077, "learning_rate": 2.4552822675099553e-05, "loss": 0.0728, "step": 18750 }, { "epoch": 13.17709065354884, "grad_norm": 0.22043494880199432, "learning_rate": 2.4552354181307097e-05, "loss": 0.0251, "step": 18751 }, { "epoch": 13.177793394237526, "grad_norm": 0.12064605951309204, "learning_rate": 2.455188568751464e-05, "loss": 0.0238, "step": 18752 }, { "epoch": 13.178496134926212, "grad_norm": 0.10551407188177109, "learning_rate": 2.4551417193722185e-05, "loss": 0.0184, "step": 18753 }, { "epoch": 13.179198875614897, "grad_norm": 0.17936058342456818, "learning_rate": 2.4550948699929725e-05, "loss": 0.0178, "step": 18754 }, { "epoch": 13.179901616303583, "grad_norm": 0.09172093868255615, "learning_rate": 2.455048020613727e-05, "loss": 0.0077, "step": 18755 }, { "epoch": 13.18060435699227, "grad_norm": 0.11333736777305603, "learning_rate": 2.4550011712344812e-05, "loss": 0.0121, "step": 18756 }, { "epoch": 13.181307097680955, "grad_norm": 0.076911561191082, "learning_rate": 2.4549543218552356e-05, "loss": 0.0146, "step": 18757 }, { "epoch": 13.182009838369641, "grad_norm": 0.13221606612205505, "learning_rate": 2.4549074724759896e-05, "loss": 0.0299, "step": 18758 }, { "epoch": 13.182712579058327, "grad_norm": 0.08574739098548889, "learning_rate": 2.454860623096744e-05, "loss": 0.0129, "step": 18759 }, { "epoch": 13.183415319747013, "grad_norm": 0.15897917747497559, "learning_rate": 2.4548137737174984e-05, "loss": 0.0243, "step": 18760 }, { "epoch": 13.184118060435699, "grad_norm": 0.13033424317836761, "learning_rate": 2.4547669243382528e-05, "loss": 0.0112, "step": 18761 }, { "epoch": 13.184820801124385, "grad_norm": 0.1268765926361084, "learning_rate": 2.4547200749590068e-05, "loss": 0.0219, "step": 18762 }, { "epoch": 13.18552354181307, "grad_norm": 0.19946549832820892, "learning_rate": 2.4546732255797612e-05, "loss": 0.0412, "step": 18763 }, { "epoch": 13.186226282501757, "grad_norm": 0.13258236646652222, "learning_rate": 2.4546263762005155e-05, "loss": 0.0231, "step": 18764 }, { "epoch": 13.186929023190443, "grad_norm": 0.21942688524723053, "learning_rate": 2.4545795268212696e-05, "loss": 0.0311, "step": 18765 }, { "epoch": 13.187631763879128, "grad_norm": 0.17103131115436554, "learning_rate": 2.454532677442024e-05, "loss": 0.0112, "step": 18766 }, { "epoch": 13.188334504567814, "grad_norm": 0.1597069501876831, "learning_rate": 2.454485828062778e-05, "loss": 0.0302, "step": 18767 }, { "epoch": 13.1890372452565, "grad_norm": 0.3912512958049774, "learning_rate": 2.4544389786835324e-05, "loss": 0.0448, "step": 18768 }, { "epoch": 13.189739985945186, "grad_norm": 0.3117997944355011, "learning_rate": 2.4543921293042867e-05, "loss": 0.0751, "step": 18769 }, { "epoch": 13.190442726633872, "grad_norm": 0.9642085433006287, "learning_rate": 2.454345279925041e-05, "loss": 0.0583, "step": 18770 }, { "epoch": 13.191145467322558, "grad_norm": 0.49815648794174194, "learning_rate": 2.454298430545795e-05, "loss": 0.0989, "step": 18771 }, { "epoch": 13.191848208011244, "grad_norm": 0.5209189057350159, "learning_rate": 2.4542515811665495e-05, "loss": 0.1362, "step": 18772 }, { "epoch": 13.19255094869993, "grad_norm": 0.9485623836517334, "learning_rate": 2.454204731787304e-05, "loss": 0.1497, "step": 18773 }, { "epoch": 13.193253689388616, "grad_norm": 0.9244195222854614, "learning_rate": 2.4541578824080583e-05, "loss": 0.1384, "step": 18774 }, { "epoch": 13.193956430077302, "grad_norm": 0.22504402697086334, "learning_rate": 2.4541110330288123e-05, "loss": 0.0436, "step": 18775 }, { "epoch": 13.194659170765988, "grad_norm": 0.0989455133676529, "learning_rate": 2.4540641836495667e-05, "loss": 0.0213, "step": 18776 }, { "epoch": 13.195361911454674, "grad_norm": 0.08034805208444595, "learning_rate": 2.454017334270321e-05, "loss": 0.0094, "step": 18777 }, { "epoch": 13.19606465214336, "grad_norm": 0.09534909576177597, "learning_rate": 2.4539704848910754e-05, "loss": 0.0129, "step": 18778 }, { "epoch": 13.196767392832045, "grad_norm": 0.11216063797473907, "learning_rate": 2.4539236355118298e-05, "loss": 0.0116, "step": 18779 }, { "epoch": 13.197470133520731, "grad_norm": 0.08909370750188828, "learning_rate": 2.4538767861325838e-05, "loss": 0.0126, "step": 18780 }, { "epoch": 13.198172874209417, "grad_norm": 0.13773417472839355, "learning_rate": 2.4538299367533382e-05, "loss": 0.0179, "step": 18781 }, { "epoch": 13.198875614898103, "grad_norm": 0.2282346487045288, "learning_rate": 2.4537830873740926e-05, "loss": 0.0304, "step": 18782 }, { "epoch": 13.19957835558679, "grad_norm": 0.21332865953445435, "learning_rate": 2.4537362379948466e-05, "loss": 0.0241, "step": 18783 }, { "epoch": 13.200281096275475, "grad_norm": 0.08212288469076157, "learning_rate": 2.4536893886156006e-05, "loss": 0.0091, "step": 18784 }, { "epoch": 13.200983836964161, "grad_norm": 0.13669128715991974, "learning_rate": 2.453642539236355e-05, "loss": 0.0197, "step": 18785 }, { "epoch": 13.201686577652847, "grad_norm": 0.08222533762454987, "learning_rate": 2.4535956898571094e-05, "loss": 0.0135, "step": 18786 }, { "epoch": 13.202389318341533, "grad_norm": 0.8082076907157898, "learning_rate": 2.4535488404778637e-05, "loss": 0.0481, "step": 18787 }, { "epoch": 13.203092059030217, "grad_norm": 0.11865933239459991, "learning_rate": 2.453501991098618e-05, "loss": 0.0147, "step": 18788 }, { "epoch": 13.203794799718903, "grad_norm": 0.19289807975292206, "learning_rate": 2.453455141719372e-05, "loss": 0.0237, "step": 18789 }, { "epoch": 13.204497540407589, "grad_norm": 0.33098304271698, "learning_rate": 2.4534082923401265e-05, "loss": 0.0262, "step": 18790 }, { "epoch": 13.205200281096275, "grad_norm": 0.079495370388031, "learning_rate": 2.453361442960881e-05, "loss": 0.0141, "step": 18791 }, { "epoch": 13.20590302178496, "grad_norm": 0.16403517127037048, "learning_rate": 2.4533145935816353e-05, "loss": 0.046, "step": 18792 }, { "epoch": 13.206605762473647, "grad_norm": 0.26895564794540405, "learning_rate": 2.4532677442023893e-05, "loss": 0.0258, "step": 18793 }, { "epoch": 13.207308503162333, "grad_norm": 0.21625486016273499, "learning_rate": 2.4532208948231437e-05, "loss": 0.0454, "step": 18794 }, { "epoch": 13.208011243851018, "grad_norm": 0.2349720001220703, "learning_rate": 2.453174045443898e-05, "loss": 0.0612, "step": 18795 }, { "epoch": 13.208713984539704, "grad_norm": 0.7412269115447998, "learning_rate": 2.4531271960646524e-05, "loss": 0.0879, "step": 18796 }, { "epoch": 13.20941672522839, "grad_norm": 0.46459463238716125, "learning_rate": 2.4530803466854064e-05, "loss": 0.1356, "step": 18797 }, { "epoch": 13.210119465917076, "grad_norm": 0.6336154937744141, "learning_rate": 2.4530334973061608e-05, "loss": 0.1395, "step": 18798 }, { "epoch": 13.210822206605762, "grad_norm": 1.1602318286895752, "learning_rate": 2.4529866479269152e-05, "loss": 0.1939, "step": 18799 }, { "epoch": 13.211524947294448, "grad_norm": 0.16553950309753418, "learning_rate": 2.4529397985476692e-05, "loss": 0.0639, "step": 18800 }, { "epoch": 13.212227687983134, "grad_norm": 0.10438697040081024, "learning_rate": 2.4528929491684236e-05, "loss": 0.0197, "step": 18801 }, { "epoch": 13.21293042867182, "grad_norm": 0.09448220580816269, "learning_rate": 2.4528460997891776e-05, "loss": 0.0201, "step": 18802 }, { "epoch": 13.213633169360506, "grad_norm": 0.10001206398010254, "learning_rate": 2.452799250409932e-05, "loss": 0.0171, "step": 18803 }, { "epoch": 13.214335910049192, "grad_norm": 0.05696793273091316, "learning_rate": 2.4527524010306864e-05, "loss": 0.0094, "step": 18804 }, { "epoch": 13.215038650737878, "grad_norm": 0.13626950979232788, "learning_rate": 2.4527055516514408e-05, "loss": 0.0187, "step": 18805 }, { "epoch": 13.215741391426564, "grad_norm": 0.08787155151367188, "learning_rate": 2.4526587022721948e-05, "loss": 0.0169, "step": 18806 }, { "epoch": 13.21644413211525, "grad_norm": 0.11566287279129028, "learning_rate": 2.452611852892949e-05, "loss": 0.0185, "step": 18807 }, { "epoch": 13.217146872803935, "grad_norm": 0.21737293899059296, "learning_rate": 2.4525650035137035e-05, "loss": 0.0191, "step": 18808 }, { "epoch": 13.217849613492621, "grad_norm": 0.08220367878675461, "learning_rate": 2.452518154134458e-05, "loss": 0.0117, "step": 18809 }, { "epoch": 13.218552354181307, "grad_norm": 0.15717586874961853, "learning_rate": 2.452471304755212e-05, "loss": 0.0308, "step": 18810 }, { "epoch": 13.219255094869993, "grad_norm": 0.07476706802845001, "learning_rate": 2.4524244553759663e-05, "loss": 0.0072, "step": 18811 }, { "epoch": 13.219957835558679, "grad_norm": 0.14736825227737427, "learning_rate": 2.4523776059967207e-05, "loss": 0.0213, "step": 18812 }, { "epoch": 13.220660576247365, "grad_norm": 0.1459694355726242, "learning_rate": 2.452330756617475e-05, "loss": 0.0224, "step": 18813 }, { "epoch": 13.221363316936051, "grad_norm": 0.1826898604631424, "learning_rate": 2.4522839072382294e-05, "loss": 0.0204, "step": 18814 }, { "epoch": 13.222066057624737, "grad_norm": 0.17489656805992126, "learning_rate": 2.4522370578589835e-05, "loss": 0.0335, "step": 18815 }, { "epoch": 13.222768798313423, "grad_norm": 0.09024020284414291, "learning_rate": 2.452190208479738e-05, "loss": 0.0105, "step": 18816 }, { "epoch": 13.223471539002109, "grad_norm": 0.1486952006816864, "learning_rate": 2.4521433591004922e-05, "loss": 0.0346, "step": 18817 }, { "epoch": 13.224174279690795, "grad_norm": 0.17663854360580444, "learning_rate": 2.4520965097212462e-05, "loss": 0.0318, "step": 18818 }, { "epoch": 13.22487702037948, "grad_norm": 0.30200013518333435, "learning_rate": 2.4520496603420003e-05, "loss": 0.0355, "step": 18819 }, { "epoch": 13.225579761068166, "grad_norm": 0.25592902302742004, "learning_rate": 2.4520028109627546e-05, "loss": 0.0833, "step": 18820 }, { "epoch": 13.226282501756852, "grad_norm": 0.218741312623024, "learning_rate": 2.451955961583509e-05, "loss": 0.0641, "step": 18821 }, { "epoch": 13.226985242445538, "grad_norm": 2.4587562084198, "learning_rate": 2.4519091122042634e-05, "loss": 0.119, "step": 18822 }, { "epoch": 13.227687983134224, "grad_norm": 0.5854198932647705, "learning_rate": 2.4518622628250174e-05, "loss": 0.1836, "step": 18823 }, { "epoch": 13.22839072382291, "grad_norm": 0.8028769493103027, "learning_rate": 2.4518154134457718e-05, "loss": 0.1565, "step": 18824 }, { "epoch": 13.229093464511596, "grad_norm": 0.4003991186618805, "learning_rate": 2.4517685640665262e-05, "loss": 0.0536, "step": 18825 }, { "epoch": 13.22979620520028, "grad_norm": 0.19219881296157837, "learning_rate": 2.4517217146872805e-05, "loss": 0.0173, "step": 18826 }, { "epoch": 13.230498945888966, "grad_norm": 0.11134620755910873, "learning_rate": 2.451674865308035e-05, "loss": 0.0205, "step": 18827 }, { "epoch": 13.231201686577652, "grad_norm": 0.08753951638936996, "learning_rate": 2.451628015928789e-05, "loss": 0.0127, "step": 18828 }, { "epoch": 13.231904427266338, "grad_norm": 0.10389197617769241, "learning_rate": 2.4515811665495433e-05, "loss": 0.0247, "step": 18829 }, { "epoch": 13.232607167955024, "grad_norm": 0.1000007838010788, "learning_rate": 2.4515343171702977e-05, "loss": 0.0134, "step": 18830 }, { "epoch": 13.23330990864371, "grad_norm": 0.08436203747987747, "learning_rate": 2.451487467791052e-05, "loss": 0.0128, "step": 18831 }, { "epoch": 13.234012649332396, "grad_norm": 0.11150457710027695, "learning_rate": 2.451440618411806e-05, "loss": 0.0299, "step": 18832 }, { "epoch": 13.234715390021082, "grad_norm": 0.09553986042737961, "learning_rate": 2.4513937690325605e-05, "loss": 0.02, "step": 18833 }, { "epoch": 13.235418130709768, "grad_norm": 0.17051714658737183, "learning_rate": 2.451346919653315e-05, "loss": 0.0239, "step": 18834 }, { "epoch": 13.236120871398454, "grad_norm": 0.11956920474767685, "learning_rate": 2.451300070274069e-05, "loss": 0.0242, "step": 18835 }, { "epoch": 13.23682361208714, "grad_norm": 0.1645825356245041, "learning_rate": 2.451253220894823e-05, "loss": 0.0167, "step": 18836 }, { "epoch": 13.237526352775825, "grad_norm": 0.27041366696357727, "learning_rate": 2.4512063715155773e-05, "loss": 0.031, "step": 18837 }, { "epoch": 13.238229093464511, "grad_norm": 0.1301763355731964, "learning_rate": 2.4511595221363317e-05, "loss": 0.0133, "step": 18838 }, { "epoch": 13.238931834153197, "grad_norm": 0.23222394287586212, "learning_rate": 2.451112672757086e-05, "loss": 0.0221, "step": 18839 }, { "epoch": 13.239634574841883, "grad_norm": 0.23082804679870605, "learning_rate": 2.4510658233778404e-05, "loss": 0.0288, "step": 18840 }, { "epoch": 13.240337315530569, "grad_norm": 0.08653126657009125, "learning_rate": 2.4510189739985944e-05, "loss": 0.0187, "step": 18841 }, { "epoch": 13.241040056219255, "grad_norm": 0.16455146670341492, "learning_rate": 2.4509721246193488e-05, "loss": 0.0428, "step": 18842 }, { "epoch": 13.24174279690794, "grad_norm": 0.1541377156972885, "learning_rate": 2.4509252752401032e-05, "loss": 0.0333, "step": 18843 }, { "epoch": 13.242445537596627, "grad_norm": 0.2974426746368408, "learning_rate": 2.4508784258608576e-05, "loss": 0.0497, "step": 18844 }, { "epoch": 13.243148278285313, "grad_norm": 0.9537526965141296, "learning_rate": 2.4508315764816116e-05, "loss": 0.0675, "step": 18845 }, { "epoch": 13.243851018973999, "grad_norm": 0.3337204158306122, "learning_rate": 2.450784727102366e-05, "loss": 0.0922, "step": 18846 }, { "epoch": 13.244553759662685, "grad_norm": 0.6056830286979675, "learning_rate": 2.4507378777231203e-05, "loss": 0.1026, "step": 18847 }, { "epoch": 13.24525650035137, "grad_norm": 1.03419828414917, "learning_rate": 2.4506910283438747e-05, "loss": 0.2039, "step": 18848 }, { "epoch": 13.245959241040056, "grad_norm": 0.5672460198402405, "learning_rate": 2.4506441789646287e-05, "loss": 0.136, "step": 18849 }, { "epoch": 13.246661981728742, "grad_norm": 0.2699672281742096, "learning_rate": 2.450597329585383e-05, "loss": 0.0682, "step": 18850 }, { "epoch": 13.247364722417428, "grad_norm": 0.126775860786438, "learning_rate": 2.4505504802061375e-05, "loss": 0.0144, "step": 18851 }, { "epoch": 13.248067463106114, "grad_norm": 0.13669852912425995, "learning_rate": 2.450503630826892e-05, "loss": 0.0216, "step": 18852 }, { "epoch": 13.2487702037948, "grad_norm": 0.08805428445339203, "learning_rate": 2.450456781447646e-05, "loss": 0.0127, "step": 18853 }, { "epoch": 13.249472944483486, "grad_norm": 0.19754591584205627, "learning_rate": 2.4504099320684e-05, "loss": 0.0088, "step": 18854 }, { "epoch": 13.250175685172172, "grad_norm": 0.7507100105285645, "learning_rate": 2.4503630826891543e-05, "loss": 0.0195, "step": 18855 }, { "epoch": 13.250878425860858, "grad_norm": 0.11015679687261581, "learning_rate": 2.4503162333099087e-05, "loss": 0.0131, "step": 18856 }, { "epoch": 13.251581166549544, "grad_norm": 0.06184772402048111, "learning_rate": 2.450269383930663e-05, "loss": 0.0073, "step": 18857 }, { "epoch": 13.25228390723823, "grad_norm": 0.18064053356647491, "learning_rate": 2.450222534551417e-05, "loss": 0.0181, "step": 18858 }, { "epoch": 13.252986647926916, "grad_norm": 0.11540228873491287, "learning_rate": 2.4501756851721714e-05, "loss": 0.011, "step": 18859 }, { "epoch": 13.253689388615602, "grad_norm": 0.07931725680828094, "learning_rate": 2.4501288357929258e-05, "loss": 0.0134, "step": 18860 }, { "epoch": 13.254392129304287, "grad_norm": 0.08991400897502899, "learning_rate": 2.4500819864136802e-05, "loss": 0.0138, "step": 18861 }, { "epoch": 13.255094869992973, "grad_norm": 0.13306835293769836, "learning_rate": 2.4500351370344342e-05, "loss": 0.0174, "step": 18862 }, { "epoch": 13.25579761068166, "grad_norm": 0.15717068314552307, "learning_rate": 2.4499882876551886e-05, "loss": 0.0121, "step": 18863 }, { "epoch": 13.256500351370345, "grad_norm": 0.24522192776203156, "learning_rate": 2.449941438275943e-05, "loss": 0.0401, "step": 18864 }, { "epoch": 13.25720309205903, "grad_norm": 0.16866259276866913, "learning_rate": 2.4498945888966973e-05, "loss": 0.0253, "step": 18865 }, { "epoch": 13.257905832747715, "grad_norm": 0.15531742572784424, "learning_rate": 2.4498477395174517e-05, "loss": 0.0136, "step": 18866 }, { "epoch": 13.258608573436401, "grad_norm": 0.18525750935077667, "learning_rate": 2.4498008901382057e-05, "loss": 0.0404, "step": 18867 }, { "epoch": 13.259311314125087, "grad_norm": 0.47813302278518677, "learning_rate": 2.44975404075896e-05, "loss": 0.0365, "step": 18868 }, { "epoch": 13.260014054813773, "grad_norm": 0.1065109372138977, "learning_rate": 2.4497071913797145e-05, "loss": 0.0212, "step": 18869 }, { "epoch": 13.260716795502459, "grad_norm": 0.25761276483535767, "learning_rate": 2.4496603420004685e-05, "loss": 0.0805, "step": 18870 }, { "epoch": 13.261419536191145, "grad_norm": 0.5631570219993591, "learning_rate": 2.4496134926212226e-05, "loss": 0.0901, "step": 18871 }, { "epoch": 13.26212227687983, "grad_norm": 0.6319758296012878, "learning_rate": 2.449566643241977e-05, "loss": 0.1311, "step": 18872 }, { "epoch": 13.262825017568517, "grad_norm": 0.7201279997825623, "learning_rate": 2.4495197938627313e-05, "loss": 0.1417, "step": 18873 }, { "epoch": 13.263527758257203, "grad_norm": 1.3092939853668213, "learning_rate": 2.4494729444834857e-05, "loss": 0.1747, "step": 18874 }, { "epoch": 13.264230498945889, "grad_norm": 0.21052716672420502, "learning_rate": 2.4494260951042397e-05, "loss": 0.0751, "step": 18875 }, { "epoch": 13.264933239634574, "grad_norm": 0.13317546248435974, "learning_rate": 2.449379245724994e-05, "loss": 0.0317, "step": 18876 }, { "epoch": 13.26563598032326, "grad_norm": 0.11745800822973251, "learning_rate": 2.4493323963457485e-05, "loss": 0.0278, "step": 18877 }, { "epoch": 13.266338721011946, "grad_norm": 0.12673015892505646, "learning_rate": 2.449285546966503e-05, "loss": 0.012, "step": 18878 }, { "epoch": 13.267041461700632, "grad_norm": 0.08340854942798615, "learning_rate": 2.4492386975872572e-05, "loss": 0.0116, "step": 18879 }, { "epoch": 13.267744202389318, "grad_norm": 0.11242488026618958, "learning_rate": 2.4491918482080112e-05, "loss": 0.0235, "step": 18880 }, { "epoch": 13.268446943078004, "grad_norm": 0.08186163753271103, "learning_rate": 2.4491449988287656e-05, "loss": 0.0131, "step": 18881 }, { "epoch": 13.26914968376669, "grad_norm": 0.08113797008991241, "learning_rate": 2.44909814944952e-05, "loss": 0.008, "step": 18882 }, { "epoch": 13.269852424455376, "grad_norm": 1.7439790964126587, "learning_rate": 2.4490513000702744e-05, "loss": 0.0359, "step": 18883 }, { "epoch": 13.270555165144062, "grad_norm": 0.11641140282154083, "learning_rate": 2.4490044506910284e-05, "loss": 0.0067, "step": 18884 }, { "epoch": 13.271257905832748, "grad_norm": 0.2782813608646393, "learning_rate": 2.4489576013117828e-05, "loss": 0.0215, "step": 18885 }, { "epoch": 13.271960646521434, "grad_norm": 0.2224089801311493, "learning_rate": 2.448910751932537e-05, "loss": 0.0104, "step": 18886 }, { "epoch": 13.27266338721012, "grad_norm": 0.10290495306253433, "learning_rate": 2.448863902553291e-05, "loss": 0.0151, "step": 18887 }, { "epoch": 13.273366127898806, "grad_norm": 0.2031429260969162, "learning_rate": 2.4488170531740452e-05, "loss": 0.0258, "step": 18888 }, { "epoch": 13.274068868587491, "grad_norm": 0.1180877760052681, "learning_rate": 2.4487702037947996e-05, "loss": 0.016, "step": 18889 }, { "epoch": 13.274771609276177, "grad_norm": 0.1312410682439804, "learning_rate": 2.448723354415554e-05, "loss": 0.0188, "step": 18890 }, { "epoch": 13.275474349964863, "grad_norm": 0.12789247930049896, "learning_rate": 2.4486765050363083e-05, "loss": 0.0184, "step": 18891 }, { "epoch": 13.27617709065355, "grad_norm": 0.16759583353996277, "learning_rate": 2.4486296556570627e-05, "loss": 0.0327, "step": 18892 }, { "epoch": 13.276879831342235, "grad_norm": 0.17542405426502228, "learning_rate": 2.4485828062778167e-05, "loss": 0.0286, "step": 18893 }, { "epoch": 13.277582572030921, "grad_norm": 0.23011896014213562, "learning_rate": 2.448535956898571e-05, "loss": 0.034, "step": 18894 }, { "epoch": 13.278285312719607, "grad_norm": 0.2870149612426758, "learning_rate": 2.4484891075193255e-05, "loss": 0.0702, "step": 18895 }, { "epoch": 13.278988053408293, "grad_norm": 0.2805873155593872, "learning_rate": 2.44844225814008e-05, "loss": 0.1007, "step": 18896 }, { "epoch": 13.279690794096979, "grad_norm": 1.8395943641662598, "learning_rate": 2.448395408760834e-05, "loss": 0.1373, "step": 18897 }, { "epoch": 13.280393534785665, "grad_norm": 0.824915885925293, "learning_rate": 2.4483485593815882e-05, "loss": 0.1483, "step": 18898 }, { "epoch": 13.28109627547435, "grad_norm": 1.3467283248901367, "learning_rate": 2.4483017100023426e-05, "loss": 0.1836, "step": 18899 }, { "epoch": 13.281799016163037, "grad_norm": 0.47086891531944275, "learning_rate": 2.448254860623097e-05, "loss": 0.0761, "step": 18900 }, { "epoch": 13.282501756851723, "grad_norm": 0.1551574021577835, "learning_rate": 2.448208011243851e-05, "loss": 0.0212, "step": 18901 }, { "epoch": 13.283204497540408, "grad_norm": 0.19553403556346893, "learning_rate": 2.4481611618646054e-05, "loss": 0.038, "step": 18902 }, { "epoch": 13.283907238229094, "grad_norm": 0.09889031201601028, "learning_rate": 2.4481143124853598e-05, "loss": 0.0157, "step": 18903 }, { "epoch": 13.284609978917779, "grad_norm": 0.06917320936918259, "learning_rate": 2.448067463106114e-05, "loss": 0.0092, "step": 18904 }, { "epoch": 13.285312719606464, "grad_norm": 0.10571927577257156, "learning_rate": 2.4480206137268682e-05, "loss": 0.0125, "step": 18905 }, { "epoch": 13.28601546029515, "grad_norm": 0.34320899844169617, "learning_rate": 2.4479737643476222e-05, "loss": 0.0157, "step": 18906 }, { "epoch": 13.286718200983836, "grad_norm": 0.10564323514699936, "learning_rate": 2.4479269149683766e-05, "loss": 0.026, "step": 18907 }, { "epoch": 13.287420941672522, "grad_norm": 0.24332286417484283, "learning_rate": 2.447880065589131e-05, "loss": 0.0197, "step": 18908 }, { "epoch": 13.288123682361208, "grad_norm": 0.09423398226499557, "learning_rate": 2.4478332162098853e-05, "loss": 0.011, "step": 18909 }, { "epoch": 13.288826423049894, "grad_norm": 0.26017114520072937, "learning_rate": 2.4477863668306394e-05, "loss": 0.0316, "step": 18910 }, { "epoch": 13.28952916373858, "grad_norm": 0.11259517073631287, "learning_rate": 2.4477395174513937e-05, "loss": 0.0094, "step": 18911 }, { "epoch": 13.290231904427266, "grad_norm": 0.2308652400970459, "learning_rate": 2.447692668072148e-05, "loss": 0.018, "step": 18912 }, { "epoch": 13.290934645115952, "grad_norm": 0.18561610579490662, "learning_rate": 2.4476458186929025e-05, "loss": 0.0119, "step": 18913 }, { "epoch": 13.291637385804638, "grad_norm": 0.2402099072933197, "learning_rate": 2.4475989693136565e-05, "loss": 0.0179, "step": 18914 }, { "epoch": 13.292340126493324, "grad_norm": 0.1421641707420349, "learning_rate": 2.447552119934411e-05, "loss": 0.0378, "step": 18915 }, { "epoch": 13.29304286718201, "grad_norm": 0.0997837483882904, "learning_rate": 2.4475052705551653e-05, "loss": 0.0138, "step": 18916 }, { "epoch": 13.293745607870695, "grad_norm": 0.35720276832580566, "learning_rate": 2.4474584211759196e-05, "loss": 0.0246, "step": 18917 }, { "epoch": 13.294448348559381, "grad_norm": 0.2814190089702606, "learning_rate": 2.447411571796674e-05, "loss": 0.0518, "step": 18918 }, { "epoch": 13.295151089248067, "grad_norm": 0.3459702432155609, "learning_rate": 2.447364722417428e-05, "loss": 0.0391, "step": 18919 }, { "epoch": 13.295853829936753, "grad_norm": 0.5461921691894531, "learning_rate": 2.4473178730381824e-05, "loss": 0.0798, "step": 18920 }, { "epoch": 13.29655657062544, "grad_norm": 0.7578876614570618, "learning_rate": 2.4472710236589368e-05, "loss": 0.0915, "step": 18921 }, { "epoch": 13.297259311314125, "grad_norm": 0.3987252414226532, "learning_rate": 2.4472241742796908e-05, "loss": 0.1029, "step": 18922 }, { "epoch": 13.297962052002811, "grad_norm": 0.5378657579421997, "learning_rate": 2.447177324900445e-05, "loss": 0.1616, "step": 18923 }, { "epoch": 13.298664792691497, "grad_norm": 1.0497452020645142, "learning_rate": 2.4471304755211992e-05, "loss": 0.1873, "step": 18924 }, { "epoch": 13.299367533380183, "grad_norm": 0.21666319668293, "learning_rate": 2.4470836261419536e-05, "loss": 0.0619, "step": 18925 }, { "epoch": 13.300070274068869, "grad_norm": 0.10376662015914917, "learning_rate": 2.447036776762708e-05, "loss": 0.0268, "step": 18926 }, { "epoch": 13.300773014757555, "grad_norm": 0.07788641005754471, "learning_rate": 2.446989927383462e-05, "loss": 0.0162, "step": 18927 }, { "epoch": 13.30147575544624, "grad_norm": 0.10205397754907608, "learning_rate": 2.4469430780042164e-05, "loss": 0.0179, "step": 18928 }, { "epoch": 13.302178496134927, "grad_norm": 0.11558033525943756, "learning_rate": 2.4468962286249707e-05, "loss": 0.0211, "step": 18929 }, { "epoch": 13.302881236823612, "grad_norm": 0.11273858696222305, "learning_rate": 2.446849379245725e-05, "loss": 0.0136, "step": 18930 }, { "epoch": 13.303583977512298, "grad_norm": 0.21734097599983215, "learning_rate": 2.4468025298664795e-05, "loss": 0.0164, "step": 18931 }, { "epoch": 13.304286718200984, "grad_norm": 0.08442564308643341, "learning_rate": 2.4467556804872335e-05, "loss": 0.0094, "step": 18932 }, { "epoch": 13.30498945888967, "grad_norm": 0.31537294387817383, "learning_rate": 2.446708831107988e-05, "loss": 0.019, "step": 18933 }, { "epoch": 13.305692199578356, "grad_norm": 0.23601622879505157, "learning_rate": 2.4466619817287423e-05, "loss": 0.0141, "step": 18934 }, { "epoch": 13.306394940267042, "grad_norm": 0.24030447006225586, "learning_rate": 2.4466151323494966e-05, "loss": 0.0228, "step": 18935 }, { "epoch": 13.307097680955728, "grad_norm": 0.08200250566005707, "learning_rate": 2.4465682829702507e-05, "loss": 0.0132, "step": 18936 }, { "epoch": 13.307800421644414, "grad_norm": 0.17786280810832977, "learning_rate": 2.446521433591005e-05, "loss": 0.0322, "step": 18937 }, { "epoch": 13.3085031623331, "grad_norm": 0.08432316035032272, "learning_rate": 2.4464745842117594e-05, "loss": 0.0175, "step": 18938 }, { "epoch": 13.309205903021786, "grad_norm": 0.25285959243774414, "learning_rate": 2.4464277348325138e-05, "loss": 0.0288, "step": 18939 }, { "epoch": 13.309908643710472, "grad_norm": 0.19805142283439636, "learning_rate": 2.4463808854532675e-05, "loss": 0.0284, "step": 18940 }, { "epoch": 13.310611384399156, "grad_norm": 0.12729008495807648, "learning_rate": 2.446334036074022e-05, "loss": 0.0264, "step": 18941 }, { "epoch": 13.311314125087842, "grad_norm": 0.1744767725467682, "learning_rate": 2.4462871866947762e-05, "loss": 0.042, "step": 18942 }, { "epoch": 13.312016865776528, "grad_norm": 0.2709301710128784, "learning_rate": 2.4462403373155306e-05, "loss": 0.0401, "step": 18943 }, { "epoch": 13.312719606465214, "grad_norm": 0.2290303111076355, "learning_rate": 2.446193487936285e-05, "loss": 0.0392, "step": 18944 }, { "epoch": 13.3134223471539, "grad_norm": 0.5445113778114319, "learning_rate": 2.446146638557039e-05, "loss": 0.0664, "step": 18945 }, { "epoch": 13.314125087842585, "grad_norm": 0.35765737295150757, "learning_rate": 2.4460997891777934e-05, "loss": 0.0781, "step": 18946 }, { "epoch": 13.314827828531271, "grad_norm": 0.7936371564865112, "learning_rate": 2.4460529397985478e-05, "loss": 0.1362, "step": 18947 }, { "epoch": 13.315530569219957, "grad_norm": 1.9262999296188354, "learning_rate": 2.446006090419302e-05, "loss": 0.1438, "step": 18948 }, { "epoch": 13.316233309908643, "grad_norm": 0.6807642579078674, "learning_rate": 2.445959241040056e-05, "loss": 0.1896, "step": 18949 }, { "epoch": 13.316936050597329, "grad_norm": 0.16038034856319427, "learning_rate": 2.4459123916608105e-05, "loss": 0.0552, "step": 18950 }, { "epoch": 13.317638791286015, "grad_norm": 0.2066752016544342, "learning_rate": 2.445865542281565e-05, "loss": 0.0205, "step": 18951 }, { "epoch": 13.318341531974701, "grad_norm": 0.12548533082008362, "learning_rate": 2.4458186929023193e-05, "loss": 0.0123, "step": 18952 }, { "epoch": 13.319044272663387, "grad_norm": 0.17417274415493011, "learning_rate": 2.4457718435230733e-05, "loss": 0.0147, "step": 18953 }, { "epoch": 13.319747013352073, "grad_norm": 0.05379454046487808, "learning_rate": 2.4457249941438277e-05, "loss": 0.0086, "step": 18954 }, { "epoch": 13.320449754040759, "grad_norm": 0.09144453704357147, "learning_rate": 2.445678144764582e-05, "loss": 0.0097, "step": 18955 }, { "epoch": 13.321152494729445, "grad_norm": 0.0799175277352333, "learning_rate": 2.4456312953853364e-05, "loss": 0.0114, "step": 18956 }, { "epoch": 13.32185523541813, "grad_norm": 0.10271790623664856, "learning_rate": 2.4455844460060905e-05, "loss": 0.0227, "step": 18957 }, { "epoch": 13.322557976106816, "grad_norm": 0.24925222992897034, "learning_rate": 2.4455375966268445e-05, "loss": 0.0173, "step": 18958 }, { "epoch": 13.323260716795502, "grad_norm": 0.10576953738927841, "learning_rate": 2.445490747247599e-05, "loss": 0.0113, "step": 18959 }, { "epoch": 13.323963457484188, "grad_norm": 0.17465414106845856, "learning_rate": 2.4454438978683532e-05, "loss": 0.0203, "step": 18960 }, { "epoch": 13.324666198172874, "grad_norm": 0.11846821010112762, "learning_rate": 2.4453970484891076e-05, "loss": 0.0107, "step": 18961 }, { "epoch": 13.32536893886156, "grad_norm": 0.12218713015317917, "learning_rate": 2.4453501991098617e-05, "loss": 0.0211, "step": 18962 }, { "epoch": 13.326071679550246, "grad_norm": 0.08682956546545029, "learning_rate": 2.445303349730616e-05, "loss": 0.011, "step": 18963 }, { "epoch": 13.326774420238932, "grad_norm": 0.15535421669483185, "learning_rate": 2.4452565003513704e-05, "loss": 0.0178, "step": 18964 }, { "epoch": 13.327477160927618, "grad_norm": 0.4607556164264679, "learning_rate": 2.4452096509721248e-05, "loss": 0.0289, "step": 18965 }, { "epoch": 13.328179901616304, "grad_norm": 0.11948300898075104, "learning_rate": 2.4451628015928788e-05, "loss": 0.0177, "step": 18966 }, { "epoch": 13.32888264230499, "grad_norm": 0.13770443201065063, "learning_rate": 2.4451159522136332e-05, "loss": 0.0255, "step": 18967 }, { "epoch": 13.329585382993676, "grad_norm": 0.15290319919586182, "learning_rate": 2.4450691028343875e-05, "loss": 0.0269, "step": 18968 }, { "epoch": 13.330288123682362, "grad_norm": 0.3175736367702484, "learning_rate": 2.445022253455142e-05, "loss": 0.0382, "step": 18969 }, { "epoch": 13.330990864371048, "grad_norm": 0.45638182759284973, "learning_rate": 2.4449754040758963e-05, "loss": 0.0634, "step": 18970 }, { "epoch": 13.331693605059733, "grad_norm": 0.6793054938316345, "learning_rate": 2.4449285546966503e-05, "loss": 0.1142, "step": 18971 }, { "epoch": 13.33239634574842, "grad_norm": 0.4043923318386078, "learning_rate": 2.4448817053174047e-05, "loss": 0.1125, "step": 18972 }, { "epoch": 13.333099086437105, "grad_norm": 0.49068140983581543, "learning_rate": 2.444834855938159e-05, "loss": 0.144, "step": 18973 }, { "epoch": 13.333801827125791, "grad_norm": 1.0630557537078857, "learning_rate": 2.444788006558913e-05, "loss": 0.2071, "step": 18974 }, { "epoch": 13.334504567814477, "grad_norm": 0.3609350919723511, "learning_rate": 2.444741157179667e-05, "loss": 0.0573, "step": 18975 }, { "epoch": 13.335207308503163, "grad_norm": 0.15399323403835297, "learning_rate": 2.4446943078004215e-05, "loss": 0.0351, "step": 18976 }, { "epoch": 13.335910049191849, "grad_norm": 0.16230061650276184, "learning_rate": 2.444647458421176e-05, "loss": 0.023, "step": 18977 }, { "epoch": 13.336612789880535, "grad_norm": 0.834996223449707, "learning_rate": 2.4446006090419303e-05, "loss": 0.0188, "step": 18978 }, { "epoch": 13.33731553056922, "grad_norm": 0.09891559928655624, "learning_rate": 2.4445537596626846e-05, "loss": 0.0154, "step": 18979 }, { "epoch": 13.338018271257905, "grad_norm": 0.12117062509059906, "learning_rate": 2.4445069102834387e-05, "loss": 0.0097, "step": 18980 }, { "epoch": 13.33872101194659, "grad_norm": 0.1750374585390091, "learning_rate": 2.444460060904193e-05, "loss": 0.0123, "step": 18981 }, { "epoch": 13.339423752635277, "grad_norm": 0.10451381653547287, "learning_rate": 2.4444132115249474e-05, "loss": 0.0154, "step": 18982 }, { "epoch": 13.340126493323963, "grad_norm": 0.06719846278429031, "learning_rate": 2.4443663621457018e-05, "loss": 0.0131, "step": 18983 }, { "epoch": 13.340829234012649, "grad_norm": 0.14236322045326233, "learning_rate": 2.4443195127664558e-05, "loss": 0.0316, "step": 18984 }, { "epoch": 13.341531974701335, "grad_norm": 0.3875453770160675, "learning_rate": 2.4442726633872102e-05, "loss": 0.0276, "step": 18985 }, { "epoch": 13.34223471539002, "grad_norm": 0.10218404233455658, "learning_rate": 2.4442258140079646e-05, "loss": 0.012, "step": 18986 }, { "epoch": 13.342937456078706, "grad_norm": 0.23909805715084076, "learning_rate": 2.444178964628719e-05, "loss": 0.0241, "step": 18987 }, { "epoch": 13.343640196767392, "grad_norm": 0.11572670936584473, "learning_rate": 2.444132115249473e-05, "loss": 0.0157, "step": 18988 }, { "epoch": 13.344342937456078, "grad_norm": 0.1059175357222557, "learning_rate": 2.4440852658702273e-05, "loss": 0.0234, "step": 18989 }, { "epoch": 13.345045678144764, "grad_norm": 0.2303805649280548, "learning_rate": 2.4440384164909817e-05, "loss": 0.0362, "step": 18990 }, { "epoch": 13.34574841883345, "grad_norm": 0.10302082449197769, "learning_rate": 2.443991567111736e-05, "loss": 0.018, "step": 18991 }, { "epoch": 13.346451159522136, "grad_norm": 0.13849781453609467, "learning_rate": 2.44394471773249e-05, "loss": 0.0264, "step": 18992 }, { "epoch": 13.347153900210822, "grad_norm": 0.5794925689697266, "learning_rate": 2.443897868353244e-05, "loss": 0.0518, "step": 18993 }, { "epoch": 13.347856640899508, "grad_norm": 0.22988243401050568, "learning_rate": 2.4438510189739985e-05, "loss": 0.0567, "step": 18994 }, { "epoch": 13.348559381588194, "grad_norm": 0.31135180592536926, "learning_rate": 2.443804169594753e-05, "loss": 0.0743, "step": 18995 }, { "epoch": 13.34926212227688, "grad_norm": 0.7889271974563599, "learning_rate": 2.4437573202155073e-05, "loss": 0.0977, "step": 18996 }, { "epoch": 13.349964862965566, "grad_norm": 0.6745681166648865, "learning_rate": 2.4437104708362613e-05, "loss": 0.1259, "step": 18997 }, { "epoch": 13.350667603654252, "grad_norm": 2.008913278579712, "learning_rate": 2.4436636214570157e-05, "loss": 0.1578, "step": 18998 }, { "epoch": 13.351370344342937, "grad_norm": 2.2302846908569336, "learning_rate": 2.44361677207777e-05, "loss": 0.1617, "step": 18999 }, { "epoch": 13.352073085031623, "grad_norm": 0.15433523058891296, "learning_rate": 2.4435699226985244e-05, "loss": 0.0564, "step": 19000 }, { "epoch": 13.352073085031623, "eval_cer": 0.19264508526725801, "eval_loss": 0.2612849175930023, "eval_runtime": 18.2617, "eval_samples_per_second": 248.498, "eval_steps_per_second": 0.821, "eval_wer": 0.3419457557051479, "step": 19000 }, { "epoch": 13.35277582572031, "grad_norm": 0.16721932590007782, "learning_rate": 2.4435230733192785e-05, "loss": 0.0146, "step": 19001 }, { "epoch": 13.353478566408995, "grad_norm": 0.13265801966190338, "learning_rate": 2.4434762239400328e-05, "loss": 0.0219, "step": 19002 }, { "epoch": 13.354181307097681, "grad_norm": 0.1651676595211029, "learning_rate": 2.4434293745607872e-05, "loss": 0.0251, "step": 19003 }, { "epoch": 13.354884047786367, "grad_norm": 0.09451150894165039, "learning_rate": 2.4433825251815416e-05, "loss": 0.0133, "step": 19004 }, { "epoch": 13.355586788475053, "grad_norm": 0.15651962161064148, "learning_rate": 2.443335675802296e-05, "loss": 0.0077, "step": 19005 }, { "epoch": 13.356289529163739, "grad_norm": 0.07012517750263214, "learning_rate": 2.44328882642305e-05, "loss": 0.01, "step": 19006 }, { "epoch": 13.356992269852425, "grad_norm": 0.26753777265548706, "learning_rate": 2.4432419770438044e-05, "loss": 0.0138, "step": 19007 }, { "epoch": 13.35769501054111, "grad_norm": 0.0832197442650795, "learning_rate": 2.4431951276645587e-05, "loss": 0.0113, "step": 19008 }, { "epoch": 13.358397751229797, "grad_norm": 0.09293626993894577, "learning_rate": 2.4431482782853128e-05, "loss": 0.0147, "step": 19009 }, { "epoch": 13.359100491918483, "grad_norm": 0.15674997866153717, "learning_rate": 2.4431014289060668e-05, "loss": 0.02, "step": 19010 }, { "epoch": 13.359803232607169, "grad_norm": 0.11500007659196854, "learning_rate": 2.443054579526821e-05, "loss": 0.0169, "step": 19011 }, { "epoch": 13.360505973295854, "grad_norm": 0.21855953335762024, "learning_rate": 2.4430077301475755e-05, "loss": 0.029, "step": 19012 }, { "epoch": 13.36120871398454, "grad_norm": 0.16770967841148376, "learning_rate": 2.44296088076833e-05, "loss": 0.0142, "step": 19013 }, { "epoch": 13.361911454673226, "grad_norm": 0.17056621611118317, "learning_rate": 2.442914031389084e-05, "loss": 0.0259, "step": 19014 }, { "epoch": 13.362614195361912, "grad_norm": 0.21689708530902863, "learning_rate": 2.4428671820098383e-05, "loss": 0.0379, "step": 19015 }, { "epoch": 13.363316936050598, "grad_norm": 0.11437561362981796, "learning_rate": 2.4428203326305927e-05, "loss": 0.016, "step": 19016 }, { "epoch": 13.364019676739284, "grad_norm": 0.36871689558029175, "learning_rate": 2.442773483251347e-05, "loss": 0.0306, "step": 19017 }, { "epoch": 13.36472241742797, "grad_norm": 0.14679576456546783, "learning_rate": 2.4427266338721014e-05, "loss": 0.0337, "step": 19018 }, { "epoch": 13.365425158116654, "grad_norm": 0.24934636056423187, "learning_rate": 2.4426797844928555e-05, "loss": 0.0414, "step": 19019 }, { "epoch": 13.36612789880534, "grad_norm": 0.3307707607746124, "learning_rate": 2.44263293511361e-05, "loss": 0.0738, "step": 19020 }, { "epoch": 13.366830639494026, "grad_norm": 0.46921804547309875, "learning_rate": 2.4425860857343642e-05, "loss": 0.0898, "step": 19021 }, { "epoch": 13.367533380182712, "grad_norm": 0.4816213846206665, "learning_rate": 2.4425392363551186e-05, "loss": 0.1179, "step": 19022 }, { "epoch": 13.368236120871398, "grad_norm": 0.7058526277542114, "learning_rate": 2.4424923869758726e-05, "loss": 0.116, "step": 19023 }, { "epoch": 13.368938861560084, "grad_norm": 1.050626516342163, "learning_rate": 2.442445537596627e-05, "loss": 0.1589, "step": 19024 }, { "epoch": 13.36964160224877, "grad_norm": 0.20578578114509583, "learning_rate": 2.4423986882173814e-05, "loss": 0.0663, "step": 19025 }, { "epoch": 13.370344342937456, "grad_norm": 0.09721942991018295, "learning_rate": 2.4423518388381357e-05, "loss": 0.025, "step": 19026 }, { "epoch": 13.371047083626141, "grad_norm": 0.19462573528289795, "learning_rate": 2.4423049894588894e-05, "loss": 0.032, "step": 19027 }, { "epoch": 13.371749824314827, "grad_norm": 0.22331570088863373, "learning_rate": 2.4422581400796438e-05, "loss": 0.0146, "step": 19028 }, { "epoch": 13.372452565003513, "grad_norm": 0.08633064478635788, "learning_rate": 2.4422112907003982e-05, "loss": 0.0161, "step": 19029 }, { "epoch": 13.3731553056922, "grad_norm": 0.13194118440151215, "learning_rate": 2.4421644413211525e-05, "loss": 0.0086, "step": 19030 }, { "epoch": 13.373858046380885, "grad_norm": 0.07419335097074509, "learning_rate": 2.442117591941907e-05, "loss": 0.0157, "step": 19031 }, { "epoch": 13.374560787069571, "grad_norm": 0.10214488953351974, "learning_rate": 2.442070742562661e-05, "loss": 0.0218, "step": 19032 }, { "epoch": 13.375263527758257, "grad_norm": 0.1637403815984726, "learning_rate": 2.4420238931834153e-05, "loss": 0.0178, "step": 19033 }, { "epoch": 13.375966268446943, "grad_norm": 0.10832717269659042, "learning_rate": 2.4419770438041697e-05, "loss": 0.0099, "step": 19034 }, { "epoch": 13.376669009135629, "grad_norm": 0.21038950979709625, "learning_rate": 2.441930194424924e-05, "loss": 0.0361, "step": 19035 }, { "epoch": 13.377371749824315, "grad_norm": 0.09176355600357056, "learning_rate": 2.441883345045678e-05, "loss": 0.0093, "step": 19036 }, { "epoch": 13.378074490513, "grad_norm": 0.10273667424917221, "learning_rate": 2.4418364956664325e-05, "loss": 0.0316, "step": 19037 }, { "epoch": 13.378777231201687, "grad_norm": 0.1384907364845276, "learning_rate": 2.441789646287187e-05, "loss": 0.0104, "step": 19038 }, { "epoch": 13.379479971890373, "grad_norm": 0.15352173149585724, "learning_rate": 2.4417427969079412e-05, "loss": 0.0237, "step": 19039 }, { "epoch": 13.380182712579058, "grad_norm": 0.18055641651153564, "learning_rate": 2.4416959475286953e-05, "loss": 0.0232, "step": 19040 }, { "epoch": 13.380885453267744, "grad_norm": 0.12378153949975967, "learning_rate": 2.4416490981494496e-05, "loss": 0.0238, "step": 19041 }, { "epoch": 13.38158819395643, "grad_norm": 0.3485654890537262, "learning_rate": 2.441602248770204e-05, "loss": 0.0453, "step": 19042 }, { "epoch": 13.382290934645116, "grad_norm": 0.2040375918149948, "learning_rate": 2.4415553993909584e-05, "loss": 0.0428, "step": 19043 }, { "epoch": 13.382993675333802, "grad_norm": 0.18189850449562073, "learning_rate": 2.4415085500117124e-05, "loss": 0.0735, "step": 19044 }, { "epoch": 13.383696416022488, "grad_norm": 0.24434976279735565, "learning_rate": 2.4414617006324664e-05, "loss": 0.0592, "step": 19045 }, { "epoch": 13.384399156711174, "grad_norm": 0.3642428517341614, "learning_rate": 2.4414148512532208e-05, "loss": 0.0958, "step": 19046 }, { "epoch": 13.38510189739986, "grad_norm": 0.4209063649177551, "learning_rate": 2.4413680018739752e-05, "loss": 0.1196, "step": 19047 }, { "epoch": 13.385804638088546, "grad_norm": 0.6984134316444397, "learning_rate": 2.4413211524947296e-05, "loss": 0.1433, "step": 19048 }, { "epoch": 13.386507378777232, "grad_norm": 0.8016669154167175, "learning_rate": 2.4412743031154836e-05, "loss": 0.1711, "step": 19049 }, { "epoch": 13.387210119465918, "grad_norm": 0.3495897948741913, "learning_rate": 2.441227453736238e-05, "loss": 0.0479, "step": 19050 }, { "epoch": 13.387912860154604, "grad_norm": 0.2599487602710724, "learning_rate": 2.4411806043569923e-05, "loss": 0.0276, "step": 19051 }, { "epoch": 13.38861560084329, "grad_norm": 0.1021048054099083, "learning_rate": 2.4411337549777467e-05, "loss": 0.022, "step": 19052 }, { "epoch": 13.389318341531975, "grad_norm": 0.15405921638011932, "learning_rate": 2.4410869055985007e-05, "loss": 0.0237, "step": 19053 }, { "epoch": 13.390021082220661, "grad_norm": 0.1733446568250656, "learning_rate": 2.441040056219255e-05, "loss": 0.0169, "step": 19054 }, { "epoch": 13.390723822909347, "grad_norm": 0.09067521244287491, "learning_rate": 2.4409932068400095e-05, "loss": 0.0122, "step": 19055 }, { "epoch": 13.391426563598033, "grad_norm": 0.09746088832616806, "learning_rate": 2.440946357460764e-05, "loss": 0.0169, "step": 19056 }, { "epoch": 13.392129304286719, "grad_norm": 0.12357941269874573, "learning_rate": 2.4408995080815182e-05, "loss": 0.016, "step": 19057 }, { "epoch": 13.392832044975403, "grad_norm": 0.058870360255241394, "learning_rate": 2.4408526587022723e-05, "loss": 0.0099, "step": 19058 }, { "epoch": 13.39353478566409, "grad_norm": 0.1758526712656021, "learning_rate": 2.4408058093230266e-05, "loss": 0.0264, "step": 19059 }, { "epoch": 13.394237526352775, "grad_norm": 0.08560138195753098, "learning_rate": 2.440758959943781e-05, "loss": 0.0151, "step": 19060 }, { "epoch": 13.394940267041461, "grad_norm": 0.08706293255090714, "learning_rate": 2.440712110564535e-05, "loss": 0.0083, "step": 19061 }, { "epoch": 13.395643007730147, "grad_norm": 0.13672472536563873, "learning_rate": 2.440665261185289e-05, "loss": 0.0304, "step": 19062 }, { "epoch": 13.396345748418833, "grad_norm": 0.6138675212860107, "learning_rate": 2.4406184118060435e-05, "loss": 0.0117, "step": 19063 }, { "epoch": 13.397048489107519, "grad_norm": 0.18218953907489777, "learning_rate": 2.4405715624267978e-05, "loss": 0.0327, "step": 19064 }, { "epoch": 13.397751229796205, "grad_norm": 0.13706359267234802, "learning_rate": 2.4405247130475522e-05, "loss": 0.0203, "step": 19065 }, { "epoch": 13.39845397048489, "grad_norm": 0.11005108058452606, "learning_rate": 2.4404778636683062e-05, "loss": 0.0091, "step": 19066 }, { "epoch": 13.399156711173577, "grad_norm": 0.24760518968105316, "learning_rate": 2.4404310142890606e-05, "loss": 0.0347, "step": 19067 }, { "epoch": 13.399859451862262, "grad_norm": 0.2983735203742981, "learning_rate": 2.440384164909815e-05, "loss": 0.0283, "step": 19068 }, { "epoch": 13.400562192550948, "grad_norm": 0.50648432970047, "learning_rate": 2.4403373155305693e-05, "loss": 0.0522, "step": 19069 }, { "epoch": 13.401264933239634, "grad_norm": 0.2539195120334625, "learning_rate": 2.4402904661513237e-05, "loss": 0.0762, "step": 19070 }, { "epoch": 13.40196767392832, "grad_norm": 0.3645954728126526, "learning_rate": 2.4402436167720778e-05, "loss": 0.0883, "step": 19071 }, { "epoch": 13.402670414617006, "grad_norm": 0.6020269989967346, "learning_rate": 2.440196767392832e-05, "loss": 0.1459, "step": 19072 }, { "epoch": 13.403373155305692, "grad_norm": 1.2928038835525513, "learning_rate": 2.4401499180135865e-05, "loss": 0.1734, "step": 19073 }, { "epoch": 13.404075895994378, "grad_norm": 1.119305968284607, "learning_rate": 2.440103068634341e-05, "loss": 0.1718, "step": 19074 }, { "epoch": 13.404778636683064, "grad_norm": 0.30037614703178406, "learning_rate": 2.440056219255095e-05, "loss": 0.0553, "step": 19075 }, { "epoch": 13.40548137737175, "grad_norm": 0.2733548879623413, "learning_rate": 2.4400093698758493e-05, "loss": 0.0237, "step": 19076 }, { "epoch": 13.406184118060436, "grad_norm": 0.3909972012042999, "learning_rate": 2.4399625204966037e-05, "loss": 0.0216, "step": 19077 }, { "epoch": 13.406886858749122, "grad_norm": 0.07562696188688278, "learning_rate": 2.439915671117358e-05, "loss": 0.0158, "step": 19078 }, { "epoch": 13.407589599437808, "grad_norm": 0.07465846836566925, "learning_rate": 2.4398688217381117e-05, "loss": 0.0145, "step": 19079 }, { "epoch": 13.408292340126494, "grad_norm": 0.1161666065454483, "learning_rate": 2.439821972358866e-05, "loss": 0.0075, "step": 19080 }, { "epoch": 13.40899508081518, "grad_norm": 0.08837452530860901, "learning_rate": 2.4397751229796205e-05, "loss": 0.0131, "step": 19081 }, { "epoch": 13.409697821503865, "grad_norm": 0.11323390901088715, "learning_rate": 2.439728273600375e-05, "loss": 0.0259, "step": 19082 }, { "epoch": 13.410400562192551, "grad_norm": 0.07739166915416718, "learning_rate": 2.4396814242211292e-05, "loss": 0.0136, "step": 19083 }, { "epoch": 13.411103302881237, "grad_norm": 0.11518898606300354, "learning_rate": 2.4396345748418832e-05, "loss": 0.0099, "step": 19084 }, { "epoch": 13.411806043569923, "grad_norm": 0.11756227165460587, "learning_rate": 2.4395877254626376e-05, "loss": 0.0261, "step": 19085 }, { "epoch": 13.412508784258609, "grad_norm": 0.09349671751260757, "learning_rate": 2.439540876083392e-05, "loss": 0.0114, "step": 19086 }, { "epoch": 13.413211524947295, "grad_norm": 0.14247173070907593, "learning_rate": 2.4394940267041464e-05, "loss": 0.0255, "step": 19087 }, { "epoch": 13.41391426563598, "grad_norm": 0.11430073529481888, "learning_rate": 2.4394471773249004e-05, "loss": 0.0105, "step": 19088 }, { "epoch": 13.414617006324667, "grad_norm": 0.1398518830537796, "learning_rate": 2.4394003279456548e-05, "loss": 0.0248, "step": 19089 }, { "epoch": 13.415319747013353, "grad_norm": 0.1579119712114334, "learning_rate": 2.439353478566409e-05, "loss": 0.0288, "step": 19090 }, { "epoch": 13.416022487702039, "grad_norm": 0.14299021661281586, "learning_rate": 2.4393066291871635e-05, "loss": 0.0226, "step": 19091 }, { "epoch": 13.416725228390725, "grad_norm": 0.17338816821575165, "learning_rate": 2.4392597798079175e-05, "loss": 0.0266, "step": 19092 }, { "epoch": 13.41742796907941, "grad_norm": 0.178477942943573, "learning_rate": 2.439212930428672e-05, "loss": 0.0472, "step": 19093 }, { "epoch": 13.418130709768096, "grad_norm": 0.4079062342643738, "learning_rate": 2.4391660810494263e-05, "loss": 0.0752, "step": 19094 }, { "epoch": 13.41883345045678, "grad_norm": 0.28379347920417786, "learning_rate": 2.4391192316701807e-05, "loss": 0.081, "step": 19095 }, { "epoch": 13.419536191145466, "grad_norm": 0.46296361088752747, "learning_rate": 2.4390723822909347e-05, "loss": 0.081, "step": 19096 }, { "epoch": 13.420238931834152, "grad_norm": 0.6120728850364685, "learning_rate": 2.4390255329116887e-05, "loss": 0.1585, "step": 19097 }, { "epoch": 13.420941672522838, "grad_norm": 0.5672067999839783, "learning_rate": 2.438978683532443e-05, "loss": 0.1457, "step": 19098 }, { "epoch": 13.421644413211524, "grad_norm": 1.1996147632598877, "learning_rate": 2.4389318341531975e-05, "loss": 0.1633, "step": 19099 }, { "epoch": 13.42234715390021, "grad_norm": 0.33719074726104736, "learning_rate": 2.438884984773952e-05, "loss": 0.059, "step": 19100 }, { "epoch": 13.423049894588896, "grad_norm": 0.3955881595611572, "learning_rate": 2.438838135394706e-05, "loss": 0.0352, "step": 19101 }, { "epoch": 13.423752635277582, "grad_norm": 0.12248215079307556, "learning_rate": 2.4387912860154603e-05, "loss": 0.0179, "step": 19102 }, { "epoch": 13.424455375966268, "grad_norm": 0.14231498539447784, "learning_rate": 2.4387444366362146e-05, "loss": 0.0188, "step": 19103 }, { "epoch": 13.425158116654954, "grad_norm": 0.064551942050457, "learning_rate": 2.438697587256969e-05, "loss": 0.0117, "step": 19104 }, { "epoch": 13.42586085734364, "grad_norm": 0.10791612416505814, "learning_rate": 2.438650737877723e-05, "loss": 0.013, "step": 19105 }, { "epoch": 13.426563598032326, "grad_norm": 0.17981067299842834, "learning_rate": 2.4386038884984774e-05, "loss": 0.015, "step": 19106 }, { "epoch": 13.427266338721012, "grad_norm": 0.08837442100048065, "learning_rate": 2.4385570391192318e-05, "loss": 0.0135, "step": 19107 }, { "epoch": 13.427969079409698, "grad_norm": 0.11045584827661514, "learning_rate": 2.438510189739986e-05, "loss": 0.0157, "step": 19108 }, { "epoch": 13.428671820098383, "grad_norm": 0.06640754640102386, "learning_rate": 2.4384633403607405e-05, "loss": 0.0087, "step": 19109 }, { "epoch": 13.42937456078707, "grad_norm": 0.22429461777210236, "learning_rate": 2.4384164909814946e-05, "loss": 0.0289, "step": 19110 }, { "epoch": 13.430077301475755, "grad_norm": 0.09914245456457138, "learning_rate": 2.438369641602249e-05, "loss": 0.012, "step": 19111 }, { "epoch": 13.430780042164441, "grad_norm": 0.3189348876476288, "learning_rate": 2.4383227922230033e-05, "loss": 0.0166, "step": 19112 }, { "epoch": 13.431482782853127, "grad_norm": 0.10200656205415726, "learning_rate": 2.4382759428437577e-05, "loss": 0.0159, "step": 19113 }, { "epoch": 13.432185523541813, "grad_norm": 0.09037864953279495, "learning_rate": 2.4382290934645114e-05, "loss": 0.0208, "step": 19114 }, { "epoch": 13.432888264230499, "grad_norm": 0.2571386992931366, "learning_rate": 2.4381822440852657e-05, "loss": 0.0319, "step": 19115 }, { "epoch": 13.433591004919185, "grad_norm": 0.3535785675048828, "learning_rate": 2.43813539470602e-05, "loss": 0.0112, "step": 19116 }, { "epoch": 13.43429374560787, "grad_norm": 0.14967180788516998, "learning_rate": 2.4380885453267745e-05, "loss": 0.0199, "step": 19117 }, { "epoch": 13.434996486296557, "grad_norm": 0.23169289529323578, "learning_rate": 2.4380416959475285e-05, "loss": 0.0455, "step": 19118 }, { "epoch": 13.435699226985243, "grad_norm": 0.3475465476512909, "learning_rate": 2.437994846568283e-05, "loss": 0.0484, "step": 19119 }, { "epoch": 13.436401967673929, "grad_norm": 0.22338946163654327, "learning_rate": 2.4379479971890373e-05, "loss": 0.0607, "step": 19120 }, { "epoch": 13.437104708362615, "grad_norm": 0.4704567790031433, "learning_rate": 2.4379011478097916e-05, "loss": 0.0952, "step": 19121 }, { "epoch": 13.4378074490513, "grad_norm": 0.9335498213768005, "learning_rate": 2.437854298430546e-05, "loss": 0.1196, "step": 19122 }, { "epoch": 13.438510189739986, "grad_norm": 0.3743354082107544, "learning_rate": 2.4378074490513e-05, "loss": 0.1424, "step": 19123 }, { "epoch": 13.439212930428672, "grad_norm": 0.5913438200950623, "learning_rate": 2.4377605996720544e-05, "loss": 0.1745, "step": 19124 }, { "epoch": 13.439915671117358, "grad_norm": 0.18635967373847961, "learning_rate": 2.4377137502928088e-05, "loss": 0.0664, "step": 19125 }, { "epoch": 13.440618411806044, "grad_norm": 0.17776010930538177, "learning_rate": 2.437666900913563e-05, "loss": 0.0187, "step": 19126 }, { "epoch": 13.44132115249473, "grad_norm": 0.20172475278377533, "learning_rate": 2.4376200515343172e-05, "loss": 0.0256, "step": 19127 }, { "epoch": 13.442023893183416, "grad_norm": 0.23047271370887756, "learning_rate": 2.4375732021550716e-05, "loss": 0.0221, "step": 19128 }, { "epoch": 13.442726633872102, "grad_norm": 0.09571246057748795, "learning_rate": 2.437526352775826e-05, "loss": 0.0117, "step": 19129 }, { "epoch": 13.443429374560788, "grad_norm": 0.06848306208848953, "learning_rate": 2.4374795033965803e-05, "loss": 0.0047, "step": 19130 }, { "epoch": 13.444132115249474, "grad_norm": 0.166648730635643, "learning_rate": 2.437432654017334e-05, "loss": 0.0177, "step": 19131 }, { "epoch": 13.44483485593816, "grad_norm": 0.13118009269237518, "learning_rate": 2.4373858046380884e-05, "loss": 0.0196, "step": 19132 }, { "epoch": 13.445537596626846, "grad_norm": 0.1673651933670044, "learning_rate": 2.4373389552588428e-05, "loss": 0.0374, "step": 19133 }, { "epoch": 13.44624033731553, "grad_norm": 0.10820555686950684, "learning_rate": 2.437292105879597e-05, "loss": 0.0201, "step": 19134 }, { "epoch": 13.446943078004216, "grad_norm": 0.17796802520751953, "learning_rate": 2.4372452565003515e-05, "loss": 0.0264, "step": 19135 }, { "epoch": 13.447645818692902, "grad_norm": 0.052648093551397324, "learning_rate": 2.4371984071211055e-05, "loss": 0.008, "step": 19136 }, { "epoch": 13.448348559381587, "grad_norm": 0.10527681559324265, "learning_rate": 2.43715155774186e-05, "loss": 0.0265, "step": 19137 }, { "epoch": 13.449051300070273, "grad_norm": 0.06737048923969269, "learning_rate": 2.4371047083626143e-05, "loss": 0.0108, "step": 19138 }, { "epoch": 13.44975404075896, "grad_norm": 0.09273774921894073, "learning_rate": 2.4370578589833686e-05, "loss": 0.0238, "step": 19139 }, { "epoch": 13.450456781447645, "grad_norm": 0.20114611089229584, "learning_rate": 2.4370110096041227e-05, "loss": 0.043, "step": 19140 }, { "epoch": 13.451159522136331, "grad_norm": 0.13828247785568237, "learning_rate": 2.436964160224877e-05, "loss": 0.0302, "step": 19141 }, { "epoch": 13.451862262825017, "grad_norm": 0.1656571477651596, "learning_rate": 2.4369173108456314e-05, "loss": 0.0276, "step": 19142 }, { "epoch": 13.452565003513703, "grad_norm": 0.12381863594055176, "learning_rate": 2.4368704614663858e-05, "loss": 0.0432, "step": 19143 }, { "epoch": 13.453267744202389, "grad_norm": 0.32944080233573914, "learning_rate": 2.43682361208714e-05, "loss": 0.0396, "step": 19144 }, { "epoch": 13.453970484891075, "grad_norm": 0.2725658118724823, "learning_rate": 2.4367767627078942e-05, "loss": 0.0576, "step": 19145 }, { "epoch": 13.45467322557976, "grad_norm": 0.3670022487640381, "learning_rate": 2.4367299133286486e-05, "loss": 0.0928, "step": 19146 }, { "epoch": 13.455375966268447, "grad_norm": 0.9928106069564819, "learning_rate": 2.436683063949403e-05, "loss": 0.1182, "step": 19147 }, { "epoch": 13.456078706957133, "grad_norm": 0.8643295764923096, "learning_rate": 2.4366362145701573e-05, "loss": 0.124, "step": 19148 }, { "epoch": 13.456781447645819, "grad_norm": 0.8749877214431763, "learning_rate": 2.436589365190911e-05, "loss": 0.1595, "step": 19149 }, { "epoch": 13.457484188334504, "grad_norm": 0.19342488050460815, "learning_rate": 2.4365425158116654e-05, "loss": 0.0711, "step": 19150 }, { "epoch": 13.45818692902319, "grad_norm": 0.1865629404783249, "learning_rate": 2.4364956664324198e-05, "loss": 0.0255, "step": 19151 }, { "epoch": 13.458889669711876, "grad_norm": 0.10844201594591141, "learning_rate": 2.436448817053174e-05, "loss": 0.0235, "step": 19152 }, { "epoch": 13.459592410400562, "grad_norm": 0.1305895745754242, "learning_rate": 2.436401967673928e-05, "loss": 0.0136, "step": 19153 }, { "epoch": 13.460295151089248, "grad_norm": 0.11554653942584991, "learning_rate": 2.4363551182946825e-05, "loss": 0.0128, "step": 19154 }, { "epoch": 13.460997891777934, "grad_norm": 0.12889108061790466, "learning_rate": 2.436308268915437e-05, "loss": 0.0203, "step": 19155 }, { "epoch": 13.46170063246662, "grad_norm": 0.15934650599956512, "learning_rate": 2.4362614195361913e-05, "loss": 0.0094, "step": 19156 }, { "epoch": 13.462403373155306, "grad_norm": 0.09009648859500885, "learning_rate": 2.4362145701569453e-05, "loss": 0.0108, "step": 19157 }, { "epoch": 13.463106113843992, "grad_norm": 0.203218474984169, "learning_rate": 2.4361677207776997e-05, "loss": 0.0263, "step": 19158 }, { "epoch": 13.463808854532678, "grad_norm": 0.20449061691761017, "learning_rate": 2.436120871398454e-05, "loss": 0.0219, "step": 19159 }, { "epoch": 13.464511595221364, "grad_norm": 0.14527209103107452, "learning_rate": 2.4360740220192084e-05, "loss": 0.0163, "step": 19160 }, { "epoch": 13.46521433591005, "grad_norm": 0.23830249905586243, "learning_rate": 2.4360271726399628e-05, "loss": 0.0181, "step": 19161 }, { "epoch": 13.465917076598735, "grad_norm": 0.17797288298606873, "learning_rate": 2.435980323260717e-05, "loss": 0.0224, "step": 19162 }, { "epoch": 13.466619817287421, "grad_norm": 0.09732317924499512, "learning_rate": 2.4359334738814712e-05, "loss": 0.0099, "step": 19163 }, { "epoch": 13.467322557976107, "grad_norm": 0.16379106044769287, "learning_rate": 2.4358866245022256e-05, "loss": 0.0328, "step": 19164 }, { "epoch": 13.468025298664793, "grad_norm": 0.15260431170463562, "learning_rate": 2.43583977512298e-05, "loss": 0.0184, "step": 19165 }, { "epoch": 13.46872803935348, "grad_norm": 0.19988854229450226, "learning_rate": 2.4357929257437337e-05, "loss": 0.0221, "step": 19166 }, { "epoch": 13.469430780042165, "grad_norm": 0.16094259917736053, "learning_rate": 2.435746076364488e-05, "loss": 0.0319, "step": 19167 }, { "epoch": 13.470133520730851, "grad_norm": 0.3552898168563843, "learning_rate": 2.4356992269852424e-05, "loss": 0.0306, "step": 19168 }, { "epoch": 13.470836261419537, "grad_norm": 0.18379715085029602, "learning_rate": 2.4356523776059968e-05, "loss": 0.0309, "step": 19169 }, { "epoch": 13.471539002108223, "grad_norm": 0.6353403329849243, "learning_rate": 2.435605528226751e-05, "loss": 0.0664, "step": 19170 }, { "epoch": 13.472241742796909, "grad_norm": 0.7286336421966553, "learning_rate": 2.4355586788475052e-05, "loss": 0.0918, "step": 19171 }, { "epoch": 13.472944483485595, "grad_norm": 0.7692383527755737, "learning_rate": 2.4355118294682596e-05, "loss": 0.1248, "step": 19172 }, { "epoch": 13.473647224174279, "grad_norm": 1.0558222532272339, "learning_rate": 2.435464980089014e-05, "loss": 0.1653, "step": 19173 }, { "epoch": 13.474349964862965, "grad_norm": 1.5727182626724243, "learning_rate": 2.4354181307097683e-05, "loss": 0.2339, "step": 19174 }, { "epoch": 13.47505270555165, "grad_norm": 0.4896696209907532, "learning_rate": 2.4353712813305223e-05, "loss": 0.0621, "step": 19175 }, { "epoch": 13.475755446240337, "grad_norm": 0.23895065486431122, "learning_rate": 2.4353244319512767e-05, "loss": 0.0198, "step": 19176 }, { "epoch": 13.476458186929023, "grad_norm": 0.2914959490299225, "learning_rate": 2.435277582572031e-05, "loss": 0.0108, "step": 19177 }, { "epoch": 13.477160927617708, "grad_norm": 0.08748028427362442, "learning_rate": 2.4352307331927854e-05, "loss": 0.0156, "step": 19178 }, { "epoch": 13.477863668306394, "grad_norm": 0.07484155148267746, "learning_rate": 2.4351838838135395e-05, "loss": 0.01, "step": 19179 }, { "epoch": 13.47856640899508, "grad_norm": 0.17039357125759125, "learning_rate": 2.435137034434294e-05, "loss": 0.0117, "step": 19180 }, { "epoch": 13.479269149683766, "grad_norm": 0.08567433059215546, "learning_rate": 2.4350901850550482e-05, "loss": 0.0165, "step": 19181 }, { "epoch": 13.479971890372452, "grad_norm": 0.11259230226278305, "learning_rate": 2.4350433356758026e-05, "loss": 0.0138, "step": 19182 }, { "epoch": 13.480674631061138, "grad_norm": 0.14334559440612793, "learning_rate": 2.4349964862965566e-05, "loss": 0.0209, "step": 19183 }, { "epoch": 13.481377371749824, "grad_norm": 0.22919712960720062, "learning_rate": 2.4349496369173107e-05, "loss": 0.007, "step": 19184 }, { "epoch": 13.48208011243851, "grad_norm": 0.2419782280921936, "learning_rate": 2.434902787538065e-05, "loss": 0.0225, "step": 19185 }, { "epoch": 13.482782853127196, "grad_norm": 0.08200158178806305, "learning_rate": 2.4348559381588194e-05, "loss": 0.0108, "step": 19186 }, { "epoch": 13.483485593815882, "grad_norm": 0.0932530090212822, "learning_rate": 2.4348090887795738e-05, "loss": 0.02, "step": 19187 }, { "epoch": 13.484188334504568, "grad_norm": 0.07244685292243958, "learning_rate": 2.4347622394003278e-05, "loss": 0.0088, "step": 19188 }, { "epoch": 13.484891075193254, "grad_norm": 0.10006754100322723, "learning_rate": 2.4347153900210822e-05, "loss": 0.0226, "step": 19189 }, { "epoch": 13.48559381588194, "grad_norm": 0.18898725509643555, "learning_rate": 2.4346685406418366e-05, "loss": 0.0442, "step": 19190 }, { "epoch": 13.486296556570625, "grad_norm": 0.1094774678349495, "learning_rate": 2.434621691262591e-05, "loss": 0.0158, "step": 19191 }, { "epoch": 13.486999297259311, "grad_norm": 0.21608249843120575, "learning_rate": 2.434574841883345e-05, "loss": 0.0341, "step": 19192 }, { "epoch": 13.487702037947997, "grad_norm": 0.16170185804367065, "learning_rate": 2.4345279925040993e-05, "loss": 0.0342, "step": 19193 }, { "epoch": 13.488404778636683, "grad_norm": 0.3102886974811554, "learning_rate": 2.4344811431248537e-05, "loss": 0.035, "step": 19194 }, { "epoch": 13.489107519325369, "grad_norm": 0.45230644941329956, "learning_rate": 2.434434293745608e-05, "loss": 0.059, "step": 19195 }, { "epoch": 13.489810260014055, "grad_norm": 0.24613057076931, "learning_rate": 2.4343874443663625e-05, "loss": 0.0723, "step": 19196 }, { "epoch": 13.490513000702741, "grad_norm": 0.5412254929542542, "learning_rate": 2.4343405949871165e-05, "loss": 0.1254, "step": 19197 }, { "epoch": 13.491215741391427, "grad_norm": 2.6427254676818848, "learning_rate": 2.434293745607871e-05, "loss": 0.1555, "step": 19198 }, { "epoch": 13.491918482080113, "grad_norm": 1.2614763975143433, "learning_rate": 2.4342468962286252e-05, "loss": 0.1873, "step": 19199 }, { "epoch": 13.492621222768799, "grad_norm": 0.3321821987628937, "learning_rate": 2.4342000468493796e-05, "loss": 0.0551, "step": 19200 }, { "epoch": 13.493323963457485, "grad_norm": 0.2845546305179596, "learning_rate": 2.4341531974701333e-05, "loss": 0.0159, "step": 19201 }, { "epoch": 13.49402670414617, "grad_norm": 0.0770506039261818, "learning_rate": 2.4341063480908877e-05, "loss": 0.0162, "step": 19202 }, { "epoch": 13.494729444834856, "grad_norm": 0.16884683072566986, "learning_rate": 2.434059498711642e-05, "loss": 0.0179, "step": 19203 }, { "epoch": 13.495432185523542, "grad_norm": 0.11588151752948761, "learning_rate": 2.4340126493323964e-05, "loss": 0.0153, "step": 19204 }, { "epoch": 13.496134926212228, "grad_norm": 0.0963737815618515, "learning_rate": 2.4339657999531505e-05, "loss": 0.0137, "step": 19205 }, { "epoch": 13.496837666900914, "grad_norm": 0.2270594984292984, "learning_rate": 2.4339189505739048e-05, "loss": 0.0195, "step": 19206 }, { "epoch": 13.4975404075896, "grad_norm": 0.31642112135887146, "learning_rate": 2.4338721011946592e-05, "loss": 0.029, "step": 19207 }, { "epoch": 13.498243148278286, "grad_norm": 0.07503878325223923, "learning_rate": 2.4338252518154136e-05, "loss": 0.0095, "step": 19208 }, { "epoch": 13.498945888966972, "grad_norm": 0.15613660216331482, "learning_rate": 2.433778402436168e-05, "loss": 0.0113, "step": 19209 }, { "epoch": 13.499648629655656, "grad_norm": 0.12333603948354721, "learning_rate": 2.433731553056922e-05, "loss": 0.0195, "step": 19210 }, { "epoch": 13.500351370344344, "grad_norm": 0.5132638812065125, "learning_rate": 2.4336847036776764e-05, "loss": 0.0123, "step": 19211 }, { "epoch": 13.501054111033028, "grad_norm": 0.12353795021772385, "learning_rate": 2.4336378542984307e-05, "loss": 0.0223, "step": 19212 }, { "epoch": 13.501756851721714, "grad_norm": 1.1644748449325562, "learning_rate": 2.433591004919185e-05, "loss": 0.0223, "step": 19213 }, { "epoch": 13.5024595924104, "grad_norm": 0.15269175171852112, "learning_rate": 2.433544155539939e-05, "loss": 0.0215, "step": 19214 }, { "epoch": 13.503162333099086, "grad_norm": 0.21151034533977509, "learning_rate": 2.4334973061606935e-05, "loss": 0.0324, "step": 19215 }, { "epoch": 13.503865073787772, "grad_norm": 0.2475365400314331, "learning_rate": 2.433450456781448e-05, "loss": 0.0281, "step": 19216 }, { "epoch": 13.504567814476458, "grad_norm": 0.2009790688753128, "learning_rate": 2.4334036074022023e-05, "loss": 0.0342, "step": 19217 }, { "epoch": 13.505270555165144, "grad_norm": 0.28845372796058655, "learning_rate": 2.433356758022956e-05, "loss": 0.0479, "step": 19218 }, { "epoch": 13.50597329585383, "grad_norm": 0.1520759016275406, "learning_rate": 2.4333099086437103e-05, "loss": 0.0424, "step": 19219 }, { "epoch": 13.506676036542515, "grad_norm": 0.43758726119995117, "learning_rate": 2.4332630592644647e-05, "loss": 0.0724, "step": 19220 }, { "epoch": 13.507378777231201, "grad_norm": 0.4454895257949829, "learning_rate": 2.433216209885219e-05, "loss": 0.0892, "step": 19221 }, { "epoch": 13.508081517919887, "grad_norm": 0.4764340817928314, "learning_rate": 2.4331693605059734e-05, "loss": 0.1401, "step": 19222 }, { "epoch": 13.508784258608573, "grad_norm": 1.217194676399231, "learning_rate": 2.4331225111267275e-05, "loss": 0.1608, "step": 19223 }, { "epoch": 13.509486999297259, "grad_norm": 2.885284900665283, "learning_rate": 2.433075661747482e-05, "loss": 0.2202, "step": 19224 }, { "epoch": 13.510189739985945, "grad_norm": 0.6127883195877075, "learning_rate": 2.4330288123682362e-05, "loss": 0.0695, "step": 19225 }, { "epoch": 13.510892480674631, "grad_norm": 0.1418105959892273, "learning_rate": 2.4329819629889906e-05, "loss": 0.0291, "step": 19226 }, { "epoch": 13.511595221363317, "grad_norm": 0.23350772261619568, "learning_rate": 2.4329351136097446e-05, "loss": 0.0342, "step": 19227 }, { "epoch": 13.512297962052003, "grad_norm": 0.07836537808179855, "learning_rate": 2.432888264230499e-05, "loss": 0.0134, "step": 19228 }, { "epoch": 13.513000702740689, "grad_norm": 0.16482917964458466, "learning_rate": 2.4328414148512534e-05, "loss": 0.0136, "step": 19229 }, { "epoch": 13.513703443429375, "grad_norm": 0.10289032012224197, "learning_rate": 2.4327945654720077e-05, "loss": 0.0112, "step": 19230 }, { "epoch": 13.51440618411806, "grad_norm": 0.1371484249830246, "learning_rate": 2.4327477160927618e-05, "loss": 0.0185, "step": 19231 }, { "epoch": 13.515108924806746, "grad_norm": 0.09751714020967484, "learning_rate": 2.432700866713516e-05, "loss": 0.0108, "step": 19232 }, { "epoch": 13.515811665495432, "grad_norm": 0.1479923278093338, "learning_rate": 2.4326540173342705e-05, "loss": 0.0317, "step": 19233 }, { "epoch": 13.516514406184118, "grad_norm": 0.10803069174289703, "learning_rate": 2.432607167955025e-05, "loss": 0.0119, "step": 19234 }, { "epoch": 13.517217146872804, "grad_norm": 0.13507434725761414, "learning_rate": 2.4325603185757793e-05, "loss": 0.0187, "step": 19235 }, { "epoch": 13.51791988756149, "grad_norm": 0.07178865373134613, "learning_rate": 2.432513469196533e-05, "loss": 0.0137, "step": 19236 }, { "epoch": 13.518622628250176, "grad_norm": 0.15770235657691956, "learning_rate": 2.4324666198172873e-05, "loss": 0.0275, "step": 19237 }, { "epoch": 13.519325368938862, "grad_norm": 0.1895618736743927, "learning_rate": 2.4324197704380417e-05, "loss": 0.0158, "step": 19238 }, { "epoch": 13.520028109627548, "grad_norm": 0.2644195258617401, "learning_rate": 2.432372921058796e-05, "loss": 0.0268, "step": 19239 }, { "epoch": 13.520730850316234, "grad_norm": 0.1619185507297516, "learning_rate": 2.43232607167955e-05, "loss": 0.0247, "step": 19240 }, { "epoch": 13.52143359100492, "grad_norm": 0.1835441291332245, "learning_rate": 2.4322792223003045e-05, "loss": 0.0198, "step": 19241 }, { "epoch": 13.522136331693606, "grad_norm": 0.21862205862998962, "learning_rate": 2.432232372921059e-05, "loss": 0.04, "step": 19242 }, { "epoch": 13.522839072382292, "grad_norm": 0.18301010131835938, "learning_rate": 2.4321855235418132e-05, "loss": 0.0419, "step": 19243 }, { "epoch": 13.523541813070977, "grad_norm": 0.24871669709682465, "learning_rate": 2.4321386741625673e-05, "loss": 0.0389, "step": 19244 }, { "epoch": 13.524244553759663, "grad_norm": 0.2489359974861145, "learning_rate": 2.4320918247833216e-05, "loss": 0.0709, "step": 19245 }, { "epoch": 13.52494729444835, "grad_norm": 0.40044498443603516, "learning_rate": 2.432044975404076e-05, "loss": 0.1048, "step": 19246 }, { "epoch": 13.525650035137035, "grad_norm": 0.5876004099845886, "learning_rate": 2.4319981260248304e-05, "loss": 0.1559, "step": 19247 }, { "epoch": 13.526352775825721, "grad_norm": 0.981228232383728, "learning_rate": 2.4319512766455847e-05, "loss": 0.1406, "step": 19248 }, { "epoch": 13.527055516514405, "grad_norm": 0.6692476272583008, "learning_rate": 2.4319044272663388e-05, "loss": 0.1644, "step": 19249 }, { "epoch": 13.527758257203093, "grad_norm": 0.19740413129329681, "learning_rate": 2.431857577887093e-05, "loss": 0.0595, "step": 19250 }, { "epoch": 13.528460997891777, "grad_norm": 0.13851678371429443, "learning_rate": 2.4318107285078475e-05, "loss": 0.0244, "step": 19251 }, { "epoch": 13.529163738580463, "grad_norm": 0.13230501115322113, "learning_rate": 2.431763879128602e-05, "loss": 0.0182, "step": 19252 }, { "epoch": 13.529866479269149, "grad_norm": 0.08542090654373169, "learning_rate": 2.4317170297493556e-05, "loss": 0.013, "step": 19253 }, { "epoch": 13.530569219957835, "grad_norm": 0.14481008052825928, "learning_rate": 2.43167018037011e-05, "loss": 0.02, "step": 19254 }, { "epoch": 13.53127196064652, "grad_norm": 0.18322604894638062, "learning_rate": 2.4316233309908643e-05, "loss": 0.0178, "step": 19255 }, { "epoch": 13.531974701335207, "grad_norm": 0.11717312783002853, "learning_rate": 2.4315764816116187e-05, "loss": 0.0129, "step": 19256 }, { "epoch": 13.532677442023893, "grad_norm": 0.11564823985099792, "learning_rate": 2.4315296322323727e-05, "loss": 0.0196, "step": 19257 }, { "epoch": 13.533380182712579, "grad_norm": 0.11852620542049408, "learning_rate": 2.431482782853127e-05, "loss": 0.015, "step": 19258 }, { "epoch": 13.534082923401265, "grad_norm": 0.1404130458831787, "learning_rate": 2.4314359334738815e-05, "loss": 0.0223, "step": 19259 }, { "epoch": 13.53478566408995, "grad_norm": 0.14256621897220612, "learning_rate": 2.431389084094636e-05, "loss": 0.0196, "step": 19260 }, { "epoch": 13.535488404778636, "grad_norm": 0.06690382212400436, "learning_rate": 2.4313422347153902e-05, "loss": 0.0111, "step": 19261 }, { "epoch": 13.536191145467322, "grad_norm": 0.12568959593772888, "learning_rate": 2.4312953853361443e-05, "loss": 0.0267, "step": 19262 }, { "epoch": 13.536893886156008, "grad_norm": 0.08922179788351059, "learning_rate": 2.4312485359568986e-05, "loss": 0.0117, "step": 19263 }, { "epoch": 13.537596626844694, "grad_norm": 0.10477198660373688, "learning_rate": 2.431201686577653e-05, "loss": 0.0162, "step": 19264 }, { "epoch": 13.53829936753338, "grad_norm": 0.17438510060310364, "learning_rate": 2.4311548371984074e-05, "loss": 0.0358, "step": 19265 }, { "epoch": 13.539002108222066, "grad_norm": 0.21355001628398895, "learning_rate": 2.4311079878191614e-05, "loss": 0.0368, "step": 19266 }, { "epoch": 13.539704848910752, "grad_norm": 0.18565720319747925, "learning_rate": 2.4310611384399158e-05, "loss": 0.0234, "step": 19267 }, { "epoch": 13.540407589599438, "grad_norm": 0.13884519040584564, "learning_rate": 2.43101428906067e-05, "loss": 0.033, "step": 19268 }, { "epoch": 13.541110330288124, "grad_norm": 0.3115456998348236, "learning_rate": 2.4309674396814245e-05, "loss": 0.0299, "step": 19269 }, { "epoch": 13.54181307097681, "grad_norm": 0.24513383209705353, "learning_rate": 2.4309205903021782e-05, "loss": 0.0511, "step": 19270 }, { "epoch": 13.542515811665496, "grad_norm": 0.2628340423107147, "learning_rate": 2.4308737409229326e-05, "loss": 0.0732, "step": 19271 }, { "epoch": 13.543218552354181, "grad_norm": 0.4899057149887085, "learning_rate": 2.430826891543687e-05, "loss": 0.1437, "step": 19272 }, { "epoch": 13.543921293042867, "grad_norm": 1.3731971979141235, "learning_rate": 2.4307800421644414e-05, "loss": 0.1909, "step": 19273 }, { "epoch": 13.544624033731553, "grad_norm": 0.6996152997016907, "learning_rate": 2.4307331927851957e-05, "loss": 0.1739, "step": 19274 }, { "epoch": 13.54532677442024, "grad_norm": 0.3509800136089325, "learning_rate": 2.4306863434059498e-05, "loss": 0.0661, "step": 19275 }, { "epoch": 13.546029515108925, "grad_norm": 0.10100075602531433, "learning_rate": 2.430639494026704e-05, "loss": 0.0187, "step": 19276 }, { "epoch": 13.546732255797611, "grad_norm": 0.27198466658592224, "learning_rate": 2.4305926446474585e-05, "loss": 0.0264, "step": 19277 }, { "epoch": 13.547434996486297, "grad_norm": 0.12044283002614975, "learning_rate": 2.430545795268213e-05, "loss": 0.0249, "step": 19278 }, { "epoch": 13.548137737174983, "grad_norm": 0.07715589553117752, "learning_rate": 2.430498945888967e-05, "loss": 0.0143, "step": 19279 }, { "epoch": 13.548840477863669, "grad_norm": 0.10823264718055725, "learning_rate": 2.4304520965097213e-05, "loss": 0.0133, "step": 19280 }, { "epoch": 13.549543218552355, "grad_norm": 0.07452595233917236, "learning_rate": 2.4304052471304757e-05, "loss": 0.0103, "step": 19281 }, { "epoch": 13.55024595924104, "grad_norm": 0.15758396685123444, "learning_rate": 2.43035839775123e-05, "loss": 0.0108, "step": 19282 }, { "epoch": 13.550948699929727, "grad_norm": 0.17498645186424255, "learning_rate": 2.430311548371984e-05, "loss": 0.0269, "step": 19283 }, { "epoch": 13.551651440618413, "grad_norm": 0.12507744133472443, "learning_rate": 2.4302646989927384e-05, "loss": 0.0111, "step": 19284 }, { "epoch": 13.552354181307098, "grad_norm": 0.08574787527322769, "learning_rate": 2.4302178496134928e-05, "loss": 0.021, "step": 19285 }, { "epoch": 13.553056921995784, "grad_norm": 0.09350284934043884, "learning_rate": 2.4301710002342472e-05, "loss": 0.0162, "step": 19286 }, { "epoch": 13.55375966268447, "grad_norm": 0.09410642087459564, "learning_rate": 2.4301241508550016e-05, "loss": 0.0232, "step": 19287 }, { "epoch": 13.554462403373154, "grad_norm": 0.17131280899047852, "learning_rate": 2.4300773014757552e-05, "loss": 0.0209, "step": 19288 }, { "epoch": 13.55516514406184, "grad_norm": 0.17683668434619904, "learning_rate": 2.4300304520965096e-05, "loss": 0.0336, "step": 19289 }, { "epoch": 13.555867884750526, "grad_norm": 0.3047155737876892, "learning_rate": 2.429983602717264e-05, "loss": 0.0458, "step": 19290 }, { "epoch": 13.556570625439212, "grad_norm": 0.0842922180891037, "learning_rate": 2.4299367533380184e-05, "loss": 0.0149, "step": 19291 }, { "epoch": 13.557273366127898, "grad_norm": 0.16946105659008026, "learning_rate": 2.4298899039587724e-05, "loss": 0.0311, "step": 19292 }, { "epoch": 13.557976106816584, "grad_norm": 0.18195027112960815, "learning_rate": 2.4298430545795268e-05, "loss": 0.0263, "step": 19293 }, { "epoch": 13.55867884750527, "grad_norm": 0.2190997153520584, "learning_rate": 2.429796205200281e-05, "loss": 0.0401, "step": 19294 }, { "epoch": 13.559381588193956, "grad_norm": 0.26092013716697693, "learning_rate": 2.4297493558210355e-05, "loss": 0.057, "step": 19295 }, { "epoch": 13.560084328882642, "grad_norm": 0.3373613655567169, "learning_rate": 2.4297025064417895e-05, "loss": 0.0721, "step": 19296 }, { "epoch": 13.560787069571328, "grad_norm": 0.4074828028678894, "learning_rate": 2.429655657062544e-05, "loss": 0.1186, "step": 19297 }, { "epoch": 13.561489810260014, "grad_norm": 0.48536691069602966, "learning_rate": 2.4296088076832983e-05, "loss": 0.1315, "step": 19298 }, { "epoch": 13.5621925509487, "grad_norm": 0.7766097784042358, "learning_rate": 2.4295619583040527e-05, "loss": 0.1739, "step": 19299 }, { "epoch": 13.562895291637385, "grad_norm": 0.1725676953792572, "learning_rate": 2.429515108924807e-05, "loss": 0.0705, "step": 19300 }, { "epoch": 13.563598032326071, "grad_norm": 0.14400900900363922, "learning_rate": 2.429468259545561e-05, "loss": 0.0295, "step": 19301 }, { "epoch": 13.564300773014757, "grad_norm": 0.07450171560049057, "learning_rate": 2.4294214101663154e-05, "loss": 0.0121, "step": 19302 }, { "epoch": 13.565003513703443, "grad_norm": 0.16190269589424133, "learning_rate": 2.4293745607870698e-05, "loss": 0.0199, "step": 19303 }, { "epoch": 13.56570625439213, "grad_norm": 0.1085495576262474, "learning_rate": 2.4293277114078242e-05, "loss": 0.0198, "step": 19304 }, { "epoch": 13.566408995080815, "grad_norm": 0.07833494991064072, "learning_rate": 2.429280862028578e-05, "loss": 0.0121, "step": 19305 }, { "epoch": 13.567111735769501, "grad_norm": 0.07114840298891068, "learning_rate": 2.4292340126493323e-05, "loss": 0.0089, "step": 19306 }, { "epoch": 13.567814476458187, "grad_norm": 0.11462618410587311, "learning_rate": 2.4291871632700866e-05, "loss": 0.0224, "step": 19307 }, { "epoch": 13.568517217146873, "grad_norm": 0.12881401181221008, "learning_rate": 2.429140313890841e-05, "loss": 0.027, "step": 19308 }, { "epoch": 13.569219957835559, "grad_norm": 0.12027726322412491, "learning_rate": 2.429093464511595e-05, "loss": 0.0116, "step": 19309 }, { "epoch": 13.569922698524245, "grad_norm": 0.13801829516887665, "learning_rate": 2.4290466151323494e-05, "loss": 0.0167, "step": 19310 }, { "epoch": 13.57062543921293, "grad_norm": 0.06860989332199097, "learning_rate": 2.4289997657531038e-05, "loss": 0.0118, "step": 19311 }, { "epoch": 13.571328179901617, "grad_norm": 0.1552620828151703, "learning_rate": 2.428952916373858e-05, "loss": 0.0221, "step": 19312 }, { "epoch": 13.572030920590302, "grad_norm": 0.09518970549106598, "learning_rate": 2.4289060669946125e-05, "loss": 0.0106, "step": 19313 }, { "epoch": 13.572733661278988, "grad_norm": 0.13070036470890045, "learning_rate": 2.4288592176153666e-05, "loss": 0.021, "step": 19314 }, { "epoch": 13.573436401967674, "grad_norm": 0.4083157181739807, "learning_rate": 2.428812368236121e-05, "loss": 0.0278, "step": 19315 }, { "epoch": 13.57413914265636, "grad_norm": 0.20290309190750122, "learning_rate": 2.4287655188568753e-05, "loss": 0.0127, "step": 19316 }, { "epoch": 13.574841883345046, "grad_norm": 0.2941715717315674, "learning_rate": 2.4287186694776297e-05, "loss": 0.04, "step": 19317 }, { "epoch": 13.575544624033732, "grad_norm": 0.2132016271352768, "learning_rate": 2.4286718200983837e-05, "loss": 0.0307, "step": 19318 }, { "epoch": 13.576247364722418, "grad_norm": 0.30397969484329224, "learning_rate": 2.428624970719138e-05, "loss": 0.0494, "step": 19319 }, { "epoch": 13.576950105411104, "grad_norm": 0.3191084861755371, "learning_rate": 2.4285781213398925e-05, "loss": 0.0599, "step": 19320 }, { "epoch": 13.57765284609979, "grad_norm": 0.3569600582122803, "learning_rate": 2.4285312719606468e-05, "loss": 0.0931, "step": 19321 }, { "epoch": 13.578355586788476, "grad_norm": 1.408583164215088, "learning_rate": 2.428484422581401e-05, "loss": 0.1352, "step": 19322 }, { "epoch": 13.579058327477162, "grad_norm": 0.5599409341812134, "learning_rate": 2.428437573202155e-05, "loss": 0.1394, "step": 19323 }, { "epoch": 13.579761068165848, "grad_norm": 1.294621229171753, "learning_rate": 2.4283907238229093e-05, "loss": 0.1651, "step": 19324 }, { "epoch": 13.580463808854532, "grad_norm": 0.15624336898326874, "learning_rate": 2.4283438744436636e-05, "loss": 0.0507, "step": 19325 }, { "epoch": 13.58116654954322, "grad_norm": 0.1524645984172821, "learning_rate": 2.428297025064418e-05, "loss": 0.0297, "step": 19326 }, { "epoch": 13.581869290231904, "grad_norm": 0.10214615613222122, "learning_rate": 2.428250175685172e-05, "loss": 0.01, "step": 19327 }, { "epoch": 13.58257203092059, "grad_norm": 0.0693984255194664, "learning_rate": 2.4282033263059264e-05, "loss": 0.012, "step": 19328 }, { "epoch": 13.583274771609275, "grad_norm": 0.08714812994003296, "learning_rate": 2.4281564769266808e-05, "loss": 0.0121, "step": 19329 }, { "epoch": 13.583977512297961, "grad_norm": 0.08713988214731216, "learning_rate": 2.428109627547435e-05, "loss": 0.0114, "step": 19330 }, { "epoch": 13.584680252986647, "grad_norm": 0.07690989971160889, "learning_rate": 2.4280627781681892e-05, "loss": 0.0087, "step": 19331 }, { "epoch": 13.585382993675333, "grad_norm": 0.09254442155361176, "learning_rate": 2.4280159287889436e-05, "loss": 0.0199, "step": 19332 }, { "epoch": 13.58608573436402, "grad_norm": 0.12855663895606995, "learning_rate": 2.427969079409698e-05, "loss": 0.0236, "step": 19333 }, { "epoch": 13.586788475052705, "grad_norm": 0.12391717731952667, "learning_rate": 2.4279222300304523e-05, "loss": 0.0206, "step": 19334 }, { "epoch": 13.587491215741391, "grad_norm": 0.09200767427682877, "learning_rate": 2.4278753806512063e-05, "loss": 0.0126, "step": 19335 }, { "epoch": 13.588193956430077, "grad_norm": 0.08761357516050339, "learning_rate": 2.4278285312719607e-05, "loss": 0.0104, "step": 19336 }, { "epoch": 13.588896697118763, "grad_norm": 0.30601516366004944, "learning_rate": 2.427781681892715e-05, "loss": 0.0281, "step": 19337 }, { "epoch": 13.589599437807449, "grad_norm": 0.14688017964363098, "learning_rate": 2.4277348325134695e-05, "loss": 0.0152, "step": 19338 }, { "epoch": 13.590302178496135, "grad_norm": 0.08225124329328537, "learning_rate": 2.427687983134224e-05, "loss": 0.0193, "step": 19339 }, { "epoch": 13.59100491918482, "grad_norm": 0.17665995657444, "learning_rate": 2.4276411337549775e-05, "loss": 0.0252, "step": 19340 }, { "epoch": 13.591707659873506, "grad_norm": 0.15065114200115204, "learning_rate": 2.427594284375732e-05, "loss": 0.0323, "step": 19341 }, { "epoch": 13.592410400562192, "grad_norm": 0.10965520888566971, "learning_rate": 2.4275474349964863e-05, "loss": 0.0191, "step": 19342 }, { "epoch": 13.593113141250878, "grad_norm": 0.3485652208328247, "learning_rate": 2.4275005856172407e-05, "loss": 0.056, "step": 19343 }, { "epoch": 13.593815881939564, "grad_norm": 0.15163230895996094, "learning_rate": 2.4274537362379947e-05, "loss": 0.0389, "step": 19344 }, { "epoch": 13.59451862262825, "grad_norm": 0.18466977775096893, "learning_rate": 2.427406886858749e-05, "loss": 0.0654, "step": 19345 }, { "epoch": 13.595221363316936, "grad_norm": 0.3519304394721985, "learning_rate": 2.4273600374795034e-05, "loss": 0.1076, "step": 19346 }, { "epoch": 13.595924104005622, "grad_norm": 0.4023314118385315, "learning_rate": 2.4273131881002578e-05, "loss": 0.1387, "step": 19347 }, { "epoch": 13.596626844694308, "grad_norm": 1.1207269430160522, "learning_rate": 2.427266338721012e-05, "loss": 0.1523, "step": 19348 }, { "epoch": 13.597329585382994, "grad_norm": 1.1204895973205566, "learning_rate": 2.4272194893417662e-05, "loss": 0.1374, "step": 19349 }, { "epoch": 13.59803232607168, "grad_norm": 0.3605693280696869, "learning_rate": 2.4271726399625206e-05, "loss": 0.0533, "step": 19350 }, { "epoch": 13.598735066760366, "grad_norm": 0.25164902210235596, "learning_rate": 2.427125790583275e-05, "loss": 0.0158, "step": 19351 }, { "epoch": 13.599437807449052, "grad_norm": 0.0733056515455246, "learning_rate": 2.4270789412040293e-05, "loss": 0.0111, "step": 19352 }, { "epoch": 13.600140548137738, "grad_norm": 0.1212446317076683, "learning_rate": 2.4270320918247834e-05, "loss": 0.0212, "step": 19353 }, { "epoch": 13.600843288826423, "grad_norm": 0.20377056300640106, "learning_rate": 2.4269852424455377e-05, "loss": 0.0095, "step": 19354 }, { "epoch": 13.60154602951511, "grad_norm": 0.10796499997377396, "learning_rate": 2.426938393066292e-05, "loss": 0.009, "step": 19355 }, { "epoch": 13.602248770203795, "grad_norm": 0.2614023983478546, "learning_rate": 2.4268915436870465e-05, "loss": 0.0236, "step": 19356 }, { "epoch": 13.602951510892481, "grad_norm": 0.077084019780159, "learning_rate": 2.4268446943078005e-05, "loss": 0.0123, "step": 19357 }, { "epoch": 13.603654251581167, "grad_norm": 0.12884649634361267, "learning_rate": 2.4267978449285545e-05, "loss": 0.017, "step": 19358 }, { "epoch": 13.604356992269853, "grad_norm": 0.23597070574760437, "learning_rate": 2.426750995549309e-05, "loss": 0.0271, "step": 19359 }, { "epoch": 13.605059732958539, "grad_norm": 0.1455698013305664, "learning_rate": 2.4267041461700633e-05, "loss": 0.0295, "step": 19360 }, { "epoch": 13.605762473647225, "grad_norm": 0.10372936725616455, "learning_rate": 2.4266572967908173e-05, "loss": 0.0162, "step": 19361 }, { "epoch": 13.60646521433591, "grad_norm": 0.07638894021511078, "learning_rate": 2.4266104474115717e-05, "loss": 0.0112, "step": 19362 }, { "epoch": 13.607167955024597, "grad_norm": 0.09174250811338425, "learning_rate": 2.426563598032326e-05, "loss": 0.0138, "step": 19363 }, { "epoch": 13.607870695713281, "grad_norm": 0.20489197969436646, "learning_rate": 2.4265167486530804e-05, "loss": 0.0202, "step": 19364 }, { "epoch": 13.608573436401969, "grad_norm": 0.21435338258743286, "learning_rate": 2.4264698992738348e-05, "loss": 0.0375, "step": 19365 }, { "epoch": 13.609276177090653, "grad_norm": 0.5305647850036621, "learning_rate": 2.426423049894589e-05, "loss": 0.0246, "step": 19366 }, { "epoch": 13.609978917779339, "grad_norm": 0.12418307363986969, "learning_rate": 2.4263762005153432e-05, "loss": 0.0184, "step": 19367 }, { "epoch": 13.610681658468025, "grad_norm": 0.29304900765419006, "learning_rate": 2.4263293511360976e-05, "loss": 0.0558, "step": 19368 }, { "epoch": 13.61138439915671, "grad_norm": 0.9369145035743713, "learning_rate": 2.426282501756852e-05, "loss": 0.0751, "step": 19369 }, { "epoch": 13.612087139845396, "grad_norm": 0.2515389919281006, "learning_rate": 2.426235652377606e-05, "loss": 0.0762, "step": 19370 }, { "epoch": 13.612789880534082, "grad_norm": 0.39190417528152466, "learning_rate": 2.4261888029983604e-05, "loss": 0.1083, "step": 19371 }, { "epoch": 13.613492621222768, "grad_norm": 0.5573078989982605, "learning_rate": 2.4261419536191147e-05, "loss": 0.1471, "step": 19372 }, { "epoch": 13.614195361911454, "grad_norm": 0.44090843200683594, "learning_rate": 2.426095104239869e-05, "loss": 0.1402, "step": 19373 }, { "epoch": 13.61489810260014, "grad_norm": 1.3457961082458496, "learning_rate": 2.426048254860623e-05, "loss": 0.195, "step": 19374 }, { "epoch": 13.615600843288826, "grad_norm": 0.14649292826652527, "learning_rate": 2.4260014054813772e-05, "loss": 0.0545, "step": 19375 }, { "epoch": 13.616303583977512, "grad_norm": 0.14283010363578796, "learning_rate": 2.4259545561021316e-05, "loss": 0.0302, "step": 19376 }, { "epoch": 13.617006324666198, "grad_norm": 0.1640767604112625, "learning_rate": 2.425907706722886e-05, "loss": 0.0173, "step": 19377 }, { "epoch": 13.617709065354884, "grad_norm": 0.10398560762405396, "learning_rate": 2.4258608573436403e-05, "loss": 0.0159, "step": 19378 }, { "epoch": 13.61841180604357, "grad_norm": 0.07853396236896515, "learning_rate": 2.4258140079643943e-05, "loss": 0.0138, "step": 19379 }, { "epoch": 13.619114546732256, "grad_norm": 0.0921185091137886, "learning_rate": 2.4257671585851487e-05, "loss": 0.0077, "step": 19380 }, { "epoch": 13.619817287420942, "grad_norm": 0.09557466953992844, "learning_rate": 2.425720309205903e-05, "loss": 0.013, "step": 19381 }, { "epoch": 13.620520028109627, "grad_norm": 0.12716828286647797, "learning_rate": 2.4256734598266575e-05, "loss": 0.0256, "step": 19382 }, { "epoch": 13.621222768798313, "grad_norm": 0.11230543255805969, "learning_rate": 2.4256266104474115e-05, "loss": 0.0186, "step": 19383 }, { "epoch": 13.621925509487, "grad_norm": 0.059128522872924805, "learning_rate": 2.425579761068166e-05, "loss": 0.0086, "step": 19384 }, { "epoch": 13.622628250175685, "grad_norm": 0.6144323945045471, "learning_rate": 2.4255329116889202e-05, "loss": 0.0219, "step": 19385 }, { "epoch": 13.623330990864371, "grad_norm": 0.14842276275157928, "learning_rate": 2.4254860623096746e-05, "loss": 0.013, "step": 19386 }, { "epoch": 13.624033731553057, "grad_norm": 0.06753616034984589, "learning_rate": 2.4254392129304286e-05, "loss": 0.0117, "step": 19387 }, { "epoch": 13.624736472241743, "grad_norm": 0.14636845886707306, "learning_rate": 2.425392363551183e-05, "loss": 0.0284, "step": 19388 }, { "epoch": 13.625439212930429, "grad_norm": 0.20067784190177917, "learning_rate": 2.4253455141719374e-05, "loss": 0.035, "step": 19389 }, { "epoch": 13.626141953619115, "grad_norm": 0.15915729105472565, "learning_rate": 2.4252986647926918e-05, "loss": 0.0235, "step": 19390 }, { "epoch": 13.6268446943078, "grad_norm": 0.20357060432434082, "learning_rate": 2.425251815413446e-05, "loss": 0.0323, "step": 19391 }, { "epoch": 13.627547434996487, "grad_norm": 0.27051353454589844, "learning_rate": 2.4252049660341998e-05, "loss": 0.0196, "step": 19392 }, { "epoch": 13.628250175685173, "grad_norm": 0.18307337164878845, "learning_rate": 2.4251581166549542e-05, "loss": 0.0286, "step": 19393 }, { "epoch": 13.628952916373859, "grad_norm": 0.3890129327774048, "learning_rate": 2.4251112672757086e-05, "loss": 0.0527, "step": 19394 }, { "epoch": 13.629655657062544, "grad_norm": 0.33358216285705566, "learning_rate": 2.425064417896463e-05, "loss": 0.0725, "step": 19395 }, { "epoch": 13.63035839775123, "grad_norm": 0.2835501432418823, "learning_rate": 2.425017568517217e-05, "loss": 0.0681, "step": 19396 }, { "epoch": 13.631061138439916, "grad_norm": 0.5057803988456726, "learning_rate": 2.4249707191379713e-05, "loss": 0.1237, "step": 19397 }, { "epoch": 13.631763879128602, "grad_norm": 0.5201452374458313, "learning_rate": 2.4249238697587257e-05, "loss": 0.1098, "step": 19398 }, { "epoch": 13.632466619817288, "grad_norm": 0.8809728622436523, "learning_rate": 2.42487702037948e-05, "loss": 0.1897, "step": 19399 }, { "epoch": 13.633169360505974, "grad_norm": 0.17841103672981262, "learning_rate": 2.4248301710002345e-05, "loss": 0.0492, "step": 19400 }, { "epoch": 13.63387210119466, "grad_norm": 0.15232528746128082, "learning_rate": 2.4247833216209885e-05, "loss": 0.0212, "step": 19401 }, { "epoch": 13.634574841883346, "grad_norm": 0.1489613950252533, "learning_rate": 2.424736472241743e-05, "loss": 0.0119, "step": 19402 }, { "epoch": 13.63527758257203, "grad_norm": 0.11163947731256485, "learning_rate": 2.4246896228624972e-05, "loss": 0.0146, "step": 19403 }, { "epoch": 13.635980323260716, "grad_norm": 0.13064122200012207, "learning_rate": 2.4246427734832516e-05, "loss": 0.0178, "step": 19404 }, { "epoch": 13.636683063949402, "grad_norm": 0.16464385390281677, "learning_rate": 2.4245959241040056e-05, "loss": 0.0084, "step": 19405 }, { "epoch": 13.637385804638088, "grad_norm": 0.15149980783462524, "learning_rate": 2.42454907472476e-05, "loss": 0.0296, "step": 19406 }, { "epoch": 13.638088545326774, "grad_norm": 0.16777245700359344, "learning_rate": 2.4245022253455144e-05, "loss": 0.0183, "step": 19407 }, { "epoch": 13.63879128601546, "grad_norm": 0.08104664087295532, "learning_rate": 2.4244553759662688e-05, "loss": 0.0198, "step": 19408 }, { "epoch": 13.639494026704146, "grad_norm": 0.0782482698559761, "learning_rate": 2.4244085265870228e-05, "loss": 0.0078, "step": 19409 }, { "epoch": 13.640196767392831, "grad_norm": 0.13844552636146545, "learning_rate": 2.424361677207777e-05, "loss": 0.0262, "step": 19410 }, { "epoch": 13.640899508081517, "grad_norm": 0.08546509593725204, "learning_rate": 2.4243148278285312e-05, "loss": 0.0125, "step": 19411 }, { "epoch": 13.641602248770203, "grad_norm": 0.1323668360710144, "learning_rate": 2.4242679784492856e-05, "loss": 0.0171, "step": 19412 }, { "epoch": 13.64230498945889, "grad_norm": 0.1340661197900772, "learning_rate": 2.42422112907004e-05, "loss": 0.0164, "step": 19413 }, { "epoch": 13.643007730147575, "grad_norm": 0.12826506793498993, "learning_rate": 2.424174279690794e-05, "loss": 0.021, "step": 19414 }, { "epoch": 13.643710470836261, "grad_norm": 0.16093169152736664, "learning_rate": 2.4241274303115484e-05, "loss": 0.0246, "step": 19415 }, { "epoch": 13.644413211524947, "grad_norm": 0.164749875664711, "learning_rate": 2.4240805809323027e-05, "loss": 0.0112, "step": 19416 }, { "epoch": 13.645115952213633, "grad_norm": 0.17034366726875305, "learning_rate": 2.424033731553057e-05, "loss": 0.0285, "step": 19417 }, { "epoch": 13.645818692902319, "grad_norm": 0.2167307734489441, "learning_rate": 2.423986882173811e-05, "loss": 0.0457, "step": 19418 }, { "epoch": 13.646521433591005, "grad_norm": 0.5751883387565613, "learning_rate": 2.4239400327945655e-05, "loss": 0.0568, "step": 19419 }, { "epoch": 13.64722417427969, "grad_norm": 0.6382379531860352, "learning_rate": 2.42389318341532e-05, "loss": 0.0607, "step": 19420 }, { "epoch": 13.647926914968377, "grad_norm": 0.4476031959056854, "learning_rate": 2.4238463340360743e-05, "loss": 0.0832, "step": 19421 }, { "epoch": 13.648629655657063, "grad_norm": 0.5177341103553772, "learning_rate": 2.4237994846568283e-05, "loss": 0.1074, "step": 19422 }, { "epoch": 13.649332396345748, "grad_norm": 0.6921364068984985, "learning_rate": 2.4237526352775827e-05, "loss": 0.1646, "step": 19423 }, { "epoch": 13.650035137034434, "grad_norm": 0.8522145748138428, "learning_rate": 2.423705785898337e-05, "loss": 0.1527, "step": 19424 }, { "epoch": 13.65073787772312, "grad_norm": 0.3767348825931549, "learning_rate": 2.4236589365190914e-05, "loss": 0.0717, "step": 19425 }, { "epoch": 13.651440618411806, "grad_norm": 0.09897706657648087, "learning_rate": 2.4236120871398458e-05, "loss": 0.0175, "step": 19426 }, { "epoch": 13.652143359100492, "grad_norm": 0.22315911948680878, "learning_rate": 2.4235652377605995e-05, "loss": 0.0489, "step": 19427 }, { "epoch": 13.652846099789178, "grad_norm": 0.10103350132703781, "learning_rate": 2.423518388381354e-05, "loss": 0.0097, "step": 19428 }, { "epoch": 13.653548840477864, "grad_norm": 0.2799360156059265, "learning_rate": 2.4234715390021082e-05, "loss": 0.0162, "step": 19429 }, { "epoch": 13.65425158116655, "grad_norm": 0.048217203468084335, "learning_rate": 2.4234246896228626e-05, "loss": 0.0062, "step": 19430 }, { "epoch": 13.654954321855236, "grad_norm": 0.15679408609867096, "learning_rate": 2.4233778402436166e-05, "loss": 0.0141, "step": 19431 }, { "epoch": 13.655657062543922, "grad_norm": 0.12622511386871338, "learning_rate": 2.423330990864371e-05, "loss": 0.0177, "step": 19432 }, { "epoch": 13.656359803232608, "grad_norm": 0.22777916491031647, "learning_rate": 2.4232841414851254e-05, "loss": 0.0239, "step": 19433 }, { "epoch": 13.657062543921294, "grad_norm": 0.05872318521142006, "learning_rate": 2.4232372921058797e-05, "loss": 0.0091, "step": 19434 }, { "epoch": 13.65776528460998, "grad_norm": 0.17717140913009644, "learning_rate": 2.4231904427266338e-05, "loss": 0.0221, "step": 19435 }, { "epoch": 13.658468025298665, "grad_norm": 0.13770760595798492, "learning_rate": 2.423143593347388e-05, "loss": 0.0169, "step": 19436 }, { "epoch": 13.659170765987351, "grad_norm": 0.5781586766242981, "learning_rate": 2.4230967439681425e-05, "loss": 0.0194, "step": 19437 }, { "epoch": 13.659873506676037, "grad_norm": 0.06638181954622269, "learning_rate": 2.423049894588897e-05, "loss": 0.0083, "step": 19438 }, { "epoch": 13.660576247364723, "grad_norm": 0.07182198017835617, "learning_rate": 2.4230030452096513e-05, "loss": 0.0151, "step": 19439 }, { "epoch": 13.66127898805341, "grad_norm": 0.23314547538757324, "learning_rate": 2.4229561958304053e-05, "loss": 0.0291, "step": 19440 }, { "epoch": 13.661981728742095, "grad_norm": 0.12450063973665237, "learning_rate": 2.4229093464511597e-05, "loss": 0.0209, "step": 19441 }, { "epoch": 13.66268446943078, "grad_norm": 0.14564339816570282, "learning_rate": 2.422862497071914e-05, "loss": 0.0238, "step": 19442 }, { "epoch": 13.663387210119465, "grad_norm": 0.31387773156166077, "learning_rate": 2.4228156476926684e-05, "loss": 0.0451, "step": 19443 }, { "epoch": 13.664089950808151, "grad_norm": 0.3768400549888611, "learning_rate": 2.4227687983134225e-05, "loss": 0.0848, "step": 19444 }, { "epoch": 13.664792691496837, "grad_norm": 0.33428946137428284, "learning_rate": 2.4227219489341765e-05, "loss": 0.0619, "step": 19445 }, { "epoch": 13.665495432185523, "grad_norm": 0.39270585775375366, "learning_rate": 2.422675099554931e-05, "loss": 0.0766, "step": 19446 }, { "epoch": 13.666198172874209, "grad_norm": 0.6959208846092224, "learning_rate": 2.4226282501756852e-05, "loss": 0.1088, "step": 19447 }, { "epoch": 13.666900913562895, "grad_norm": 0.6853111386299133, "learning_rate": 2.4225814007964393e-05, "loss": 0.162, "step": 19448 }, { "epoch": 13.66760365425158, "grad_norm": 0.9829486012458801, "learning_rate": 2.4225345514171936e-05, "loss": 0.1628, "step": 19449 }, { "epoch": 13.668306394940267, "grad_norm": 0.4449198544025421, "learning_rate": 2.422487702037948e-05, "loss": 0.0682, "step": 19450 }, { "epoch": 13.669009135628952, "grad_norm": 0.08548089861869812, "learning_rate": 2.4224408526587024e-05, "loss": 0.0142, "step": 19451 }, { "epoch": 13.669711876317638, "grad_norm": 0.19793665409088135, "learning_rate": 2.4223940032794568e-05, "loss": 0.0249, "step": 19452 }, { "epoch": 13.670414617006324, "grad_norm": 0.11596330255270004, "learning_rate": 2.4223471539002108e-05, "loss": 0.0138, "step": 19453 }, { "epoch": 13.67111735769501, "grad_norm": 0.11430017650127411, "learning_rate": 2.422300304520965e-05, "loss": 0.0184, "step": 19454 }, { "epoch": 13.671820098383696, "grad_norm": 0.08674411475658417, "learning_rate": 2.4222534551417195e-05, "loss": 0.0087, "step": 19455 }, { "epoch": 13.672522839072382, "grad_norm": 0.14137393236160278, "learning_rate": 2.422206605762474e-05, "loss": 0.0162, "step": 19456 }, { "epoch": 13.673225579761068, "grad_norm": 0.0833824872970581, "learning_rate": 2.422159756383228e-05, "loss": 0.0134, "step": 19457 }, { "epoch": 13.673928320449754, "grad_norm": 0.413402259349823, "learning_rate": 2.4221129070039823e-05, "loss": 0.0171, "step": 19458 }, { "epoch": 13.67463106113844, "grad_norm": 0.07505536079406738, "learning_rate": 2.4220660576247367e-05, "loss": 0.013, "step": 19459 }, { "epoch": 13.675333801827126, "grad_norm": 0.09632351249456406, "learning_rate": 2.422019208245491e-05, "loss": 0.0233, "step": 19460 }, { "epoch": 13.676036542515812, "grad_norm": 0.15645506978034973, "learning_rate": 2.421972358866245e-05, "loss": 0.0252, "step": 19461 }, { "epoch": 13.676739283204498, "grad_norm": 0.1013646051287651, "learning_rate": 2.421925509486999e-05, "loss": 0.0186, "step": 19462 }, { "epoch": 13.677442023893184, "grad_norm": 1.075089931488037, "learning_rate": 2.4218786601077535e-05, "loss": 0.0087, "step": 19463 }, { "epoch": 13.67814476458187, "grad_norm": 0.17454947531223297, "learning_rate": 2.421831810728508e-05, "loss": 0.0221, "step": 19464 }, { "epoch": 13.678847505270555, "grad_norm": 0.17632488906383514, "learning_rate": 2.4217849613492622e-05, "loss": 0.0275, "step": 19465 }, { "epoch": 13.679550245959241, "grad_norm": 0.15906581282615662, "learning_rate": 2.4217381119700163e-05, "loss": 0.0232, "step": 19466 }, { "epoch": 13.680252986647927, "grad_norm": 0.2295469492673874, "learning_rate": 2.4216912625907706e-05, "loss": 0.0335, "step": 19467 }, { "epoch": 13.680955727336613, "grad_norm": 0.3281409740447998, "learning_rate": 2.421644413211525e-05, "loss": 0.0283, "step": 19468 }, { "epoch": 13.681658468025299, "grad_norm": 0.6409674286842346, "learning_rate": 2.4215975638322794e-05, "loss": 0.0293, "step": 19469 }, { "epoch": 13.682361208713985, "grad_norm": 0.2763574421405792, "learning_rate": 2.4215507144530334e-05, "loss": 0.0667, "step": 19470 }, { "epoch": 13.683063949402671, "grad_norm": 0.7223895192146301, "learning_rate": 2.4215038650737878e-05, "loss": 0.0869, "step": 19471 }, { "epoch": 13.683766690091357, "grad_norm": 0.5375245213508606, "learning_rate": 2.4214570156945422e-05, "loss": 0.1559, "step": 19472 }, { "epoch": 13.684469430780043, "grad_norm": 0.5239374041557312, "learning_rate": 2.4214101663152965e-05, "loss": 0.1627, "step": 19473 }, { "epoch": 13.685172171468729, "grad_norm": 0.9902624487876892, "learning_rate": 2.4213633169360506e-05, "loss": 0.148, "step": 19474 }, { "epoch": 13.685874912157415, "grad_norm": 1.5534240007400513, "learning_rate": 2.421316467556805e-05, "loss": 0.0585, "step": 19475 }, { "epoch": 13.6865776528461, "grad_norm": 0.1490614414215088, "learning_rate": 2.4212696181775593e-05, "loss": 0.0237, "step": 19476 }, { "epoch": 13.687280393534786, "grad_norm": 0.20815509557724, "learning_rate": 2.4212227687983137e-05, "loss": 0.0214, "step": 19477 }, { "epoch": 13.687983134223472, "grad_norm": 0.10763143002986908, "learning_rate": 2.421175919419068e-05, "loss": 0.0155, "step": 19478 }, { "epoch": 13.688685874912156, "grad_norm": 0.08449848741292953, "learning_rate": 2.4211290700398218e-05, "loss": 0.0074, "step": 19479 }, { "epoch": 13.689388615600844, "grad_norm": 0.13245181739330292, "learning_rate": 2.421082220660576e-05, "loss": 0.0164, "step": 19480 }, { "epoch": 13.690091356289528, "grad_norm": 0.08149047195911407, "learning_rate": 2.4210353712813305e-05, "loss": 0.0107, "step": 19481 }, { "epoch": 13.690794096978214, "grad_norm": 0.1267922818660736, "learning_rate": 2.420988521902085e-05, "loss": 0.0171, "step": 19482 }, { "epoch": 13.6914968376669, "grad_norm": 0.13251589238643646, "learning_rate": 2.420941672522839e-05, "loss": 0.0221, "step": 19483 }, { "epoch": 13.692199578355586, "grad_norm": 0.08713109791278839, "learning_rate": 2.4208948231435933e-05, "loss": 0.0091, "step": 19484 }, { "epoch": 13.692902319044272, "grad_norm": 0.13103896379470825, "learning_rate": 2.4208479737643477e-05, "loss": 0.0182, "step": 19485 }, { "epoch": 13.693605059732958, "grad_norm": 0.09308698773384094, "learning_rate": 2.420801124385102e-05, "loss": 0.0177, "step": 19486 }, { "epoch": 13.694307800421644, "grad_norm": 0.16280949115753174, "learning_rate": 2.420754275005856e-05, "loss": 0.0208, "step": 19487 }, { "epoch": 13.69501054111033, "grad_norm": 0.12559175491333008, "learning_rate": 2.4207074256266104e-05, "loss": 0.0186, "step": 19488 }, { "epoch": 13.695713281799016, "grad_norm": 0.12139808386564255, "learning_rate": 2.4206605762473648e-05, "loss": 0.0229, "step": 19489 }, { "epoch": 13.696416022487702, "grad_norm": 0.13759852945804596, "learning_rate": 2.4206137268681192e-05, "loss": 0.024, "step": 19490 }, { "epoch": 13.697118763176388, "grad_norm": 0.25600507855415344, "learning_rate": 2.4205668774888736e-05, "loss": 0.0377, "step": 19491 }, { "epoch": 13.697821503865073, "grad_norm": 0.09115168452262878, "learning_rate": 2.4205200281096276e-05, "loss": 0.0213, "step": 19492 }, { "epoch": 13.69852424455376, "grad_norm": 0.7286381125450134, "learning_rate": 2.420473178730382e-05, "loss": 0.0596, "step": 19493 }, { "epoch": 13.699226985242445, "grad_norm": 0.28140732645988464, "learning_rate": 2.4204263293511363e-05, "loss": 0.0496, "step": 19494 }, { "epoch": 13.699929725931131, "grad_norm": 0.9454573392868042, "learning_rate": 2.4203794799718907e-05, "loss": 0.0669, "step": 19495 }, { "epoch": 13.700632466619817, "grad_norm": 0.3049350380897522, "learning_rate": 2.4203326305926447e-05, "loss": 0.0943, "step": 19496 }, { "epoch": 13.701335207308503, "grad_norm": 0.4490078091621399, "learning_rate": 2.4202857812133988e-05, "loss": 0.1444, "step": 19497 }, { "epoch": 13.702037947997189, "grad_norm": 0.5301024913787842, "learning_rate": 2.420238931834153e-05, "loss": 0.144, "step": 19498 }, { "epoch": 13.702740688685875, "grad_norm": 1.3501747846603394, "learning_rate": 2.4201920824549075e-05, "loss": 0.2036, "step": 19499 }, { "epoch": 13.70344342937456, "grad_norm": 0.13520634174346924, "learning_rate": 2.4201452330756616e-05, "loss": 0.0472, "step": 19500 }, { "epoch": 13.704146170063247, "grad_norm": 0.12714727222919464, "learning_rate": 2.420098383696416e-05, "loss": 0.0202, "step": 19501 }, { "epoch": 13.704848910751933, "grad_norm": 0.10607148706912994, "learning_rate": 2.4200515343171703e-05, "loss": 0.0241, "step": 19502 }, { "epoch": 13.705551651440619, "grad_norm": 0.1917288601398468, "learning_rate": 2.4200046849379247e-05, "loss": 0.031, "step": 19503 }, { "epoch": 13.706254392129305, "grad_norm": 0.14626547694206238, "learning_rate": 2.419957835558679e-05, "loss": 0.0154, "step": 19504 }, { "epoch": 13.70695713281799, "grad_norm": 0.071380116045475, "learning_rate": 2.419910986179433e-05, "loss": 0.013, "step": 19505 }, { "epoch": 13.707659873506676, "grad_norm": 0.06523582339286804, "learning_rate": 2.4198641368001874e-05, "loss": 0.0113, "step": 19506 }, { "epoch": 13.708362614195362, "grad_norm": 0.1508311927318573, "learning_rate": 2.4198172874209418e-05, "loss": 0.0162, "step": 19507 }, { "epoch": 13.709065354884048, "grad_norm": 0.09786343574523926, "learning_rate": 2.4197704380416962e-05, "loss": 0.0212, "step": 19508 }, { "epoch": 13.709768095572734, "grad_norm": 0.06632980704307556, "learning_rate": 2.4197235886624502e-05, "loss": 0.0101, "step": 19509 }, { "epoch": 13.71047083626142, "grad_norm": 0.1411363035440445, "learning_rate": 2.4196767392832046e-05, "loss": 0.0143, "step": 19510 }, { "epoch": 13.711173576950106, "grad_norm": 0.09302274137735367, "learning_rate": 2.419629889903959e-05, "loss": 0.0096, "step": 19511 }, { "epoch": 13.711876317638792, "grad_norm": 0.16243240237236023, "learning_rate": 2.4195830405247133e-05, "loss": 0.019, "step": 19512 }, { "epoch": 13.712579058327478, "grad_norm": 0.10765058547258377, "learning_rate": 2.4195361911454674e-05, "loss": 0.018, "step": 19513 }, { "epoch": 13.713281799016164, "grad_norm": 0.18477848172187805, "learning_rate": 2.4194893417662214e-05, "loss": 0.0402, "step": 19514 }, { "epoch": 13.71398453970485, "grad_norm": 0.19889752566814423, "learning_rate": 2.4194424923869758e-05, "loss": 0.0221, "step": 19515 }, { "epoch": 13.714687280393536, "grad_norm": 0.157241553068161, "learning_rate": 2.41939564300773e-05, "loss": 0.0167, "step": 19516 }, { "epoch": 13.715390021082221, "grad_norm": 0.21116036176681519, "learning_rate": 2.4193487936284845e-05, "loss": 0.0382, "step": 19517 }, { "epoch": 13.716092761770906, "grad_norm": 0.37098705768585205, "learning_rate": 2.4193019442492386e-05, "loss": 0.0452, "step": 19518 }, { "epoch": 13.716795502459593, "grad_norm": 0.19531309604644775, "learning_rate": 2.419255094869993e-05, "loss": 0.0462, "step": 19519 }, { "epoch": 13.717498243148277, "grad_norm": 0.5510464310646057, "learning_rate": 2.4192082454907473e-05, "loss": 0.0697, "step": 19520 }, { "epoch": 13.718200983836963, "grad_norm": 0.9108975529670715, "learning_rate": 2.4191613961115017e-05, "loss": 0.0982, "step": 19521 }, { "epoch": 13.71890372452565, "grad_norm": 0.7740064263343811, "learning_rate": 2.4191145467322557e-05, "loss": 0.1283, "step": 19522 }, { "epoch": 13.719606465214335, "grad_norm": 0.920396089553833, "learning_rate": 2.41906769735301e-05, "loss": 0.1516, "step": 19523 }, { "epoch": 13.720309205903021, "grad_norm": 0.7671899795532227, "learning_rate": 2.4190208479737645e-05, "loss": 0.1843, "step": 19524 }, { "epoch": 13.721011946591707, "grad_norm": 0.30531126260757446, "learning_rate": 2.418973998594519e-05, "loss": 0.0729, "step": 19525 }, { "epoch": 13.721714687280393, "grad_norm": 0.6261196136474609, "learning_rate": 2.418927149215273e-05, "loss": 0.0209, "step": 19526 }, { "epoch": 13.722417427969079, "grad_norm": 0.09068962931632996, "learning_rate": 2.4188802998360272e-05, "loss": 0.0165, "step": 19527 }, { "epoch": 13.723120168657765, "grad_norm": 0.10981105268001556, "learning_rate": 2.4188334504567816e-05, "loss": 0.0178, "step": 19528 }, { "epoch": 13.72382290934645, "grad_norm": 0.09261438250541687, "learning_rate": 2.418786601077536e-05, "loss": 0.0212, "step": 19529 }, { "epoch": 13.724525650035137, "grad_norm": 0.20939798653125763, "learning_rate": 2.4187397516982904e-05, "loss": 0.0131, "step": 19530 }, { "epoch": 13.725228390723823, "grad_norm": 0.08498059213161469, "learning_rate": 2.4186929023190444e-05, "loss": 0.0168, "step": 19531 }, { "epoch": 13.725931131412509, "grad_norm": 0.24458815157413483, "learning_rate": 2.4186460529397984e-05, "loss": 0.0169, "step": 19532 }, { "epoch": 13.726633872101194, "grad_norm": 0.09534633159637451, "learning_rate": 2.4185992035605528e-05, "loss": 0.0117, "step": 19533 }, { "epoch": 13.72733661278988, "grad_norm": 0.1171659454703331, "learning_rate": 2.418552354181307e-05, "loss": 0.0141, "step": 19534 }, { "epoch": 13.728039353478566, "grad_norm": 0.8929540514945984, "learning_rate": 2.4185055048020612e-05, "loss": 0.0158, "step": 19535 }, { "epoch": 13.728742094167252, "grad_norm": 0.07467657327651978, "learning_rate": 2.4184586554228156e-05, "loss": 0.0099, "step": 19536 }, { "epoch": 13.729444834855938, "grad_norm": 0.3901980221271515, "learning_rate": 2.41841180604357e-05, "loss": 0.02, "step": 19537 }, { "epoch": 13.730147575544624, "grad_norm": 0.24403399229049683, "learning_rate": 2.4183649566643243e-05, "loss": 0.0144, "step": 19538 }, { "epoch": 13.73085031623331, "grad_norm": 0.31515154242515564, "learning_rate": 2.4183181072850784e-05, "loss": 0.028, "step": 19539 }, { "epoch": 13.731553056921996, "grad_norm": 0.12215621769428253, "learning_rate": 2.4182712579058327e-05, "loss": 0.0209, "step": 19540 }, { "epoch": 13.732255797610682, "grad_norm": 0.1553562879562378, "learning_rate": 2.418224408526587e-05, "loss": 0.0311, "step": 19541 }, { "epoch": 13.732958538299368, "grad_norm": 0.13354915380477905, "learning_rate": 2.4181775591473415e-05, "loss": 0.0359, "step": 19542 }, { "epoch": 13.733661278988054, "grad_norm": 0.2553670108318329, "learning_rate": 2.418130709768096e-05, "loss": 0.0282, "step": 19543 }, { "epoch": 13.73436401967674, "grad_norm": 0.4967522621154785, "learning_rate": 2.41808386038885e-05, "loss": 0.0411, "step": 19544 }, { "epoch": 13.735066760365426, "grad_norm": 2.0664706230163574, "learning_rate": 2.4180370110096042e-05, "loss": 0.0759, "step": 19545 }, { "epoch": 13.735769501054111, "grad_norm": 0.3104909062385559, "learning_rate": 2.4179901616303586e-05, "loss": 0.1033, "step": 19546 }, { "epoch": 13.736472241742797, "grad_norm": 0.5957318544387817, "learning_rate": 2.417943312251113e-05, "loss": 0.1653, "step": 19547 }, { "epoch": 13.737174982431483, "grad_norm": 0.7159046530723572, "learning_rate": 2.417896462871867e-05, "loss": 0.1586, "step": 19548 }, { "epoch": 13.73787772312017, "grad_norm": 1.7248656749725342, "learning_rate": 2.417849613492621e-05, "loss": 0.2015, "step": 19549 }, { "epoch": 13.738580463808855, "grad_norm": 0.17336072027683258, "learning_rate": 2.4178027641133754e-05, "loss": 0.0652, "step": 19550 }, { "epoch": 13.739283204497541, "grad_norm": 0.10914354026317596, "learning_rate": 2.4177559147341298e-05, "loss": 0.0173, "step": 19551 }, { "epoch": 13.739985945186227, "grad_norm": 0.11246008425951004, "learning_rate": 2.417709065354884e-05, "loss": 0.0119, "step": 19552 }, { "epoch": 13.740688685874913, "grad_norm": 0.20989976823329926, "learning_rate": 2.4176622159756382e-05, "loss": 0.0223, "step": 19553 }, { "epoch": 13.741391426563599, "grad_norm": 0.06749560683965683, "learning_rate": 2.4176153665963926e-05, "loss": 0.0085, "step": 19554 }, { "epoch": 13.742094167252285, "grad_norm": 0.07793789356946945, "learning_rate": 2.417568517217147e-05, "loss": 0.0092, "step": 19555 }, { "epoch": 13.74279690794097, "grad_norm": 0.13362225890159607, "learning_rate": 2.4175216678379013e-05, "loss": 0.0269, "step": 19556 }, { "epoch": 13.743499648629655, "grad_norm": 0.22781860828399658, "learning_rate": 2.4174748184586554e-05, "loss": 0.0141, "step": 19557 }, { "epoch": 13.74420238931834, "grad_norm": 0.11110887676477432, "learning_rate": 2.4174279690794097e-05, "loss": 0.0156, "step": 19558 }, { "epoch": 13.744905130007027, "grad_norm": 0.0827345922589302, "learning_rate": 2.417381119700164e-05, "loss": 0.0107, "step": 19559 }, { "epoch": 13.745607870695713, "grad_norm": 0.4010419547557831, "learning_rate": 2.4173342703209185e-05, "loss": 0.0148, "step": 19560 }, { "epoch": 13.746310611384398, "grad_norm": 0.18953388929367065, "learning_rate": 2.4172874209416725e-05, "loss": 0.0176, "step": 19561 }, { "epoch": 13.747013352073084, "grad_norm": 0.6807948350906372, "learning_rate": 2.417240571562427e-05, "loss": 0.0446, "step": 19562 }, { "epoch": 13.74771609276177, "grad_norm": 0.20324452221393585, "learning_rate": 2.4171937221831813e-05, "loss": 0.0135, "step": 19563 }, { "epoch": 13.748418833450456, "grad_norm": 0.2687673270702362, "learning_rate": 2.4171468728039356e-05, "loss": 0.0206, "step": 19564 }, { "epoch": 13.749121574139142, "grad_norm": 0.334347665309906, "learning_rate": 2.4171000234246897e-05, "loss": 0.0371, "step": 19565 }, { "epoch": 13.749824314827828, "grad_norm": 0.14838357269763947, "learning_rate": 2.417053174045444e-05, "loss": 0.0247, "step": 19566 }, { "epoch": 13.750527055516514, "grad_norm": 0.2678598165512085, "learning_rate": 2.417006324666198e-05, "loss": 0.0476, "step": 19567 }, { "epoch": 13.7512297962052, "grad_norm": 0.14081387221813202, "learning_rate": 2.4169594752869524e-05, "loss": 0.0236, "step": 19568 }, { "epoch": 13.751932536893886, "grad_norm": 0.6347100734710693, "learning_rate": 2.4169126259077068e-05, "loss": 0.0347, "step": 19569 }, { "epoch": 13.752635277582572, "grad_norm": 0.2866072654724121, "learning_rate": 2.416865776528461e-05, "loss": 0.0639, "step": 19570 }, { "epoch": 13.753338018271258, "grad_norm": 0.43371590971946716, "learning_rate": 2.4168189271492152e-05, "loss": 0.0851, "step": 19571 }, { "epoch": 13.754040758959944, "grad_norm": 0.6203228831291199, "learning_rate": 2.4167720777699696e-05, "loss": 0.1469, "step": 19572 }, { "epoch": 13.75474349964863, "grad_norm": 1.3831454515457153, "learning_rate": 2.416725228390724e-05, "loss": 0.161, "step": 19573 }, { "epoch": 13.755446240337315, "grad_norm": 1.2387109994888306, "learning_rate": 2.416678379011478e-05, "loss": 0.184, "step": 19574 }, { "epoch": 13.756148981026001, "grad_norm": 0.2118256837129593, "learning_rate": 2.4166315296322324e-05, "loss": 0.067, "step": 19575 }, { "epoch": 13.756851721714687, "grad_norm": 0.15109294652938843, "learning_rate": 2.4165846802529867e-05, "loss": 0.022, "step": 19576 }, { "epoch": 13.757554462403373, "grad_norm": 0.07924724370241165, "learning_rate": 2.416537830873741e-05, "loss": 0.0125, "step": 19577 }, { "epoch": 13.75825720309206, "grad_norm": 0.10679908096790314, "learning_rate": 2.416490981494495e-05, "loss": 0.0216, "step": 19578 }, { "epoch": 13.758959943780745, "grad_norm": 0.06837289780378342, "learning_rate": 2.4164441321152495e-05, "loss": 0.0156, "step": 19579 }, { "epoch": 13.759662684469431, "grad_norm": 0.06426502019166946, "learning_rate": 2.416397282736004e-05, "loss": 0.0108, "step": 19580 }, { "epoch": 13.760365425158117, "grad_norm": 0.07657855749130249, "learning_rate": 2.4163504333567583e-05, "loss": 0.0144, "step": 19581 }, { "epoch": 13.761068165846803, "grad_norm": 0.30299368500709534, "learning_rate": 2.4163035839775126e-05, "loss": 0.0309, "step": 19582 }, { "epoch": 13.761770906535489, "grad_norm": 0.16089065372943878, "learning_rate": 2.4162567345982667e-05, "loss": 0.0182, "step": 19583 }, { "epoch": 13.762473647224175, "grad_norm": 0.09886962175369263, "learning_rate": 2.4162098852190207e-05, "loss": 0.0114, "step": 19584 }, { "epoch": 13.76317638791286, "grad_norm": 0.0926152765750885, "learning_rate": 2.416163035839775e-05, "loss": 0.0215, "step": 19585 }, { "epoch": 13.763879128601546, "grad_norm": 0.10242333263158798, "learning_rate": 2.4161161864605295e-05, "loss": 0.0132, "step": 19586 }, { "epoch": 13.764581869290232, "grad_norm": 0.14637689292430878, "learning_rate": 2.4160693370812835e-05, "loss": 0.0319, "step": 19587 }, { "epoch": 13.765284609978918, "grad_norm": 0.2151026427745819, "learning_rate": 2.416022487702038e-05, "loss": 0.0151, "step": 19588 }, { "epoch": 13.765987350667604, "grad_norm": 0.19653727114200592, "learning_rate": 2.4159756383227922e-05, "loss": 0.03, "step": 19589 }, { "epoch": 13.76669009135629, "grad_norm": 0.1978393942117691, "learning_rate": 2.4159287889435466e-05, "loss": 0.0303, "step": 19590 }, { "epoch": 13.767392832044976, "grad_norm": 0.12654180824756622, "learning_rate": 2.415881939564301e-05, "loss": 0.0196, "step": 19591 }, { "epoch": 13.768095572733662, "grad_norm": 0.17188455164432526, "learning_rate": 2.415835090185055e-05, "loss": 0.0322, "step": 19592 }, { "epoch": 13.768798313422348, "grad_norm": 0.28077244758605957, "learning_rate": 2.4157882408058094e-05, "loss": 0.0414, "step": 19593 }, { "epoch": 13.769501054111032, "grad_norm": 0.19932638108730316, "learning_rate": 2.4157413914265638e-05, "loss": 0.051, "step": 19594 }, { "epoch": 13.77020379479972, "grad_norm": 0.40819305181503296, "learning_rate": 2.415694542047318e-05, "loss": 0.1, "step": 19595 }, { "epoch": 13.770906535488404, "grad_norm": 0.34468916058540344, "learning_rate": 2.415647692668072e-05, "loss": 0.0902, "step": 19596 }, { "epoch": 13.77160927617709, "grad_norm": 0.7352005243301392, "learning_rate": 2.4156008432888265e-05, "loss": 0.1358, "step": 19597 }, { "epoch": 13.772312016865776, "grad_norm": 0.6122785210609436, "learning_rate": 2.415553993909581e-05, "loss": 0.1562, "step": 19598 }, { "epoch": 13.773014757554462, "grad_norm": 1.2875165939331055, "learning_rate": 2.4155071445303353e-05, "loss": 0.2193, "step": 19599 }, { "epoch": 13.773717498243148, "grad_norm": 0.17845775187015533, "learning_rate": 2.4154602951510893e-05, "loss": 0.0597, "step": 19600 }, { "epoch": 13.774420238931834, "grad_norm": 0.19767358899116516, "learning_rate": 2.4154134457718434e-05, "loss": 0.041, "step": 19601 }, { "epoch": 13.77512297962052, "grad_norm": 0.10428831726312637, "learning_rate": 2.4153665963925977e-05, "loss": 0.0222, "step": 19602 }, { "epoch": 13.775825720309205, "grad_norm": 0.08598463237285614, "learning_rate": 2.415319747013352e-05, "loss": 0.0138, "step": 19603 }, { "epoch": 13.776528460997891, "grad_norm": 0.10195107012987137, "learning_rate": 2.4152728976341065e-05, "loss": 0.0288, "step": 19604 }, { "epoch": 13.777231201686577, "grad_norm": 0.14387886226177216, "learning_rate": 2.4152260482548605e-05, "loss": 0.0072, "step": 19605 }, { "epoch": 13.777933942375263, "grad_norm": 0.09867526590824127, "learning_rate": 2.415179198875615e-05, "loss": 0.0165, "step": 19606 }, { "epoch": 13.778636683063949, "grad_norm": 0.17336319386959076, "learning_rate": 2.4151323494963692e-05, "loss": 0.0141, "step": 19607 }, { "epoch": 13.779339423752635, "grad_norm": 0.08046445995569229, "learning_rate": 2.4150855001171236e-05, "loss": 0.0205, "step": 19608 }, { "epoch": 13.780042164441321, "grad_norm": 0.15390625596046448, "learning_rate": 2.4150386507378777e-05, "loss": 0.016, "step": 19609 }, { "epoch": 13.780744905130007, "grad_norm": 0.08104508370161057, "learning_rate": 2.414991801358632e-05, "loss": 0.0185, "step": 19610 }, { "epoch": 13.781447645818693, "grad_norm": 0.12182538956403732, "learning_rate": 2.4149449519793864e-05, "loss": 0.0138, "step": 19611 }, { "epoch": 13.782150386507379, "grad_norm": 0.33495089411735535, "learning_rate": 2.4148981026001408e-05, "loss": 0.0226, "step": 19612 }, { "epoch": 13.782853127196065, "grad_norm": 0.07882894575595856, "learning_rate": 2.4148512532208948e-05, "loss": 0.0214, "step": 19613 }, { "epoch": 13.78355586788475, "grad_norm": 0.36681437492370605, "learning_rate": 2.4148044038416492e-05, "loss": 0.019, "step": 19614 }, { "epoch": 13.784258608573436, "grad_norm": 0.17378155887126923, "learning_rate": 2.4147575544624035e-05, "loss": 0.0256, "step": 19615 }, { "epoch": 13.784961349262122, "grad_norm": 0.4038929343223572, "learning_rate": 2.414710705083158e-05, "loss": 0.0138, "step": 19616 }, { "epoch": 13.785664089950808, "grad_norm": 0.1409052461385727, "learning_rate": 2.4146638557039123e-05, "loss": 0.0246, "step": 19617 }, { "epoch": 13.786366830639494, "grad_norm": 0.1840071976184845, "learning_rate": 2.4146170063246663e-05, "loss": 0.0276, "step": 19618 }, { "epoch": 13.78706957132818, "grad_norm": 0.5253867506980896, "learning_rate": 2.4145701569454204e-05, "loss": 0.0304, "step": 19619 }, { "epoch": 13.787772312016866, "grad_norm": 0.6104207634925842, "learning_rate": 2.4145233075661747e-05, "loss": 0.0757, "step": 19620 }, { "epoch": 13.788475052705552, "grad_norm": 0.40094995498657227, "learning_rate": 2.414476458186929e-05, "loss": 0.1012, "step": 19621 }, { "epoch": 13.789177793394238, "grad_norm": 1.5786875486373901, "learning_rate": 2.414429608807683e-05, "loss": 0.1101, "step": 19622 }, { "epoch": 13.789880534082924, "grad_norm": 0.6581251621246338, "learning_rate": 2.4143827594284375e-05, "loss": 0.1671, "step": 19623 }, { "epoch": 13.79058327477161, "grad_norm": 0.7615600228309631, "learning_rate": 2.414335910049192e-05, "loss": 0.2004, "step": 19624 }, { "epoch": 13.791286015460296, "grad_norm": 0.16114477813243866, "learning_rate": 2.4142890606699463e-05, "loss": 0.0488, "step": 19625 }, { "epoch": 13.791988756148982, "grad_norm": 0.13348332047462463, "learning_rate": 2.4142422112907003e-05, "loss": 0.014, "step": 19626 }, { "epoch": 13.792691496837667, "grad_norm": 0.21118266880512238, "learning_rate": 2.4141953619114547e-05, "loss": 0.0424, "step": 19627 }, { "epoch": 13.793394237526353, "grad_norm": 0.15254053473472595, "learning_rate": 2.414148512532209e-05, "loss": 0.0202, "step": 19628 }, { "epoch": 13.79409697821504, "grad_norm": 0.44919300079345703, "learning_rate": 2.4141016631529634e-05, "loss": 0.0212, "step": 19629 }, { "epoch": 13.794799718903725, "grad_norm": 0.05680595338344574, "learning_rate": 2.4140548137737178e-05, "loss": 0.0101, "step": 19630 }, { "epoch": 13.795502459592411, "grad_norm": 0.084080271422863, "learning_rate": 2.4140079643944718e-05, "loss": 0.0151, "step": 19631 }, { "epoch": 13.796205200281097, "grad_norm": 0.3560958206653595, "learning_rate": 2.4139611150152262e-05, "loss": 0.0184, "step": 19632 }, { "epoch": 13.796907940969781, "grad_norm": 0.10855790227651596, "learning_rate": 2.4139142656359806e-05, "loss": 0.0178, "step": 19633 }, { "epoch": 13.797610681658469, "grad_norm": 0.11262431740760803, "learning_rate": 2.413867416256735e-05, "loss": 0.012, "step": 19634 }, { "epoch": 13.798313422347153, "grad_norm": 0.11330810934305191, "learning_rate": 2.413820566877489e-05, "loss": 0.0163, "step": 19635 }, { "epoch": 13.799016163035839, "grad_norm": 0.09706398844718933, "learning_rate": 2.413773717498243e-05, "loss": 0.0137, "step": 19636 }, { "epoch": 13.799718903724525, "grad_norm": 0.2326674461364746, "learning_rate": 2.4137268681189974e-05, "loss": 0.0193, "step": 19637 }, { "epoch": 13.80042164441321, "grad_norm": 0.1574329137802124, "learning_rate": 2.4136800187397517e-05, "loss": 0.0193, "step": 19638 }, { "epoch": 13.801124385101897, "grad_norm": 0.14997516572475433, "learning_rate": 2.4136331693605058e-05, "loss": 0.0273, "step": 19639 }, { "epoch": 13.801827125790583, "grad_norm": 0.16548947989940643, "learning_rate": 2.41358631998126e-05, "loss": 0.0404, "step": 19640 }, { "epoch": 13.802529866479269, "grad_norm": 0.10338294506072998, "learning_rate": 2.4135394706020145e-05, "loss": 0.0125, "step": 19641 }, { "epoch": 13.803232607167955, "grad_norm": 0.1982795000076294, "learning_rate": 2.413492621222769e-05, "loss": 0.0335, "step": 19642 }, { "epoch": 13.80393534785664, "grad_norm": 0.3950537443161011, "learning_rate": 2.4134457718435233e-05, "loss": 0.0168, "step": 19643 }, { "epoch": 13.804638088545326, "grad_norm": 0.22918406128883362, "learning_rate": 2.4133989224642773e-05, "loss": 0.0694, "step": 19644 }, { "epoch": 13.805340829234012, "grad_norm": 0.21670465171337128, "learning_rate": 2.4133520730850317e-05, "loss": 0.0592, "step": 19645 }, { "epoch": 13.806043569922698, "grad_norm": 0.26055294275283813, "learning_rate": 2.413305223705786e-05, "loss": 0.0811, "step": 19646 }, { "epoch": 13.806746310611384, "grad_norm": 0.4357536733150482, "learning_rate": 2.4132583743265404e-05, "loss": 0.1308, "step": 19647 }, { "epoch": 13.80744905130007, "grad_norm": 0.6756681203842163, "learning_rate": 2.4132115249472945e-05, "loss": 0.1624, "step": 19648 }, { "epoch": 13.808151791988756, "grad_norm": 0.9400915503501892, "learning_rate": 2.4131646755680488e-05, "loss": 0.212, "step": 19649 }, { "epoch": 13.808854532677442, "grad_norm": 0.25528591871261597, "learning_rate": 2.4131178261888032e-05, "loss": 0.046, "step": 19650 }, { "epoch": 13.809557273366128, "grad_norm": 0.10510699450969696, "learning_rate": 2.4130709768095576e-05, "loss": 0.0232, "step": 19651 }, { "epoch": 13.810260014054814, "grad_norm": 0.10526569932699203, "learning_rate": 2.4130241274303116e-05, "loss": 0.0182, "step": 19652 }, { "epoch": 13.8109627547435, "grad_norm": 0.055117782205343246, "learning_rate": 2.412977278051066e-05, "loss": 0.0079, "step": 19653 }, { "epoch": 13.811665495432186, "grad_norm": 0.20431146025657654, "learning_rate": 2.41293042867182e-05, "loss": 0.0228, "step": 19654 }, { "epoch": 13.812368236120872, "grad_norm": 0.07244065403938293, "learning_rate": 2.4128835792925744e-05, "loss": 0.013, "step": 19655 }, { "epoch": 13.813070976809557, "grad_norm": 0.06323237717151642, "learning_rate": 2.4128367299133288e-05, "loss": 0.0101, "step": 19656 }, { "epoch": 13.813773717498243, "grad_norm": 0.2598039209842682, "learning_rate": 2.4127898805340828e-05, "loss": 0.0307, "step": 19657 }, { "epoch": 13.81447645818693, "grad_norm": 0.0971146896481514, "learning_rate": 2.412743031154837e-05, "loss": 0.0157, "step": 19658 }, { "epoch": 13.815179198875615, "grad_norm": 0.13671888411045074, "learning_rate": 2.4126961817755915e-05, "loss": 0.0087, "step": 19659 }, { "epoch": 13.815881939564301, "grad_norm": 0.2349073588848114, "learning_rate": 2.412649332396346e-05, "loss": 0.028, "step": 19660 }, { "epoch": 13.816584680252987, "grad_norm": 0.17902064323425293, "learning_rate": 2.4126024830171e-05, "loss": 0.0162, "step": 19661 }, { "epoch": 13.817287420941673, "grad_norm": 0.17018745839595795, "learning_rate": 2.4125556336378543e-05, "loss": 0.0225, "step": 19662 }, { "epoch": 13.817990161630359, "grad_norm": 0.12259484827518463, "learning_rate": 2.4125087842586087e-05, "loss": 0.0207, "step": 19663 }, { "epoch": 13.818692902319045, "grad_norm": 0.1973901242017746, "learning_rate": 2.412461934879363e-05, "loss": 0.0276, "step": 19664 }, { "epoch": 13.81939564300773, "grad_norm": 0.09330601990222931, "learning_rate": 2.412415085500117e-05, "loss": 0.0255, "step": 19665 }, { "epoch": 13.820098383696417, "grad_norm": 0.19161099195480347, "learning_rate": 2.4123682361208715e-05, "loss": 0.0183, "step": 19666 }, { "epoch": 13.820801124385103, "grad_norm": 0.4232977330684662, "learning_rate": 2.412321386741626e-05, "loss": 0.0327, "step": 19667 }, { "epoch": 13.821503865073788, "grad_norm": 0.2336534559726715, "learning_rate": 2.4122745373623802e-05, "loss": 0.0576, "step": 19668 }, { "epoch": 13.822206605762474, "grad_norm": 0.44683489203453064, "learning_rate": 2.4122276879831346e-05, "loss": 0.0466, "step": 19669 }, { "epoch": 13.82290934645116, "grad_norm": 0.24281366169452667, "learning_rate": 2.4121808386038886e-05, "loss": 0.0546, "step": 19670 }, { "epoch": 13.823612087139846, "grad_norm": 0.5414309501647949, "learning_rate": 2.4121339892246427e-05, "loss": 0.0982, "step": 19671 }, { "epoch": 13.82431482782853, "grad_norm": 1.2612078189849854, "learning_rate": 2.412087139845397e-05, "loss": 0.1443, "step": 19672 }, { "epoch": 13.825017568517218, "grad_norm": 0.5494444370269775, "learning_rate": 2.4120402904661514e-05, "loss": 0.1582, "step": 19673 }, { "epoch": 13.825720309205902, "grad_norm": 1.1827956438064575, "learning_rate": 2.4119934410869054e-05, "loss": 0.1554, "step": 19674 }, { "epoch": 13.826423049894588, "grad_norm": 0.21401184797286987, "learning_rate": 2.4119465917076598e-05, "loss": 0.0582, "step": 19675 }, { "epoch": 13.827125790583274, "grad_norm": 0.13371963798999786, "learning_rate": 2.4118997423284142e-05, "loss": 0.0224, "step": 19676 }, { "epoch": 13.82782853127196, "grad_norm": 0.12674804031848907, "learning_rate": 2.4118528929491685e-05, "loss": 0.0143, "step": 19677 }, { "epoch": 13.828531271960646, "grad_norm": 0.13616229593753815, "learning_rate": 2.4118060435699226e-05, "loss": 0.0163, "step": 19678 }, { "epoch": 13.829234012649332, "grad_norm": 0.08075684309005737, "learning_rate": 2.411759194190677e-05, "loss": 0.012, "step": 19679 }, { "epoch": 13.829936753338018, "grad_norm": 0.06899375468492508, "learning_rate": 2.4117123448114313e-05, "loss": 0.0126, "step": 19680 }, { "epoch": 13.830639494026704, "grad_norm": 0.14448444545269012, "learning_rate": 2.4116654954321857e-05, "loss": 0.0283, "step": 19681 }, { "epoch": 13.83134223471539, "grad_norm": 0.19488798081874847, "learning_rate": 2.41161864605294e-05, "loss": 0.0168, "step": 19682 }, { "epoch": 13.832044975404076, "grad_norm": 0.17622001469135284, "learning_rate": 2.411571796673694e-05, "loss": 0.0139, "step": 19683 }, { "epoch": 13.832747716092761, "grad_norm": 0.07363208383321762, "learning_rate": 2.4115249472944485e-05, "loss": 0.0092, "step": 19684 }, { "epoch": 13.833450456781447, "grad_norm": 0.19217655062675476, "learning_rate": 2.411478097915203e-05, "loss": 0.0184, "step": 19685 }, { "epoch": 13.834153197470133, "grad_norm": 0.09053945541381836, "learning_rate": 2.4114312485359572e-05, "loss": 0.0116, "step": 19686 }, { "epoch": 13.83485593815882, "grad_norm": 0.14101770520210266, "learning_rate": 2.4113843991567113e-05, "loss": 0.0293, "step": 19687 }, { "epoch": 13.835558678847505, "grad_norm": 0.0928146094083786, "learning_rate": 2.4113375497774653e-05, "loss": 0.0134, "step": 19688 }, { "epoch": 13.836261419536191, "grad_norm": 0.09846428036689758, "learning_rate": 2.4112907003982197e-05, "loss": 0.0257, "step": 19689 }, { "epoch": 13.836964160224877, "grad_norm": 0.17543353140354156, "learning_rate": 2.411243851018974e-05, "loss": 0.029, "step": 19690 }, { "epoch": 13.837666900913563, "grad_norm": 0.13649944961071014, "learning_rate": 2.411197001639728e-05, "loss": 0.0152, "step": 19691 }, { "epoch": 13.838369641602249, "grad_norm": 0.08746741712093353, "learning_rate": 2.4111501522604824e-05, "loss": 0.017, "step": 19692 }, { "epoch": 13.839072382290935, "grad_norm": 0.13190430402755737, "learning_rate": 2.4111033028812368e-05, "loss": 0.0299, "step": 19693 }, { "epoch": 13.83977512297962, "grad_norm": 0.4515990614891052, "learning_rate": 2.4110564535019912e-05, "loss": 0.0437, "step": 19694 }, { "epoch": 13.840477863668307, "grad_norm": 0.24865369498729706, "learning_rate": 2.4110096041227456e-05, "loss": 0.0562, "step": 19695 }, { "epoch": 13.841180604356992, "grad_norm": 0.5687583088874817, "learning_rate": 2.4109627547434996e-05, "loss": 0.0897, "step": 19696 }, { "epoch": 13.841883345045678, "grad_norm": 0.4136126935482025, "learning_rate": 2.410915905364254e-05, "loss": 0.1291, "step": 19697 }, { "epoch": 13.842586085734364, "grad_norm": 1.2081170082092285, "learning_rate": 2.4108690559850083e-05, "loss": 0.1664, "step": 19698 }, { "epoch": 13.84328882642305, "grad_norm": 0.7237892150878906, "learning_rate": 2.4108222066057627e-05, "loss": 0.1865, "step": 19699 }, { "epoch": 13.843991567111736, "grad_norm": 0.22688642144203186, "learning_rate": 2.4107753572265167e-05, "loss": 0.0735, "step": 19700 }, { "epoch": 13.844694307800422, "grad_norm": 0.27937376499176025, "learning_rate": 2.410728507847271e-05, "loss": 0.0254, "step": 19701 }, { "epoch": 13.845397048489108, "grad_norm": 0.09834986180067062, "learning_rate": 2.4106816584680255e-05, "loss": 0.0193, "step": 19702 }, { "epoch": 13.846099789177794, "grad_norm": 0.26268279552459717, "learning_rate": 2.41063480908878e-05, "loss": 0.0148, "step": 19703 }, { "epoch": 13.84680252986648, "grad_norm": 0.18835024535655975, "learning_rate": 2.410587959709534e-05, "loss": 0.0132, "step": 19704 }, { "epoch": 13.847505270555166, "grad_norm": 0.15027928352355957, "learning_rate": 2.4105411103302883e-05, "loss": 0.02, "step": 19705 }, { "epoch": 13.848208011243852, "grad_norm": 0.14832939207553864, "learning_rate": 2.4104942609510423e-05, "loss": 0.0202, "step": 19706 }, { "epoch": 13.848910751932538, "grad_norm": 0.0971500501036644, "learning_rate": 2.4104474115717967e-05, "loss": 0.0188, "step": 19707 }, { "epoch": 13.849613492621224, "grad_norm": 0.11708415299654007, "learning_rate": 2.410400562192551e-05, "loss": 0.0111, "step": 19708 }, { "epoch": 13.85031623330991, "grad_norm": 0.08833737671375275, "learning_rate": 2.410353712813305e-05, "loss": 0.0194, "step": 19709 }, { "epoch": 13.851018973998595, "grad_norm": 0.11218526214361191, "learning_rate": 2.4103068634340595e-05, "loss": 0.0179, "step": 19710 }, { "epoch": 13.85172171468728, "grad_norm": 0.11129115521907806, "learning_rate": 2.4102600140548138e-05, "loss": 0.0156, "step": 19711 }, { "epoch": 13.852424455375965, "grad_norm": 0.2537301778793335, "learning_rate": 2.4102131646755682e-05, "loss": 0.0463, "step": 19712 }, { "epoch": 13.853127196064651, "grad_norm": 0.0675962045788765, "learning_rate": 2.4101663152963222e-05, "loss": 0.0147, "step": 19713 }, { "epoch": 13.853829936753337, "grad_norm": 0.28686004877090454, "learning_rate": 2.4101194659170766e-05, "loss": 0.0276, "step": 19714 }, { "epoch": 13.854532677442023, "grad_norm": 0.14448827505111694, "learning_rate": 2.410072616537831e-05, "loss": 0.0269, "step": 19715 }, { "epoch": 13.85523541813071, "grad_norm": 0.10814709961414337, "learning_rate": 2.4100257671585853e-05, "loss": 0.0143, "step": 19716 }, { "epoch": 13.855938158819395, "grad_norm": 0.13959510624408722, "learning_rate": 2.4099789177793394e-05, "loss": 0.0325, "step": 19717 }, { "epoch": 13.856640899508081, "grad_norm": 0.1476123332977295, "learning_rate": 2.4099320684000938e-05, "loss": 0.0293, "step": 19718 }, { "epoch": 13.857343640196767, "grad_norm": 0.2576214075088501, "learning_rate": 2.409885219020848e-05, "loss": 0.0481, "step": 19719 }, { "epoch": 13.858046380885453, "grad_norm": 0.23583301901817322, "learning_rate": 2.4098383696416025e-05, "loss": 0.0559, "step": 19720 }, { "epoch": 13.858749121574139, "grad_norm": 1.1265599727630615, "learning_rate": 2.409791520262357e-05, "loss": 0.1065, "step": 19721 }, { "epoch": 13.859451862262825, "grad_norm": 0.43272164463996887, "learning_rate": 2.409744670883111e-05, "loss": 0.1057, "step": 19722 }, { "epoch": 13.86015460295151, "grad_norm": 0.8232231736183167, "learning_rate": 2.409697821503865e-05, "loss": 0.1563, "step": 19723 }, { "epoch": 13.860857343640197, "grad_norm": 1.1538957357406616, "learning_rate": 2.4096509721246193e-05, "loss": 0.1458, "step": 19724 }, { "epoch": 13.861560084328882, "grad_norm": 0.23221436142921448, "learning_rate": 2.4096041227453737e-05, "loss": 0.0569, "step": 19725 }, { "epoch": 13.862262825017568, "grad_norm": 0.12590713798999786, "learning_rate": 2.4095572733661277e-05, "loss": 0.0198, "step": 19726 }, { "epoch": 13.862965565706254, "grad_norm": 0.09805195033550262, "learning_rate": 2.409510423986882e-05, "loss": 0.0188, "step": 19727 }, { "epoch": 13.86366830639494, "grad_norm": 0.07300663739442825, "learning_rate": 2.4094635746076365e-05, "loss": 0.0097, "step": 19728 }, { "epoch": 13.864371047083626, "grad_norm": 0.10220903158187866, "learning_rate": 2.409416725228391e-05, "loss": 0.013, "step": 19729 }, { "epoch": 13.865073787772312, "grad_norm": 0.16817989945411682, "learning_rate": 2.409369875849145e-05, "loss": 0.0167, "step": 19730 }, { "epoch": 13.865776528460998, "grad_norm": 0.15892662107944489, "learning_rate": 2.4093230264698992e-05, "loss": 0.0109, "step": 19731 }, { "epoch": 13.866479269149684, "grad_norm": 0.11267038434743881, "learning_rate": 2.4092761770906536e-05, "loss": 0.0106, "step": 19732 }, { "epoch": 13.86718200983837, "grad_norm": 0.130309596657753, "learning_rate": 2.409229327711408e-05, "loss": 0.0205, "step": 19733 }, { "epoch": 13.867884750527056, "grad_norm": 0.08154097944498062, "learning_rate": 2.4091824783321624e-05, "loss": 0.0091, "step": 19734 }, { "epoch": 13.868587491215742, "grad_norm": 0.1360715925693512, "learning_rate": 2.4091356289529164e-05, "loss": 0.0241, "step": 19735 }, { "epoch": 13.869290231904428, "grad_norm": 0.13113704323768616, "learning_rate": 2.4090887795736708e-05, "loss": 0.0143, "step": 19736 }, { "epoch": 13.869992972593113, "grad_norm": 0.13216276466846466, "learning_rate": 2.409041930194425e-05, "loss": 0.024, "step": 19737 }, { "epoch": 13.8706957132818, "grad_norm": 0.13953079283237457, "learning_rate": 2.4089950808151795e-05, "loss": 0.0126, "step": 19738 }, { "epoch": 13.871398453970485, "grad_norm": 0.1522992104291916, "learning_rate": 2.4089482314359335e-05, "loss": 0.02, "step": 19739 }, { "epoch": 13.872101194659171, "grad_norm": 0.11097712069749832, "learning_rate": 2.408901382056688e-05, "loss": 0.0192, "step": 19740 }, { "epoch": 13.872803935347857, "grad_norm": 0.1520128697156906, "learning_rate": 2.408854532677442e-05, "loss": 0.0143, "step": 19741 }, { "epoch": 13.873506676036543, "grad_norm": 0.21646076440811157, "learning_rate": 2.4088076832981963e-05, "loss": 0.0314, "step": 19742 }, { "epoch": 13.874209416725229, "grad_norm": 0.21623964607715607, "learning_rate": 2.4087608339189504e-05, "loss": 0.0457, "step": 19743 }, { "epoch": 13.874912157413915, "grad_norm": 0.3158988654613495, "learning_rate": 2.4087139845397047e-05, "loss": 0.0387, "step": 19744 }, { "epoch": 13.8756148981026, "grad_norm": 1.4780957698822021, "learning_rate": 2.408667135160459e-05, "loss": 0.0885, "step": 19745 }, { "epoch": 13.876317638791287, "grad_norm": 0.9558231234550476, "learning_rate": 2.4086202857812135e-05, "loss": 0.1122, "step": 19746 }, { "epoch": 13.877020379479973, "grad_norm": 0.5601719617843628, "learning_rate": 2.408573436401968e-05, "loss": 0.148, "step": 19747 }, { "epoch": 13.877723120168657, "grad_norm": 1.0552115440368652, "learning_rate": 2.408526587022722e-05, "loss": 0.1373, "step": 19748 }, { "epoch": 13.878425860857345, "grad_norm": 2.7489216327667236, "learning_rate": 2.4084797376434763e-05, "loss": 0.1685, "step": 19749 }, { "epoch": 13.879128601546029, "grad_norm": 0.18566328287124634, "learning_rate": 2.4084328882642306e-05, "loss": 0.061, "step": 19750 }, { "epoch": 13.879831342234715, "grad_norm": 0.2019505649805069, "learning_rate": 2.408386038884985e-05, "loss": 0.0359, "step": 19751 }, { "epoch": 13.8805340829234, "grad_norm": 0.276197612285614, "learning_rate": 2.408339189505739e-05, "loss": 0.0169, "step": 19752 }, { "epoch": 13.881236823612086, "grad_norm": 0.1101454421877861, "learning_rate": 2.4082923401264934e-05, "loss": 0.0167, "step": 19753 }, { "epoch": 13.881939564300772, "grad_norm": 0.11026876419782639, "learning_rate": 2.4082454907472478e-05, "loss": 0.0254, "step": 19754 }, { "epoch": 13.882642304989458, "grad_norm": 0.08867249637842178, "learning_rate": 2.408198641368002e-05, "loss": 0.0125, "step": 19755 }, { "epoch": 13.883345045678144, "grad_norm": 0.07576682418584824, "learning_rate": 2.4081517919887562e-05, "loss": 0.0069, "step": 19756 }, { "epoch": 13.88404778636683, "grad_norm": 0.11380297690629959, "learning_rate": 2.4081049426095106e-05, "loss": 0.0203, "step": 19757 }, { "epoch": 13.884750527055516, "grad_norm": 0.1173318400979042, "learning_rate": 2.4080580932302646e-05, "loss": 0.0251, "step": 19758 }, { "epoch": 13.885453267744202, "grad_norm": 0.13073229789733887, "learning_rate": 2.408011243851019e-05, "loss": 0.0114, "step": 19759 }, { "epoch": 13.886156008432888, "grad_norm": 0.12963561713695526, "learning_rate": 2.4079643944717733e-05, "loss": 0.0195, "step": 19760 }, { "epoch": 13.886858749121574, "grad_norm": 0.6946240067481995, "learning_rate": 2.4079175450925274e-05, "loss": 0.0104, "step": 19761 }, { "epoch": 13.88756148981026, "grad_norm": 0.13692443072795868, "learning_rate": 2.4078706957132817e-05, "loss": 0.031, "step": 19762 }, { "epoch": 13.888264230498946, "grad_norm": 0.094776451587677, "learning_rate": 2.407823846334036e-05, "loss": 0.0102, "step": 19763 }, { "epoch": 13.888966971187632, "grad_norm": 0.16888904571533203, "learning_rate": 2.4077769969547905e-05, "loss": 0.0321, "step": 19764 }, { "epoch": 13.889669711876317, "grad_norm": 0.13283227384090424, "learning_rate": 2.4077301475755445e-05, "loss": 0.0202, "step": 19765 }, { "epoch": 13.890372452565003, "grad_norm": 0.1618868112564087, "learning_rate": 2.407683298196299e-05, "loss": 0.0145, "step": 19766 }, { "epoch": 13.89107519325369, "grad_norm": 0.17693690955638885, "learning_rate": 2.4076364488170533e-05, "loss": 0.0396, "step": 19767 }, { "epoch": 13.891777933942375, "grad_norm": 0.14839531481266022, "learning_rate": 2.4075895994378076e-05, "loss": 0.0192, "step": 19768 }, { "epoch": 13.892480674631061, "grad_norm": 0.3643525540828705, "learning_rate": 2.4075427500585617e-05, "loss": 0.0586, "step": 19769 }, { "epoch": 13.893183415319747, "grad_norm": 0.3168383836746216, "learning_rate": 2.407495900679316e-05, "loss": 0.0692, "step": 19770 }, { "epoch": 13.893886156008433, "grad_norm": 0.8059732913970947, "learning_rate": 2.4074490513000704e-05, "loss": 0.099, "step": 19771 }, { "epoch": 13.894588896697119, "grad_norm": 0.4999401271343231, "learning_rate": 2.4074022019208248e-05, "loss": 0.1251, "step": 19772 }, { "epoch": 13.895291637385805, "grad_norm": 0.4984346032142639, "learning_rate": 2.407355352541579e-05, "loss": 0.1811, "step": 19773 }, { "epoch": 13.89599437807449, "grad_norm": 0.6776517629623413, "learning_rate": 2.4073085031623332e-05, "loss": 0.1629, "step": 19774 }, { "epoch": 13.896697118763177, "grad_norm": 0.2202468365430832, "learning_rate": 2.4072616537830876e-05, "loss": 0.0615, "step": 19775 }, { "epoch": 13.897399859451863, "grad_norm": 0.18765129148960114, "learning_rate": 2.4072148044038416e-05, "loss": 0.0264, "step": 19776 }, { "epoch": 13.898102600140549, "grad_norm": 0.0943775326013565, "learning_rate": 2.407167955024596e-05, "loss": 0.0172, "step": 19777 }, { "epoch": 13.898805340829234, "grad_norm": 0.12346626073122025, "learning_rate": 2.40712110564535e-05, "loss": 0.0164, "step": 19778 }, { "epoch": 13.89950808151792, "grad_norm": 0.10434160381555557, "learning_rate": 2.4070742562661044e-05, "loss": 0.0172, "step": 19779 }, { "epoch": 13.900210822206606, "grad_norm": 0.12682975828647614, "learning_rate": 2.4070274068868588e-05, "loss": 0.0153, "step": 19780 }, { "epoch": 13.900913562895292, "grad_norm": 0.213674396276474, "learning_rate": 2.406980557507613e-05, "loss": 0.013, "step": 19781 }, { "epoch": 13.901616303583978, "grad_norm": 0.1319790929555893, "learning_rate": 2.406933708128367e-05, "loss": 0.0235, "step": 19782 }, { "epoch": 13.902319044272664, "grad_norm": 0.10902001708745956, "learning_rate": 2.4068868587491215e-05, "loss": 0.0162, "step": 19783 }, { "epoch": 13.90302178496135, "grad_norm": 0.0881136879324913, "learning_rate": 2.406840009369876e-05, "loss": 0.0087, "step": 19784 }, { "epoch": 13.903724525650036, "grad_norm": 0.2599794268608093, "learning_rate": 2.4067931599906303e-05, "loss": 0.0277, "step": 19785 }, { "epoch": 13.904427266338722, "grad_norm": 0.15497782826423645, "learning_rate": 2.4067463106113846e-05, "loss": 0.0102, "step": 19786 }, { "epoch": 13.905130007027406, "grad_norm": 0.10476499050855637, "learning_rate": 2.4066994612321387e-05, "loss": 0.0184, "step": 19787 }, { "epoch": 13.905832747716094, "grad_norm": 0.23681245744228363, "learning_rate": 2.406652611852893e-05, "loss": 0.0169, "step": 19788 }, { "epoch": 13.906535488404778, "grad_norm": 0.3360663056373596, "learning_rate": 2.4066057624736474e-05, "loss": 0.0231, "step": 19789 }, { "epoch": 13.907238229093464, "grad_norm": 0.16916558146476746, "learning_rate": 2.4065589130944018e-05, "loss": 0.0152, "step": 19790 }, { "epoch": 13.90794096978215, "grad_norm": 0.08062522113323212, "learning_rate": 2.406512063715156e-05, "loss": 0.0137, "step": 19791 }, { "epoch": 13.908643710470836, "grad_norm": 0.1547934114933014, "learning_rate": 2.4064652143359102e-05, "loss": 0.0284, "step": 19792 }, { "epoch": 13.909346451159522, "grad_norm": 0.2372039407491684, "learning_rate": 2.4064183649566642e-05, "loss": 0.0292, "step": 19793 }, { "epoch": 13.910049191848207, "grad_norm": 0.8438183665275574, "learning_rate": 2.4063715155774186e-05, "loss": 0.0428, "step": 19794 }, { "epoch": 13.910751932536893, "grad_norm": 0.19253864884376526, "learning_rate": 2.406324666198173e-05, "loss": 0.0526, "step": 19795 }, { "epoch": 13.91145467322558, "grad_norm": 1.2673399448394775, "learning_rate": 2.406277816818927e-05, "loss": 0.0977, "step": 19796 }, { "epoch": 13.912157413914265, "grad_norm": 1.2296228408813477, "learning_rate": 2.4062309674396814e-05, "loss": 0.1574, "step": 19797 }, { "epoch": 13.912860154602951, "grad_norm": 0.5897495746612549, "learning_rate": 2.4061841180604358e-05, "loss": 0.145, "step": 19798 }, { "epoch": 13.913562895291637, "grad_norm": 0.7565795183181763, "learning_rate": 2.40613726868119e-05, "loss": 0.1511, "step": 19799 }, { "epoch": 13.914265635980323, "grad_norm": 0.2721706032752991, "learning_rate": 2.4060904193019442e-05, "loss": 0.0787, "step": 19800 }, { "epoch": 13.914968376669009, "grad_norm": 0.16386352479457855, "learning_rate": 2.4060435699226985e-05, "loss": 0.039, "step": 19801 }, { "epoch": 13.915671117357695, "grad_norm": 0.18342743813991547, "learning_rate": 2.405996720543453e-05, "loss": 0.0383, "step": 19802 }, { "epoch": 13.91637385804638, "grad_norm": 0.12287687510251999, "learning_rate": 2.4059498711642073e-05, "loss": 0.0125, "step": 19803 }, { "epoch": 13.917076598735067, "grad_norm": 0.16861885786056519, "learning_rate": 2.4059030217849613e-05, "loss": 0.0176, "step": 19804 }, { "epoch": 13.917779339423753, "grad_norm": 0.3312164545059204, "learning_rate": 2.4058561724057157e-05, "loss": 0.0132, "step": 19805 }, { "epoch": 13.918482080112438, "grad_norm": 0.09603225439786911, "learning_rate": 2.40580932302647e-05, "loss": 0.0074, "step": 19806 }, { "epoch": 13.919184820801124, "grad_norm": 0.1774633824825287, "learning_rate": 2.4057624736472244e-05, "loss": 0.0278, "step": 19807 }, { "epoch": 13.91988756148981, "grad_norm": 0.18002009391784668, "learning_rate": 2.4057156242679788e-05, "loss": 0.0191, "step": 19808 }, { "epoch": 13.920590302178496, "grad_norm": 0.13888955116271973, "learning_rate": 2.405668774888733e-05, "loss": 0.0106, "step": 19809 }, { "epoch": 13.921293042867182, "grad_norm": 0.23085977137088776, "learning_rate": 2.405621925509487e-05, "loss": 0.0305, "step": 19810 }, { "epoch": 13.921995783555868, "grad_norm": 0.24734213948249817, "learning_rate": 2.4055750761302413e-05, "loss": 0.0112, "step": 19811 }, { "epoch": 13.922698524244554, "grad_norm": 0.2256254404783249, "learning_rate": 2.4055282267509956e-05, "loss": 0.0287, "step": 19812 }, { "epoch": 13.92340126493324, "grad_norm": 0.2870666980743408, "learning_rate": 2.4054813773717497e-05, "loss": 0.0129, "step": 19813 }, { "epoch": 13.924104005621926, "grad_norm": 0.15212219953536987, "learning_rate": 2.405434527992504e-05, "loss": 0.0221, "step": 19814 }, { "epoch": 13.924806746310612, "grad_norm": 0.37009578943252563, "learning_rate": 2.4053876786132584e-05, "loss": 0.0283, "step": 19815 }, { "epoch": 13.925509486999298, "grad_norm": 0.09052634239196777, "learning_rate": 2.4053408292340128e-05, "loss": 0.0149, "step": 19816 }, { "epoch": 13.926212227687984, "grad_norm": 0.24496479332447052, "learning_rate": 2.4052939798547668e-05, "loss": 0.028, "step": 19817 }, { "epoch": 13.92691496837667, "grad_norm": 0.19741849601268768, "learning_rate": 2.4052471304755212e-05, "loss": 0.0459, "step": 19818 }, { "epoch": 13.927617709065355, "grad_norm": 0.18434903025627136, "learning_rate": 2.4052002810962756e-05, "loss": 0.0356, "step": 19819 }, { "epoch": 13.928320449754041, "grad_norm": 0.3254314959049225, "learning_rate": 2.40515343171703e-05, "loss": 0.0636, "step": 19820 }, { "epoch": 13.929023190442727, "grad_norm": 0.3349728584289551, "learning_rate": 2.4051065823377843e-05, "loss": 0.0904, "step": 19821 }, { "epoch": 13.929725931131413, "grad_norm": 0.5521634817123413, "learning_rate": 2.4050597329585383e-05, "loss": 0.1336, "step": 19822 }, { "epoch": 13.9304286718201, "grad_norm": 0.7704699039459229, "learning_rate": 2.4050128835792927e-05, "loss": 0.148, "step": 19823 }, { "epoch": 13.931131412508785, "grad_norm": 0.9890546202659607, "learning_rate": 2.404966034200047e-05, "loss": 0.156, "step": 19824 }, { "epoch": 13.931834153197471, "grad_norm": 0.16912366449832916, "learning_rate": 2.4049191848208014e-05, "loss": 0.0544, "step": 19825 }, { "epoch": 13.932536893886155, "grad_norm": 0.10864580422639847, "learning_rate": 2.4048723354415555e-05, "loss": 0.019, "step": 19826 }, { "epoch": 13.933239634574843, "grad_norm": 0.2789142429828644, "learning_rate": 2.40482548606231e-05, "loss": 0.0287, "step": 19827 }, { "epoch": 13.933942375263527, "grad_norm": 0.0687095895409584, "learning_rate": 2.404778636683064e-05, "loss": 0.0159, "step": 19828 }, { "epoch": 13.934645115952213, "grad_norm": 0.13934525847434998, "learning_rate": 2.4047317873038183e-05, "loss": 0.0158, "step": 19829 }, { "epoch": 13.935347856640899, "grad_norm": 0.1019878089427948, "learning_rate": 2.4046849379245723e-05, "loss": 0.0126, "step": 19830 }, { "epoch": 13.936050597329585, "grad_norm": 0.226989284157753, "learning_rate": 2.4046380885453267e-05, "loss": 0.0278, "step": 19831 }, { "epoch": 13.93675333801827, "grad_norm": 0.12594686448574066, "learning_rate": 2.404591239166081e-05, "loss": 0.0115, "step": 19832 }, { "epoch": 13.937456078706957, "grad_norm": 0.11077636480331421, "learning_rate": 2.4045443897868354e-05, "loss": 0.0155, "step": 19833 }, { "epoch": 13.938158819395642, "grad_norm": 0.06148171052336693, "learning_rate": 2.4044975404075898e-05, "loss": 0.0069, "step": 19834 }, { "epoch": 13.938861560084328, "grad_norm": 0.26567497849464417, "learning_rate": 2.4044506910283438e-05, "loss": 0.0158, "step": 19835 }, { "epoch": 13.939564300773014, "grad_norm": 0.1181068867444992, "learning_rate": 2.4044038416490982e-05, "loss": 0.0074, "step": 19836 }, { "epoch": 13.9402670414617, "grad_norm": 0.17978225648403168, "learning_rate": 2.4043569922698526e-05, "loss": 0.0235, "step": 19837 }, { "epoch": 13.940969782150386, "grad_norm": 0.18991146981716156, "learning_rate": 2.404310142890607e-05, "loss": 0.0128, "step": 19838 }, { "epoch": 13.941672522839072, "grad_norm": 0.19052530825138092, "learning_rate": 2.404263293511361e-05, "loss": 0.0182, "step": 19839 }, { "epoch": 13.942375263527758, "grad_norm": 0.1376393586397171, "learning_rate": 2.4042164441321153e-05, "loss": 0.0206, "step": 19840 }, { "epoch": 13.943078004216444, "grad_norm": 0.06577260047197342, "learning_rate": 2.4041695947528697e-05, "loss": 0.0118, "step": 19841 }, { "epoch": 13.94378074490513, "grad_norm": 0.2864771783351898, "learning_rate": 2.404122745373624e-05, "loss": 0.0331, "step": 19842 }, { "epoch": 13.944483485593816, "grad_norm": 0.14694952964782715, "learning_rate": 2.404075895994378e-05, "loss": 0.0321, "step": 19843 }, { "epoch": 13.945186226282502, "grad_norm": 0.23228062689304352, "learning_rate": 2.4040290466151325e-05, "loss": 0.0601, "step": 19844 }, { "epoch": 13.945888966971188, "grad_norm": 0.3481791317462921, "learning_rate": 2.4039821972358865e-05, "loss": 0.0622, "step": 19845 }, { "epoch": 13.946591707659874, "grad_norm": 0.9388554692268372, "learning_rate": 2.403935347856641e-05, "loss": 0.0977, "step": 19846 }, { "epoch": 13.94729444834856, "grad_norm": 0.3435627818107605, "learning_rate": 2.4038884984773953e-05, "loss": 0.1061, "step": 19847 }, { "epoch": 13.947997189037245, "grad_norm": 0.5698347091674805, "learning_rate": 2.4038416490981493e-05, "loss": 0.1461, "step": 19848 }, { "epoch": 13.948699929725931, "grad_norm": 1.1564123630523682, "learning_rate": 2.4037947997189037e-05, "loss": 0.176, "step": 19849 }, { "epoch": 13.949402670414617, "grad_norm": 0.2260427474975586, "learning_rate": 2.403747950339658e-05, "loss": 0.0653, "step": 19850 }, { "epoch": 13.950105411103303, "grad_norm": 0.2690390944480896, "learning_rate": 2.4037011009604124e-05, "loss": 0.0427, "step": 19851 }, { "epoch": 13.950808151791989, "grad_norm": 0.1829919070005417, "learning_rate": 2.4036542515811665e-05, "loss": 0.0221, "step": 19852 }, { "epoch": 13.951510892480675, "grad_norm": 0.24734118580818176, "learning_rate": 2.403607402201921e-05, "loss": 0.0282, "step": 19853 }, { "epoch": 13.952213633169361, "grad_norm": 0.2721351087093353, "learning_rate": 2.4035605528226752e-05, "loss": 0.0115, "step": 19854 }, { "epoch": 13.952916373858047, "grad_norm": 0.08885771036148071, "learning_rate": 2.4035137034434296e-05, "loss": 0.0161, "step": 19855 }, { "epoch": 13.953619114546733, "grad_norm": 0.07603408396244049, "learning_rate": 2.4034668540641836e-05, "loss": 0.0104, "step": 19856 }, { "epoch": 13.954321855235419, "grad_norm": 0.11450011283159256, "learning_rate": 2.403420004684938e-05, "loss": 0.0217, "step": 19857 }, { "epoch": 13.955024595924105, "grad_norm": 0.061047859489917755, "learning_rate": 2.4033731553056924e-05, "loss": 0.0127, "step": 19858 }, { "epoch": 13.95572733661279, "grad_norm": 0.22135761380195618, "learning_rate": 2.4033263059264467e-05, "loss": 0.0122, "step": 19859 }, { "epoch": 13.956430077301476, "grad_norm": 0.11516232043504715, "learning_rate": 2.403279456547201e-05, "loss": 0.0173, "step": 19860 }, { "epoch": 13.957132817990162, "grad_norm": 0.08120307326316833, "learning_rate": 2.403232607167955e-05, "loss": 0.0134, "step": 19861 }, { "epoch": 13.957835558678848, "grad_norm": 0.34969520568847656, "learning_rate": 2.4031857577887095e-05, "loss": 0.0269, "step": 19862 }, { "epoch": 13.958538299367534, "grad_norm": 0.12239591777324677, "learning_rate": 2.4031389084094635e-05, "loss": 0.0141, "step": 19863 }, { "epoch": 13.95924104005622, "grad_norm": 0.08779245615005493, "learning_rate": 2.403092059030218e-05, "loss": 0.0164, "step": 19864 }, { "epoch": 13.959943780744904, "grad_norm": 0.24649621546268463, "learning_rate": 2.403045209650972e-05, "loss": 0.0295, "step": 19865 }, { "epoch": 13.96064652143359, "grad_norm": 0.17164403200149536, "learning_rate": 2.4029983602717263e-05, "loss": 0.0189, "step": 19866 }, { "epoch": 13.961349262122276, "grad_norm": 0.2836344242095947, "learning_rate": 2.4029515108924807e-05, "loss": 0.028, "step": 19867 }, { "epoch": 13.962052002810962, "grad_norm": 0.16256900131702423, "learning_rate": 2.402904661513235e-05, "loss": 0.035, "step": 19868 }, { "epoch": 13.962754743499648, "grad_norm": 0.7390093803405762, "learning_rate": 2.402857812133989e-05, "loss": 0.0392, "step": 19869 }, { "epoch": 13.963457484188334, "grad_norm": 0.5884645581245422, "learning_rate": 2.4028109627547435e-05, "loss": 0.0628, "step": 19870 }, { "epoch": 13.96416022487702, "grad_norm": 8.625732421875, "learning_rate": 2.402764113375498e-05, "loss": 0.0888, "step": 19871 }, { "epoch": 13.964862965565706, "grad_norm": 0.5974254012107849, "learning_rate": 2.4027172639962522e-05, "loss": 0.1347, "step": 19872 }, { "epoch": 13.965565706254392, "grad_norm": 1.217165231704712, "learning_rate": 2.4026704146170066e-05, "loss": 0.1576, "step": 19873 }, { "epoch": 13.966268446943078, "grad_norm": 1.4361088275909424, "learning_rate": 2.4026235652377606e-05, "loss": 0.1747, "step": 19874 }, { "epoch": 13.966971187631763, "grad_norm": 0.2451496720314026, "learning_rate": 2.402576715858515e-05, "loss": 0.0666, "step": 19875 }, { "epoch": 13.96767392832045, "grad_norm": 0.18082088232040405, "learning_rate": 2.4025298664792694e-05, "loss": 0.033, "step": 19876 }, { "epoch": 13.968376669009135, "grad_norm": 0.15089556574821472, "learning_rate": 2.4024830171000237e-05, "loss": 0.0215, "step": 19877 }, { "epoch": 13.969079409697821, "grad_norm": 0.11350042372941971, "learning_rate": 2.4024361677207778e-05, "loss": 0.0157, "step": 19878 }, { "epoch": 13.969782150386507, "grad_norm": 0.09362884610891342, "learning_rate": 2.402389318341532e-05, "loss": 0.013, "step": 19879 }, { "epoch": 13.970484891075193, "grad_norm": 0.7987640500068665, "learning_rate": 2.4023424689622862e-05, "loss": 0.0098, "step": 19880 }, { "epoch": 13.971187631763879, "grad_norm": 0.14082182943820953, "learning_rate": 2.4022956195830406e-05, "loss": 0.0068, "step": 19881 }, { "epoch": 13.971890372452565, "grad_norm": 0.2700801491737366, "learning_rate": 2.4022487702037946e-05, "loss": 0.0151, "step": 19882 }, { "epoch": 13.97259311314125, "grad_norm": 0.21620389819145203, "learning_rate": 2.402201920824549e-05, "loss": 0.0266, "step": 19883 }, { "epoch": 13.973295853829937, "grad_norm": 0.07879181951284409, "learning_rate": 2.4021550714453033e-05, "loss": 0.0061, "step": 19884 }, { "epoch": 13.973998594518623, "grad_norm": 0.10143020749092102, "learning_rate": 2.4021082220660577e-05, "loss": 0.0236, "step": 19885 }, { "epoch": 13.974701335207309, "grad_norm": 0.13291601836681366, "learning_rate": 2.402061372686812e-05, "loss": 0.0172, "step": 19886 }, { "epoch": 13.975404075895995, "grad_norm": 0.13253161311149597, "learning_rate": 2.402014523307566e-05, "loss": 0.0248, "step": 19887 }, { "epoch": 13.97610681658468, "grad_norm": 0.10178674757480621, "learning_rate": 2.4019676739283205e-05, "loss": 0.0292, "step": 19888 }, { "epoch": 13.976809557273366, "grad_norm": 0.17791177332401276, "learning_rate": 2.401920824549075e-05, "loss": 0.0243, "step": 19889 }, { "epoch": 13.977512297962052, "grad_norm": 0.3457508087158203, "learning_rate": 2.4018739751698292e-05, "loss": 0.0301, "step": 19890 }, { "epoch": 13.978215038650738, "grad_norm": 0.1673051416873932, "learning_rate": 2.4018271257905833e-05, "loss": 0.0539, "step": 19891 }, { "epoch": 13.978917779339424, "grad_norm": 0.2271781861782074, "learning_rate": 2.4017802764113376e-05, "loss": 0.0166, "step": 19892 }, { "epoch": 13.97962052002811, "grad_norm": 0.19062870740890503, "learning_rate": 2.401733427032092e-05, "loss": 0.0491, "step": 19893 }, { "epoch": 13.980323260716796, "grad_norm": 0.21594633162021637, "learning_rate": 2.4016865776528464e-05, "loss": 0.0395, "step": 19894 }, { "epoch": 13.981026001405482, "grad_norm": 0.30503636598587036, "learning_rate": 2.4016397282736004e-05, "loss": 0.0545, "step": 19895 }, { "epoch": 13.981728742094168, "grad_norm": 0.4035845696926117, "learning_rate": 2.4015928788943548e-05, "loss": 0.1192, "step": 19896 }, { "epoch": 13.982431482782854, "grad_norm": 0.42423152923583984, "learning_rate": 2.4015460295151088e-05, "loss": 0.1213, "step": 19897 }, { "epoch": 13.98313422347154, "grad_norm": 0.5322735905647278, "learning_rate": 2.4014991801358632e-05, "loss": 0.1734, "step": 19898 }, { "epoch": 13.983836964160226, "grad_norm": 1.8184218406677246, "learning_rate": 2.4014523307566176e-05, "loss": 0.2053, "step": 19899 }, { "epoch": 13.984539704848912, "grad_norm": 0.1912672370672226, "learning_rate": 2.4014054813773716e-05, "loss": 0.0633, "step": 19900 }, { "epoch": 13.985242445537597, "grad_norm": 0.3043409585952759, "learning_rate": 2.401358631998126e-05, "loss": 0.02, "step": 19901 }, { "epoch": 13.985945186226282, "grad_norm": 0.08368702977895737, "learning_rate": 2.4013117826188803e-05, "loss": 0.015, "step": 19902 }, { "epoch": 13.98664792691497, "grad_norm": 0.3870909512042999, "learning_rate": 2.4012649332396347e-05, "loss": 0.0239, "step": 19903 }, { "epoch": 13.987350667603653, "grad_norm": 0.09180998057126999, "learning_rate": 2.4012180838603887e-05, "loss": 0.0152, "step": 19904 }, { "epoch": 13.98805340829234, "grad_norm": 0.13847436010837555, "learning_rate": 2.401171234481143e-05, "loss": 0.0187, "step": 19905 }, { "epoch": 13.988756148981025, "grad_norm": 0.1094093844294548, "learning_rate": 2.4011243851018975e-05, "loss": 0.0224, "step": 19906 }, { "epoch": 13.989458889669711, "grad_norm": 0.11056600511074066, "learning_rate": 2.401077535722652e-05, "loss": 0.02, "step": 19907 }, { "epoch": 13.990161630358397, "grad_norm": 0.14824740588665009, "learning_rate": 2.401030686343406e-05, "loss": 0.0165, "step": 19908 }, { "epoch": 13.990864371047083, "grad_norm": 0.09154009073972702, "learning_rate": 2.4009838369641603e-05, "loss": 0.0129, "step": 19909 }, { "epoch": 13.991567111735769, "grad_norm": 0.24122354388237, "learning_rate": 2.4009369875849146e-05, "loss": 0.021, "step": 19910 }, { "epoch": 13.992269852424455, "grad_norm": 0.2674899399280548, "learning_rate": 2.400890138205669e-05, "loss": 0.019, "step": 19911 }, { "epoch": 13.99297259311314, "grad_norm": 0.1437475085258484, "learning_rate": 2.4008432888264234e-05, "loss": 0.0138, "step": 19912 }, { "epoch": 13.993675333801827, "grad_norm": 0.24527642130851746, "learning_rate": 2.4007964394471774e-05, "loss": 0.0337, "step": 19913 }, { "epoch": 13.994378074490513, "grad_norm": 0.09022224694490433, "learning_rate": 2.4007495900679318e-05, "loss": 0.0097, "step": 19914 }, { "epoch": 13.995080815179199, "grad_norm": 0.17968055605888367, "learning_rate": 2.4007027406886858e-05, "loss": 0.0305, "step": 19915 }, { "epoch": 13.995783555867884, "grad_norm": 0.1987980604171753, "learning_rate": 2.4006558913094402e-05, "loss": 0.0438, "step": 19916 }, { "epoch": 13.99648629655657, "grad_norm": 0.27076879143714905, "learning_rate": 2.4006090419301942e-05, "loss": 0.0433, "step": 19917 }, { "epoch": 13.997189037245256, "grad_norm": 0.2674843966960907, "learning_rate": 2.4005621925509486e-05, "loss": 0.0531, "step": 19918 }, { "epoch": 13.997891777933942, "grad_norm": 0.39347612857818604, "learning_rate": 2.400515343171703e-05, "loss": 0.1497, "step": 19919 }, { "epoch": 13.998594518622628, "grad_norm": 0.4615651071071625, "learning_rate": 2.4004684937924574e-05, "loss": 0.1513, "step": 19920 }, { "epoch": 13.999297259311314, "grad_norm": 0.62629234790802, "learning_rate": 2.4004216444132114e-05, "loss": 0.1396, "step": 19921 }, { "epoch": 14.0, "grad_norm": 3.086205005645752, "learning_rate": 2.4003747950339658e-05, "loss": 0.1321, "step": 19922 }, { "epoch": 14.000702740688686, "grad_norm": 0.18236485123634338, "learning_rate": 2.40032794565472e-05, "loss": 0.0592, "step": 19923 }, { "epoch": 14.001405481377372, "grad_norm": 0.16246436536312103, "learning_rate": 2.4002810962754745e-05, "loss": 0.0204, "step": 19924 }, { "epoch": 14.002108222066058, "grad_norm": 0.12174730747938156, "learning_rate": 2.400234246896229e-05, "loss": 0.022, "step": 19925 }, { "epoch": 14.002810962754744, "grad_norm": 0.18843427300453186, "learning_rate": 2.400187397516983e-05, "loss": 0.0181, "step": 19926 }, { "epoch": 14.00351370344343, "grad_norm": 0.06526410579681396, "learning_rate": 2.4001405481377373e-05, "loss": 0.0123, "step": 19927 }, { "epoch": 14.004216444132116, "grad_norm": 0.07745984196662903, "learning_rate": 2.4000936987584917e-05, "loss": 0.0151, "step": 19928 }, { "epoch": 14.004919184820801, "grad_norm": 0.10203225910663605, "learning_rate": 2.400046849379246e-05, "loss": 0.0164, "step": 19929 }, { "epoch": 14.005621925509487, "grad_norm": 0.09586460888385773, "learning_rate": 2.4e-05, "loss": 0.0189, "step": 19930 }, { "epoch": 14.006324666198173, "grad_norm": 0.07712601125240326, "learning_rate": 2.3999531506207544e-05, "loss": 0.0157, "step": 19931 }, { "epoch": 14.00702740688686, "grad_norm": 0.091753751039505, "learning_rate": 2.3999063012415085e-05, "loss": 0.0104, "step": 19932 }, { "epoch": 14.007730147575545, "grad_norm": 0.09431610256433487, "learning_rate": 2.399859451862263e-05, "loss": 0.0151, "step": 19933 }, { "epoch": 14.008432888264231, "grad_norm": 0.101099893450737, "learning_rate": 2.399812602483017e-05, "loss": 0.0093, "step": 19934 }, { "epoch": 14.009135628952917, "grad_norm": 0.11783977597951889, "learning_rate": 2.3997657531037712e-05, "loss": 0.0208, "step": 19935 }, { "epoch": 14.009838369641603, "grad_norm": 0.06393953412771225, "learning_rate": 2.3997189037245256e-05, "loss": 0.0081, "step": 19936 }, { "epoch": 14.010541110330289, "grad_norm": 0.11454515904188156, "learning_rate": 2.39967205434528e-05, "loss": 0.0286, "step": 19937 }, { "epoch": 14.011243851018975, "grad_norm": 0.7987282276153564, "learning_rate": 2.3996252049660344e-05, "loss": 0.0225, "step": 19938 }, { "epoch": 14.01194659170766, "grad_norm": 0.15735934674739838, "learning_rate": 2.3995783555867884e-05, "loss": 0.0353, "step": 19939 }, { "epoch": 14.012649332396347, "grad_norm": 0.17402422428131104, "learning_rate": 2.3995315062075428e-05, "loss": 0.0231, "step": 19940 }, { "epoch": 14.013352073085033, "grad_norm": 0.20690570771694183, "learning_rate": 2.399484656828297e-05, "loss": 0.0387, "step": 19941 }, { "epoch": 14.014054813773717, "grad_norm": 0.4144883453845978, "learning_rate": 2.3994378074490515e-05, "loss": 0.0383, "step": 19942 }, { "epoch": 14.014757554462403, "grad_norm": 0.2272852510213852, "learning_rate": 2.3993909580698055e-05, "loss": 0.0566, "step": 19943 }, { "epoch": 14.015460295151088, "grad_norm": 0.28130584955215454, "learning_rate": 2.39934410869056e-05, "loss": 0.0873, "step": 19944 }, { "epoch": 14.016163035839774, "grad_norm": 0.5350891947746277, "learning_rate": 2.3992972593113143e-05, "loss": 0.1253, "step": 19945 }, { "epoch": 14.01686577652846, "grad_norm": 0.6598955392837524, "learning_rate": 2.3992504099320687e-05, "loss": 0.146, "step": 19946 }, { "epoch": 14.017568517217146, "grad_norm": 1.2549664974212646, "learning_rate": 2.3992035605528227e-05, "loss": 0.1694, "step": 19947 }, { "epoch": 14.018271257905832, "grad_norm": 0.2271174043416977, "learning_rate": 2.399156711173577e-05, "loss": 0.0561, "step": 19948 }, { "epoch": 14.018973998594518, "grad_norm": 0.13529673218727112, "learning_rate": 2.3991098617943314e-05, "loss": 0.0218, "step": 19949 }, { "epoch": 14.019676739283204, "grad_norm": 0.11020512878894806, "learning_rate": 2.3990630124150855e-05, "loss": 0.0119, "step": 19950 }, { "epoch": 14.02037947997189, "grad_norm": 0.05964279547333717, "learning_rate": 2.39901616303584e-05, "loss": 0.0121, "step": 19951 }, { "epoch": 14.021082220660576, "grad_norm": 0.08201934397220612, "learning_rate": 2.398969313656594e-05, "loss": 0.0142, "step": 19952 }, { "epoch": 14.021784961349262, "grad_norm": 0.08004949986934662, "learning_rate": 2.3989224642773483e-05, "loss": 0.0115, "step": 19953 }, { "epoch": 14.022487702037948, "grad_norm": 0.15532203018665314, "learning_rate": 2.3988756148981026e-05, "loss": 0.0088, "step": 19954 }, { "epoch": 14.023190442726634, "grad_norm": 0.10688994824886322, "learning_rate": 2.398828765518857e-05, "loss": 0.011, "step": 19955 }, { "epoch": 14.02389318341532, "grad_norm": 0.15516498684883118, "learning_rate": 2.398781916139611e-05, "loss": 0.0225, "step": 19956 }, { "epoch": 14.024595924104005, "grad_norm": 0.10031995922327042, "learning_rate": 2.3987350667603654e-05, "loss": 0.0132, "step": 19957 }, { "epoch": 14.025298664792691, "grad_norm": 0.13512052595615387, "learning_rate": 2.3986882173811198e-05, "loss": 0.0171, "step": 19958 }, { "epoch": 14.026001405481377, "grad_norm": 0.209456667304039, "learning_rate": 2.398641368001874e-05, "loss": 0.0105, "step": 19959 }, { "epoch": 14.026704146170063, "grad_norm": 0.20149116218090057, "learning_rate": 2.3985945186226282e-05, "loss": 0.0329, "step": 19960 }, { "epoch": 14.02740688685875, "grad_norm": 0.15169553458690643, "learning_rate": 2.3985476692433826e-05, "loss": 0.0132, "step": 19961 }, { "epoch": 14.028109627547435, "grad_norm": 0.10810372978448868, "learning_rate": 2.398500819864137e-05, "loss": 0.0209, "step": 19962 }, { "epoch": 14.028812368236121, "grad_norm": 0.15579599142074585, "learning_rate": 2.3984539704848913e-05, "loss": 0.038, "step": 19963 }, { "epoch": 14.029515108924807, "grad_norm": 0.1311326026916504, "learning_rate": 2.3984071211056457e-05, "loss": 0.0168, "step": 19964 }, { "epoch": 14.030217849613493, "grad_norm": 0.16121436655521393, "learning_rate": 2.3983602717263997e-05, "loss": 0.0236, "step": 19965 }, { "epoch": 14.030920590302179, "grad_norm": 0.24413084983825684, "learning_rate": 2.398313422347154e-05, "loss": 0.0332, "step": 19966 }, { "epoch": 14.031623330990865, "grad_norm": 0.1453397423028946, "learning_rate": 2.398266572967908e-05, "loss": 0.0302, "step": 19967 }, { "epoch": 14.03232607167955, "grad_norm": 0.24069809913635254, "learning_rate": 2.3982197235886625e-05, "loss": 0.0617, "step": 19968 }, { "epoch": 14.033028812368237, "grad_norm": 0.4162367284297943, "learning_rate": 2.3981728742094165e-05, "loss": 0.0969, "step": 19969 }, { "epoch": 14.033731553056922, "grad_norm": 0.42211753129959106, "learning_rate": 2.398126024830171e-05, "loss": 0.125, "step": 19970 }, { "epoch": 14.034434293745608, "grad_norm": 0.8630192279815674, "learning_rate": 2.3980791754509253e-05, "loss": 0.1289, "step": 19971 }, { "epoch": 14.035137034434294, "grad_norm": 0.7296954989433289, "learning_rate": 2.3980323260716796e-05, "loss": 0.1733, "step": 19972 }, { "epoch": 14.03583977512298, "grad_norm": 0.3365560472011566, "learning_rate": 2.3979854766924337e-05, "loss": 0.0418, "step": 19973 }, { "epoch": 14.036542515811666, "grad_norm": 0.16379952430725098, "learning_rate": 2.397938627313188e-05, "loss": 0.0166, "step": 19974 }, { "epoch": 14.037245256500352, "grad_norm": 0.08939728885889053, "learning_rate": 2.3978917779339424e-05, "loss": 0.0128, "step": 19975 }, { "epoch": 14.037947997189038, "grad_norm": 0.1028406172990799, "learning_rate": 2.3978449285546968e-05, "loss": 0.0179, "step": 19976 }, { "epoch": 14.038650737877724, "grad_norm": 0.10926453024148941, "learning_rate": 2.397798079175451e-05, "loss": 0.0231, "step": 19977 }, { "epoch": 14.03935347856641, "grad_norm": 0.09861686080694199, "learning_rate": 2.3977512297962052e-05, "loss": 0.0161, "step": 19978 }, { "epoch": 14.040056219255096, "grad_norm": 0.3304503262042999, "learning_rate": 2.3977043804169596e-05, "loss": 0.0125, "step": 19979 }, { "epoch": 14.04075895994378, "grad_norm": 0.15020851790905, "learning_rate": 2.397657531037714e-05, "loss": 0.0224, "step": 19980 }, { "epoch": 14.041461700632466, "grad_norm": 0.10585269331932068, "learning_rate": 2.3976106816584683e-05, "loss": 0.0193, "step": 19981 }, { "epoch": 14.042164441321152, "grad_norm": 0.08538749068975449, "learning_rate": 2.3975638322792223e-05, "loss": 0.0116, "step": 19982 }, { "epoch": 14.042867182009838, "grad_norm": 0.15745003521442413, "learning_rate": 2.3975169828999767e-05, "loss": 0.0212, "step": 19983 }, { "epoch": 14.043569922698524, "grad_norm": 0.09593665599822998, "learning_rate": 2.3974701335207308e-05, "loss": 0.0162, "step": 19984 }, { "epoch": 14.04427266338721, "grad_norm": 0.1593731790781021, "learning_rate": 2.397423284141485e-05, "loss": 0.016, "step": 19985 }, { "epoch": 14.044975404075895, "grad_norm": 0.06264904141426086, "learning_rate": 2.3973764347622395e-05, "loss": 0.0069, "step": 19986 }, { "epoch": 14.045678144764581, "grad_norm": 0.3383054733276367, "learning_rate": 2.3973295853829935e-05, "loss": 0.0333, "step": 19987 }, { "epoch": 14.046380885453267, "grad_norm": 1.269317626953125, "learning_rate": 2.397282736003748e-05, "loss": 0.0167, "step": 19988 }, { "epoch": 14.047083626141953, "grad_norm": 0.26891833543777466, "learning_rate": 2.3972358866245023e-05, "loss": 0.02, "step": 19989 }, { "epoch": 14.047786366830639, "grad_norm": 0.16315102577209473, "learning_rate": 2.3971890372452567e-05, "loss": 0.0265, "step": 19990 }, { "epoch": 14.048489107519325, "grad_norm": 0.1716143786907196, "learning_rate": 2.3971421878660107e-05, "loss": 0.039, "step": 19991 }, { "epoch": 14.049191848208011, "grad_norm": 0.1396399438381195, "learning_rate": 2.397095338486765e-05, "loss": 0.0342, "step": 19992 }, { "epoch": 14.049894588896697, "grad_norm": 0.30019208788871765, "learning_rate": 2.3970484891075194e-05, "loss": 0.0601, "step": 19993 }, { "epoch": 14.050597329585383, "grad_norm": 0.4595549404621124, "learning_rate": 2.3970016397282738e-05, "loss": 0.0948, "step": 19994 }, { "epoch": 14.051300070274069, "grad_norm": 0.4574408233165741, "learning_rate": 2.396954790349028e-05, "loss": 0.116, "step": 19995 }, { "epoch": 14.052002810962755, "grad_norm": 0.5171664953231812, "learning_rate": 2.3969079409697822e-05, "loss": 0.1179, "step": 19996 }, { "epoch": 14.05270555165144, "grad_norm": 1.3741800785064697, "learning_rate": 2.3968610915905366e-05, "loss": 0.1383, "step": 19997 }, { "epoch": 14.053408292340126, "grad_norm": 0.1580260843038559, "learning_rate": 2.396814242211291e-05, "loss": 0.05, "step": 19998 }, { "epoch": 14.054111033028812, "grad_norm": 0.16012364625930786, "learning_rate": 2.396767392832045e-05, "loss": 0.0179, "step": 19999 }, { "epoch": 14.054813773717498, "grad_norm": 0.13201993703842163, "learning_rate": 2.3967205434527994e-05, "loss": 0.0147, "step": 20000 }, { "epoch": 14.054813773717498, "eval_cer": 0.19333791583472446, "eval_loss": 0.2781136631965637, "eval_runtime": 18.5945, "eval_samples_per_second": 244.051, "eval_steps_per_second": 0.807, "eval_wer": 0.3450182955783358, "step": 20000 }, { "epoch": 14.055516514406184, "grad_norm": 0.16789866983890533, "learning_rate": 2.3966736940735537e-05, "loss": 0.0216, "step": 20001 }, { "epoch": 14.05621925509487, "grad_norm": 0.10369962453842163, "learning_rate": 2.3966268446943078e-05, "loss": 0.0136, "step": 20002 }, { "epoch": 14.056921995783556, "grad_norm": 0.6630900502204895, "learning_rate": 2.396579995315062e-05, "loss": 0.0101, "step": 20003 }, { "epoch": 14.057624736472242, "grad_norm": 0.2310725450515747, "learning_rate": 2.3965331459358162e-05, "loss": 0.0175, "step": 20004 }, { "epoch": 14.058327477160928, "grad_norm": 0.16991026699543, "learning_rate": 2.3964862965565705e-05, "loss": 0.0262, "step": 20005 }, { "epoch": 14.059030217849614, "grad_norm": 0.12422703206539154, "learning_rate": 2.396439447177325e-05, "loss": 0.0126, "step": 20006 }, { "epoch": 14.0597329585383, "grad_norm": 0.10454226285219193, "learning_rate": 2.3963925977980793e-05, "loss": 0.0129, "step": 20007 }, { "epoch": 14.060435699226986, "grad_norm": 0.18917113542556763, "learning_rate": 2.3963457484188333e-05, "loss": 0.0183, "step": 20008 }, { "epoch": 14.061138439915672, "grad_norm": 0.07449109107255936, "learning_rate": 2.3962988990395877e-05, "loss": 0.01, "step": 20009 }, { "epoch": 14.061841180604358, "grad_norm": 0.09478110074996948, "learning_rate": 2.396252049660342e-05, "loss": 0.0175, "step": 20010 }, { "epoch": 14.062543921293043, "grad_norm": 0.09764118492603302, "learning_rate": 2.3962052002810964e-05, "loss": 0.0099, "step": 20011 }, { "epoch": 14.06324666198173, "grad_norm": 0.1548468917608261, "learning_rate": 2.3961583509018508e-05, "loss": 0.019, "step": 20012 }, { "epoch": 14.063949402670415, "grad_norm": 0.3709402084350586, "learning_rate": 2.396111501522605e-05, "loss": 0.0312, "step": 20013 }, { "epoch": 14.064652143359101, "grad_norm": 0.08030660450458527, "learning_rate": 2.3960646521433592e-05, "loss": 0.0088, "step": 20014 }, { "epoch": 14.065354884047787, "grad_norm": 0.47159841656684875, "learning_rate": 2.3960178027641136e-05, "loss": 0.0385, "step": 20015 }, { "epoch": 14.066057624736473, "grad_norm": 0.2294704169034958, "learning_rate": 2.395970953384868e-05, "loss": 0.0343, "step": 20016 }, { "epoch": 14.066760365425159, "grad_norm": 0.20708423852920532, "learning_rate": 2.395924104005622e-05, "loss": 0.0481, "step": 20017 }, { "epoch": 14.067463106113845, "grad_norm": 0.344588965177536, "learning_rate": 2.3958772546263764e-05, "loss": 0.0554, "step": 20018 }, { "epoch": 14.068165846802529, "grad_norm": 0.6589415073394775, "learning_rate": 2.3958304052471304e-05, "loss": 0.0939, "step": 20019 }, { "epoch": 14.068868587491215, "grad_norm": 0.4416063129901886, "learning_rate": 2.3957835558678848e-05, "loss": 0.119, "step": 20020 }, { "epoch": 14.0695713281799, "grad_norm": 0.57834392786026, "learning_rate": 2.3957367064886388e-05, "loss": 0.1409, "step": 20021 }, { "epoch": 14.070274068868587, "grad_norm": 0.7962194085121155, "learning_rate": 2.3956898571093932e-05, "loss": 0.1703, "step": 20022 }, { "epoch": 14.070976809557273, "grad_norm": 0.19775277376174927, "learning_rate": 2.3956430077301476e-05, "loss": 0.0631, "step": 20023 }, { "epoch": 14.071679550245959, "grad_norm": 0.09214898943901062, "learning_rate": 2.395596158350902e-05, "loss": 0.0209, "step": 20024 }, { "epoch": 14.072382290934645, "grad_norm": 0.06372988969087601, "learning_rate": 2.3955493089716563e-05, "loss": 0.0119, "step": 20025 }, { "epoch": 14.07308503162333, "grad_norm": 0.31099340319633484, "learning_rate": 2.3955024595924103e-05, "loss": 0.0147, "step": 20026 }, { "epoch": 14.073787772312016, "grad_norm": 0.11427249014377594, "learning_rate": 2.3954556102131647e-05, "loss": 0.0109, "step": 20027 }, { "epoch": 14.074490513000702, "grad_norm": 0.14174242317676544, "learning_rate": 2.395408760833919e-05, "loss": 0.0131, "step": 20028 }, { "epoch": 14.075193253689388, "grad_norm": 0.15180468559265137, "learning_rate": 2.3953619114546735e-05, "loss": 0.0168, "step": 20029 }, { "epoch": 14.075895994378074, "grad_norm": 0.0793839618563652, "learning_rate": 2.3953150620754275e-05, "loss": 0.006, "step": 20030 }, { "epoch": 14.07659873506676, "grad_norm": 0.16155967116355896, "learning_rate": 2.395268212696182e-05, "loss": 0.018, "step": 20031 }, { "epoch": 14.077301475755446, "grad_norm": 0.1214975044131279, "learning_rate": 2.3952213633169362e-05, "loss": 0.0102, "step": 20032 }, { "epoch": 14.078004216444132, "grad_norm": 0.1715998351573944, "learning_rate": 2.3951745139376906e-05, "loss": 0.0149, "step": 20033 }, { "epoch": 14.078706957132818, "grad_norm": 0.10342246294021606, "learning_rate": 2.3951276645584446e-05, "loss": 0.0122, "step": 20034 }, { "epoch": 14.079409697821504, "grad_norm": 0.12016598880290985, "learning_rate": 2.395080815179199e-05, "loss": 0.0227, "step": 20035 }, { "epoch": 14.08011243851019, "grad_norm": 0.07518216967582703, "learning_rate": 2.3950339657999534e-05, "loss": 0.011, "step": 20036 }, { "epoch": 14.080815179198876, "grad_norm": 0.4318265914916992, "learning_rate": 2.3949871164207074e-05, "loss": 0.0332, "step": 20037 }, { "epoch": 14.081517919887562, "grad_norm": 0.31698745489120483, "learning_rate": 2.3949402670414618e-05, "loss": 0.0233, "step": 20038 }, { "epoch": 14.082220660576247, "grad_norm": 0.09591562300920486, "learning_rate": 2.3948934176622158e-05, "loss": 0.0106, "step": 20039 }, { "epoch": 14.082923401264933, "grad_norm": 0.7422814965248108, "learning_rate": 2.3948465682829702e-05, "loss": 0.0383, "step": 20040 }, { "epoch": 14.08362614195362, "grad_norm": 0.2788209021091461, "learning_rate": 2.3947997189037246e-05, "loss": 0.0451, "step": 20041 }, { "epoch": 14.084328882642305, "grad_norm": 0.3718322515487671, "learning_rate": 2.394752869524479e-05, "loss": 0.0533, "step": 20042 }, { "epoch": 14.085031623330991, "grad_norm": 1.040625810623169, "learning_rate": 2.394706020145233e-05, "loss": 0.0636, "step": 20043 }, { "epoch": 14.085734364019677, "grad_norm": 0.6880691647529602, "learning_rate": 2.3946591707659873e-05, "loss": 0.0746, "step": 20044 }, { "epoch": 14.086437104708363, "grad_norm": 0.710306704044342, "learning_rate": 2.3946123213867417e-05, "loss": 0.1379, "step": 20045 }, { "epoch": 14.087139845397049, "grad_norm": 0.567396342754364, "learning_rate": 2.394565472007496e-05, "loss": 0.1356, "step": 20046 }, { "epoch": 14.087842586085735, "grad_norm": 5.2283759117126465, "learning_rate": 2.39451862262825e-05, "loss": 0.1772, "step": 20047 }, { "epoch": 14.08854532677442, "grad_norm": 0.2493610382080078, "learning_rate": 2.3944717732490045e-05, "loss": 0.0608, "step": 20048 }, { "epoch": 14.089248067463107, "grad_norm": 0.12450768798589706, "learning_rate": 2.394424923869759e-05, "loss": 0.0218, "step": 20049 }, { "epoch": 14.089950808151793, "grad_norm": 0.12166015058755875, "learning_rate": 2.3943780744905132e-05, "loss": 0.0243, "step": 20050 }, { "epoch": 14.090653548840478, "grad_norm": 0.33973681926727295, "learning_rate": 2.3943312251112676e-05, "loss": 0.0181, "step": 20051 }, { "epoch": 14.091356289529164, "grad_norm": 0.09875734895467758, "learning_rate": 2.3942843757320216e-05, "loss": 0.0102, "step": 20052 }, { "epoch": 14.09205903021785, "grad_norm": 0.09465963393449783, "learning_rate": 2.394237526352776e-05, "loss": 0.0114, "step": 20053 }, { "epoch": 14.092761770906536, "grad_norm": 0.11670218408107758, "learning_rate": 2.39419067697353e-05, "loss": 0.019, "step": 20054 }, { "epoch": 14.093464511595222, "grad_norm": 0.08676613867282867, "learning_rate": 2.3941438275942844e-05, "loss": 0.0094, "step": 20055 }, { "epoch": 14.094167252283908, "grad_norm": 0.13973531126976013, "learning_rate": 2.3940969782150385e-05, "loss": 0.0197, "step": 20056 }, { "epoch": 14.094869992972592, "grad_norm": 0.09428029507398605, "learning_rate": 2.394050128835793e-05, "loss": 0.0139, "step": 20057 }, { "epoch": 14.095572733661278, "grad_norm": 0.15912824869155884, "learning_rate": 2.3940032794565472e-05, "loss": 0.0214, "step": 20058 }, { "epoch": 14.096275474349964, "grad_norm": 0.0794076919555664, "learning_rate": 2.3939564300773016e-05, "loss": 0.0036, "step": 20059 }, { "epoch": 14.09697821503865, "grad_norm": 0.10582926124334335, "learning_rate": 2.3939095806980556e-05, "loss": 0.0163, "step": 20060 }, { "epoch": 14.097680955727336, "grad_norm": 0.2503780126571655, "learning_rate": 2.39386273131881e-05, "loss": 0.0298, "step": 20061 }, { "epoch": 14.098383696416022, "grad_norm": 0.5802837014198303, "learning_rate": 2.3938158819395644e-05, "loss": 0.0204, "step": 20062 }, { "epoch": 14.099086437104708, "grad_norm": 0.20775017142295837, "learning_rate": 2.3937690325603187e-05, "loss": 0.0378, "step": 20063 }, { "epoch": 14.099789177793394, "grad_norm": 0.10893553495407104, "learning_rate": 2.393722183181073e-05, "loss": 0.0176, "step": 20064 }, { "epoch": 14.10049191848208, "grad_norm": 0.17495949566364288, "learning_rate": 2.393675333801827e-05, "loss": 0.029, "step": 20065 }, { "epoch": 14.101194659170766, "grad_norm": 0.20033811032772064, "learning_rate": 2.3936284844225815e-05, "loss": 0.0282, "step": 20066 }, { "epoch": 14.101897399859451, "grad_norm": 0.20849204063415527, "learning_rate": 2.393581635043336e-05, "loss": 0.0444, "step": 20067 }, { "epoch": 14.102600140548137, "grad_norm": 0.23203814029693604, "learning_rate": 2.3935347856640903e-05, "loss": 0.0642, "step": 20068 }, { "epoch": 14.103302881236823, "grad_norm": 0.32006731629371643, "learning_rate": 2.3934879362848443e-05, "loss": 0.0812, "step": 20069 }, { "epoch": 14.10400562192551, "grad_norm": 0.4251028299331665, "learning_rate": 2.3934410869055987e-05, "loss": 0.12, "step": 20070 }, { "epoch": 14.104708362614195, "grad_norm": 0.7506887912750244, "learning_rate": 2.393394237526353e-05, "loss": 0.1461, "step": 20071 }, { "epoch": 14.105411103302881, "grad_norm": 1.1291580200195312, "learning_rate": 2.393347388147107e-05, "loss": 0.1671, "step": 20072 }, { "epoch": 14.106113843991567, "grad_norm": 0.49956652522087097, "learning_rate": 2.393300538767861e-05, "loss": 0.0531, "step": 20073 }, { "epoch": 14.106816584680253, "grad_norm": 0.14169077575206757, "learning_rate": 2.3932536893886155e-05, "loss": 0.0332, "step": 20074 }, { "epoch": 14.107519325368939, "grad_norm": 0.14940859377384186, "learning_rate": 2.39320684000937e-05, "loss": 0.0241, "step": 20075 }, { "epoch": 14.108222066057625, "grad_norm": 0.09813232719898224, "learning_rate": 2.3931599906301242e-05, "loss": 0.0237, "step": 20076 }, { "epoch": 14.10892480674631, "grad_norm": 0.18110191822052002, "learning_rate": 2.3931131412508786e-05, "loss": 0.0124, "step": 20077 }, { "epoch": 14.109627547434997, "grad_norm": 0.2808976471424103, "learning_rate": 2.3930662918716326e-05, "loss": 0.0215, "step": 20078 }, { "epoch": 14.110330288123683, "grad_norm": 0.11997523158788681, "learning_rate": 2.393019442492387e-05, "loss": 0.0128, "step": 20079 }, { "epoch": 14.111033028812368, "grad_norm": 0.43531516194343567, "learning_rate": 2.3929725931131414e-05, "loss": 0.013, "step": 20080 }, { "epoch": 14.111735769501054, "grad_norm": 0.18548980355262756, "learning_rate": 2.3929257437338957e-05, "loss": 0.0359, "step": 20081 }, { "epoch": 14.11243851018974, "grad_norm": 0.07479996979236603, "learning_rate": 2.3928788943546498e-05, "loss": 0.0127, "step": 20082 }, { "epoch": 14.113141250878426, "grad_norm": 0.1260923594236374, "learning_rate": 2.392832044975404e-05, "loss": 0.0151, "step": 20083 }, { "epoch": 14.113843991567112, "grad_norm": 0.08510049432516098, "learning_rate": 2.3927851955961585e-05, "loss": 0.009, "step": 20084 }, { "epoch": 14.114546732255798, "grad_norm": 0.5014881491661072, "learning_rate": 2.392738346216913e-05, "loss": 0.0199, "step": 20085 }, { "epoch": 14.115249472944484, "grad_norm": 0.0868879109621048, "learning_rate": 2.392691496837667e-05, "loss": 0.013, "step": 20086 }, { "epoch": 14.11595221363317, "grad_norm": 0.6344323754310608, "learning_rate": 2.3926446474584213e-05, "loss": 0.0242, "step": 20087 }, { "epoch": 14.116654954321856, "grad_norm": 0.16241109371185303, "learning_rate": 2.3925977980791757e-05, "loss": 0.0416, "step": 20088 }, { "epoch": 14.117357695010542, "grad_norm": 0.11035100370645523, "learning_rate": 2.3925509486999297e-05, "loss": 0.0137, "step": 20089 }, { "epoch": 14.118060435699228, "grad_norm": 0.14582356810569763, "learning_rate": 2.392504099320684e-05, "loss": 0.0326, "step": 20090 }, { "epoch": 14.118763176387914, "grad_norm": 0.11365632712841034, "learning_rate": 2.392457249941438e-05, "loss": 0.0214, "step": 20091 }, { "epoch": 14.1194659170766, "grad_norm": 0.193778857588768, "learning_rate": 2.3924104005621925e-05, "loss": 0.0389, "step": 20092 }, { "epoch": 14.120168657765285, "grad_norm": 0.3143807351589203, "learning_rate": 2.392363551182947e-05, "loss": 0.064, "step": 20093 }, { "epoch": 14.120871398453971, "grad_norm": 0.4045582413673401, "learning_rate": 2.3923167018037012e-05, "loss": 0.1057, "step": 20094 }, { "epoch": 14.121574139142655, "grad_norm": 0.558294415473938, "learning_rate": 2.3922698524244553e-05, "loss": 0.1184, "step": 20095 }, { "epoch": 14.122276879831341, "grad_norm": 0.9303790330886841, "learning_rate": 2.3922230030452096e-05, "loss": 0.1565, "step": 20096 }, { "epoch": 14.122979620520027, "grad_norm": 1.1016510725021362, "learning_rate": 2.392176153665964e-05, "loss": 0.163, "step": 20097 }, { "epoch": 14.123682361208713, "grad_norm": 0.2389446347951889, "learning_rate": 2.3921293042867184e-05, "loss": 0.0617, "step": 20098 }, { "epoch": 14.1243851018974, "grad_norm": 0.13777098059654236, "learning_rate": 2.3920824549074724e-05, "loss": 0.0218, "step": 20099 }, { "epoch": 14.125087842586085, "grad_norm": 0.1200014054775238, "learning_rate": 2.3920356055282268e-05, "loss": 0.0126, "step": 20100 }, { "epoch": 14.125790583274771, "grad_norm": 0.09202230721712112, "learning_rate": 2.391988756148981e-05, "loss": 0.02, "step": 20101 }, { "epoch": 14.126493323963457, "grad_norm": 0.1488308161497116, "learning_rate": 2.3919419067697355e-05, "loss": 0.0163, "step": 20102 }, { "epoch": 14.127196064652143, "grad_norm": 0.09635940939188004, "learning_rate": 2.39189505739049e-05, "loss": 0.0141, "step": 20103 }, { "epoch": 14.127898805340829, "grad_norm": 0.08896428346633911, "learning_rate": 2.391848208011244e-05, "loss": 0.0133, "step": 20104 }, { "epoch": 14.128601546029515, "grad_norm": 0.13381168246269226, "learning_rate": 2.3918013586319983e-05, "loss": 0.016, "step": 20105 }, { "epoch": 14.1293042867182, "grad_norm": 0.3208480179309845, "learning_rate": 2.3917545092527523e-05, "loss": 0.0156, "step": 20106 }, { "epoch": 14.130007027406887, "grad_norm": 0.2064480036497116, "learning_rate": 2.3917076598735067e-05, "loss": 0.0169, "step": 20107 }, { "epoch": 14.130709768095572, "grad_norm": 0.1627654880285263, "learning_rate": 2.3916608104942608e-05, "loss": 0.028, "step": 20108 }, { "epoch": 14.131412508784258, "grad_norm": 0.11548063904047012, "learning_rate": 2.391613961115015e-05, "loss": 0.0109, "step": 20109 }, { "epoch": 14.132115249472944, "grad_norm": 0.1283436268568039, "learning_rate": 2.3915671117357695e-05, "loss": 0.0158, "step": 20110 }, { "epoch": 14.13281799016163, "grad_norm": 0.16690093278884888, "learning_rate": 2.391520262356524e-05, "loss": 0.0159, "step": 20111 }, { "epoch": 14.133520730850316, "grad_norm": 0.28162819147109985, "learning_rate": 2.391473412977278e-05, "loss": 0.0346, "step": 20112 }, { "epoch": 14.134223471539002, "grad_norm": 0.10505246371030807, "learning_rate": 2.3914265635980323e-05, "loss": 0.0203, "step": 20113 }, { "epoch": 14.134926212227688, "grad_norm": 0.17672240734100342, "learning_rate": 2.3913797142187866e-05, "loss": 0.017, "step": 20114 }, { "epoch": 14.135628952916374, "grad_norm": 0.14029282331466675, "learning_rate": 2.391332864839541e-05, "loss": 0.0172, "step": 20115 }, { "epoch": 14.13633169360506, "grad_norm": 0.14002662897109985, "learning_rate": 2.3912860154602954e-05, "loss": 0.0225, "step": 20116 }, { "epoch": 14.137034434293746, "grad_norm": 0.3427453637123108, "learning_rate": 2.3912391660810494e-05, "loss": 0.0685, "step": 20117 }, { "epoch": 14.137737174982432, "grad_norm": 0.21555308997631073, "learning_rate": 2.3911923167018038e-05, "loss": 0.0483, "step": 20118 }, { "epoch": 14.138439915671118, "grad_norm": 0.33287015557289124, "learning_rate": 2.3911454673225582e-05, "loss": 0.0819, "step": 20119 }, { "epoch": 14.139142656359803, "grad_norm": 0.5361321568489075, "learning_rate": 2.3910986179433125e-05, "loss": 0.1126, "step": 20120 }, { "epoch": 14.13984539704849, "grad_norm": 0.4944854974746704, "learning_rate": 2.3910517685640666e-05, "loss": 0.1598, "step": 20121 }, { "epoch": 14.140548137737175, "grad_norm": 0.6955603957176208, "learning_rate": 2.391004919184821e-05, "loss": 0.153, "step": 20122 }, { "epoch": 14.141250878425861, "grad_norm": 0.19011296331882477, "learning_rate": 2.3909580698055753e-05, "loss": 0.0601, "step": 20123 }, { "epoch": 14.141953619114547, "grad_norm": 0.2739824652671814, "learning_rate": 2.3909112204263294e-05, "loss": 0.03, "step": 20124 }, { "epoch": 14.142656359803233, "grad_norm": 0.11960139870643616, "learning_rate": 2.3908643710470834e-05, "loss": 0.0153, "step": 20125 }, { "epoch": 14.143359100491919, "grad_norm": 0.0886632427573204, "learning_rate": 2.3908175216678378e-05, "loss": 0.0158, "step": 20126 }, { "epoch": 14.144061841180605, "grad_norm": 0.1086919829249382, "learning_rate": 2.390770672288592e-05, "loss": 0.0198, "step": 20127 }, { "epoch": 14.14476458186929, "grad_norm": 0.13531075417995453, "learning_rate": 2.3907238229093465e-05, "loss": 0.0197, "step": 20128 }, { "epoch": 14.145467322557977, "grad_norm": 0.10009389370679855, "learning_rate": 2.390676973530101e-05, "loss": 0.0163, "step": 20129 }, { "epoch": 14.146170063246663, "grad_norm": 0.04785371944308281, "learning_rate": 2.390630124150855e-05, "loss": 0.006, "step": 20130 }, { "epoch": 14.146872803935349, "grad_norm": 0.1492847502231598, "learning_rate": 2.3905832747716093e-05, "loss": 0.0271, "step": 20131 }, { "epoch": 14.147575544624035, "grad_norm": 0.08055595308542252, "learning_rate": 2.3905364253923637e-05, "loss": 0.0075, "step": 20132 }, { "epoch": 14.14827828531272, "grad_norm": 0.11279221624135971, "learning_rate": 2.390489576013118e-05, "loss": 0.0246, "step": 20133 }, { "epoch": 14.148981026001405, "grad_norm": 0.12376682460308075, "learning_rate": 2.390442726633872e-05, "loss": 0.0136, "step": 20134 }, { "epoch": 14.14968376669009, "grad_norm": 0.17156372964382172, "learning_rate": 2.3903958772546264e-05, "loss": 0.0275, "step": 20135 }, { "epoch": 14.150386507378776, "grad_norm": 0.1491321474313736, "learning_rate": 2.3903490278753808e-05, "loss": 0.0272, "step": 20136 }, { "epoch": 14.151089248067462, "grad_norm": 0.060941893607378006, "learning_rate": 2.3903021784961352e-05, "loss": 0.0106, "step": 20137 }, { "epoch": 14.151791988756148, "grad_norm": 0.4330153465270996, "learning_rate": 2.3902553291168892e-05, "loss": 0.0275, "step": 20138 }, { "epoch": 14.152494729444834, "grad_norm": 0.14732135832309723, "learning_rate": 2.3902084797376436e-05, "loss": 0.0185, "step": 20139 }, { "epoch": 14.15319747013352, "grad_norm": 0.30931657552719116, "learning_rate": 2.390161630358398e-05, "loss": 0.0444, "step": 20140 }, { "epoch": 14.153900210822206, "grad_norm": 0.3188689947128296, "learning_rate": 2.390114780979152e-05, "loss": 0.0335, "step": 20141 }, { "epoch": 14.154602951510892, "grad_norm": 0.2571829855442047, "learning_rate": 2.3900679315999064e-05, "loss": 0.056, "step": 20142 }, { "epoch": 14.155305692199578, "grad_norm": 0.450595498085022, "learning_rate": 2.3900210822206604e-05, "loss": 0.0768, "step": 20143 }, { "epoch": 14.156008432888264, "grad_norm": 0.34891828894615173, "learning_rate": 2.3899742328414148e-05, "loss": 0.086, "step": 20144 }, { "epoch": 14.15671117357695, "grad_norm": 0.7165782451629639, "learning_rate": 2.389927383462169e-05, "loss": 0.1521, "step": 20145 }, { "epoch": 14.157413914265636, "grad_norm": 0.948326051235199, "learning_rate": 2.3898805340829235e-05, "loss": 0.1701, "step": 20146 }, { "epoch": 14.158116654954322, "grad_norm": 0.5445332527160645, "learning_rate": 2.3898336847036776e-05, "loss": 0.1438, "step": 20147 }, { "epoch": 14.158819395643008, "grad_norm": 0.3845028281211853, "learning_rate": 2.389786835324432e-05, "loss": 0.0514, "step": 20148 }, { "epoch": 14.159522136331693, "grad_norm": 0.15142907202243805, "learning_rate": 2.3897399859451863e-05, "loss": 0.0291, "step": 20149 }, { "epoch": 14.16022487702038, "grad_norm": 0.2558826804161072, "learning_rate": 2.3896931365659407e-05, "loss": 0.0338, "step": 20150 }, { "epoch": 14.160927617709065, "grad_norm": 0.3848450183868408, "learning_rate": 2.3896462871866947e-05, "loss": 0.0149, "step": 20151 }, { "epoch": 14.161630358397751, "grad_norm": 0.09092894941568375, "learning_rate": 2.389599437807449e-05, "loss": 0.0126, "step": 20152 }, { "epoch": 14.162333099086437, "grad_norm": 0.0920664370059967, "learning_rate": 2.3895525884282034e-05, "loss": 0.0107, "step": 20153 }, { "epoch": 14.163035839775123, "grad_norm": 0.18974192440509796, "learning_rate": 2.3895057390489578e-05, "loss": 0.0235, "step": 20154 }, { "epoch": 14.163738580463809, "grad_norm": 0.2345607727766037, "learning_rate": 2.3894588896697122e-05, "loss": 0.0187, "step": 20155 }, { "epoch": 14.164441321152495, "grad_norm": 0.10321680456399918, "learning_rate": 2.3894120402904662e-05, "loss": 0.0157, "step": 20156 }, { "epoch": 14.16514406184118, "grad_norm": 0.06784182786941528, "learning_rate": 2.3893651909112206e-05, "loss": 0.0068, "step": 20157 }, { "epoch": 14.165846802529867, "grad_norm": 0.14983482658863068, "learning_rate": 2.389318341531975e-05, "loss": 0.0219, "step": 20158 }, { "epoch": 14.166549543218553, "grad_norm": 0.10034804046154022, "learning_rate": 2.389271492152729e-05, "loss": 0.0128, "step": 20159 }, { "epoch": 14.167252283907239, "grad_norm": 0.11561068147420883, "learning_rate": 2.389224642773483e-05, "loss": 0.0154, "step": 20160 }, { "epoch": 14.167955024595924, "grad_norm": 0.103561632335186, "learning_rate": 2.3891777933942374e-05, "loss": 0.0139, "step": 20161 }, { "epoch": 14.16865776528461, "grad_norm": 0.10224147140979767, "learning_rate": 2.3891309440149918e-05, "loss": 0.0227, "step": 20162 }, { "epoch": 14.169360505973296, "grad_norm": 0.3033964931964874, "learning_rate": 2.389084094635746e-05, "loss": 0.0259, "step": 20163 }, { "epoch": 14.170063246661982, "grad_norm": 0.16899235546588898, "learning_rate": 2.3890372452565002e-05, "loss": 0.0216, "step": 20164 }, { "epoch": 14.170765987350668, "grad_norm": 0.2159811407327652, "learning_rate": 2.3889903958772546e-05, "loss": 0.0305, "step": 20165 }, { "epoch": 14.171468728039354, "grad_norm": 0.27726390957832336, "learning_rate": 2.388943546498009e-05, "loss": 0.0452, "step": 20166 }, { "epoch": 14.17217146872804, "grad_norm": 0.26540321111679077, "learning_rate": 2.3888966971187633e-05, "loss": 0.0525, "step": 20167 }, { "epoch": 14.172874209416726, "grad_norm": 0.323578417301178, "learning_rate": 2.3888498477395177e-05, "loss": 0.0524, "step": 20168 }, { "epoch": 14.173576950105412, "grad_norm": 0.4096430540084839, "learning_rate": 2.3888029983602717e-05, "loss": 0.0753, "step": 20169 }, { "epoch": 14.174279690794098, "grad_norm": 0.31658411026000977, "learning_rate": 2.388756148981026e-05, "loss": 0.112, "step": 20170 }, { "epoch": 14.174982431482784, "grad_norm": 0.7792525887489319, "learning_rate": 2.3887092996017805e-05, "loss": 0.1516, "step": 20171 }, { "epoch": 14.17568517217147, "grad_norm": 1.569029450416565, "learning_rate": 2.388662450222535e-05, "loss": 0.17, "step": 20172 }, { "epoch": 14.176387912860154, "grad_norm": 0.19594749808311462, "learning_rate": 2.388615600843289e-05, "loss": 0.0649, "step": 20173 }, { "epoch": 14.17709065354884, "grad_norm": 0.0992153063416481, "learning_rate": 2.3885687514640432e-05, "loss": 0.0226, "step": 20174 }, { "epoch": 14.177793394237526, "grad_norm": 0.08444428443908691, "learning_rate": 2.3885219020847976e-05, "loss": 0.0197, "step": 20175 }, { "epoch": 14.178496134926212, "grad_norm": 0.08240630477666855, "learning_rate": 2.3884750527055516e-05, "loss": 0.0111, "step": 20176 }, { "epoch": 14.179198875614897, "grad_norm": 0.3641519546508789, "learning_rate": 2.388428203326306e-05, "loss": 0.0141, "step": 20177 }, { "epoch": 14.179901616303583, "grad_norm": 0.10587888956069946, "learning_rate": 2.38838135394706e-05, "loss": 0.0168, "step": 20178 }, { "epoch": 14.18060435699227, "grad_norm": 0.2347843050956726, "learning_rate": 2.3883345045678144e-05, "loss": 0.0197, "step": 20179 }, { "epoch": 14.181307097680955, "grad_norm": 0.1871812343597412, "learning_rate": 2.3882876551885688e-05, "loss": 0.0192, "step": 20180 }, { "epoch": 14.182009838369641, "grad_norm": 0.12135475128889084, "learning_rate": 2.388240805809323e-05, "loss": 0.0158, "step": 20181 }, { "epoch": 14.182712579058327, "grad_norm": 0.07939602434635162, "learning_rate": 2.3881939564300772e-05, "loss": 0.0132, "step": 20182 }, { "epoch": 14.183415319747013, "grad_norm": 0.14822858572006226, "learning_rate": 2.3881471070508316e-05, "loss": 0.0144, "step": 20183 }, { "epoch": 14.184118060435699, "grad_norm": 0.21082523465156555, "learning_rate": 2.388100257671586e-05, "loss": 0.0078, "step": 20184 }, { "epoch": 14.184820801124385, "grad_norm": 0.1491096466779709, "learning_rate": 2.3880534082923403e-05, "loss": 0.0201, "step": 20185 }, { "epoch": 14.18552354181307, "grad_norm": 0.6130048036575317, "learning_rate": 2.3880065589130944e-05, "loss": 0.0178, "step": 20186 }, { "epoch": 14.186226282501757, "grad_norm": 0.14972710609436035, "learning_rate": 2.3879597095338487e-05, "loss": 0.0226, "step": 20187 }, { "epoch": 14.186929023190443, "grad_norm": 0.17710985243320465, "learning_rate": 2.387912860154603e-05, "loss": 0.0235, "step": 20188 }, { "epoch": 14.187631763879128, "grad_norm": 0.14501361548900604, "learning_rate": 2.3878660107753575e-05, "loss": 0.0106, "step": 20189 }, { "epoch": 14.188334504567814, "grad_norm": 0.1471310406923294, "learning_rate": 2.3878191613961115e-05, "loss": 0.0399, "step": 20190 }, { "epoch": 14.1890372452565, "grad_norm": 0.13278059661388397, "learning_rate": 2.387772312016866e-05, "loss": 0.0265, "step": 20191 }, { "epoch": 14.189739985945186, "grad_norm": 0.1315135955810547, "learning_rate": 2.3877254626376202e-05, "loss": 0.0327, "step": 20192 }, { "epoch": 14.190442726633872, "grad_norm": 0.17666566371917725, "learning_rate": 2.3876786132583743e-05, "loss": 0.0595, "step": 20193 }, { "epoch": 14.191145467322558, "grad_norm": 0.42814901471138, "learning_rate": 2.3876317638791287e-05, "loss": 0.0794, "step": 20194 }, { "epoch": 14.191848208011244, "grad_norm": 0.45561766624450684, "learning_rate": 2.3875849144998827e-05, "loss": 0.1279, "step": 20195 }, { "epoch": 14.19255094869993, "grad_norm": 0.5100744962692261, "learning_rate": 2.387538065120637e-05, "loss": 0.1231, "step": 20196 }, { "epoch": 14.193253689388616, "grad_norm": 0.89448082447052, "learning_rate": 2.3874912157413914e-05, "loss": 0.1495, "step": 20197 }, { "epoch": 14.193956430077302, "grad_norm": 0.21851180493831635, "learning_rate": 2.3874443663621458e-05, "loss": 0.0563, "step": 20198 }, { "epoch": 14.194659170765988, "grad_norm": 0.1107100248336792, "learning_rate": 2.3873975169829e-05, "loss": 0.0262, "step": 20199 }, { "epoch": 14.195361911454674, "grad_norm": 0.13447432219982147, "learning_rate": 2.3873506676036542e-05, "loss": 0.0192, "step": 20200 }, { "epoch": 14.19606465214336, "grad_norm": 0.07142747193574905, "learning_rate": 2.3873038182244086e-05, "loss": 0.0131, "step": 20201 }, { "epoch": 14.196767392832045, "grad_norm": 0.10199880599975586, "learning_rate": 2.387256968845163e-05, "loss": 0.0154, "step": 20202 }, { "epoch": 14.197470133520731, "grad_norm": 0.08332019299268723, "learning_rate": 2.3872101194659173e-05, "loss": 0.0054, "step": 20203 }, { "epoch": 14.198172874209417, "grad_norm": 0.19464512169361115, "learning_rate": 2.3871632700866714e-05, "loss": 0.0104, "step": 20204 }, { "epoch": 14.198875614898103, "grad_norm": 0.08024762570858002, "learning_rate": 2.3871164207074257e-05, "loss": 0.0141, "step": 20205 }, { "epoch": 14.19957835558679, "grad_norm": 0.15292511880397797, "learning_rate": 2.38706957132818e-05, "loss": 0.0297, "step": 20206 }, { "epoch": 14.200281096275475, "grad_norm": 0.061653874814510345, "learning_rate": 2.3870227219489345e-05, "loss": 0.008, "step": 20207 }, { "epoch": 14.200983836964161, "grad_norm": 0.18000870943069458, "learning_rate": 2.3869758725696885e-05, "loss": 0.022, "step": 20208 }, { "epoch": 14.201686577652847, "grad_norm": 0.12474004179239273, "learning_rate": 2.386929023190443e-05, "loss": 0.0171, "step": 20209 }, { "epoch": 14.202389318341533, "grad_norm": 0.12205181270837784, "learning_rate": 2.3868821738111973e-05, "loss": 0.0204, "step": 20210 }, { "epoch": 14.203092059030217, "grad_norm": 0.07463721185922623, "learning_rate": 2.3868353244319513e-05, "loss": 0.0059, "step": 20211 }, { "epoch": 14.203794799718903, "grad_norm": 0.20528745651245117, "learning_rate": 2.3867884750527053e-05, "loss": 0.0285, "step": 20212 }, { "epoch": 14.204497540407589, "grad_norm": 0.15808753669261932, "learning_rate": 2.3867416256734597e-05, "loss": 0.0248, "step": 20213 }, { "epoch": 14.205200281096275, "grad_norm": 0.13878989219665527, "learning_rate": 2.386694776294214e-05, "loss": 0.0194, "step": 20214 }, { "epoch": 14.20590302178496, "grad_norm": 0.2378474622964859, "learning_rate": 2.3866479269149684e-05, "loss": 0.0378, "step": 20215 }, { "epoch": 14.206605762473647, "grad_norm": 0.17635920643806458, "learning_rate": 2.3866010775357228e-05, "loss": 0.035, "step": 20216 }, { "epoch": 14.207308503162333, "grad_norm": 0.2844506800174713, "learning_rate": 2.386554228156477e-05, "loss": 0.051, "step": 20217 }, { "epoch": 14.208011243851018, "grad_norm": 0.33811917901039124, "learning_rate": 2.3865073787772312e-05, "loss": 0.0643, "step": 20218 }, { "epoch": 14.208713984539704, "grad_norm": 0.6828439831733704, "learning_rate": 2.3864605293979856e-05, "loss": 0.0868, "step": 20219 }, { "epoch": 14.20941672522839, "grad_norm": 1.9986103773117065, "learning_rate": 2.38641368001874e-05, "loss": 0.1059, "step": 20220 }, { "epoch": 14.210119465917076, "grad_norm": 0.6926116943359375, "learning_rate": 2.386366830639494e-05, "loss": 0.1525, "step": 20221 }, { "epoch": 14.210822206605762, "grad_norm": 1.61179780960083, "learning_rate": 2.3863199812602484e-05, "loss": 0.1602, "step": 20222 }, { "epoch": 14.211524947294448, "grad_norm": 0.37105005979537964, "learning_rate": 2.3862731318810027e-05, "loss": 0.0692, "step": 20223 }, { "epoch": 14.212227687983134, "grad_norm": 0.3196246027946472, "learning_rate": 2.386226282501757e-05, "loss": 0.0222, "step": 20224 }, { "epoch": 14.21293042867182, "grad_norm": 0.1616595983505249, "learning_rate": 2.386179433122511e-05, "loss": 0.0147, "step": 20225 }, { "epoch": 14.213633169360506, "grad_norm": 0.06790407001972198, "learning_rate": 2.3861325837432655e-05, "loss": 0.0142, "step": 20226 }, { "epoch": 14.214335910049192, "grad_norm": 0.11969456821680069, "learning_rate": 2.38608573436402e-05, "loss": 0.0183, "step": 20227 }, { "epoch": 14.215038650737878, "grad_norm": 0.06563840061426163, "learning_rate": 2.386038884984774e-05, "loss": 0.0116, "step": 20228 }, { "epoch": 14.215741391426564, "grad_norm": 0.09661093354225159, "learning_rate": 2.3859920356055283e-05, "loss": 0.0209, "step": 20229 }, { "epoch": 14.21644413211525, "grad_norm": 0.3846755027770996, "learning_rate": 2.3859451862262823e-05, "loss": 0.0201, "step": 20230 }, { "epoch": 14.217146872803935, "grad_norm": 0.8238300085067749, "learning_rate": 2.3858983368470367e-05, "loss": 0.0148, "step": 20231 }, { "epoch": 14.217849613492621, "grad_norm": 0.14363563060760498, "learning_rate": 2.385851487467791e-05, "loss": 0.0086, "step": 20232 }, { "epoch": 14.218552354181307, "grad_norm": 0.13934464752674103, "learning_rate": 2.3858046380885455e-05, "loss": 0.0268, "step": 20233 }, { "epoch": 14.219255094869993, "grad_norm": 0.08698838204145432, "learning_rate": 2.3857577887092995e-05, "loss": 0.0063, "step": 20234 }, { "epoch": 14.219957835558679, "grad_norm": 0.1967359036207199, "learning_rate": 2.385710939330054e-05, "loss": 0.03, "step": 20235 }, { "epoch": 14.220660576247365, "grad_norm": 0.1264631301164627, "learning_rate": 2.3856640899508082e-05, "loss": 0.0105, "step": 20236 }, { "epoch": 14.221363316936051, "grad_norm": 0.1062546968460083, "learning_rate": 2.3856172405715626e-05, "loss": 0.0203, "step": 20237 }, { "epoch": 14.222066057624737, "grad_norm": 0.15721964836120605, "learning_rate": 2.3855703911923166e-05, "loss": 0.035, "step": 20238 }, { "epoch": 14.222768798313423, "grad_norm": 0.17331640422344208, "learning_rate": 2.385523541813071e-05, "loss": 0.0116, "step": 20239 }, { "epoch": 14.223471539002109, "grad_norm": 0.18564988672733307, "learning_rate": 2.3854766924338254e-05, "loss": 0.0193, "step": 20240 }, { "epoch": 14.224174279690795, "grad_norm": 0.28210631012916565, "learning_rate": 2.3854298430545798e-05, "loss": 0.0663, "step": 20241 }, { "epoch": 14.22487702037948, "grad_norm": 0.21120132505893707, "learning_rate": 2.385382993675334e-05, "loss": 0.0553, "step": 20242 }, { "epoch": 14.225579761068166, "grad_norm": 0.294719398021698, "learning_rate": 2.385336144296088e-05, "loss": 0.0697, "step": 20243 }, { "epoch": 14.226282501756852, "grad_norm": 0.8907017111778259, "learning_rate": 2.3852892949168425e-05, "loss": 0.087, "step": 20244 }, { "epoch": 14.226985242445538, "grad_norm": 0.39037373661994934, "learning_rate": 2.385242445537597e-05, "loss": 0.1346, "step": 20245 }, { "epoch": 14.227687983134224, "grad_norm": 0.542759358882904, "learning_rate": 2.385195596158351e-05, "loss": 0.1297, "step": 20246 }, { "epoch": 14.22839072382291, "grad_norm": 0.5275744795799255, "learning_rate": 2.385148746779105e-05, "loss": 0.1501, "step": 20247 }, { "epoch": 14.229093464511596, "grad_norm": 0.19002781808376312, "learning_rate": 2.3851018973998594e-05, "loss": 0.0587, "step": 20248 }, { "epoch": 14.22979620520028, "grad_norm": 0.2333148568868637, "learning_rate": 2.3850550480206137e-05, "loss": 0.033, "step": 20249 }, { "epoch": 14.230498945888966, "grad_norm": 0.1518315225839615, "learning_rate": 2.385008198641368e-05, "loss": 0.0305, "step": 20250 }, { "epoch": 14.231201686577652, "grad_norm": 0.07896586507558823, "learning_rate": 2.384961349262122e-05, "loss": 0.0136, "step": 20251 }, { "epoch": 14.231904427266338, "grad_norm": 0.08018305897712708, "learning_rate": 2.3849144998828765e-05, "loss": 0.0109, "step": 20252 }, { "epoch": 14.232607167955024, "grad_norm": 0.12110225111246109, "learning_rate": 2.384867650503631e-05, "loss": 0.0099, "step": 20253 }, { "epoch": 14.23330990864371, "grad_norm": 0.08358367532491684, "learning_rate": 2.3848208011243852e-05, "loss": 0.0067, "step": 20254 }, { "epoch": 14.234012649332396, "grad_norm": 0.07277050614356995, "learning_rate": 2.3847739517451396e-05, "loss": 0.021, "step": 20255 }, { "epoch": 14.234715390021082, "grad_norm": 0.09519484639167786, "learning_rate": 2.3847271023658937e-05, "loss": 0.0146, "step": 20256 }, { "epoch": 14.235418130709768, "grad_norm": 0.08284791558980942, "learning_rate": 2.384680252986648e-05, "loss": 0.0098, "step": 20257 }, { "epoch": 14.236120871398454, "grad_norm": 0.15052364766597748, "learning_rate": 2.3846334036074024e-05, "loss": 0.019, "step": 20258 }, { "epoch": 14.23682361208714, "grad_norm": 0.12726415693759918, "learning_rate": 2.3845865542281568e-05, "loss": 0.02, "step": 20259 }, { "epoch": 14.237526352775825, "grad_norm": 0.141597718000412, "learning_rate": 2.3845397048489108e-05, "loss": 0.0247, "step": 20260 }, { "epoch": 14.238229093464511, "grad_norm": 0.08093201369047165, "learning_rate": 2.3844928554696652e-05, "loss": 0.0244, "step": 20261 }, { "epoch": 14.238931834153197, "grad_norm": 0.14677591621875763, "learning_rate": 2.3844460060904195e-05, "loss": 0.0272, "step": 20262 }, { "epoch": 14.239634574841883, "grad_norm": 0.13776320219039917, "learning_rate": 2.3843991567111736e-05, "loss": 0.0343, "step": 20263 }, { "epoch": 14.240337315530569, "grad_norm": 0.07519947737455368, "learning_rate": 2.3843523073319276e-05, "loss": 0.0107, "step": 20264 }, { "epoch": 14.241040056219255, "grad_norm": 0.16187182068824768, "learning_rate": 2.384305457952682e-05, "loss": 0.0279, "step": 20265 }, { "epoch": 14.24174279690794, "grad_norm": 0.3229066729545593, "learning_rate": 2.3842586085734364e-05, "loss": 0.0346, "step": 20266 }, { "epoch": 14.242445537596627, "grad_norm": 0.4619206488132477, "learning_rate": 2.3842117591941907e-05, "loss": 0.0383, "step": 20267 }, { "epoch": 14.243148278285313, "grad_norm": 0.21053895354270935, "learning_rate": 2.384164909814945e-05, "loss": 0.0611, "step": 20268 }, { "epoch": 14.243851018973999, "grad_norm": 0.31490617990493774, "learning_rate": 2.384118060435699e-05, "loss": 0.0871, "step": 20269 }, { "epoch": 14.244553759662685, "grad_norm": 0.4403224289417267, "learning_rate": 2.3840712110564535e-05, "loss": 0.1289, "step": 20270 }, { "epoch": 14.24525650035137, "grad_norm": 0.7813166379928589, "learning_rate": 2.384024361677208e-05, "loss": 0.1808, "step": 20271 }, { "epoch": 14.245959241040056, "grad_norm": 0.7934972047805786, "learning_rate": 2.3839775122979623e-05, "loss": 0.1522, "step": 20272 }, { "epoch": 14.246661981728742, "grad_norm": 0.19977469742298126, "learning_rate": 2.3839306629187163e-05, "loss": 0.0586, "step": 20273 }, { "epoch": 14.247364722417428, "grad_norm": 0.0992901623249054, "learning_rate": 2.3838838135394707e-05, "loss": 0.0193, "step": 20274 }, { "epoch": 14.248067463106114, "grad_norm": 0.11234894394874573, "learning_rate": 2.383836964160225e-05, "loss": 0.0245, "step": 20275 }, { "epoch": 14.2487702037948, "grad_norm": 0.3235313296318054, "learning_rate": 2.3837901147809794e-05, "loss": 0.0208, "step": 20276 }, { "epoch": 14.249472944483486, "grad_norm": 0.1564541608095169, "learning_rate": 2.3837432654017334e-05, "loss": 0.0147, "step": 20277 }, { "epoch": 14.250175685172172, "grad_norm": 0.0722459927201271, "learning_rate": 2.3836964160224878e-05, "loss": 0.0095, "step": 20278 }, { "epoch": 14.250878425860858, "grad_norm": 0.10521698743104935, "learning_rate": 2.3836495666432422e-05, "loss": 0.0114, "step": 20279 }, { "epoch": 14.251581166549544, "grad_norm": 0.6738610863685608, "learning_rate": 2.3836027172639966e-05, "loss": 0.0246, "step": 20280 }, { "epoch": 14.25228390723823, "grad_norm": 0.09593109041452408, "learning_rate": 2.3835558678847506e-05, "loss": 0.014, "step": 20281 }, { "epoch": 14.252986647926916, "grad_norm": 0.0633460059762001, "learning_rate": 2.3835090185055046e-05, "loss": 0.0078, "step": 20282 }, { "epoch": 14.253689388615602, "grad_norm": 0.20631633698940277, "learning_rate": 2.383462169126259e-05, "loss": 0.016, "step": 20283 }, { "epoch": 14.254392129304287, "grad_norm": 0.08770939707756042, "learning_rate": 2.3834153197470134e-05, "loss": 0.0093, "step": 20284 }, { "epoch": 14.255094869992973, "grad_norm": 0.14144928753376007, "learning_rate": 2.3833684703677677e-05, "loss": 0.0236, "step": 20285 }, { "epoch": 14.25579761068166, "grad_norm": 0.12490189075469971, "learning_rate": 2.3833216209885218e-05, "loss": 0.0121, "step": 20286 }, { "epoch": 14.256500351370345, "grad_norm": 0.11655110120773315, "learning_rate": 2.383274771609276e-05, "loss": 0.0161, "step": 20287 }, { "epoch": 14.25720309205903, "grad_norm": 0.7732633948326111, "learning_rate": 2.3832279222300305e-05, "loss": 0.0279, "step": 20288 }, { "epoch": 14.257905832747715, "grad_norm": 0.23248261213302612, "learning_rate": 2.383181072850785e-05, "loss": 0.0103, "step": 20289 }, { "epoch": 14.258608573436401, "grad_norm": 0.44333338737487793, "learning_rate": 2.383134223471539e-05, "loss": 0.0387, "step": 20290 }, { "epoch": 14.259311314125087, "grad_norm": 0.2573631703853607, "learning_rate": 2.3830873740922933e-05, "loss": 0.0372, "step": 20291 }, { "epoch": 14.260014054813773, "grad_norm": 0.4140416979789734, "learning_rate": 2.3830405247130477e-05, "loss": 0.0539, "step": 20292 }, { "epoch": 14.260716795502459, "grad_norm": 0.5596375465393066, "learning_rate": 2.382993675333802e-05, "loss": 0.0752, "step": 20293 }, { "epoch": 14.261419536191145, "grad_norm": 0.4071947932243347, "learning_rate": 2.3829468259545564e-05, "loss": 0.1091, "step": 20294 }, { "epoch": 14.26212227687983, "grad_norm": 0.3204643428325653, "learning_rate": 2.3828999765753105e-05, "loss": 0.1042, "step": 20295 }, { "epoch": 14.262825017568517, "grad_norm": 0.9230095744132996, "learning_rate": 2.3828531271960648e-05, "loss": 0.1326, "step": 20296 }, { "epoch": 14.263527758257203, "grad_norm": 0.7986630797386169, "learning_rate": 2.3828062778168192e-05, "loss": 0.1782, "step": 20297 }, { "epoch": 14.264230498945889, "grad_norm": 0.2614133358001709, "learning_rate": 2.3827594284375732e-05, "loss": 0.0736, "step": 20298 }, { "epoch": 14.264933239634574, "grad_norm": 0.14872893691062927, "learning_rate": 2.3827125790583273e-05, "loss": 0.0295, "step": 20299 }, { "epoch": 14.26563598032326, "grad_norm": 0.11376695334911346, "learning_rate": 2.3826657296790816e-05, "loss": 0.025, "step": 20300 }, { "epoch": 14.266338721011946, "grad_norm": 0.2151585817337036, "learning_rate": 2.382618880299836e-05, "loss": 0.017, "step": 20301 }, { "epoch": 14.267041461700632, "grad_norm": 0.10343703627586365, "learning_rate": 2.3825720309205904e-05, "loss": 0.0178, "step": 20302 }, { "epoch": 14.267744202389318, "grad_norm": 0.13965235650539398, "learning_rate": 2.3825251815413444e-05, "loss": 0.0114, "step": 20303 }, { "epoch": 14.268446943078004, "grad_norm": 0.07216345518827438, "learning_rate": 2.3824783321620988e-05, "loss": 0.0084, "step": 20304 }, { "epoch": 14.26914968376669, "grad_norm": 0.1492266207933426, "learning_rate": 2.382431482782853e-05, "loss": 0.0176, "step": 20305 }, { "epoch": 14.269852424455376, "grad_norm": 0.2120705246925354, "learning_rate": 2.3823846334036075e-05, "loss": 0.0206, "step": 20306 }, { "epoch": 14.270555165144062, "grad_norm": 0.15541619062423706, "learning_rate": 2.382337784024362e-05, "loss": 0.0187, "step": 20307 }, { "epoch": 14.271257905832748, "grad_norm": 0.11753380298614502, "learning_rate": 2.382290934645116e-05, "loss": 0.0286, "step": 20308 }, { "epoch": 14.271960646521434, "grad_norm": 0.1148313656449318, "learning_rate": 2.3822440852658703e-05, "loss": 0.0078, "step": 20309 }, { "epoch": 14.27266338721012, "grad_norm": 0.1271013617515564, "learning_rate": 2.3821972358866247e-05, "loss": 0.025, "step": 20310 }, { "epoch": 14.273366127898806, "grad_norm": 0.06429806351661682, "learning_rate": 2.382150386507379e-05, "loss": 0.0096, "step": 20311 }, { "epoch": 14.274068868587491, "grad_norm": 0.24668176472187042, "learning_rate": 2.382103537128133e-05, "loss": 0.0183, "step": 20312 }, { "epoch": 14.274771609276177, "grad_norm": 0.16402484476566315, "learning_rate": 2.3820566877488875e-05, "loss": 0.0239, "step": 20313 }, { "epoch": 14.275474349964863, "grad_norm": 0.8626152873039246, "learning_rate": 2.382009838369642e-05, "loss": 0.0255, "step": 20314 }, { "epoch": 14.27617709065355, "grad_norm": 0.3106938302516937, "learning_rate": 2.381962988990396e-05, "loss": 0.0406, "step": 20315 }, { "epoch": 14.276879831342235, "grad_norm": 0.48170503973960876, "learning_rate": 2.38191613961115e-05, "loss": 0.0377, "step": 20316 }, { "epoch": 14.277582572030921, "grad_norm": 0.16069693863391876, "learning_rate": 2.3818692902319043e-05, "loss": 0.0371, "step": 20317 }, { "epoch": 14.278285312719607, "grad_norm": 0.34383589029312134, "learning_rate": 2.3818224408526587e-05, "loss": 0.0646, "step": 20318 }, { "epoch": 14.278988053408293, "grad_norm": 0.33176442980766296, "learning_rate": 2.381775591473413e-05, "loss": 0.1125, "step": 20319 }, { "epoch": 14.279690794096979, "grad_norm": 0.49775511026382446, "learning_rate": 2.3817287420941674e-05, "loss": 0.1183, "step": 20320 }, { "epoch": 14.280393534785665, "grad_norm": 0.5891982913017273, "learning_rate": 2.3816818927149214e-05, "loss": 0.1351, "step": 20321 }, { "epoch": 14.28109627547435, "grad_norm": 1.5667840242385864, "learning_rate": 2.3816350433356758e-05, "loss": 0.18, "step": 20322 }, { "epoch": 14.281799016163037, "grad_norm": 0.2085961550474167, "learning_rate": 2.3815881939564302e-05, "loss": 0.0602, "step": 20323 }, { "epoch": 14.282501756851723, "grad_norm": 0.17854702472686768, "learning_rate": 2.3815413445771845e-05, "loss": 0.0236, "step": 20324 }, { "epoch": 14.283204497540408, "grad_norm": 0.09850586205720901, "learning_rate": 2.3814944951979386e-05, "loss": 0.0144, "step": 20325 }, { "epoch": 14.283907238229094, "grad_norm": 0.08129309117794037, "learning_rate": 2.381447645818693e-05, "loss": 0.0091, "step": 20326 }, { "epoch": 14.284609978917779, "grad_norm": 0.1844538450241089, "learning_rate": 2.3814007964394473e-05, "loss": 0.0251, "step": 20327 }, { "epoch": 14.285312719606464, "grad_norm": 0.08766555041074753, "learning_rate": 2.3813539470602017e-05, "loss": 0.0082, "step": 20328 }, { "epoch": 14.28601546029515, "grad_norm": 0.2681579291820526, "learning_rate": 2.3813070976809557e-05, "loss": 0.0156, "step": 20329 }, { "epoch": 14.286718200983836, "grad_norm": 0.7089790105819702, "learning_rate": 2.38126024830171e-05, "loss": 0.024, "step": 20330 }, { "epoch": 14.287420941672522, "grad_norm": 0.13020914793014526, "learning_rate": 2.3812133989224645e-05, "loss": 0.0285, "step": 20331 }, { "epoch": 14.288123682361208, "grad_norm": 0.0915037989616394, "learning_rate": 2.381166549543219e-05, "loss": 0.0154, "step": 20332 }, { "epoch": 14.288826423049894, "grad_norm": 0.21556071937084198, "learning_rate": 2.381119700163973e-05, "loss": 0.0314, "step": 20333 }, { "epoch": 14.28952916373858, "grad_norm": 0.10622908920049667, "learning_rate": 2.381072850784727e-05, "loss": 0.0119, "step": 20334 }, { "epoch": 14.290231904427266, "grad_norm": 0.07981160283088684, "learning_rate": 2.3810260014054813e-05, "loss": 0.0204, "step": 20335 }, { "epoch": 14.290934645115952, "grad_norm": 0.08781401067972183, "learning_rate": 2.3809791520262357e-05, "loss": 0.0113, "step": 20336 }, { "epoch": 14.291637385804638, "grad_norm": 0.15087656676769257, "learning_rate": 2.38093230264699e-05, "loss": 0.0161, "step": 20337 }, { "epoch": 14.292340126493324, "grad_norm": 0.1300537884235382, "learning_rate": 2.380885453267744e-05, "loss": 0.0236, "step": 20338 }, { "epoch": 14.29304286718201, "grad_norm": 0.1419857144355774, "learning_rate": 2.3808386038884984e-05, "loss": 0.0125, "step": 20339 }, { "epoch": 14.293745607870695, "grad_norm": 0.10937174409627914, "learning_rate": 2.3807917545092528e-05, "loss": 0.0239, "step": 20340 }, { "epoch": 14.294448348559381, "grad_norm": 0.36843717098236084, "learning_rate": 2.3807449051300072e-05, "loss": 0.0495, "step": 20341 }, { "epoch": 14.295151089248067, "grad_norm": 1.0779880285263062, "learning_rate": 2.3806980557507612e-05, "loss": 0.0366, "step": 20342 }, { "epoch": 14.295853829936753, "grad_norm": 0.24109259247779846, "learning_rate": 2.3806512063715156e-05, "loss": 0.0662, "step": 20343 }, { "epoch": 14.29655657062544, "grad_norm": 0.7446858286857605, "learning_rate": 2.38060435699227e-05, "loss": 0.0929, "step": 20344 }, { "epoch": 14.297259311314125, "grad_norm": 0.5694371461868286, "learning_rate": 2.3805575076130243e-05, "loss": 0.1035, "step": 20345 }, { "epoch": 14.297962052002811, "grad_norm": 0.8966894149780273, "learning_rate": 2.3805106582337787e-05, "loss": 0.1637, "step": 20346 }, { "epoch": 14.298664792691497, "grad_norm": 0.9124488830566406, "learning_rate": 2.3804638088545327e-05, "loss": 0.144, "step": 20347 }, { "epoch": 14.299367533380183, "grad_norm": 0.23907971382141113, "learning_rate": 2.380416959475287e-05, "loss": 0.0643, "step": 20348 }, { "epoch": 14.300070274068869, "grad_norm": 0.10816743224859238, "learning_rate": 2.3803701100960415e-05, "loss": 0.0185, "step": 20349 }, { "epoch": 14.300773014757555, "grad_norm": 0.08680886030197144, "learning_rate": 2.3803232607167955e-05, "loss": 0.0176, "step": 20350 }, { "epoch": 14.30147575544624, "grad_norm": 0.12097484618425369, "learning_rate": 2.3802764113375496e-05, "loss": 0.0143, "step": 20351 }, { "epoch": 14.302178496134927, "grad_norm": 0.10891874879598618, "learning_rate": 2.380229561958304e-05, "loss": 0.0149, "step": 20352 }, { "epoch": 14.302881236823612, "grad_norm": 0.14723338186740875, "learning_rate": 2.3801827125790583e-05, "loss": 0.0099, "step": 20353 }, { "epoch": 14.303583977512298, "grad_norm": 0.48017826676368713, "learning_rate": 2.3801358631998127e-05, "loss": 0.015, "step": 20354 }, { "epoch": 14.304286718200984, "grad_norm": 0.18421560525894165, "learning_rate": 2.3800890138205667e-05, "loss": 0.0167, "step": 20355 }, { "epoch": 14.30498945888967, "grad_norm": 0.13752901554107666, "learning_rate": 2.380042164441321e-05, "loss": 0.0161, "step": 20356 }, { "epoch": 14.305692199578356, "grad_norm": 0.10224936902523041, "learning_rate": 2.3799953150620755e-05, "loss": 0.0159, "step": 20357 }, { "epoch": 14.306394940267042, "grad_norm": 0.11785265803337097, "learning_rate": 2.3799484656828298e-05, "loss": 0.0184, "step": 20358 }, { "epoch": 14.307097680955728, "grad_norm": 0.11516820639371872, "learning_rate": 2.3799016163035842e-05, "loss": 0.0063, "step": 20359 }, { "epoch": 14.307800421644414, "grad_norm": 0.17476852238178253, "learning_rate": 2.3798547669243382e-05, "loss": 0.0232, "step": 20360 }, { "epoch": 14.3085031623331, "grad_norm": 0.11907438188791275, "learning_rate": 2.3798079175450926e-05, "loss": 0.0187, "step": 20361 }, { "epoch": 14.309205903021786, "grad_norm": 0.11539826542139053, "learning_rate": 2.379761068165847e-05, "loss": 0.0211, "step": 20362 }, { "epoch": 14.309908643710472, "grad_norm": 0.5024051070213318, "learning_rate": 2.3797142187866013e-05, "loss": 0.0484, "step": 20363 }, { "epoch": 14.310611384399156, "grad_norm": 0.3001421391963959, "learning_rate": 2.3796673694073554e-05, "loss": 0.0275, "step": 20364 }, { "epoch": 14.311314125087842, "grad_norm": 0.3627249300479889, "learning_rate": 2.3796205200281098e-05, "loss": 0.0213, "step": 20365 }, { "epoch": 14.312016865776528, "grad_norm": 0.9316980838775635, "learning_rate": 2.379573670648864e-05, "loss": 0.0441, "step": 20366 }, { "epoch": 14.312719606465214, "grad_norm": 0.2885231673717499, "learning_rate": 2.3795268212696185e-05, "loss": 0.0479, "step": 20367 }, { "epoch": 14.3134223471539, "grad_norm": 0.37829622626304626, "learning_rate": 2.3794799718903722e-05, "loss": 0.0665, "step": 20368 }, { "epoch": 14.314125087842585, "grad_norm": 0.41681787371635437, "learning_rate": 2.3794331225111266e-05, "loss": 0.0777, "step": 20369 }, { "epoch": 14.314827828531271, "grad_norm": 1.0715703964233398, "learning_rate": 2.379386273131881e-05, "loss": 0.1262, "step": 20370 }, { "epoch": 14.315530569219957, "grad_norm": 0.4777955710887909, "learning_rate": 2.3793394237526353e-05, "loss": 0.1482, "step": 20371 }, { "epoch": 14.316233309908643, "grad_norm": 0.876335084438324, "learning_rate": 2.3792925743733897e-05, "loss": 0.1886, "step": 20372 }, { "epoch": 14.316936050597329, "grad_norm": 0.4599555730819702, "learning_rate": 2.3792457249941437e-05, "loss": 0.0621, "step": 20373 }, { "epoch": 14.317638791286015, "grad_norm": 0.15346336364746094, "learning_rate": 2.379198875614898e-05, "loss": 0.0163, "step": 20374 }, { "epoch": 14.318341531974701, "grad_norm": 0.12076757848262787, "learning_rate": 2.3791520262356525e-05, "loss": 0.0282, "step": 20375 }, { "epoch": 14.319044272663387, "grad_norm": 0.17656046152114868, "learning_rate": 2.379105176856407e-05, "loss": 0.0183, "step": 20376 }, { "epoch": 14.319747013352073, "grad_norm": 0.04980003833770752, "learning_rate": 2.379058327477161e-05, "loss": 0.0063, "step": 20377 }, { "epoch": 14.320449754040759, "grad_norm": 0.05804057791829109, "learning_rate": 2.3790114780979152e-05, "loss": 0.0105, "step": 20378 }, { "epoch": 14.321152494729445, "grad_norm": 0.17418606579303741, "learning_rate": 2.3789646287186696e-05, "loss": 0.0175, "step": 20379 }, { "epoch": 14.32185523541813, "grad_norm": 0.13653643429279327, "learning_rate": 2.378917779339424e-05, "loss": 0.0099, "step": 20380 }, { "epoch": 14.322557976106816, "grad_norm": 0.14228492975234985, "learning_rate": 2.378870929960178e-05, "loss": 0.0203, "step": 20381 }, { "epoch": 14.323260716795502, "grad_norm": 0.17835640907287598, "learning_rate": 2.3788240805809324e-05, "loss": 0.0123, "step": 20382 }, { "epoch": 14.323963457484188, "grad_norm": 0.13382269442081451, "learning_rate": 2.3787772312016868e-05, "loss": 0.0171, "step": 20383 }, { "epoch": 14.324666198172874, "grad_norm": 0.19458848237991333, "learning_rate": 2.378730381822441e-05, "loss": 0.0242, "step": 20384 }, { "epoch": 14.32536893886156, "grad_norm": 0.12621554732322693, "learning_rate": 2.3786835324431952e-05, "loss": 0.0222, "step": 20385 }, { "epoch": 14.326071679550246, "grad_norm": 0.1320570409297943, "learning_rate": 2.3786366830639492e-05, "loss": 0.0241, "step": 20386 }, { "epoch": 14.326774420238932, "grad_norm": 0.24775756895542145, "learning_rate": 2.3785898336847036e-05, "loss": 0.0142, "step": 20387 }, { "epoch": 14.327477160927618, "grad_norm": 0.2333458960056305, "learning_rate": 2.378542984305458e-05, "loss": 0.0206, "step": 20388 }, { "epoch": 14.328179901616304, "grad_norm": 0.13591843843460083, "learning_rate": 2.3784961349262123e-05, "loss": 0.0146, "step": 20389 }, { "epoch": 14.32888264230499, "grad_norm": 0.18370845913887024, "learning_rate": 2.3784492855469664e-05, "loss": 0.0464, "step": 20390 }, { "epoch": 14.329585382993676, "grad_norm": 0.12361980974674225, "learning_rate": 2.3784024361677207e-05, "loss": 0.0257, "step": 20391 }, { "epoch": 14.330288123682362, "grad_norm": 0.1178225576877594, "learning_rate": 2.378355586788475e-05, "loss": 0.0288, "step": 20392 }, { "epoch": 14.330990864371048, "grad_norm": 0.385017454624176, "learning_rate": 2.3783087374092295e-05, "loss": 0.0541, "step": 20393 }, { "epoch": 14.331693605059733, "grad_norm": 0.8957279920578003, "learning_rate": 2.3782618880299835e-05, "loss": 0.065, "step": 20394 }, { "epoch": 14.33239634574842, "grad_norm": 1.5973985195159912, "learning_rate": 2.378215038650738e-05, "loss": 0.1183, "step": 20395 }, { "epoch": 14.333099086437105, "grad_norm": 0.6274470686912537, "learning_rate": 2.3781681892714923e-05, "loss": 0.1331, "step": 20396 }, { "epoch": 14.333801827125791, "grad_norm": 0.7454769015312195, "learning_rate": 2.3781213398922466e-05, "loss": 0.1447, "step": 20397 }, { "epoch": 14.334504567814477, "grad_norm": 0.3218628764152527, "learning_rate": 2.378074490513001e-05, "loss": 0.0705, "step": 20398 }, { "epoch": 14.335207308503163, "grad_norm": 0.17204515635967255, "learning_rate": 2.378027641133755e-05, "loss": 0.0188, "step": 20399 }, { "epoch": 14.335910049191849, "grad_norm": 0.1747194528579712, "learning_rate": 2.3779807917545094e-05, "loss": 0.0277, "step": 20400 }, { "epoch": 14.336612789880535, "grad_norm": 0.10326934605836868, "learning_rate": 2.3779339423752638e-05, "loss": 0.0174, "step": 20401 }, { "epoch": 14.33731553056922, "grad_norm": 0.09919680655002594, "learning_rate": 2.3778870929960178e-05, "loss": 0.0122, "step": 20402 }, { "epoch": 14.338018271257905, "grad_norm": 0.06700798124074936, "learning_rate": 2.377840243616772e-05, "loss": 0.013, "step": 20403 }, { "epoch": 14.33872101194659, "grad_norm": 0.1205156221985817, "learning_rate": 2.3777933942375262e-05, "loss": 0.0107, "step": 20404 }, { "epoch": 14.339423752635277, "grad_norm": 0.06927991658449173, "learning_rate": 2.3777465448582806e-05, "loss": 0.0062, "step": 20405 }, { "epoch": 14.340126493323963, "grad_norm": 0.1456010639667511, "learning_rate": 2.377699695479035e-05, "loss": 0.0275, "step": 20406 }, { "epoch": 14.340829234012649, "grad_norm": 0.18428784608840942, "learning_rate": 2.3776528460997893e-05, "loss": 0.0113, "step": 20407 }, { "epoch": 14.341531974701335, "grad_norm": 0.49725157022476196, "learning_rate": 2.3776059967205434e-05, "loss": 0.017, "step": 20408 }, { "epoch": 14.34223471539002, "grad_norm": 0.10052033513784409, "learning_rate": 2.3775591473412977e-05, "loss": 0.0085, "step": 20409 }, { "epoch": 14.342937456078706, "grad_norm": 0.1299896240234375, "learning_rate": 2.377512297962052e-05, "loss": 0.0265, "step": 20410 }, { "epoch": 14.343640196767392, "grad_norm": 0.13157787919044495, "learning_rate": 2.3774654485828065e-05, "loss": 0.0126, "step": 20411 }, { "epoch": 14.344342937456078, "grad_norm": 0.12370006740093231, "learning_rate": 2.3774185992035605e-05, "loss": 0.0269, "step": 20412 }, { "epoch": 14.345045678144764, "grad_norm": 0.2609400451183319, "learning_rate": 2.377371749824315e-05, "loss": 0.0355, "step": 20413 }, { "epoch": 14.34574841883345, "grad_norm": 0.2301512509584427, "learning_rate": 2.3773249004450693e-05, "loss": 0.0168, "step": 20414 }, { "epoch": 14.346451159522136, "grad_norm": 0.30521127581596375, "learning_rate": 2.3772780510658236e-05, "loss": 0.0293, "step": 20415 }, { "epoch": 14.347153900210822, "grad_norm": 0.5951382517814636, "learning_rate": 2.3772312016865777e-05, "loss": 0.0272, "step": 20416 }, { "epoch": 14.347856640899508, "grad_norm": 0.18505387008190155, "learning_rate": 2.377184352307332e-05, "loss": 0.0302, "step": 20417 }, { "epoch": 14.348559381588194, "grad_norm": 0.3775826394557953, "learning_rate": 2.3771375029280864e-05, "loss": 0.067, "step": 20418 }, { "epoch": 14.34926212227688, "grad_norm": 0.39625418186187744, "learning_rate": 2.3770906535488408e-05, "loss": 0.0858, "step": 20419 }, { "epoch": 14.349964862965566, "grad_norm": 0.45430347323417664, "learning_rate": 2.3770438041695948e-05, "loss": 0.121, "step": 20420 }, { "epoch": 14.350667603654252, "grad_norm": 1.6482436656951904, "learning_rate": 2.376996954790349e-05, "loss": 0.1409, "step": 20421 }, { "epoch": 14.351370344342937, "grad_norm": 0.811961829662323, "learning_rate": 2.3769501054111032e-05, "loss": 0.153, "step": 20422 }, { "epoch": 14.352073085031623, "grad_norm": 0.2187490612268448, "learning_rate": 2.3769032560318576e-05, "loss": 0.0608, "step": 20423 }, { "epoch": 14.35277582572031, "grad_norm": 0.1079246774315834, "learning_rate": 2.376856406652612e-05, "loss": 0.0195, "step": 20424 }, { "epoch": 14.353478566408995, "grad_norm": 0.22248665988445282, "learning_rate": 2.376809557273366e-05, "loss": 0.0208, "step": 20425 }, { "epoch": 14.354181307097681, "grad_norm": 0.7256488800048828, "learning_rate": 2.3767627078941204e-05, "loss": 0.0202, "step": 20426 }, { "epoch": 14.354884047786367, "grad_norm": 0.11917201429605484, "learning_rate": 2.3767158585148748e-05, "loss": 0.0173, "step": 20427 }, { "epoch": 14.355586788475053, "grad_norm": 0.06288705766201019, "learning_rate": 2.376669009135629e-05, "loss": 0.0086, "step": 20428 }, { "epoch": 14.356289529163739, "grad_norm": 0.0636749342083931, "learning_rate": 2.376622159756383e-05, "loss": 0.004, "step": 20429 }, { "epoch": 14.356992269852425, "grad_norm": 0.131113201379776, "learning_rate": 2.3765753103771375e-05, "loss": 0.016, "step": 20430 }, { "epoch": 14.35769501054111, "grad_norm": 0.09362679719924927, "learning_rate": 2.376528460997892e-05, "loss": 0.0108, "step": 20431 }, { "epoch": 14.358397751229797, "grad_norm": 0.18309961259365082, "learning_rate": 2.3764816116186463e-05, "loss": 0.0126, "step": 20432 }, { "epoch": 14.359100491918483, "grad_norm": 0.19295020401477814, "learning_rate": 2.3764347622394006e-05, "loss": 0.0305, "step": 20433 }, { "epoch": 14.359803232607169, "grad_norm": 0.29915040731430054, "learning_rate": 2.3763879128601547e-05, "loss": 0.0199, "step": 20434 }, { "epoch": 14.360505973295854, "grad_norm": 0.1881340891122818, "learning_rate": 2.376341063480909e-05, "loss": 0.0267, "step": 20435 }, { "epoch": 14.36120871398454, "grad_norm": 0.14083264768123627, "learning_rate": 2.3762942141016634e-05, "loss": 0.0153, "step": 20436 }, { "epoch": 14.361911454673226, "grad_norm": 0.14507664740085602, "learning_rate": 2.3762473647224175e-05, "loss": 0.0191, "step": 20437 }, { "epoch": 14.362614195361912, "grad_norm": 0.24463388323783875, "learning_rate": 2.3762005153431715e-05, "loss": 0.0468, "step": 20438 }, { "epoch": 14.363316936050598, "grad_norm": 0.26740387082099915, "learning_rate": 2.376153665963926e-05, "loss": 0.0185, "step": 20439 }, { "epoch": 14.364019676739284, "grad_norm": 0.10281834006309509, "learning_rate": 2.3761068165846802e-05, "loss": 0.0239, "step": 20440 }, { "epoch": 14.36472241742797, "grad_norm": 0.16314145922660828, "learning_rate": 2.3760599672054346e-05, "loss": 0.0311, "step": 20441 }, { "epoch": 14.365425158116654, "grad_norm": 0.277391642332077, "learning_rate": 2.3760131178261886e-05, "loss": 0.0444, "step": 20442 }, { "epoch": 14.36612789880534, "grad_norm": 0.5123058557510376, "learning_rate": 2.375966268446943e-05, "loss": 0.0897, "step": 20443 }, { "epoch": 14.366830639494026, "grad_norm": 0.3653400242328644, "learning_rate": 2.3759194190676974e-05, "loss": 0.0924, "step": 20444 }, { "epoch": 14.367533380182712, "grad_norm": 0.3555225133895874, "learning_rate": 2.3758725696884518e-05, "loss": 0.116, "step": 20445 }, { "epoch": 14.368236120871398, "grad_norm": 0.5370108485221863, "learning_rate": 2.375825720309206e-05, "loss": 0.1356, "step": 20446 }, { "epoch": 14.368938861560084, "grad_norm": 0.9367749094963074, "learning_rate": 2.3757788709299602e-05, "loss": 0.2213, "step": 20447 }, { "epoch": 14.36964160224877, "grad_norm": 0.18994154036045074, "learning_rate": 2.3757320215507145e-05, "loss": 0.0595, "step": 20448 }, { "epoch": 14.370344342937456, "grad_norm": 0.12949597835540771, "learning_rate": 2.375685172171469e-05, "loss": 0.0157, "step": 20449 }, { "epoch": 14.371047083626141, "grad_norm": 0.23871201276779175, "learning_rate": 2.3756383227922233e-05, "loss": 0.0134, "step": 20450 }, { "epoch": 14.371749824314827, "grad_norm": 0.2856038510799408, "learning_rate": 2.3755914734129773e-05, "loss": 0.0175, "step": 20451 }, { "epoch": 14.372452565003513, "grad_norm": 0.08978644758462906, "learning_rate": 2.3755446240337317e-05, "loss": 0.0103, "step": 20452 }, { "epoch": 14.3731553056922, "grad_norm": 0.09329761564731598, "learning_rate": 2.375497774654486e-05, "loss": 0.0135, "step": 20453 }, { "epoch": 14.373858046380885, "grad_norm": 0.06610386073589325, "learning_rate": 2.3754509252752404e-05, "loss": 0.0129, "step": 20454 }, { "epoch": 14.374560787069571, "grad_norm": 0.14270441234111786, "learning_rate": 2.375404075895994e-05, "loss": 0.0187, "step": 20455 }, { "epoch": 14.375263527758257, "grad_norm": 0.07406899333000183, "learning_rate": 2.3753572265167485e-05, "loss": 0.014, "step": 20456 }, { "epoch": 14.375966268446943, "grad_norm": 0.1640392243862152, "learning_rate": 2.375310377137503e-05, "loss": 0.0124, "step": 20457 }, { "epoch": 14.376669009135629, "grad_norm": 0.13150976598262787, "learning_rate": 2.3752635277582573e-05, "loss": 0.0257, "step": 20458 }, { "epoch": 14.377371749824315, "grad_norm": 0.16125065088272095, "learning_rate": 2.3752166783790116e-05, "loss": 0.009, "step": 20459 }, { "epoch": 14.378074490513, "grad_norm": 0.1624949872493744, "learning_rate": 2.3751698289997657e-05, "loss": 0.0249, "step": 20460 }, { "epoch": 14.378777231201687, "grad_norm": 0.07179860770702362, "learning_rate": 2.37512297962052e-05, "loss": 0.0134, "step": 20461 }, { "epoch": 14.379479971890373, "grad_norm": 0.17459386587142944, "learning_rate": 2.3750761302412744e-05, "loss": 0.023, "step": 20462 }, { "epoch": 14.380182712579058, "grad_norm": 0.18726004660129547, "learning_rate": 2.3750292808620288e-05, "loss": 0.0397, "step": 20463 }, { "epoch": 14.380885453267744, "grad_norm": 0.27390486001968384, "learning_rate": 2.3749824314827828e-05, "loss": 0.0171, "step": 20464 }, { "epoch": 14.38158819395643, "grad_norm": 0.6096564531326294, "learning_rate": 2.3749355821035372e-05, "loss": 0.0314, "step": 20465 }, { "epoch": 14.382290934645116, "grad_norm": 0.20369786024093628, "learning_rate": 2.3748887327242916e-05, "loss": 0.0338, "step": 20466 }, { "epoch": 14.382993675333802, "grad_norm": 0.19252994656562805, "learning_rate": 2.374841883345046e-05, "loss": 0.0401, "step": 20467 }, { "epoch": 14.383696416022488, "grad_norm": 0.8624212741851807, "learning_rate": 2.3747950339658e-05, "loss": 0.0646, "step": 20468 }, { "epoch": 14.384399156711174, "grad_norm": 0.4985280930995941, "learning_rate": 2.3747481845865543e-05, "loss": 0.1119, "step": 20469 }, { "epoch": 14.38510189739986, "grad_norm": 0.6255429983139038, "learning_rate": 2.3747013352073087e-05, "loss": 0.1146, "step": 20470 }, { "epoch": 14.385804638088546, "grad_norm": 0.8196946978569031, "learning_rate": 2.374654485828063e-05, "loss": 0.1227, "step": 20471 }, { "epoch": 14.386507378777232, "grad_norm": 1.4775484800338745, "learning_rate": 2.374607636448817e-05, "loss": 0.1712, "step": 20472 }, { "epoch": 14.387210119465918, "grad_norm": 0.27853313088417053, "learning_rate": 2.374560787069571e-05, "loss": 0.0647, "step": 20473 }, { "epoch": 14.387912860154604, "grad_norm": 0.0859784483909607, "learning_rate": 2.3745139376903255e-05, "loss": 0.0209, "step": 20474 }, { "epoch": 14.38861560084329, "grad_norm": 0.11568311601877213, "learning_rate": 2.37446708831108e-05, "loss": 0.0254, "step": 20475 }, { "epoch": 14.389318341531975, "grad_norm": 0.11829785257577896, "learning_rate": 2.3744202389318343e-05, "loss": 0.014, "step": 20476 }, { "epoch": 14.390021082220661, "grad_norm": 0.07562298327684402, "learning_rate": 2.3743733895525883e-05, "loss": 0.0098, "step": 20477 }, { "epoch": 14.390723822909347, "grad_norm": 0.06994006782770157, "learning_rate": 2.3743265401733427e-05, "loss": 0.0044, "step": 20478 }, { "epoch": 14.391426563598033, "grad_norm": 0.10448511689901352, "learning_rate": 2.374279690794097e-05, "loss": 0.0139, "step": 20479 }, { "epoch": 14.392129304286719, "grad_norm": 0.23934894800186157, "learning_rate": 2.3742328414148514e-05, "loss": 0.0341, "step": 20480 }, { "epoch": 14.392832044975403, "grad_norm": 0.11833016574382782, "learning_rate": 2.3741859920356054e-05, "loss": 0.0178, "step": 20481 }, { "epoch": 14.39353478566409, "grad_norm": 0.15745016932487488, "learning_rate": 2.3741391426563598e-05, "loss": 0.0125, "step": 20482 }, { "epoch": 14.394237526352775, "grad_norm": 0.09474314749240875, "learning_rate": 2.3740922932771142e-05, "loss": 0.0166, "step": 20483 }, { "epoch": 14.394940267041461, "grad_norm": 0.08987967669963837, "learning_rate": 2.3740454438978686e-05, "loss": 0.0075, "step": 20484 }, { "epoch": 14.395643007730147, "grad_norm": 0.18296657502651215, "learning_rate": 2.373998594518623e-05, "loss": 0.0376, "step": 20485 }, { "epoch": 14.396345748418833, "grad_norm": 0.13118045032024384, "learning_rate": 2.373951745139377e-05, "loss": 0.0113, "step": 20486 }, { "epoch": 14.397048489107519, "grad_norm": 0.16599103808403015, "learning_rate": 2.3739048957601313e-05, "loss": 0.0331, "step": 20487 }, { "epoch": 14.397751229796205, "grad_norm": 0.12859676778316498, "learning_rate": 2.3738580463808857e-05, "loss": 0.0235, "step": 20488 }, { "epoch": 14.39845397048489, "grad_norm": 0.08598877489566803, "learning_rate": 2.37381119700164e-05, "loss": 0.0092, "step": 20489 }, { "epoch": 14.399156711173577, "grad_norm": 0.24419419467449188, "learning_rate": 2.3737643476223938e-05, "loss": 0.0315, "step": 20490 }, { "epoch": 14.399859451862262, "grad_norm": 0.3470710515975952, "learning_rate": 2.373717498243148e-05, "loss": 0.0369, "step": 20491 }, { "epoch": 14.400562192550948, "grad_norm": 0.24182796478271484, "learning_rate": 2.3736706488639025e-05, "loss": 0.041, "step": 20492 }, { "epoch": 14.401264933239634, "grad_norm": 0.401315838098526, "learning_rate": 2.373623799484657e-05, "loss": 0.0614, "step": 20493 }, { "epoch": 14.40196767392832, "grad_norm": 0.781689465045929, "learning_rate": 2.373576950105411e-05, "loss": 0.0838, "step": 20494 }, { "epoch": 14.402670414617006, "grad_norm": 0.549705445766449, "learning_rate": 2.3735301007261653e-05, "loss": 0.1155, "step": 20495 }, { "epoch": 14.403373155305692, "grad_norm": 1.1447138786315918, "learning_rate": 2.3734832513469197e-05, "loss": 0.1482, "step": 20496 }, { "epoch": 14.404075895994378, "grad_norm": 1.181355595588684, "learning_rate": 2.373436401967674e-05, "loss": 0.1363, "step": 20497 }, { "epoch": 14.404778636683064, "grad_norm": 0.18103311955928802, "learning_rate": 2.3733895525884284e-05, "loss": 0.0537, "step": 20498 }, { "epoch": 14.40548137737175, "grad_norm": 0.1443619579076767, "learning_rate": 2.3733427032091825e-05, "loss": 0.0329, "step": 20499 }, { "epoch": 14.406184118060436, "grad_norm": 0.17356087267398834, "learning_rate": 2.373295853829937e-05, "loss": 0.045, "step": 20500 }, { "epoch": 14.406886858749122, "grad_norm": 0.08523741364479065, "learning_rate": 2.3732490044506912e-05, "loss": 0.0108, "step": 20501 }, { "epoch": 14.407589599437808, "grad_norm": 0.14557814598083496, "learning_rate": 2.3732021550714456e-05, "loss": 0.0183, "step": 20502 }, { "epoch": 14.408292340126494, "grad_norm": 0.16348592936992645, "learning_rate": 2.3731553056921996e-05, "loss": 0.0098, "step": 20503 }, { "epoch": 14.40899508081518, "grad_norm": 0.06599780917167664, "learning_rate": 2.373108456312954e-05, "loss": 0.0077, "step": 20504 }, { "epoch": 14.409697821503865, "grad_norm": 0.09632927924394608, "learning_rate": 2.3730616069337084e-05, "loss": 0.0165, "step": 20505 }, { "epoch": 14.410400562192551, "grad_norm": 0.2658715546131134, "learning_rate": 2.3730147575544627e-05, "loss": 0.0155, "step": 20506 }, { "epoch": 14.411103302881237, "grad_norm": 0.10128217935562134, "learning_rate": 2.3729679081752164e-05, "loss": 0.01, "step": 20507 }, { "epoch": 14.411806043569923, "grad_norm": 0.2424546182155609, "learning_rate": 2.3729210587959708e-05, "loss": 0.0232, "step": 20508 }, { "epoch": 14.412508784258609, "grad_norm": 0.10904072970151901, "learning_rate": 2.372874209416725e-05, "loss": 0.0124, "step": 20509 }, { "epoch": 14.413211524947295, "grad_norm": 0.1753416657447815, "learning_rate": 2.3728273600374795e-05, "loss": 0.0369, "step": 20510 }, { "epoch": 14.41391426563598, "grad_norm": 0.15592733025550842, "learning_rate": 2.372780510658234e-05, "loss": 0.0097, "step": 20511 }, { "epoch": 14.414617006324667, "grad_norm": 0.1048106774687767, "learning_rate": 2.372733661278988e-05, "loss": 0.0181, "step": 20512 }, { "epoch": 14.415319747013353, "grad_norm": 0.13093878328800201, "learning_rate": 2.3726868118997423e-05, "loss": 0.0336, "step": 20513 }, { "epoch": 14.416022487702039, "grad_norm": 0.15894941985607147, "learning_rate": 2.3726399625204967e-05, "loss": 0.0158, "step": 20514 }, { "epoch": 14.416725228390725, "grad_norm": 0.16375648975372314, "learning_rate": 2.372593113141251e-05, "loss": 0.0258, "step": 20515 }, { "epoch": 14.41742796907941, "grad_norm": 0.37545323371887207, "learning_rate": 2.372546263762005e-05, "loss": 0.039, "step": 20516 }, { "epoch": 14.418130709768096, "grad_norm": 1.1961127519607544, "learning_rate": 2.3724994143827595e-05, "loss": 0.0443, "step": 20517 }, { "epoch": 14.41883345045678, "grad_norm": 0.19535286724567413, "learning_rate": 2.372452565003514e-05, "loss": 0.0788, "step": 20518 }, { "epoch": 14.419536191145466, "grad_norm": 0.6697192788124084, "learning_rate": 2.3724057156242682e-05, "loss": 0.09, "step": 20519 }, { "epoch": 14.420238931834152, "grad_norm": 0.5787156820297241, "learning_rate": 2.3723588662450222e-05, "loss": 0.1305, "step": 20520 }, { "epoch": 14.420941672522838, "grad_norm": 0.49752387404441833, "learning_rate": 2.3723120168657766e-05, "loss": 0.1697, "step": 20521 }, { "epoch": 14.421644413211524, "grad_norm": 1.4113774299621582, "learning_rate": 2.372265167486531e-05, "loss": 0.1797, "step": 20522 }, { "epoch": 14.42234715390021, "grad_norm": 0.2677002251148224, "learning_rate": 2.3722183181072854e-05, "loss": 0.0579, "step": 20523 }, { "epoch": 14.423049894588896, "grad_norm": 0.08110800385475159, "learning_rate": 2.3721714687280394e-05, "loss": 0.0158, "step": 20524 }, { "epoch": 14.423752635277582, "grad_norm": 0.1580856889486313, "learning_rate": 2.3721246193487934e-05, "loss": 0.0191, "step": 20525 }, { "epoch": 14.424455375966268, "grad_norm": 0.17269036173820496, "learning_rate": 2.3720777699695478e-05, "loss": 0.0115, "step": 20526 }, { "epoch": 14.425158116654954, "grad_norm": 0.16288641095161438, "learning_rate": 2.3720309205903022e-05, "loss": 0.0244, "step": 20527 }, { "epoch": 14.42586085734364, "grad_norm": 0.07990420609712601, "learning_rate": 2.3719840712110566e-05, "loss": 0.019, "step": 20528 }, { "epoch": 14.426563598032326, "grad_norm": 0.13204680383205414, "learning_rate": 2.3719372218318106e-05, "loss": 0.0095, "step": 20529 }, { "epoch": 14.427266338721012, "grad_norm": 0.12879876792430878, "learning_rate": 2.371890372452565e-05, "loss": 0.0226, "step": 20530 }, { "epoch": 14.427969079409698, "grad_norm": 0.14629167318344116, "learning_rate": 2.3718435230733193e-05, "loss": 0.031, "step": 20531 }, { "epoch": 14.428671820098383, "grad_norm": 0.1470000147819519, "learning_rate": 2.3717966736940737e-05, "loss": 0.0062, "step": 20532 }, { "epoch": 14.42937456078707, "grad_norm": 0.23326827585697174, "learning_rate": 2.3717498243148277e-05, "loss": 0.0198, "step": 20533 }, { "epoch": 14.430077301475755, "grad_norm": 0.16013990342617035, "learning_rate": 2.371702974935582e-05, "loss": 0.0162, "step": 20534 }, { "epoch": 14.430780042164441, "grad_norm": 0.21014158427715302, "learning_rate": 2.3716561255563365e-05, "loss": 0.0299, "step": 20535 }, { "epoch": 14.431482782853127, "grad_norm": 0.3553679287433624, "learning_rate": 2.371609276177091e-05, "loss": 0.0188, "step": 20536 }, { "epoch": 14.432185523541813, "grad_norm": 0.10203681886196136, "learning_rate": 2.3715624267978452e-05, "loss": 0.0143, "step": 20537 }, { "epoch": 14.432888264230499, "grad_norm": 0.19060200452804565, "learning_rate": 2.3715155774185993e-05, "loss": 0.0412, "step": 20538 }, { "epoch": 14.433591004919185, "grad_norm": 0.1839427649974823, "learning_rate": 2.3714687280393536e-05, "loss": 0.0245, "step": 20539 }, { "epoch": 14.43429374560787, "grad_norm": 0.3192102015018463, "learning_rate": 2.371421878660108e-05, "loss": 0.031, "step": 20540 }, { "epoch": 14.434996486296557, "grad_norm": 0.23149727284908295, "learning_rate": 2.3713750292808624e-05, "loss": 0.032, "step": 20541 }, { "epoch": 14.435699226985243, "grad_norm": 0.6220602989196777, "learning_rate": 2.371328179901616e-05, "loss": 0.044, "step": 20542 }, { "epoch": 14.436401967673929, "grad_norm": 0.47090059518814087, "learning_rate": 2.3712813305223704e-05, "loss": 0.0646, "step": 20543 }, { "epoch": 14.437104708362615, "grad_norm": 0.4678957462310791, "learning_rate": 2.3712344811431248e-05, "loss": 0.0912, "step": 20544 }, { "epoch": 14.4378074490513, "grad_norm": 0.8212683200836182, "learning_rate": 2.3711876317638792e-05, "loss": 0.1319, "step": 20545 }, { "epoch": 14.438510189739986, "grad_norm": 1.48190176486969, "learning_rate": 2.3711407823846332e-05, "loss": 0.1567, "step": 20546 }, { "epoch": 14.439212930428672, "grad_norm": 1.2351895570755005, "learning_rate": 2.3710939330053876e-05, "loss": 0.2095, "step": 20547 }, { "epoch": 14.439915671117358, "grad_norm": 0.3908681869506836, "learning_rate": 2.371047083626142e-05, "loss": 0.0489, "step": 20548 }, { "epoch": 14.440618411806044, "grad_norm": 0.10070295631885529, "learning_rate": 2.3710002342468963e-05, "loss": 0.0217, "step": 20549 }, { "epoch": 14.44132115249473, "grad_norm": 0.22503265738487244, "learning_rate": 2.3709533848676507e-05, "loss": 0.015, "step": 20550 }, { "epoch": 14.442023893183416, "grad_norm": 0.10391383618116379, "learning_rate": 2.3709065354884047e-05, "loss": 0.0164, "step": 20551 }, { "epoch": 14.442726633872102, "grad_norm": 0.12500052154064178, "learning_rate": 2.370859686109159e-05, "loss": 0.0312, "step": 20552 }, { "epoch": 14.443429374560788, "grad_norm": 0.10416525602340698, "learning_rate": 2.3708128367299135e-05, "loss": 0.0178, "step": 20553 }, { "epoch": 14.444132115249474, "grad_norm": 0.1171952560544014, "learning_rate": 2.370765987350668e-05, "loss": 0.0136, "step": 20554 }, { "epoch": 14.44483485593816, "grad_norm": 0.09291645884513855, "learning_rate": 2.370719137971422e-05, "loss": 0.0101, "step": 20555 }, { "epoch": 14.445537596626846, "grad_norm": 0.13980960845947266, "learning_rate": 2.3706722885921763e-05, "loss": 0.0179, "step": 20556 }, { "epoch": 14.44624033731553, "grad_norm": 0.6607435941696167, "learning_rate": 2.3706254392129306e-05, "loss": 0.0254, "step": 20557 }, { "epoch": 14.446943078004216, "grad_norm": 0.21560519933700562, "learning_rate": 2.370578589833685e-05, "loss": 0.0248, "step": 20558 }, { "epoch": 14.447645818692902, "grad_norm": 0.15103772282600403, "learning_rate": 2.3705317404544387e-05, "loss": 0.017, "step": 20559 }, { "epoch": 14.448348559381587, "grad_norm": 0.13762477040290833, "learning_rate": 2.370484891075193e-05, "loss": 0.0311, "step": 20560 }, { "epoch": 14.449051300070273, "grad_norm": 0.20663806796073914, "learning_rate": 2.3704380416959475e-05, "loss": 0.0145, "step": 20561 }, { "epoch": 14.44975404075896, "grad_norm": 0.15769898891448975, "learning_rate": 2.3703911923167018e-05, "loss": 0.0257, "step": 20562 }, { "epoch": 14.450456781447645, "grad_norm": 0.19707909226417542, "learning_rate": 2.3703443429374562e-05, "loss": 0.0263, "step": 20563 }, { "epoch": 14.451159522136331, "grad_norm": 0.09824106842279434, "learning_rate": 2.3702974935582102e-05, "loss": 0.0231, "step": 20564 }, { "epoch": 14.451862262825017, "grad_norm": 0.1051153615117073, "learning_rate": 2.3702506441789646e-05, "loss": 0.0256, "step": 20565 }, { "epoch": 14.452565003513703, "grad_norm": 0.29575562477111816, "learning_rate": 2.370203794799719e-05, "loss": 0.0342, "step": 20566 }, { "epoch": 14.453267744202389, "grad_norm": 0.48123404383659363, "learning_rate": 2.3701569454204734e-05, "loss": 0.0635, "step": 20567 }, { "epoch": 14.453970484891075, "grad_norm": 0.5469598174095154, "learning_rate": 2.3701100960412274e-05, "loss": 0.0399, "step": 20568 }, { "epoch": 14.45467322557976, "grad_norm": 0.5282831788063049, "learning_rate": 2.3700632466619818e-05, "loss": 0.1026, "step": 20569 }, { "epoch": 14.455375966268447, "grad_norm": 0.6131752133369446, "learning_rate": 2.370016397282736e-05, "loss": 0.1162, "step": 20570 }, { "epoch": 14.456078706957133, "grad_norm": 0.7489452362060547, "learning_rate": 2.3699695479034905e-05, "loss": 0.1493, "step": 20571 }, { "epoch": 14.456781447645819, "grad_norm": 0.7085833549499512, "learning_rate": 2.3699226985242445e-05, "loss": 0.1824, "step": 20572 }, { "epoch": 14.457484188334504, "grad_norm": 0.7723261117935181, "learning_rate": 2.369875849144999e-05, "loss": 0.084, "step": 20573 }, { "epoch": 14.45818692902319, "grad_norm": 0.19754774868488312, "learning_rate": 2.3698289997657533e-05, "loss": 0.0309, "step": 20574 }, { "epoch": 14.458889669711876, "grad_norm": 0.070857472717762, "learning_rate": 2.3697821503865077e-05, "loss": 0.0101, "step": 20575 }, { "epoch": 14.459592410400562, "grad_norm": 0.1141519695520401, "learning_rate": 2.369735301007262e-05, "loss": 0.0168, "step": 20576 }, { "epoch": 14.460295151089248, "grad_norm": 0.22915300726890564, "learning_rate": 2.3696884516280157e-05, "loss": 0.0189, "step": 20577 }, { "epoch": 14.460997891777934, "grad_norm": 0.28442278504371643, "learning_rate": 2.36964160224877e-05, "loss": 0.0175, "step": 20578 }, { "epoch": 14.46170063246662, "grad_norm": 0.23569510877132416, "learning_rate": 2.3695947528695245e-05, "loss": 0.0239, "step": 20579 }, { "epoch": 14.462403373155306, "grad_norm": 0.12377749383449554, "learning_rate": 2.369547903490279e-05, "loss": 0.0181, "step": 20580 }, { "epoch": 14.463106113843992, "grad_norm": 0.2081107497215271, "learning_rate": 2.369501054111033e-05, "loss": 0.0256, "step": 20581 }, { "epoch": 14.463808854532678, "grad_norm": 0.1479056179523468, "learning_rate": 2.3694542047317872e-05, "loss": 0.0107, "step": 20582 }, { "epoch": 14.464511595221364, "grad_norm": 0.2801207900047302, "learning_rate": 2.3694073553525416e-05, "loss": 0.0273, "step": 20583 }, { "epoch": 14.46521433591005, "grad_norm": 0.0740121454000473, "learning_rate": 2.369360505973296e-05, "loss": 0.0086, "step": 20584 }, { "epoch": 14.465917076598735, "grad_norm": 0.13742311298847198, "learning_rate": 2.36931365659405e-05, "loss": 0.0175, "step": 20585 }, { "epoch": 14.466619817287421, "grad_norm": 0.0708884745836258, "learning_rate": 2.3692668072148044e-05, "loss": 0.0146, "step": 20586 }, { "epoch": 14.467322557976107, "grad_norm": 0.1080377921462059, "learning_rate": 2.3692199578355588e-05, "loss": 0.017, "step": 20587 }, { "epoch": 14.468025298664793, "grad_norm": 0.1662980318069458, "learning_rate": 2.369173108456313e-05, "loss": 0.0324, "step": 20588 }, { "epoch": 14.46872803935348, "grad_norm": 0.16179050505161285, "learning_rate": 2.3691262590770675e-05, "loss": 0.0204, "step": 20589 }, { "epoch": 14.469430780042165, "grad_norm": 0.17685961723327637, "learning_rate": 2.3690794096978215e-05, "loss": 0.0293, "step": 20590 }, { "epoch": 14.470133520730851, "grad_norm": 0.13478721678256989, "learning_rate": 2.369032560318576e-05, "loss": 0.0221, "step": 20591 }, { "epoch": 14.470836261419537, "grad_norm": 0.2849726676940918, "learning_rate": 2.3689857109393303e-05, "loss": 0.064, "step": 20592 }, { "epoch": 14.471539002108223, "grad_norm": 0.19215697050094604, "learning_rate": 2.3689388615600847e-05, "loss": 0.0561, "step": 20593 }, { "epoch": 14.472241742796909, "grad_norm": 0.7024012207984924, "learning_rate": 2.3688920121808384e-05, "loss": 0.1108, "step": 20594 }, { "epoch": 14.472944483485595, "grad_norm": 0.4666334092617035, "learning_rate": 2.3688451628015927e-05, "loss": 0.146, "step": 20595 }, { "epoch": 14.473647224174279, "grad_norm": 0.5569031238555908, "learning_rate": 2.368798313422347e-05, "loss": 0.1338, "step": 20596 }, { "epoch": 14.474349964862965, "grad_norm": 0.971000075340271, "learning_rate": 2.3687514640431015e-05, "loss": 0.1624, "step": 20597 }, { "epoch": 14.47505270555165, "grad_norm": 0.2664334177970886, "learning_rate": 2.368704614663856e-05, "loss": 0.0698, "step": 20598 }, { "epoch": 14.475755446240337, "grad_norm": 0.20394940674304962, "learning_rate": 2.36865776528461e-05, "loss": 0.0216, "step": 20599 }, { "epoch": 14.476458186929023, "grad_norm": 0.16448312997817993, "learning_rate": 2.3686109159053643e-05, "loss": 0.02, "step": 20600 }, { "epoch": 14.477160927617708, "grad_norm": 0.07795067876577377, "learning_rate": 2.3685640665261186e-05, "loss": 0.0149, "step": 20601 }, { "epoch": 14.477863668306394, "grad_norm": 0.10047990083694458, "learning_rate": 2.368517217146873e-05, "loss": 0.0098, "step": 20602 }, { "epoch": 14.47856640899508, "grad_norm": 0.262666255235672, "learning_rate": 2.368470367767627e-05, "loss": 0.0193, "step": 20603 }, { "epoch": 14.479269149683766, "grad_norm": 0.1362610012292862, "learning_rate": 2.3684235183883814e-05, "loss": 0.0213, "step": 20604 }, { "epoch": 14.479971890372452, "grad_norm": 0.098471999168396, "learning_rate": 2.3683766690091358e-05, "loss": 0.0132, "step": 20605 }, { "epoch": 14.480674631061138, "grad_norm": 0.13930226862430573, "learning_rate": 2.36832981962989e-05, "loss": 0.0169, "step": 20606 }, { "epoch": 14.481377371749824, "grad_norm": 0.19495652616024017, "learning_rate": 2.3682829702506442e-05, "loss": 0.0118, "step": 20607 }, { "epoch": 14.48208011243851, "grad_norm": 0.22772477567195892, "learning_rate": 2.3682361208713986e-05, "loss": 0.0144, "step": 20608 }, { "epoch": 14.482782853127196, "grad_norm": 0.07155930995941162, "learning_rate": 2.368189271492153e-05, "loss": 0.0091, "step": 20609 }, { "epoch": 14.483485593815882, "grad_norm": 0.16025415062904358, "learning_rate": 2.3681424221129073e-05, "loss": 0.0271, "step": 20610 }, { "epoch": 14.484188334504568, "grad_norm": 0.3702148199081421, "learning_rate": 2.3680955727336613e-05, "loss": 0.0256, "step": 20611 }, { "epoch": 14.484891075193254, "grad_norm": 0.18797317147254944, "learning_rate": 2.3680487233544154e-05, "loss": 0.0303, "step": 20612 }, { "epoch": 14.48559381588194, "grad_norm": 0.3911376893520355, "learning_rate": 2.3680018739751697e-05, "loss": 0.0427, "step": 20613 }, { "epoch": 14.486296556570625, "grad_norm": 0.08705076575279236, "learning_rate": 2.367955024595924e-05, "loss": 0.0133, "step": 20614 }, { "epoch": 14.486999297259311, "grad_norm": 0.23865793645381927, "learning_rate": 2.3679081752166785e-05, "loss": 0.0325, "step": 20615 }, { "epoch": 14.487702037947997, "grad_norm": 0.23289217054843903, "learning_rate": 2.3678613258374325e-05, "loss": 0.0301, "step": 20616 }, { "epoch": 14.488404778636683, "grad_norm": 0.2608634829521179, "learning_rate": 2.367814476458187e-05, "loss": 0.0343, "step": 20617 }, { "epoch": 14.489107519325369, "grad_norm": 0.5167599320411682, "learning_rate": 2.3677676270789413e-05, "loss": 0.0489, "step": 20618 }, { "epoch": 14.489810260014055, "grad_norm": 0.6230285167694092, "learning_rate": 2.3677207776996956e-05, "loss": 0.1156, "step": 20619 }, { "epoch": 14.490513000702741, "grad_norm": 1.0529340505599976, "learning_rate": 2.3676739283204497e-05, "loss": 0.1581, "step": 20620 }, { "epoch": 14.491215741391427, "grad_norm": 1.0191515684127808, "learning_rate": 2.367627078941204e-05, "loss": 0.1544, "step": 20621 }, { "epoch": 14.491918482080113, "grad_norm": 1.3998844623565674, "learning_rate": 2.3675802295619584e-05, "loss": 0.2049, "step": 20622 }, { "epoch": 14.492621222768799, "grad_norm": 0.35982149839401245, "learning_rate": 2.3675333801827128e-05, "loss": 0.0563, "step": 20623 }, { "epoch": 14.493323963457485, "grad_norm": 0.09741070121526718, "learning_rate": 2.367486530803467e-05, "loss": 0.0186, "step": 20624 }, { "epoch": 14.49402670414617, "grad_norm": 0.10497429966926575, "learning_rate": 2.3674396814242212e-05, "loss": 0.0173, "step": 20625 }, { "epoch": 14.494729444834856, "grad_norm": 0.15597961843013763, "learning_rate": 2.3673928320449756e-05, "loss": 0.0143, "step": 20626 }, { "epoch": 14.495432185523542, "grad_norm": 0.19689279794692993, "learning_rate": 2.36734598266573e-05, "loss": 0.0165, "step": 20627 }, { "epoch": 14.496134926212228, "grad_norm": 0.14323967695236206, "learning_rate": 2.3672991332864843e-05, "loss": 0.0176, "step": 20628 }, { "epoch": 14.496837666900914, "grad_norm": 0.13835273683071136, "learning_rate": 2.367252283907238e-05, "loss": 0.0157, "step": 20629 }, { "epoch": 14.4975404075896, "grad_norm": 0.05237996578216553, "learning_rate": 2.3672054345279924e-05, "loss": 0.0068, "step": 20630 }, { "epoch": 14.498243148278286, "grad_norm": 0.28547051548957825, "learning_rate": 2.3671585851487468e-05, "loss": 0.0234, "step": 20631 }, { "epoch": 14.498945888966972, "grad_norm": 0.10781751573085785, "learning_rate": 2.367111735769501e-05, "loss": 0.0134, "step": 20632 }, { "epoch": 14.499648629655656, "grad_norm": 0.10335037112236023, "learning_rate": 2.367064886390255e-05, "loss": 0.0082, "step": 20633 }, { "epoch": 14.500351370344344, "grad_norm": 0.11574602872133255, "learning_rate": 2.3670180370110095e-05, "loss": 0.0161, "step": 20634 }, { "epoch": 14.501054111033028, "grad_norm": 0.1538262665271759, "learning_rate": 2.366971187631764e-05, "loss": 0.0238, "step": 20635 }, { "epoch": 14.501756851721714, "grad_norm": 0.13168089091777802, "learning_rate": 2.3669243382525183e-05, "loss": 0.0229, "step": 20636 }, { "epoch": 14.5024595924104, "grad_norm": 0.1394047886133194, "learning_rate": 2.3668774888732727e-05, "loss": 0.0242, "step": 20637 }, { "epoch": 14.503162333099086, "grad_norm": 0.6268937587738037, "learning_rate": 2.3668306394940267e-05, "loss": 0.0399, "step": 20638 }, { "epoch": 14.503865073787772, "grad_norm": 0.17804588377475739, "learning_rate": 2.366783790114781e-05, "loss": 0.0221, "step": 20639 }, { "epoch": 14.504567814476458, "grad_norm": 0.12500207126140594, "learning_rate": 2.3667369407355354e-05, "loss": 0.0283, "step": 20640 }, { "epoch": 14.505270555165144, "grad_norm": 0.150804340839386, "learning_rate": 2.3666900913562898e-05, "loss": 0.0267, "step": 20641 }, { "epoch": 14.50597329585383, "grad_norm": 0.1997969001531601, "learning_rate": 2.366643241977044e-05, "loss": 0.0396, "step": 20642 }, { "epoch": 14.506676036542515, "grad_norm": 0.5890460014343262, "learning_rate": 2.3665963925977982e-05, "loss": 0.1022, "step": 20643 }, { "epoch": 14.507378777231201, "grad_norm": 0.4715115427970886, "learning_rate": 2.3665495432185526e-05, "loss": 0.0861, "step": 20644 }, { "epoch": 14.508081517919887, "grad_norm": 0.6378836631774902, "learning_rate": 2.366502693839307e-05, "loss": 0.129, "step": 20645 }, { "epoch": 14.508784258608573, "grad_norm": 0.9170595407485962, "learning_rate": 2.3664558444600607e-05, "loss": 0.183, "step": 20646 }, { "epoch": 14.509486999297259, "grad_norm": 1.416050672531128, "learning_rate": 2.366408995080815e-05, "loss": 0.175, "step": 20647 }, { "epoch": 14.510189739985945, "grad_norm": 0.23392756283283234, "learning_rate": 2.3663621457015694e-05, "loss": 0.0465, "step": 20648 }, { "epoch": 14.510892480674631, "grad_norm": 0.0969986543059349, "learning_rate": 2.3663152963223238e-05, "loss": 0.0251, "step": 20649 }, { "epoch": 14.511595221363317, "grad_norm": 0.15844735503196716, "learning_rate": 2.366268446943078e-05, "loss": 0.0157, "step": 20650 }, { "epoch": 14.512297962052003, "grad_norm": 0.08656666427850723, "learning_rate": 2.3662215975638322e-05, "loss": 0.0109, "step": 20651 }, { "epoch": 14.513000702740689, "grad_norm": 0.10933815687894821, "learning_rate": 2.3661747481845865e-05, "loss": 0.0201, "step": 20652 }, { "epoch": 14.513703443429375, "grad_norm": 0.11027073860168457, "learning_rate": 2.366127898805341e-05, "loss": 0.0051, "step": 20653 }, { "epoch": 14.51440618411806, "grad_norm": 0.07683099061250687, "learning_rate": 2.3660810494260953e-05, "loss": 0.0117, "step": 20654 }, { "epoch": 14.515108924806746, "grad_norm": 0.1556500792503357, "learning_rate": 2.3660342000468493e-05, "loss": 0.0178, "step": 20655 }, { "epoch": 14.515811665495432, "grad_norm": 0.37641796469688416, "learning_rate": 2.3659873506676037e-05, "loss": 0.0162, "step": 20656 }, { "epoch": 14.516514406184118, "grad_norm": 0.237439826130867, "learning_rate": 2.365940501288358e-05, "loss": 0.0113, "step": 20657 }, { "epoch": 14.517217146872804, "grad_norm": 0.12205138057470322, "learning_rate": 2.3658936519091124e-05, "loss": 0.022, "step": 20658 }, { "epoch": 14.51791988756149, "grad_norm": 0.11865083873271942, "learning_rate": 2.3658468025298665e-05, "loss": 0.0237, "step": 20659 }, { "epoch": 14.518622628250176, "grad_norm": 0.13631711900234222, "learning_rate": 2.365799953150621e-05, "loss": 0.0142, "step": 20660 }, { "epoch": 14.519325368938862, "grad_norm": 0.22856701910495758, "learning_rate": 2.3657531037713752e-05, "loss": 0.0166, "step": 20661 }, { "epoch": 14.520028109627548, "grad_norm": 0.19207549095153809, "learning_rate": 2.3657062543921296e-05, "loss": 0.0187, "step": 20662 }, { "epoch": 14.520730850316234, "grad_norm": 0.24394400417804718, "learning_rate": 2.365659405012884e-05, "loss": 0.0318, "step": 20663 }, { "epoch": 14.52143359100492, "grad_norm": 0.3394234776496887, "learning_rate": 2.3656125556336377e-05, "loss": 0.0193, "step": 20664 }, { "epoch": 14.522136331693606, "grad_norm": 0.21289511024951935, "learning_rate": 2.365565706254392e-05, "loss": 0.0453, "step": 20665 }, { "epoch": 14.522839072382292, "grad_norm": 0.36098605394363403, "learning_rate": 2.3655188568751464e-05, "loss": 0.0296, "step": 20666 }, { "epoch": 14.523541813070977, "grad_norm": 0.3307998776435852, "learning_rate": 2.3654720074959008e-05, "loss": 0.0351, "step": 20667 }, { "epoch": 14.524244553759663, "grad_norm": 0.5610581636428833, "learning_rate": 2.3654251581166548e-05, "loss": 0.072, "step": 20668 }, { "epoch": 14.52494729444835, "grad_norm": 0.4952525794506073, "learning_rate": 2.3653783087374092e-05, "loss": 0.098, "step": 20669 }, { "epoch": 14.525650035137035, "grad_norm": 0.4576973617076874, "learning_rate": 2.3653314593581636e-05, "loss": 0.1117, "step": 20670 }, { "epoch": 14.526352775825721, "grad_norm": 0.7697075009346008, "learning_rate": 2.365284609978918e-05, "loss": 0.1616, "step": 20671 }, { "epoch": 14.527055516514405, "grad_norm": 0.849112868309021, "learning_rate": 2.365237760599672e-05, "loss": 0.1785, "step": 20672 }, { "epoch": 14.527758257203093, "grad_norm": 1.4305975437164307, "learning_rate": 2.3651909112204263e-05, "loss": 0.0728, "step": 20673 }, { "epoch": 14.528460997891777, "grad_norm": 0.08358428627252579, "learning_rate": 2.3651440618411807e-05, "loss": 0.0229, "step": 20674 }, { "epoch": 14.529163738580463, "grad_norm": 0.4787827134132385, "learning_rate": 2.365097212461935e-05, "loss": 0.0235, "step": 20675 }, { "epoch": 14.529866479269149, "grad_norm": 0.2760178744792938, "learning_rate": 2.3650503630826895e-05, "loss": 0.0207, "step": 20676 }, { "epoch": 14.530569219957835, "grad_norm": 0.09183915704488754, "learning_rate": 2.3650035137034435e-05, "loss": 0.0162, "step": 20677 }, { "epoch": 14.53127196064652, "grad_norm": 0.10223598778247833, "learning_rate": 2.364956664324198e-05, "loss": 0.0112, "step": 20678 }, { "epoch": 14.531974701335207, "grad_norm": 0.11542507261037827, "learning_rate": 2.3649098149449522e-05, "loss": 0.0268, "step": 20679 }, { "epoch": 14.532677442023893, "grad_norm": 0.1932884007692337, "learning_rate": 2.3648629655657066e-05, "loss": 0.0124, "step": 20680 }, { "epoch": 14.533380182712579, "grad_norm": 0.07709822058677673, "learning_rate": 2.3648161161864603e-05, "loss": 0.0151, "step": 20681 }, { "epoch": 14.534082923401265, "grad_norm": 0.10976807773113251, "learning_rate": 2.3647692668072147e-05, "loss": 0.0084, "step": 20682 }, { "epoch": 14.53478566408995, "grad_norm": 0.18060290813446045, "learning_rate": 2.364722417427969e-05, "loss": 0.0236, "step": 20683 }, { "epoch": 14.535488404778636, "grad_norm": 0.14862127602100372, "learning_rate": 2.3646755680487234e-05, "loss": 0.0163, "step": 20684 }, { "epoch": 14.536191145467322, "grad_norm": 0.16207942366600037, "learning_rate": 2.3646287186694775e-05, "loss": 0.0337, "step": 20685 }, { "epoch": 14.536893886156008, "grad_norm": 0.08833937346935272, "learning_rate": 2.3645818692902318e-05, "loss": 0.0123, "step": 20686 }, { "epoch": 14.537596626844694, "grad_norm": 0.1569838672876358, "learning_rate": 2.3645350199109862e-05, "loss": 0.0153, "step": 20687 }, { "epoch": 14.53829936753338, "grad_norm": 0.1934671849012375, "learning_rate": 2.3644881705317406e-05, "loss": 0.0455, "step": 20688 }, { "epoch": 14.539002108222066, "grad_norm": 0.3899466097354889, "learning_rate": 2.364441321152495e-05, "loss": 0.0168, "step": 20689 }, { "epoch": 14.539704848910752, "grad_norm": 0.13868854939937592, "learning_rate": 2.364394471773249e-05, "loss": 0.029, "step": 20690 }, { "epoch": 14.540407589599438, "grad_norm": 0.3786075711250305, "learning_rate": 2.3643476223940033e-05, "loss": 0.0264, "step": 20691 }, { "epoch": 14.541110330288124, "grad_norm": 0.8548966646194458, "learning_rate": 2.3643007730147577e-05, "loss": 0.0367, "step": 20692 }, { "epoch": 14.54181307097681, "grad_norm": 0.6392034292221069, "learning_rate": 2.364253923635512e-05, "loss": 0.0523, "step": 20693 }, { "epoch": 14.542515811665496, "grad_norm": 0.3978797197341919, "learning_rate": 2.364207074256266e-05, "loss": 0.0813, "step": 20694 }, { "epoch": 14.543218552354181, "grad_norm": 0.6353722214698792, "learning_rate": 2.3641602248770205e-05, "loss": 0.1189, "step": 20695 }, { "epoch": 14.543921293042867, "grad_norm": 0.6363232135772705, "learning_rate": 2.364113375497775e-05, "loss": 0.1538, "step": 20696 }, { "epoch": 14.544624033731553, "grad_norm": 0.8660137057304382, "learning_rate": 2.3640665261185292e-05, "loss": 0.1906, "step": 20697 }, { "epoch": 14.54532677442024, "grad_norm": 0.20101875066757202, "learning_rate": 2.3640196767392833e-05, "loss": 0.0605, "step": 20698 }, { "epoch": 14.546029515108925, "grad_norm": 0.1489589810371399, "learning_rate": 2.3639728273600373e-05, "loss": 0.0261, "step": 20699 }, { "epoch": 14.546732255797611, "grad_norm": 0.3305470049381256, "learning_rate": 2.3639259779807917e-05, "loss": 0.0242, "step": 20700 }, { "epoch": 14.547434996486297, "grad_norm": 0.15590786933898926, "learning_rate": 2.363879128601546e-05, "loss": 0.0201, "step": 20701 }, { "epoch": 14.548137737174983, "grad_norm": 0.36778581142425537, "learning_rate": 2.3638322792223004e-05, "loss": 0.015, "step": 20702 }, { "epoch": 14.548840477863669, "grad_norm": 0.10091472417116165, "learning_rate": 2.3637854298430545e-05, "loss": 0.0134, "step": 20703 }, { "epoch": 14.549543218552355, "grad_norm": 0.11955205351114273, "learning_rate": 2.363738580463809e-05, "loss": 0.0137, "step": 20704 }, { "epoch": 14.55024595924104, "grad_norm": 0.0775504931807518, "learning_rate": 2.3636917310845632e-05, "loss": 0.0132, "step": 20705 }, { "epoch": 14.550948699929727, "grad_norm": 0.138371542096138, "learning_rate": 2.3636448817053176e-05, "loss": 0.0157, "step": 20706 }, { "epoch": 14.551651440618413, "grad_norm": 0.10605636239051819, "learning_rate": 2.3635980323260716e-05, "loss": 0.0129, "step": 20707 }, { "epoch": 14.552354181307098, "grad_norm": 0.14361698925495148, "learning_rate": 2.363551182946826e-05, "loss": 0.0156, "step": 20708 }, { "epoch": 14.553056921995784, "grad_norm": 0.838157057762146, "learning_rate": 2.3635043335675804e-05, "loss": 0.0157, "step": 20709 }, { "epoch": 14.55375966268447, "grad_norm": 0.18245813250541687, "learning_rate": 2.3634574841883347e-05, "loss": 0.0195, "step": 20710 }, { "epoch": 14.554462403373154, "grad_norm": 0.16020816564559937, "learning_rate": 2.3634106348090888e-05, "loss": 0.0181, "step": 20711 }, { "epoch": 14.55516514406184, "grad_norm": 0.12146176397800446, "learning_rate": 2.363363785429843e-05, "loss": 0.0212, "step": 20712 }, { "epoch": 14.555867884750526, "grad_norm": 0.5027256011962891, "learning_rate": 2.3633169360505975e-05, "loss": 0.0331, "step": 20713 }, { "epoch": 14.556570625439212, "grad_norm": 0.1893787980079651, "learning_rate": 2.363270086671352e-05, "loss": 0.0174, "step": 20714 }, { "epoch": 14.557273366127898, "grad_norm": 0.13759098947048187, "learning_rate": 2.3632232372921063e-05, "loss": 0.0324, "step": 20715 }, { "epoch": 14.557976106816584, "grad_norm": 0.33004915714263916, "learning_rate": 2.36317638791286e-05, "loss": 0.0434, "step": 20716 }, { "epoch": 14.55867884750527, "grad_norm": 0.3166162669658661, "learning_rate": 2.3631295385336143e-05, "loss": 0.041, "step": 20717 }, { "epoch": 14.559381588193956, "grad_norm": 0.49863794445991516, "learning_rate": 2.3630826891543687e-05, "loss": 0.0533, "step": 20718 }, { "epoch": 14.560084328882642, "grad_norm": 0.327880322933197, "learning_rate": 2.363035839775123e-05, "loss": 0.1016, "step": 20719 }, { "epoch": 14.560787069571328, "grad_norm": 5.087981700897217, "learning_rate": 2.362988990395877e-05, "loss": 0.1349, "step": 20720 }, { "epoch": 14.561489810260014, "grad_norm": 0.41934654116630554, "learning_rate": 2.3629421410166315e-05, "loss": 0.1449, "step": 20721 }, { "epoch": 14.5621925509487, "grad_norm": NaN, "learning_rate": 2.3629421410166315e-05, "loss": 0.1815, "step": 20722 }, { "epoch": 14.562895291637385, "grad_norm": 0.18519502878189087, "learning_rate": 2.362895291637386e-05, "loss": 0.0719, "step": 20723 }, { "epoch": 14.563598032326071, "grad_norm": 0.10546135157346725, "learning_rate": 2.3628484422581402e-05, "loss": 0.023, "step": 20724 }, { "epoch": 14.564300773014757, "grad_norm": 0.19886109232902527, "learning_rate": 2.3628015928788943e-05, "loss": 0.0223, "step": 20725 }, { "epoch": 14.565003513703443, "grad_norm": 0.1316065639257431, "learning_rate": 2.3627547434996486e-05, "loss": 0.014, "step": 20726 }, { "epoch": 14.56570625439213, "grad_norm": 0.3656293749809265, "learning_rate": 2.362707894120403e-05, "loss": 0.0222, "step": 20727 }, { "epoch": 14.566408995080815, "grad_norm": 0.08648591488599777, "learning_rate": 2.3626610447411574e-05, "loss": 0.0077, "step": 20728 }, { "epoch": 14.567111735769501, "grad_norm": 0.23683056235313416, "learning_rate": 2.3626141953619117e-05, "loss": 0.0139, "step": 20729 }, { "epoch": 14.567814476458187, "grad_norm": 0.172846257686615, "learning_rate": 2.3625673459826658e-05, "loss": 0.0181, "step": 20730 }, { "epoch": 14.568517217146873, "grad_norm": 0.09847433120012283, "learning_rate": 2.36252049660342e-05, "loss": 0.0178, "step": 20731 }, { "epoch": 14.569219957835559, "grad_norm": 0.18682782351970673, "learning_rate": 2.3624736472241745e-05, "loss": 0.0297, "step": 20732 }, { "epoch": 14.569922698524245, "grad_norm": 0.14692671597003937, "learning_rate": 2.362426797844929e-05, "loss": 0.0256, "step": 20733 }, { "epoch": 14.57062543921293, "grad_norm": 0.2544858753681183, "learning_rate": 2.3623799484656826e-05, "loss": 0.01, "step": 20734 }, { "epoch": 14.571328179901617, "grad_norm": 0.20864655077457428, "learning_rate": 2.362333099086437e-05, "loss": 0.0198, "step": 20735 }, { "epoch": 14.572030920590302, "grad_norm": 0.10575582832098007, "learning_rate": 2.3622862497071913e-05, "loss": 0.017, "step": 20736 }, { "epoch": 14.572733661278988, "grad_norm": 0.15268641710281372, "learning_rate": 2.3622394003279457e-05, "loss": 0.0234, "step": 20737 }, { "epoch": 14.573436401967674, "grad_norm": 0.2860507071018219, "learning_rate": 2.3621925509486997e-05, "loss": 0.0377, "step": 20738 }, { "epoch": 14.57413914265636, "grad_norm": 0.11973027884960175, "learning_rate": 2.362145701569454e-05, "loss": 0.0124, "step": 20739 }, { "epoch": 14.574841883345046, "grad_norm": 0.18963229656219482, "learning_rate": 2.3620988521902085e-05, "loss": 0.0342, "step": 20740 }, { "epoch": 14.575544624033732, "grad_norm": 0.1840481162071228, "learning_rate": 2.362052002810963e-05, "loss": 0.03, "step": 20741 }, { "epoch": 14.576247364722418, "grad_norm": 0.3872523903846741, "learning_rate": 2.3620051534317172e-05, "loss": 0.0553, "step": 20742 }, { "epoch": 14.576950105411104, "grad_norm": 0.29248982667922974, "learning_rate": 2.3619583040524713e-05, "loss": 0.0598, "step": 20743 }, { "epoch": 14.57765284609979, "grad_norm": 0.3187919855117798, "learning_rate": 2.3619114546732256e-05, "loss": 0.1047, "step": 20744 }, { "epoch": 14.578355586788476, "grad_norm": 0.725665807723999, "learning_rate": 2.36186460529398e-05, "loss": 0.1095, "step": 20745 }, { "epoch": 14.579058327477162, "grad_norm": 0.7542288899421692, "learning_rate": 2.3618177559147344e-05, "loss": 0.161, "step": 20746 }, { "epoch": 14.579761068165848, "grad_norm": 1.8181308507919312, "learning_rate": 2.3617709065354884e-05, "loss": 0.1887, "step": 20747 }, { "epoch": 14.580463808854532, "grad_norm": 0.3582738935947418, "learning_rate": 2.3617240571562428e-05, "loss": 0.0493, "step": 20748 }, { "epoch": 14.58116654954322, "grad_norm": 0.13867728412151337, "learning_rate": 2.361677207776997e-05, "loss": 0.026, "step": 20749 }, { "epoch": 14.581869290231904, "grad_norm": 0.09163034707307816, "learning_rate": 2.3616303583977515e-05, "loss": 0.0165, "step": 20750 }, { "epoch": 14.58257203092059, "grad_norm": 0.17181047797203064, "learning_rate": 2.3615835090185056e-05, "loss": 0.0218, "step": 20751 }, { "epoch": 14.583274771609275, "grad_norm": 0.0931035727262497, "learning_rate": 2.3615366596392596e-05, "loss": 0.0151, "step": 20752 }, { "epoch": 14.583977512297961, "grad_norm": 0.08866184949874878, "learning_rate": 2.361489810260014e-05, "loss": 0.0096, "step": 20753 }, { "epoch": 14.584680252986647, "grad_norm": 0.09531056135892868, "learning_rate": 2.3614429608807683e-05, "loss": 0.011, "step": 20754 }, { "epoch": 14.585382993675333, "grad_norm": 0.25058406591415405, "learning_rate": 2.3613961115015227e-05, "loss": 0.0373, "step": 20755 }, { "epoch": 14.58608573436402, "grad_norm": 0.08754090964794159, "learning_rate": 2.3613492621222768e-05, "loss": 0.0164, "step": 20756 }, { "epoch": 14.586788475052705, "grad_norm": 0.06457922607660294, "learning_rate": 2.361302412743031e-05, "loss": 0.006, "step": 20757 }, { "epoch": 14.587491215741391, "grad_norm": 0.16264782845973969, "learning_rate": 2.3612555633637855e-05, "loss": 0.0172, "step": 20758 }, { "epoch": 14.588193956430077, "grad_norm": 0.08070384711027145, "learning_rate": 2.36120871398454e-05, "loss": 0.0091, "step": 20759 }, { "epoch": 14.588896697118763, "grad_norm": 0.14929866790771484, "learning_rate": 2.361161864605294e-05, "loss": 0.0282, "step": 20760 }, { "epoch": 14.589599437807449, "grad_norm": 0.09656207263469696, "learning_rate": 2.3611150152260483e-05, "loss": 0.008, "step": 20761 }, { "epoch": 14.590302178496135, "grad_norm": 0.4303331971168518, "learning_rate": 2.3610681658468026e-05, "loss": 0.0268, "step": 20762 }, { "epoch": 14.59100491918482, "grad_norm": 0.16276146471500397, "learning_rate": 2.361021316467557e-05, "loss": 0.0251, "step": 20763 }, { "epoch": 14.591707659873506, "grad_norm": 0.20253817737102509, "learning_rate": 2.360974467088311e-05, "loss": 0.0178, "step": 20764 }, { "epoch": 14.592410400562192, "grad_norm": 0.2735811769962311, "learning_rate": 2.3609276177090654e-05, "loss": 0.0355, "step": 20765 }, { "epoch": 14.593113141250878, "grad_norm": 0.17336787283420563, "learning_rate": 2.3608807683298198e-05, "loss": 0.0382, "step": 20766 }, { "epoch": 14.593815881939564, "grad_norm": 0.13303793966770172, "learning_rate": 2.3608339189505742e-05, "loss": 0.0412, "step": 20767 }, { "epoch": 14.59451862262825, "grad_norm": 0.43022042512893677, "learning_rate": 2.3607870695713285e-05, "loss": 0.0888, "step": 20768 }, { "epoch": 14.595221363316936, "grad_norm": 0.4396066963672638, "learning_rate": 2.3607402201920822e-05, "loss": 0.0834, "step": 20769 }, { "epoch": 14.595924104005622, "grad_norm": 0.7584880590438843, "learning_rate": 2.3606933708128366e-05, "loss": 0.1322, "step": 20770 }, { "epoch": 14.596626844694308, "grad_norm": 0.7019749283790588, "learning_rate": 2.360646521433591e-05, "loss": 0.1542, "step": 20771 }, { "epoch": 14.597329585382994, "grad_norm": 1.4525644779205322, "learning_rate": 2.3605996720543454e-05, "loss": 0.1677, "step": 20772 }, { "epoch": 14.59803232607168, "grad_norm": 0.4339057505130768, "learning_rate": 2.3605528226750994e-05, "loss": 0.0615, "step": 20773 }, { "epoch": 14.598735066760366, "grad_norm": 0.6467993855476379, "learning_rate": 2.3605059732958538e-05, "loss": 0.0193, "step": 20774 }, { "epoch": 14.599437807449052, "grad_norm": 0.1215587854385376, "learning_rate": 2.360459123916608e-05, "loss": 0.0221, "step": 20775 }, { "epoch": 14.600140548137738, "grad_norm": 0.10070011764764786, "learning_rate": 2.3604122745373625e-05, "loss": 0.0141, "step": 20776 }, { "epoch": 14.600843288826423, "grad_norm": 0.12155269086360931, "learning_rate": 2.3603654251581165e-05, "loss": 0.0207, "step": 20777 }, { "epoch": 14.60154602951511, "grad_norm": 0.07187346369028091, "learning_rate": 2.360318575778871e-05, "loss": 0.0096, "step": 20778 }, { "epoch": 14.602248770203795, "grad_norm": 0.07142636924982071, "learning_rate": 2.3602717263996253e-05, "loss": 0.0103, "step": 20779 }, { "epoch": 14.602951510892481, "grad_norm": 0.1605173945426941, "learning_rate": 2.3602248770203797e-05, "loss": 0.0134, "step": 20780 }, { "epoch": 14.603654251581167, "grad_norm": 0.10130812227725983, "learning_rate": 2.360178027641134e-05, "loss": 0.0176, "step": 20781 }, { "epoch": 14.604356992269853, "grad_norm": 0.09299761056900024, "learning_rate": 2.360131178261888e-05, "loss": 0.0153, "step": 20782 }, { "epoch": 14.605059732958539, "grad_norm": 0.21912018954753876, "learning_rate": 2.3600843288826424e-05, "loss": 0.0136, "step": 20783 }, { "epoch": 14.605762473647225, "grad_norm": 0.10314501076936722, "learning_rate": 2.3600374795033968e-05, "loss": 0.0113, "step": 20784 }, { "epoch": 14.60646521433591, "grad_norm": 0.12134265154600143, "learning_rate": 2.3599906301241512e-05, "loss": 0.0217, "step": 20785 }, { "epoch": 14.607167955024597, "grad_norm": 0.11437743902206421, "learning_rate": 2.3599437807449052e-05, "loss": 0.0127, "step": 20786 }, { "epoch": 14.607870695713281, "grad_norm": 1.0163471698760986, "learning_rate": 2.3598969313656593e-05, "loss": 0.0235, "step": 20787 }, { "epoch": 14.608573436401969, "grad_norm": 0.13642776012420654, "learning_rate": 2.3598500819864136e-05, "loss": 0.0231, "step": 20788 }, { "epoch": 14.609276177090653, "grad_norm": 0.676819384098053, "learning_rate": 2.359803232607168e-05, "loss": 0.0172, "step": 20789 }, { "epoch": 14.609978917779339, "grad_norm": 0.529350757598877, "learning_rate": 2.359756383227922e-05, "loss": 0.0297, "step": 20790 }, { "epoch": 14.610681658468025, "grad_norm": 0.2909548282623291, "learning_rate": 2.3597095338486764e-05, "loss": 0.0267, "step": 20791 }, { "epoch": 14.61138439915671, "grad_norm": 0.25538599491119385, "learning_rate": 2.3596626844694308e-05, "loss": 0.0449, "step": 20792 }, { "epoch": 14.612087139845396, "grad_norm": 0.2615352272987366, "learning_rate": 2.359615835090185e-05, "loss": 0.0645, "step": 20793 }, { "epoch": 14.612789880534082, "grad_norm": 0.7656745314598083, "learning_rate": 2.3595689857109395e-05, "loss": 0.1085, "step": 20794 }, { "epoch": 14.613492621222768, "grad_norm": 0.7936549186706543, "learning_rate": 2.3595221363316936e-05, "loss": 0.1023, "step": 20795 }, { "epoch": 14.614195361911454, "grad_norm": 0.5427066087722778, "learning_rate": 2.359475286952448e-05, "loss": 0.1575, "step": 20796 }, { "epoch": 14.61489810260014, "grad_norm": 1.0240130424499512, "learning_rate": 2.3594284375732023e-05, "loss": 0.1804, "step": 20797 }, { "epoch": 14.615600843288826, "grad_norm": 0.2452680617570877, "learning_rate": 2.3593815881939567e-05, "loss": 0.0488, "step": 20798 }, { "epoch": 14.616303583977512, "grad_norm": 0.10844893753528595, "learning_rate": 2.3593347388147107e-05, "loss": 0.0254, "step": 20799 }, { "epoch": 14.617006324666198, "grad_norm": 0.10648851096630096, "learning_rate": 2.359287889435465e-05, "loss": 0.0123, "step": 20800 }, { "epoch": 14.617709065354884, "grad_norm": 0.32601699233055115, "learning_rate": 2.3592410400562194e-05, "loss": 0.0172, "step": 20801 }, { "epoch": 14.61841180604357, "grad_norm": 0.10381408035755157, "learning_rate": 2.3591941906769738e-05, "loss": 0.0196, "step": 20802 }, { "epoch": 14.619114546732256, "grad_norm": 0.36608871817588806, "learning_rate": 2.359147341297728e-05, "loss": 0.01, "step": 20803 }, { "epoch": 14.619817287420942, "grad_norm": 0.07128497958183289, "learning_rate": 2.359100491918482e-05, "loss": 0.0158, "step": 20804 }, { "epoch": 14.620520028109627, "grad_norm": 0.13037031888961792, "learning_rate": 2.3590536425392363e-05, "loss": 0.0199, "step": 20805 }, { "epoch": 14.621222768798313, "grad_norm": 0.1031404584646225, "learning_rate": 2.3590067931599906e-05, "loss": 0.013, "step": 20806 }, { "epoch": 14.621925509487, "grad_norm": 0.09619602560997009, "learning_rate": 2.358959943780745e-05, "loss": 0.0139, "step": 20807 }, { "epoch": 14.622628250175685, "grad_norm": 0.10648408532142639, "learning_rate": 2.358913094401499e-05, "loss": 0.0168, "step": 20808 }, { "epoch": 14.623330990864371, "grad_norm": 0.14765483140945435, "learning_rate": 2.3588662450222534e-05, "loss": 0.0131, "step": 20809 }, { "epoch": 14.624033731553057, "grad_norm": 0.275485634803772, "learning_rate": 2.3588193956430078e-05, "loss": 0.0308, "step": 20810 }, { "epoch": 14.624736472241743, "grad_norm": 0.19830921292304993, "learning_rate": 2.358772546263762e-05, "loss": 0.0234, "step": 20811 }, { "epoch": 14.625439212930429, "grad_norm": 0.1989426165819168, "learning_rate": 2.3587256968845162e-05, "loss": 0.0352, "step": 20812 }, { "epoch": 14.626141953619115, "grad_norm": 0.15808366239070892, "learning_rate": 2.3586788475052706e-05, "loss": 0.0281, "step": 20813 }, { "epoch": 14.6268446943078, "grad_norm": 0.10188212245702744, "learning_rate": 2.358631998126025e-05, "loss": 0.011, "step": 20814 }, { "epoch": 14.627547434996487, "grad_norm": 0.2539919316768646, "learning_rate": 2.3585851487467793e-05, "loss": 0.0429, "step": 20815 }, { "epoch": 14.628250175685173, "grad_norm": 0.16601692140102386, "learning_rate": 2.3585382993675337e-05, "loss": 0.0311, "step": 20816 }, { "epoch": 14.628952916373859, "grad_norm": 0.372522234916687, "learning_rate": 2.3584914499882877e-05, "loss": 0.049, "step": 20817 }, { "epoch": 14.629655657062544, "grad_norm": 0.21842060983181, "learning_rate": 2.358444600609042e-05, "loss": 0.0491, "step": 20818 }, { "epoch": 14.63035839775123, "grad_norm": 0.5228503346443176, "learning_rate": 2.3583977512297965e-05, "loss": 0.0907, "step": 20819 }, { "epoch": 14.631061138439916, "grad_norm": 0.336075097322464, "learning_rate": 2.358350901850551e-05, "loss": 0.1039, "step": 20820 }, { "epoch": 14.631763879128602, "grad_norm": 0.7175515294075012, "learning_rate": 2.3583040524713045e-05, "loss": 0.1488, "step": 20821 }, { "epoch": 14.632466619817288, "grad_norm": 5.112446308135986, "learning_rate": 2.358257203092059e-05, "loss": 0.189, "step": 20822 }, { "epoch": 14.633169360505974, "grad_norm": 0.5566748976707458, "learning_rate": 2.3582103537128133e-05, "loss": 0.0957, "step": 20823 }, { "epoch": 14.63387210119466, "grad_norm": 0.1437271237373352, "learning_rate": 2.3581635043335676e-05, "loss": 0.0234, "step": 20824 }, { "epoch": 14.634574841883346, "grad_norm": 0.13294008374214172, "learning_rate": 2.3581166549543217e-05, "loss": 0.0127, "step": 20825 }, { "epoch": 14.63527758257203, "grad_norm": 0.11067486554384232, "learning_rate": 2.358069805575076e-05, "loss": 0.026, "step": 20826 }, { "epoch": 14.635980323260716, "grad_norm": 0.10996387153863907, "learning_rate": 2.3580229561958304e-05, "loss": 0.0158, "step": 20827 }, { "epoch": 14.636683063949402, "grad_norm": 0.09373804926872253, "learning_rate": 2.3579761068165848e-05, "loss": 0.0112, "step": 20828 }, { "epoch": 14.637385804638088, "grad_norm": 0.16835379600524902, "learning_rate": 2.357929257437339e-05, "loss": 0.0116, "step": 20829 }, { "epoch": 14.638088545326774, "grad_norm": 0.07202640920877457, "learning_rate": 2.3578824080580932e-05, "loss": 0.0068, "step": 20830 }, { "epoch": 14.63879128601546, "grad_norm": 1.127203345298767, "learning_rate": 2.3578355586788476e-05, "loss": 0.0262, "step": 20831 }, { "epoch": 14.639494026704146, "grad_norm": 0.1500367820262909, "learning_rate": 2.357788709299602e-05, "loss": 0.0086, "step": 20832 }, { "epoch": 14.640196767392831, "grad_norm": 0.22150899469852448, "learning_rate": 2.3577418599203563e-05, "loss": 0.032, "step": 20833 }, { "epoch": 14.640899508081517, "grad_norm": 0.12652595341205597, "learning_rate": 2.3576950105411104e-05, "loss": 0.0199, "step": 20834 }, { "epoch": 14.641602248770203, "grad_norm": 0.13568899035453796, "learning_rate": 2.3576481611618647e-05, "loss": 0.0168, "step": 20835 }, { "epoch": 14.64230498945889, "grad_norm": 0.08575214445590973, "learning_rate": 2.357601311782619e-05, "loss": 0.0224, "step": 20836 }, { "epoch": 14.643007730147575, "grad_norm": 0.23276835680007935, "learning_rate": 2.3575544624033735e-05, "loss": 0.029, "step": 20837 }, { "epoch": 14.643710470836261, "grad_norm": 0.21680642664432526, "learning_rate": 2.3575076130241275e-05, "loss": 0.0257, "step": 20838 }, { "epoch": 14.644413211524947, "grad_norm": 0.12613892555236816, "learning_rate": 2.3574607636448815e-05, "loss": 0.0219, "step": 20839 }, { "epoch": 14.645115952213633, "grad_norm": 0.20392730832099915, "learning_rate": 2.357413914265636e-05, "loss": 0.0328, "step": 20840 }, { "epoch": 14.645818692902319, "grad_norm": 0.1970987617969513, "learning_rate": 2.3573670648863903e-05, "loss": 0.0301, "step": 20841 }, { "epoch": 14.646521433591005, "grad_norm": 0.21036113798618317, "learning_rate": 2.3573202155071447e-05, "loss": 0.0468, "step": 20842 }, { "epoch": 14.64722417427969, "grad_norm": 0.2777903079986572, "learning_rate": 2.3572733661278987e-05, "loss": 0.0508, "step": 20843 }, { "epoch": 14.647926914968377, "grad_norm": 0.2562773525714874, "learning_rate": 2.357226516748653e-05, "loss": 0.0649, "step": 20844 }, { "epoch": 14.648629655657063, "grad_norm": 0.7313152551651001, "learning_rate": 2.3571796673694074e-05, "loss": 0.121, "step": 20845 }, { "epoch": 14.649332396345748, "grad_norm": 0.6535728573799133, "learning_rate": 2.3571328179901618e-05, "loss": 0.1515, "step": 20846 }, { "epoch": 14.650035137034434, "grad_norm": 1.3576292991638184, "learning_rate": 2.357085968610916e-05, "loss": 0.1785, "step": 20847 }, { "epoch": 14.65073787772312, "grad_norm": 0.23128019273281097, "learning_rate": 2.3570391192316702e-05, "loss": 0.0611, "step": 20848 }, { "epoch": 14.651440618411806, "grad_norm": 0.12091303616762161, "learning_rate": 2.3569922698524246e-05, "loss": 0.018, "step": 20849 }, { "epoch": 14.652143359100492, "grad_norm": 0.13111841678619385, "learning_rate": 2.356945420473179e-05, "loss": 0.0212, "step": 20850 }, { "epoch": 14.652846099789178, "grad_norm": 0.1720174103975296, "learning_rate": 2.356898571093933e-05, "loss": 0.0165, "step": 20851 }, { "epoch": 14.653548840477864, "grad_norm": 0.09681583940982819, "learning_rate": 2.3568517217146874e-05, "loss": 0.0117, "step": 20852 }, { "epoch": 14.65425158116655, "grad_norm": 0.0735422819852829, "learning_rate": 2.3568048723354417e-05, "loss": 0.0079, "step": 20853 }, { "epoch": 14.654954321855236, "grad_norm": 0.12186414003372192, "learning_rate": 2.356758022956196e-05, "loss": 0.0154, "step": 20854 }, { "epoch": 14.655657062543922, "grad_norm": 0.2588309347629547, "learning_rate": 2.3567111735769505e-05, "loss": 0.0168, "step": 20855 }, { "epoch": 14.656359803232608, "grad_norm": 0.22962459921836853, "learning_rate": 2.3566643241977042e-05, "loss": 0.0132, "step": 20856 }, { "epoch": 14.657062543921294, "grad_norm": 0.08687550574541092, "learning_rate": 2.3566174748184586e-05, "loss": 0.0143, "step": 20857 }, { "epoch": 14.65776528460998, "grad_norm": 0.1915227174758911, "learning_rate": 2.356570625439213e-05, "loss": 0.0299, "step": 20858 }, { "epoch": 14.658468025298665, "grad_norm": 0.2290605753660202, "learning_rate": 2.3565237760599673e-05, "loss": 0.011, "step": 20859 }, { "epoch": 14.659170765987351, "grad_norm": 0.3003217577934265, "learning_rate": 2.3564769266807213e-05, "loss": 0.0339, "step": 20860 }, { "epoch": 14.659873506676037, "grad_norm": 0.13511119782924652, "learning_rate": 2.3564300773014757e-05, "loss": 0.0124, "step": 20861 }, { "epoch": 14.660576247364723, "grad_norm": 0.2402341216802597, "learning_rate": 2.35638322792223e-05, "loss": 0.0395, "step": 20862 }, { "epoch": 14.66127898805341, "grad_norm": 0.2538892924785614, "learning_rate": 2.3563363785429844e-05, "loss": 0.0392, "step": 20863 }, { "epoch": 14.661981728742095, "grad_norm": 0.12646283209323883, "learning_rate": 2.3562895291637385e-05, "loss": 0.0135, "step": 20864 }, { "epoch": 14.66268446943078, "grad_norm": 0.17472562193870544, "learning_rate": 2.356242679784493e-05, "loss": 0.0316, "step": 20865 }, { "epoch": 14.663387210119465, "grad_norm": 0.4238770306110382, "learning_rate": 2.3561958304052472e-05, "loss": 0.0571, "step": 20866 }, { "epoch": 14.664089950808151, "grad_norm": 0.22391654551029205, "learning_rate": 2.3561489810260016e-05, "loss": 0.0422, "step": 20867 }, { "epoch": 14.664792691496837, "grad_norm": 0.46570244431495667, "learning_rate": 2.356102131646756e-05, "loss": 0.0596, "step": 20868 }, { "epoch": 14.665495432185523, "grad_norm": 0.3542748987674713, "learning_rate": 2.35605528226751e-05, "loss": 0.0935, "step": 20869 }, { "epoch": 14.666198172874209, "grad_norm": 0.4969163239002228, "learning_rate": 2.3560084328882644e-05, "loss": 0.11, "step": 20870 }, { "epoch": 14.666900913562895, "grad_norm": 0.7859424948692322, "learning_rate": 2.3559615835090187e-05, "loss": 0.1424, "step": 20871 }, { "epoch": 14.66760365425158, "grad_norm": 0.7728527188301086, "learning_rate": 2.355914734129773e-05, "loss": 0.1526, "step": 20872 }, { "epoch": 14.668306394940267, "grad_norm": 0.3452056646347046, "learning_rate": 2.355867884750527e-05, "loss": 0.0626, "step": 20873 }, { "epoch": 14.669009135628952, "grad_norm": 0.14712339639663696, "learning_rate": 2.3558210353712812e-05, "loss": 0.0273, "step": 20874 }, { "epoch": 14.669711876317638, "grad_norm": 0.1800346076488495, "learning_rate": 2.3557741859920356e-05, "loss": 0.0136, "step": 20875 }, { "epoch": 14.670414617006324, "grad_norm": 0.14215490221977234, "learning_rate": 2.35572733661279e-05, "loss": 0.0236, "step": 20876 }, { "epoch": 14.67111735769501, "grad_norm": 0.19231750071048737, "learning_rate": 2.355680487233544e-05, "loss": 0.0176, "step": 20877 }, { "epoch": 14.671820098383696, "grad_norm": 0.07477416843175888, "learning_rate": 2.3556336378542983e-05, "loss": 0.0114, "step": 20878 }, { "epoch": 14.672522839072382, "grad_norm": 0.2893056273460388, "learning_rate": 2.3555867884750527e-05, "loss": 0.0185, "step": 20879 }, { "epoch": 14.673225579761068, "grad_norm": 0.32723745703697205, "learning_rate": 2.355539939095807e-05, "loss": 0.0195, "step": 20880 }, { "epoch": 14.673928320449754, "grad_norm": 0.10364210605621338, "learning_rate": 2.3554930897165615e-05, "loss": 0.0268, "step": 20881 }, { "epoch": 14.67463106113844, "grad_norm": 0.0788143202662468, "learning_rate": 2.3554462403373155e-05, "loss": 0.0147, "step": 20882 }, { "epoch": 14.675333801827126, "grad_norm": 0.10462413728237152, "learning_rate": 2.35539939095807e-05, "loss": 0.0156, "step": 20883 }, { "epoch": 14.676036542515812, "grad_norm": 0.12693911790847778, "learning_rate": 2.3553525415788242e-05, "loss": 0.0173, "step": 20884 }, { "epoch": 14.676739283204498, "grad_norm": 0.12171417474746704, "learning_rate": 2.3553056921995786e-05, "loss": 0.0207, "step": 20885 }, { "epoch": 14.677442023893184, "grad_norm": 0.19666343927383423, "learning_rate": 2.3552588428203326e-05, "loss": 0.0179, "step": 20886 }, { "epoch": 14.67814476458187, "grad_norm": 0.17251846194267273, "learning_rate": 2.355211993441087e-05, "loss": 0.021, "step": 20887 }, { "epoch": 14.678847505270555, "grad_norm": 0.10886968672275543, "learning_rate": 2.3551651440618414e-05, "loss": 0.0216, "step": 20888 }, { "epoch": 14.679550245959241, "grad_norm": 0.1635119467973709, "learning_rate": 2.3551182946825958e-05, "loss": 0.0383, "step": 20889 }, { "epoch": 14.680252986647927, "grad_norm": 0.17521023750305176, "learning_rate": 2.3550714453033498e-05, "loss": 0.0246, "step": 20890 }, { "epoch": 14.680955727336613, "grad_norm": 0.3765561282634735, "learning_rate": 2.3550245959241038e-05, "loss": 0.0598, "step": 20891 }, { "epoch": 14.681658468025299, "grad_norm": 0.1571061611175537, "learning_rate": 2.3549777465448582e-05, "loss": 0.0384, "step": 20892 }, { "epoch": 14.682361208713985, "grad_norm": 0.24712513387203217, "learning_rate": 2.3549308971656126e-05, "loss": 0.0643, "step": 20893 }, { "epoch": 14.683063949402671, "grad_norm": 0.7665297389030457, "learning_rate": 2.354884047786367e-05, "loss": 0.0793, "step": 20894 }, { "epoch": 14.683766690091357, "grad_norm": 0.514760434627533, "learning_rate": 2.354837198407121e-05, "loss": 0.1484, "step": 20895 }, { "epoch": 14.684469430780043, "grad_norm": 0.6067773103713989, "learning_rate": 2.3547903490278754e-05, "loss": 0.1541, "step": 20896 }, { "epoch": 14.685172171468729, "grad_norm": 0.9342012405395508, "learning_rate": 2.3547434996486297e-05, "loss": 0.1791, "step": 20897 }, { "epoch": 14.685874912157415, "grad_norm": 0.4781350791454315, "learning_rate": 2.354696650269384e-05, "loss": 0.0466, "step": 20898 }, { "epoch": 14.6865776528461, "grad_norm": 0.08544307947158813, "learning_rate": 2.354649800890138e-05, "loss": 0.0103, "step": 20899 }, { "epoch": 14.687280393534786, "grad_norm": 0.08465065062046051, "learning_rate": 2.3546029515108925e-05, "loss": 0.0142, "step": 20900 }, { "epoch": 14.687983134223472, "grad_norm": 0.11955807358026505, "learning_rate": 2.354556102131647e-05, "loss": 0.021, "step": 20901 }, { "epoch": 14.688685874912156, "grad_norm": 0.15009625256061554, "learning_rate": 2.3545092527524012e-05, "loss": 0.0111, "step": 20902 }, { "epoch": 14.689388615600844, "grad_norm": 0.2541091740131378, "learning_rate": 2.3544624033731553e-05, "loss": 0.0224, "step": 20903 }, { "epoch": 14.690091356289528, "grad_norm": 0.12113554775714874, "learning_rate": 2.3544155539939097e-05, "loss": 0.0214, "step": 20904 }, { "epoch": 14.690794096978214, "grad_norm": 0.10696946829557419, "learning_rate": 2.354368704614664e-05, "loss": 0.0178, "step": 20905 }, { "epoch": 14.6914968376669, "grad_norm": 0.21083270013332367, "learning_rate": 2.3543218552354184e-05, "loss": 0.0266, "step": 20906 }, { "epoch": 14.692199578355586, "grad_norm": 0.3558436632156372, "learning_rate": 2.3542750058561728e-05, "loss": 0.0092, "step": 20907 }, { "epoch": 14.692902319044272, "grad_norm": 0.08723429590463638, "learning_rate": 2.3542281564769265e-05, "loss": 0.0198, "step": 20908 }, { "epoch": 14.693605059732958, "grad_norm": 0.08842069655656815, "learning_rate": 2.354181307097681e-05, "loss": 0.0108, "step": 20909 }, { "epoch": 14.694307800421644, "grad_norm": 0.14138364791870117, "learning_rate": 2.3541344577184352e-05, "loss": 0.0223, "step": 20910 }, { "epoch": 14.69501054111033, "grad_norm": 0.10380424559116364, "learning_rate": 2.3540876083391896e-05, "loss": 0.0124, "step": 20911 }, { "epoch": 14.695713281799016, "grad_norm": 0.20699509978294373, "learning_rate": 2.3540407589599436e-05, "loss": 0.0231, "step": 20912 }, { "epoch": 14.696416022487702, "grad_norm": 0.23978669941425323, "learning_rate": 2.353993909580698e-05, "loss": 0.0331, "step": 20913 }, { "epoch": 14.697118763176388, "grad_norm": 0.15626035630702972, "learning_rate": 2.3539470602014524e-05, "loss": 0.013, "step": 20914 }, { "epoch": 14.697821503865073, "grad_norm": 0.1399749368429184, "learning_rate": 2.3539002108222067e-05, "loss": 0.0244, "step": 20915 }, { "epoch": 14.69852424455376, "grad_norm": 0.2835853695869446, "learning_rate": 2.3538533614429608e-05, "loss": 0.0606, "step": 20916 }, { "epoch": 14.699226985242445, "grad_norm": 0.16207902133464813, "learning_rate": 2.353806512063715e-05, "loss": 0.0246, "step": 20917 }, { "epoch": 14.699929725931131, "grad_norm": 0.3459773063659668, "learning_rate": 2.3537596626844695e-05, "loss": 0.0599, "step": 20918 }, { "epoch": 14.700632466619817, "grad_norm": 0.4620515704154968, "learning_rate": 2.353712813305224e-05, "loss": 0.0878, "step": 20919 }, { "epoch": 14.701335207308503, "grad_norm": 0.4471827745437622, "learning_rate": 2.3536659639259783e-05, "loss": 0.1474, "step": 20920 }, { "epoch": 14.702037947997189, "grad_norm": 1.305479884147644, "learning_rate": 2.3536191145467323e-05, "loss": 0.1551, "step": 20921 }, { "epoch": 14.702740688685875, "grad_norm": 0.7093528509140015, "learning_rate": 2.3535722651674867e-05, "loss": 0.1695, "step": 20922 }, { "epoch": 14.70344342937456, "grad_norm": 0.15434515476226807, "learning_rate": 2.353525415788241e-05, "loss": 0.0585, "step": 20923 }, { "epoch": 14.704146170063247, "grad_norm": 0.16087481379508972, "learning_rate": 2.3534785664089954e-05, "loss": 0.0177, "step": 20924 }, { "epoch": 14.704848910751933, "grad_norm": 0.07028649002313614, "learning_rate": 2.3534317170297494e-05, "loss": 0.0127, "step": 20925 }, { "epoch": 14.705551651440619, "grad_norm": 0.09230080991983414, "learning_rate": 2.3533848676505035e-05, "loss": 0.0119, "step": 20926 }, { "epoch": 14.706254392129305, "grad_norm": 0.08839662373065948, "learning_rate": 2.353338018271258e-05, "loss": 0.0141, "step": 20927 }, { "epoch": 14.70695713281799, "grad_norm": 0.10530024766921997, "learning_rate": 2.3532911688920122e-05, "loss": 0.0112, "step": 20928 }, { "epoch": 14.707659873506676, "grad_norm": 0.0600300133228302, "learning_rate": 2.3532443195127663e-05, "loss": 0.0061, "step": 20929 }, { "epoch": 14.708362614195362, "grad_norm": 0.10599471628665924, "learning_rate": 2.3531974701335206e-05, "loss": 0.0191, "step": 20930 }, { "epoch": 14.709065354884048, "grad_norm": 0.09446390718221664, "learning_rate": 2.353150620754275e-05, "loss": 0.0154, "step": 20931 }, { "epoch": 14.709768095572734, "grad_norm": 0.05247168615460396, "learning_rate": 2.3531037713750294e-05, "loss": 0.0068, "step": 20932 }, { "epoch": 14.71047083626142, "grad_norm": 0.07086877524852753, "learning_rate": 2.3530569219957837e-05, "loss": 0.0212, "step": 20933 }, { "epoch": 14.711173576950106, "grad_norm": 0.12153928726911545, "learning_rate": 2.3530100726165378e-05, "loss": 0.0095, "step": 20934 }, { "epoch": 14.711876317638792, "grad_norm": 0.10336338728666306, "learning_rate": 2.352963223237292e-05, "loss": 0.0266, "step": 20935 }, { "epoch": 14.712579058327478, "grad_norm": 0.1378771811723709, "learning_rate": 2.3529163738580465e-05, "loss": 0.0113, "step": 20936 }, { "epoch": 14.713281799016164, "grad_norm": 0.08639997243881226, "learning_rate": 2.352869524478801e-05, "loss": 0.0201, "step": 20937 }, { "epoch": 14.71398453970485, "grad_norm": 0.13353948295116425, "learning_rate": 2.352822675099555e-05, "loss": 0.0281, "step": 20938 }, { "epoch": 14.714687280393536, "grad_norm": 0.14786876738071442, "learning_rate": 2.3527758257203093e-05, "loss": 0.0195, "step": 20939 }, { "epoch": 14.715390021082221, "grad_norm": 0.4016188979148865, "learning_rate": 2.3527289763410637e-05, "loss": 0.0423, "step": 20940 }, { "epoch": 14.716092761770906, "grad_norm": 0.1386888176202774, "learning_rate": 2.352682126961818e-05, "loss": 0.0232, "step": 20941 }, { "epoch": 14.716795502459593, "grad_norm": 0.2647458612918854, "learning_rate": 2.352635277582572e-05, "loss": 0.0531, "step": 20942 }, { "epoch": 14.717498243148277, "grad_norm": 0.4120634198188782, "learning_rate": 2.352588428203326e-05, "loss": 0.0696, "step": 20943 }, { "epoch": 14.718200983836963, "grad_norm": 0.719511866569519, "learning_rate": 2.3525415788240805e-05, "loss": 0.0906, "step": 20944 }, { "epoch": 14.71890372452565, "grad_norm": 0.41855379939079285, "learning_rate": 2.352494729444835e-05, "loss": 0.1171, "step": 20945 }, { "epoch": 14.719606465214335, "grad_norm": 0.6349735260009766, "learning_rate": 2.3524478800655892e-05, "loss": 0.1283, "step": 20946 }, { "epoch": 14.720309205903021, "grad_norm": 1.4138410091400146, "learning_rate": 2.3524010306863433e-05, "loss": 0.15, "step": 20947 }, { "epoch": 14.721011946591707, "grad_norm": 0.5008849501609802, "learning_rate": 2.3523541813070976e-05, "loss": 0.0663, "step": 20948 }, { "epoch": 14.721714687280393, "grad_norm": 0.18866586685180664, "learning_rate": 2.352307331927852e-05, "loss": 0.0229, "step": 20949 }, { "epoch": 14.722417427969079, "grad_norm": 0.14213496446609497, "learning_rate": 2.3522604825486064e-05, "loss": 0.0207, "step": 20950 }, { "epoch": 14.723120168657765, "grad_norm": 0.0999387800693512, "learning_rate": 2.3522136331693604e-05, "loss": 0.0098, "step": 20951 }, { "epoch": 14.72382290934645, "grad_norm": 0.09210578352212906, "learning_rate": 2.3521667837901148e-05, "loss": 0.0101, "step": 20952 }, { "epoch": 14.724525650035137, "grad_norm": 0.08046314120292664, "learning_rate": 2.352119934410869e-05, "loss": 0.0124, "step": 20953 }, { "epoch": 14.725228390723823, "grad_norm": 0.19286462664604187, "learning_rate": 2.3520730850316235e-05, "loss": 0.0169, "step": 20954 }, { "epoch": 14.725931131412509, "grad_norm": 0.10221754014492035, "learning_rate": 2.3520262356523776e-05, "loss": 0.0197, "step": 20955 }, { "epoch": 14.726633872101194, "grad_norm": 0.16451099514961243, "learning_rate": 2.351979386273132e-05, "loss": 0.0134, "step": 20956 }, { "epoch": 14.72733661278988, "grad_norm": 0.1617024689912796, "learning_rate": 2.3519325368938863e-05, "loss": 0.0131, "step": 20957 }, { "epoch": 14.728039353478566, "grad_norm": 0.12012573331594467, "learning_rate": 2.3518856875146407e-05, "loss": 0.0181, "step": 20958 }, { "epoch": 14.728742094167252, "grad_norm": 0.11093191802501678, "learning_rate": 2.351838838135395e-05, "loss": 0.0077, "step": 20959 }, { "epoch": 14.729444834855938, "grad_norm": 0.11340195685625076, "learning_rate": 2.351791988756149e-05, "loss": 0.0187, "step": 20960 }, { "epoch": 14.730147575544624, "grad_norm": 0.11309690028429031, "learning_rate": 2.351745139376903e-05, "loss": 0.0175, "step": 20961 }, { "epoch": 14.73085031623331, "grad_norm": 0.23568736016750336, "learning_rate": 2.3516982899976575e-05, "loss": 0.0239, "step": 20962 }, { "epoch": 14.731553056921996, "grad_norm": 0.12218668311834335, "learning_rate": 2.351651440618412e-05, "loss": 0.0102, "step": 20963 }, { "epoch": 14.732255797610682, "grad_norm": 0.49507856369018555, "learning_rate": 2.351604591239166e-05, "loss": 0.0405, "step": 20964 }, { "epoch": 14.732958538299368, "grad_norm": 0.3980897068977356, "learning_rate": 2.3515577418599203e-05, "loss": 0.0276, "step": 20965 }, { "epoch": 14.733661278988054, "grad_norm": 0.20962487161159515, "learning_rate": 2.3515108924806747e-05, "loss": 0.034, "step": 20966 }, { "epoch": 14.73436401967674, "grad_norm": 0.25967344641685486, "learning_rate": 2.351464043101429e-05, "loss": 0.0322, "step": 20967 }, { "epoch": 14.735066760365426, "grad_norm": 0.2610488831996918, "learning_rate": 2.351417193722183e-05, "loss": 0.0566, "step": 20968 }, { "epoch": 14.735769501054111, "grad_norm": 0.28708118200302124, "learning_rate": 2.3513703443429374e-05, "loss": 0.0911, "step": 20969 }, { "epoch": 14.736472241742797, "grad_norm": 0.7226487994194031, "learning_rate": 2.3513234949636918e-05, "loss": 0.1268, "step": 20970 }, { "epoch": 14.737174982431483, "grad_norm": 0.6984308362007141, "learning_rate": 2.3512766455844462e-05, "loss": 0.131, "step": 20971 }, { "epoch": 14.73787772312017, "grad_norm": 1.090170979499817, "learning_rate": 2.3512297962052005e-05, "loss": 0.1618, "step": 20972 }, { "epoch": 14.738580463808855, "grad_norm": 0.1688976287841797, "learning_rate": 2.3511829468259546e-05, "loss": 0.0561, "step": 20973 }, { "epoch": 14.739283204497541, "grad_norm": 0.11286932975053787, "learning_rate": 2.351136097446709e-05, "loss": 0.029, "step": 20974 }, { "epoch": 14.739985945186227, "grad_norm": 0.2110302597284317, "learning_rate": 2.3510892480674633e-05, "loss": 0.0294, "step": 20975 }, { "epoch": 14.740688685874913, "grad_norm": 0.15462951362133026, "learning_rate": 2.3510423986882177e-05, "loss": 0.0173, "step": 20976 }, { "epoch": 14.741391426563599, "grad_norm": 0.1068359836935997, "learning_rate": 2.3509955493089717e-05, "loss": 0.0148, "step": 20977 }, { "epoch": 14.742094167252285, "grad_norm": 0.1670617163181305, "learning_rate": 2.3509486999297258e-05, "loss": 0.0101, "step": 20978 }, { "epoch": 14.74279690794097, "grad_norm": 0.09958931058645248, "learning_rate": 2.35090185055048e-05, "loss": 0.0161, "step": 20979 }, { "epoch": 14.743499648629655, "grad_norm": 0.19640196859836578, "learning_rate": 2.3508550011712345e-05, "loss": 0.0214, "step": 20980 }, { "epoch": 14.74420238931834, "grad_norm": 0.12121637910604477, "learning_rate": 2.3508081517919885e-05, "loss": 0.0151, "step": 20981 }, { "epoch": 14.744905130007027, "grad_norm": 0.12988165020942688, "learning_rate": 2.350761302412743e-05, "loss": 0.0119, "step": 20982 }, { "epoch": 14.745607870695713, "grad_norm": 0.10679295659065247, "learning_rate": 2.3507144530334973e-05, "loss": 0.021, "step": 20983 }, { "epoch": 14.746310611384398, "grad_norm": 0.06853233277797699, "learning_rate": 2.3506676036542517e-05, "loss": 0.0126, "step": 20984 }, { "epoch": 14.747013352073084, "grad_norm": 0.2802310287952423, "learning_rate": 2.350620754275006e-05, "loss": 0.0199, "step": 20985 }, { "epoch": 14.74771609276177, "grad_norm": 0.11565755307674408, "learning_rate": 2.35057390489576e-05, "loss": 0.009, "step": 20986 }, { "epoch": 14.748418833450456, "grad_norm": 0.3470202088356018, "learning_rate": 2.3505270555165144e-05, "loss": 0.0265, "step": 20987 }, { "epoch": 14.749121574139142, "grad_norm": 0.1229303851723671, "learning_rate": 2.3504802061372688e-05, "loss": 0.023, "step": 20988 }, { "epoch": 14.749824314827828, "grad_norm": 0.1772243231534958, "learning_rate": 2.3504333567580232e-05, "loss": 0.0318, "step": 20989 }, { "epoch": 14.750527055516514, "grad_norm": 0.10199970006942749, "learning_rate": 2.3503865073787772e-05, "loss": 0.0173, "step": 20990 }, { "epoch": 14.7512297962052, "grad_norm": 0.29401108622550964, "learning_rate": 2.3503396579995316e-05, "loss": 0.0409, "step": 20991 }, { "epoch": 14.751932536893886, "grad_norm": 0.19320166110992432, "learning_rate": 2.350292808620286e-05, "loss": 0.0326, "step": 20992 }, { "epoch": 14.752635277582572, "grad_norm": 0.9426783323287964, "learning_rate": 2.3502459592410403e-05, "loss": 0.072, "step": 20993 }, { "epoch": 14.753338018271258, "grad_norm": 0.4011540412902832, "learning_rate": 2.3501991098617944e-05, "loss": 0.0972, "step": 20994 }, { "epoch": 14.754040758959944, "grad_norm": 0.7152203917503357, "learning_rate": 2.3501522604825487e-05, "loss": 0.1079, "step": 20995 }, { "epoch": 14.75474349964863, "grad_norm": 0.6340525150299072, "learning_rate": 2.3501054111033028e-05, "loss": 0.1643, "step": 20996 }, { "epoch": 14.755446240337315, "grad_norm": 0.9361770153045654, "learning_rate": 2.350058561724057e-05, "loss": 0.1521, "step": 20997 }, { "epoch": 14.756148981026001, "grad_norm": 0.16304732859134674, "learning_rate": 2.3500117123448115e-05, "loss": 0.0465, "step": 20998 }, { "epoch": 14.756851721714687, "grad_norm": 0.1461978256702423, "learning_rate": 2.3499648629655656e-05, "loss": 0.0214, "step": 20999 }, { "epoch": 14.757554462403373, "grad_norm": 0.11631020903587341, "learning_rate": 2.34991801358632e-05, "loss": 0.0173, "step": 21000 }, { "epoch": 14.757554462403373, "eval_cer": 0.19237777268610956, "eval_loss": 0.2696112394332886, "eval_runtime": 17.9487, "eval_samples_per_second": 252.832, "eval_steps_per_second": 0.836, "eval_wer": 0.34267199240244683, "step": 21000 }, { "epoch": 14.75825720309206, "grad_norm": 0.1792164444923401, "learning_rate": 2.3498711642070743e-05, "loss": 0.0319, "step": 21001 }, { "epoch": 14.758959943780745, "grad_norm": 0.08654583990573883, "learning_rate": 2.3498243148278287e-05, "loss": 0.0167, "step": 21002 }, { "epoch": 14.759662684469431, "grad_norm": 0.21318966150283813, "learning_rate": 2.3497774654485827e-05, "loss": 0.0232, "step": 21003 }, { "epoch": 14.760365425158117, "grad_norm": 0.0674646645784378, "learning_rate": 2.349730616069337e-05, "loss": 0.0056, "step": 21004 }, { "epoch": 14.761068165846803, "grad_norm": 0.1053958311676979, "learning_rate": 2.3496837666900915e-05, "loss": 0.0147, "step": 21005 }, { "epoch": 14.761770906535489, "grad_norm": 0.14837384223937988, "learning_rate": 2.3496369173108458e-05, "loss": 0.0121, "step": 21006 }, { "epoch": 14.762473647224175, "grad_norm": 0.0804685428738594, "learning_rate": 2.3495900679316e-05, "loss": 0.0138, "step": 21007 }, { "epoch": 14.76317638791286, "grad_norm": 0.12493891268968582, "learning_rate": 2.3495432185523542e-05, "loss": 0.0193, "step": 21008 }, { "epoch": 14.763879128601546, "grad_norm": 0.08621075749397278, "learning_rate": 2.3494963691731086e-05, "loss": 0.0143, "step": 21009 }, { "epoch": 14.764581869290232, "grad_norm": 0.13024544715881348, "learning_rate": 2.349449519793863e-05, "loss": 0.0185, "step": 21010 }, { "epoch": 14.765284609978918, "grad_norm": 0.19041454792022705, "learning_rate": 2.3494026704146173e-05, "loss": 0.013, "step": 21011 }, { "epoch": 14.765987350667604, "grad_norm": 0.12381424009799957, "learning_rate": 2.3493558210353714e-05, "loss": 0.0353, "step": 21012 }, { "epoch": 14.76669009135629, "grad_norm": 0.3383874297142029, "learning_rate": 2.3493089716561254e-05, "loss": 0.022, "step": 21013 }, { "epoch": 14.767392832044976, "grad_norm": 0.10677580535411835, "learning_rate": 2.3492621222768798e-05, "loss": 0.0157, "step": 21014 }, { "epoch": 14.768095572733662, "grad_norm": 0.22880801558494568, "learning_rate": 2.349215272897634e-05, "loss": 0.0423, "step": 21015 }, { "epoch": 14.768798313422348, "grad_norm": 0.2541350722312927, "learning_rate": 2.3491684235183882e-05, "loss": 0.0514, "step": 21016 }, { "epoch": 14.769501054111032, "grad_norm": 0.1710882931947708, "learning_rate": 2.3491215741391426e-05, "loss": 0.0366, "step": 21017 }, { "epoch": 14.77020379479972, "grad_norm": 0.3728960454463959, "learning_rate": 2.349074724759897e-05, "loss": 0.0505, "step": 21018 }, { "epoch": 14.770906535488404, "grad_norm": 0.30031442642211914, "learning_rate": 2.3490278753806513e-05, "loss": 0.1039, "step": 21019 }, { "epoch": 14.77160927617709, "grad_norm": 0.6956123113632202, "learning_rate": 2.3489810260014057e-05, "loss": 0.1314, "step": 21020 }, { "epoch": 14.772312016865776, "grad_norm": 0.6389318108558655, "learning_rate": 2.3489341766221597e-05, "loss": 0.1704, "step": 21021 }, { "epoch": 14.773014757554462, "grad_norm": 0.7899965643882751, "learning_rate": 2.348887327242914e-05, "loss": 0.1664, "step": 21022 }, { "epoch": 14.773717498243148, "grad_norm": 0.2509218454360962, "learning_rate": 2.3488404778636685e-05, "loss": 0.0577, "step": 21023 }, { "epoch": 14.774420238931834, "grad_norm": 0.12573246657848358, "learning_rate": 2.348793628484423e-05, "loss": 0.0177, "step": 21024 }, { "epoch": 14.77512297962052, "grad_norm": 0.17753276228904724, "learning_rate": 2.348746779105177e-05, "loss": 0.0288, "step": 21025 }, { "epoch": 14.775825720309205, "grad_norm": 0.10886112600564957, "learning_rate": 2.3486999297259312e-05, "loss": 0.0146, "step": 21026 }, { "epoch": 14.776528460997891, "grad_norm": 0.09133841842412949, "learning_rate": 2.3486530803466856e-05, "loss": 0.0218, "step": 21027 }, { "epoch": 14.777231201686577, "grad_norm": 0.06743406504392624, "learning_rate": 2.34860623096744e-05, "loss": 0.0054, "step": 21028 }, { "epoch": 14.777933942375263, "grad_norm": 0.10304291546344757, "learning_rate": 2.348559381588194e-05, "loss": 0.0194, "step": 21029 }, { "epoch": 14.778636683063949, "grad_norm": 0.1139165386557579, "learning_rate": 2.348512532208948e-05, "loss": 0.0119, "step": 21030 }, { "epoch": 14.779339423752635, "grad_norm": 0.15670457482337952, "learning_rate": 2.3484656828297024e-05, "loss": 0.0206, "step": 21031 }, { "epoch": 14.780042164441321, "grad_norm": 0.0809573084115982, "learning_rate": 2.3484188334504568e-05, "loss": 0.007, "step": 21032 }, { "epoch": 14.780744905130007, "grad_norm": 0.1738843470811844, "learning_rate": 2.3483719840712112e-05, "loss": 0.025, "step": 21033 }, { "epoch": 14.781447645818693, "grad_norm": 0.07364822179079056, "learning_rate": 2.3483251346919652e-05, "loss": 0.0082, "step": 21034 }, { "epoch": 14.782150386507379, "grad_norm": 0.1705455482006073, "learning_rate": 2.3482782853127196e-05, "loss": 0.0237, "step": 21035 }, { "epoch": 14.782853127196065, "grad_norm": 0.2658602297306061, "learning_rate": 2.348231435933474e-05, "loss": 0.0128, "step": 21036 }, { "epoch": 14.78355586788475, "grad_norm": 0.16743192076683044, "learning_rate": 2.3481845865542283e-05, "loss": 0.0365, "step": 21037 }, { "epoch": 14.784258608573436, "grad_norm": 0.110728420317173, "learning_rate": 2.3481377371749824e-05, "loss": 0.0155, "step": 21038 }, { "epoch": 14.784961349262122, "grad_norm": 0.09317714720964432, "learning_rate": 2.3480908877957367e-05, "loss": 0.0129, "step": 21039 }, { "epoch": 14.785664089950808, "grad_norm": 0.2511570155620575, "learning_rate": 2.348044038416491e-05, "loss": 0.0427, "step": 21040 }, { "epoch": 14.786366830639494, "grad_norm": 0.21602456271648407, "learning_rate": 2.3479971890372455e-05, "loss": 0.0251, "step": 21041 }, { "epoch": 14.78706957132818, "grad_norm": 0.32413557171821594, "learning_rate": 2.3479503396579995e-05, "loss": 0.0518, "step": 21042 }, { "epoch": 14.787772312016866, "grad_norm": 0.4574090242385864, "learning_rate": 2.347903490278754e-05, "loss": 0.0502, "step": 21043 }, { "epoch": 14.788475052705552, "grad_norm": 0.3956606090068817, "learning_rate": 2.3478566408995083e-05, "loss": 0.0896, "step": 21044 }, { "epoch": 14.789177793394238, "grad_norm": 0.3561159074306488, "learning_rate": 2.3478097915202626e-05, "loss": 0.1153, "step": 21045 }, { "epoch": 14.789880534082924, "grad_norm": 0.7137964963912964, "learning_rate": 2.347762942141017e-05, "loss": 0.1727, "step": 21046 }, { "epoch": 14.79058327477161, "grad_norm": 2.7761690616607666, "learning_rate": 2.347716092761771e-05, "loss": 0.1982, "step": 21047 }, { "epoch": 14.791286015460296, "grad_norm": 0.24460504949092865, "learning_rate": 2.347669243382525e-05, "loss": 0.0508, "step": 21048 }, { "epoch": 14.791988756148982, "grad_norm": 0.1889750361442566, "learning_rate": 2.3476223940032794e-05, "loss": 0.0321, "step": 21049 }, { "epoch": 14.792691496837667, "grad_norm": 0.10035303980112076, "learning_rate": 2.3475755446240338e-05, "loss": 0.0222, "step": 21050 }, { "epoch": 14.793394237526353, "grad_norm": 0.11233252286911011, "learning_rate": 2.347528695244788e-05, "loss": 0.0141, "step": 21051 }, { "epoch": 14.79409697821504, "grad_norm": 0.09251218289136887, "learning_rate": 2.3474818458655422e-05, "loss": 0.013, "step": 21052 }, { "epoch": 14.794799718903725, "grad_norm": 0.09380286186933517, "learning_rate": 2.3474349964862966e-05, "loss": 0.0115, "step": 21053 }, { "epoch": 14.795502459592411, "grad_norm": 0.060247208923101425, "learning_rate": 2.347388147107051e-05, "loss": 0.0121, "step": 21054 }, { "epoch": 14.796205200281097, "grad_norm": 0.17858432233333588, "learning_rate": 2.347341297727805e-05, "loss": 0.0169, "step": 21055 }, { "epoch": 14.796907940969781, "grad_norm": 0.2732556462287903, "learning_rate": 2.3472944483485594e-05, "loss": 0.0301, "step": 21056 }, { "epoch": 14.797610681658469, "grad_norm": 0.07996634393930435, "learning_rate": 2.3472475989693137e-05, "loss": 0.0117, "step": 21057 }, { "epoch": 14.798313422347153, "grad_norm": 0.14062827825546265, "learning_rate": 2.347200749590068e-05, "loss": 0.0247, "step": 21058 }, { "epoch": 14.799016163035839, "grad_norm": 0.14229676127433777, "learning_rate": 2.3471539002108225e-05, "loss": 0.0213, "step": 21059 }, { "epoch": 14.799718903724525, "grad_norm": 0.18924817442893982, "learning_rate": 2.3471070508315765e-05, "loss": 0.0173, "step": 21060 }, { "epoch": 14.80042164441321, "grad_norm": 0.11973082274198532, "learning_rate": 2.347060201452331e-05, "loss": 0.0152, "step": 21061 }, { "epoch": 14.801124385101897, "grad_norm": 0.19681614637374878, "learning_rate": 2.3470133520730853e-05, "loss": 0.0298, "step": 21062 }, { "epoch": 14.801827125790583, "grad_norm": 0.15349158644676208, "learning_rate": 2.3469665026938396e-05, "loss": 0.0297, "step": 21063 }, { "epoch": 14.802529866479269, "grad_norm": 0.0938732847571373, "learning_rate": 2.3469196533145937e-05, "loss": 0.0142, "step": 21064 }, { "epoch": 14.803232607167955, "grad_norm": 0.6644953489303589, "learning_rate": 2.3468728039353477e-05, "loss": 0.0242, "step": 21065 }, { "epoch": 14.80393534785664, "grad_norm": 0.32452619075775146, "learning_rate": 2.346825954556102e-05, "loss": 0.0588, "step": 21066 }, { "epoch": 14.804638088545326, "grad_norm": 0.17671620845794678, "learning_rate": 2.3467791051768565e-05, "loss": 0.0452, "step": 21067 }, { "epoch": 14.805340829234012, "grad_norm": 0.260908305644989, "learning_rate": 2.3467322557976105e-05, "loss": 0.081, "step": 21068 }, { "epoch": 14.806043569922698, "grad_norm": 0.2977422773838043, "learning_rate": 2.346685406418365e-05, "loss": 0.0863, "step": 21069 }, { "epoch": 14.806746310611384, "grad_norm": 0.41525495052337646, "learning_rate": 2.3466385570391192e-05, "loss": 0.1419, "step": 21070 }, { "epoch": 14.80744905130007, "grad_norm": 0.7333936095237732, "learning_rate": 2.3465917076598736e-05, "loss": 0.1659, "step": 21071 }, { "epoch": 14.808151791988756, "grad_norm": 0.6408944129943848, "learning_rate": 2.346544858280628e-05, "loss": 0.1439, "step": 21072 }, { "epoch": 14.808854532677442, "grad_norm": 0.26587650179862976, "learning_rate": 2.346498008901382e-05, "loss": 0.0671, "step": 21073 }, { "epoch": 14.809557273366128, "grad_norm": 0.09623522311449051, "learning_rate": 2.3464511595221364e-05, "loss": 0.0211, "step": 21074 }, { "epoch": 14.810260014054814, "grad_norm": 0.12092465162277222, "learning_rate": 2.3464043101428908e-05, "loss": 0.0195, "step": 21075 }, { "epoch": 14.8109627547435, "grad_norm": 0.106229767203331, "learning_rate": 2.346357460763645e-05, "loss": 0.0142, "step": 21076 }, { "epoch": 14.811665495432186, "grad_norm": 0.19349829852581024, "learning_rate": 2.346310611384399e-05, "loss": 0.0166, "step": 21077 }, { "epoch": 14.812368236120872, "grad_norm": 0.06652036309242249, "learning_rate": 2.3462637620051535e-05, "loss": 0.0065, "step": 21078 }, { "epoch": 14.813070976809557, "grad_norm": 0.09788981825113297, "learning_rate": 2.346216912625908e-05, "loss": 0.0131, "step": 21079 }, { "epoch": 14.813773717498243, "grad_norm": 0.4438803791999817, "learning_rate": 2.3461700632466623e-05, "loss": 0.0241, "step": 21080 }, { "epoch": 14.81447645818693, "grad_norm": 0.08543440699577332, "learning_rate": 2.3461232138674163e-05, "loss": 0.0152, "step": 21081 }, { "epoch": 14.815179198875615, "grad_norm": 0.10702335834503174, "learning_rate": 2.3460763644881707e-05, "loss": 0.0093, "step": 21082 }, { "epoch": 14.815881939564301, "grad_norm": 0.19740910828113556, "learning_rate": 2.3460295151089247e-05, "loss": 0.0223, "step": 21083 }, { "epoch": 14.816584680252987, "grad_norm": 0.07808905839920044, "learning_rate": 2.345982665729679e-05, "loss": 0.0076, "step": 21084 }, { "epoch": 14.817287420941673, "grad_norm": 0.1304645985364914, "learning_rate": 2.3459358163504335e-05, "loss": 0.0247, "step": 21085 }, { "epoch": 14.817990161630359, "grad_norm": 0.109051913022995, "learning_rate": 2.3458889669711875e-05, "loss": 0.0147, "step": 21086 }, { "epoch": 14.818692902319045, "grad_norm": 0.2240278571844101, "learning_rate": 2.345842117591942e-05, "loss": 0.029, "step": 21087 }, { "epoch": 14.81939564300773, "grad_norm": 0.332268089056015, "learning_rate": 2.3457952682126962e-05, "loss": 0.0265, "step": 21088 }, { "epoch": 14.820098383696417, "grad_norm": 0.19733652472496033, "learning_rate": 2.3457484188334506e-05, "loss": 0.0239, "step": 21089 }, { "epoch": 14.820801124385103, "grad_norm": 0.10766611993312836, "learning_rate": 2.3457015694542046e-05, "loss": 0.0208, "step": 21090 }, { "epoch": 14.821503865073788, "grad_norm": 0.2015179544687271, "learning_rate": 2.345654720074959e-05, "loss": 0.0461, "step": 21091 }, { "epoch": 14.822206605762474, "grad_norm": 0.2474711537361145, "learning_rate": 2.3456078706957134e-05, "loss": 0.0446, "step": 21092 }, { "epoch": 14.82290934645116, "grad_norm": 0.4826856553554535, "learning_rate": 2.3455610213164678e-05, "loss": 0.0665, "step": 21093 }, { "epoch": 14.823612087139846, "grad_norm": 0.3056175410747528, "learning_rate": 2.3455141719372218e-05, "loss": 0.0928, "step": 21094 }, { "epoch": 14.82431482782853, "grad_norm": 1.115769624710083, "learning_rate": 2.3454673225579762e-05, "loss": 0.1349, "step": 21095 }, { "epoch": 14.825017568517218, "grad_norm": 0.5211876630783081, "learning_rate": 2.3454204731787305e-05, "loss": 0.1348, "step": 21096 }, { "epoch": 14.825720309205902, "grad_norm": 0.7300803065299988, "learning_rate": 2.345373623799485e-05, "loss": 0.1733, "step": 21097 }, { "epoch": 14.826423049894588, "grad_norm": 0.17877961695194244, "learning_rate": 2.3453267744202393e-05, "loss": 0.0578, "step": 21098 }, { "epoch": 14.827125790583274, "grad_norm": 0.14485472440719604, "learning_rate": 2.3452799250409933e-05, "loss": 0.0163, "step": 21099 }, { "epoch": 14.82782853127196, "grad_norm": 0.11307749152183533, "learning_rate": 2.3452330756617474e-05, "loss": 0.0136, "step": 21100 }, { "epoch": 14.828531271960646, "grad_norm": 0.1992575228214264, "learning_rate": 2.3451862262825017e-05, "loss": 0.0142, "step": 21101 }, { "epoch": 14.829234012649332, "grad_norm": 0.12842875719070435, "learning_rate": 2.345139376903256e-05, "loss": 0.0197, "step": 21102 }, { "epoch": 14.829936753338018, "grad_norm": 0.13106130063533783, "learning_rate": 2.34509252752401e-05, "loss": 0.0145, "step": 21103 }, { "epoch": 14.830639494026704, "grad_norm": 0.07424556463956833, "learning_rate": 2.3450456781447645e-05, "loss": 0.0111, "step": 21104 }, { "epoch": 14.83134223471539, "grad_norm": 0.12292109429836273, "learning_rate": 2.344998828765519e-05, "loss": 0.0173, "step": 21105 }, { "epoch": 14.832044975404076, "grad_norm": 0.2062045782804489, "learning_rate": 2.3449519793862733e-05, "loss": 0.0159, "step": 21106 }, { "epoch": 14.832747716092761, "grad_norm": 0.09728997945785522, "learning_rate": 2.3449051300070273e-05, "loss": 0.0116, "step": 21107 }, { "epoch": 14.833450456781447, "grad_norm": 0.13451191782951355, "learning_rate": 2.3448582806277817e-05, "loss": 0.0161, "step": 21108 }, { "epoch": 14.834153197470133, "grad_norm": 0.09760154038667679, "learning_rate": 2.344811431248536e-05, "loss": 0.0117, "step": 21109 }, { "epoch": 14.83485593815882, "grad_norm": 0.13472485542297363, "learning_rate": 2.3447645818692904e-05, "loss": 0.0194, "step": 21110 }, { "epoch": 14.835558678847505, "grad_norm": 0.09937111288309097, "learning_rate": 2.3447177324900448e-05, "loss": 0.021, "step": 21111 }, { "epoch": 14.836261419536191, "grad_norm": 0.1463432013988495, "learning_rate": 2.3446708831107988e-05, "loss": 0.0278, "step": 21112 }, { "epoch": 14.836964160224877, "grad_norm": 0.1900453120470047, "learning_rate": 2.3446240337315532e-05, "loss": 0.0218, "step": 21113 }, { "epoch": 14.837666900913563, "grad_norm": 0.16961419582366943, "learning_rate": 2.3445771843523076e-05, "loss": 0.0116, "step": 21114 }, { "epoch": 14.838369641602249, "grad_norm": 0.15064288675785065, "learning_rate": 2.344530334973062e-05, "loss": 0.028, "step": 21115 }, { "epoch": 14.839072382290935, "grad_norm": 0.18364408612251282, "learning_rate": 2.344483485593816e-05, "loss": 0.0416, "step": 21116 }, { "epoch": 14.83977512297962, "grad_norm": 0.1515890508890152, "learning_rate": 2.34443663621457e-05, "loss": 0.0347, "step": 21117 }, { "epoch": 14.840477863668307, "grad_norm": 0.25196123123168945, "learning_rate": 2.3443897868353244e-05, "loss": 0.0533, "step": 21118 }, { "epoch": 14.841180604356992, "grad_norm": 0.5499275326728821, "learning_rate": 2.3443429374560787e-05, "loss": 0.0932, "step": 21119 }, { "epoch": 14.841883345045678, "grad_norm": 0.43183228373527527, "learning_rate": 2.3442960880768328e-05, "loss": 0.1322, "step": 21120 }, { "epoch": 14.842586085734364, "grad_norm": 0.7476230263710022, "learning_rate": 2.344249238697587e-05, "loss": 0.1439, "step": 21121 }, { "epoch": 14.84328882642305, "grad_norm": 0.6538933515548706, "learning_rate": 2.3442023893183415e-05, "loss": 0.1594, "step": 21122 }, { "epoch": 14.843991567111736, "grad_norm": 0.23743335902690887, "learning_rate": 2.344155539939096e-05, "loss": 0.0599, "step": 21123 }, { "epoch": 14.844694307800422, "grad_norm": 0.13899993896484375, "learning_rate": 2.3441086905598503e-05, "loss": 0.0326, "step": 21124 }, { "epoch": 14.845397048489108, "grad_norm": 0.1264781951904297, "learning_rate": 2.3440618411806043e-05, "loss": 0.0148, "step": 21125 }, { "epoch": 14.846099789177794, "grad_norm": 0.1539008617401123, "learning_rate": 2.3440149918013587e-05, "loss": 0.0142, "step": 21126 }, { "epoch": 14.84680252986648, "grad_norm": 0.04554169997572899, "learning_rate": 2.343968142422113e-05, "loss": 0.0083, "step": 21127 }, { "epoch": 14.847505270555166, "grad_norm": 0.08583630621433258, "learning_rate": 2.3439212930428674e-05, "loss": 0.0101, "step": 21128 }, { "epoch": 14.848208011243852, "grad_norm": 0.5935643315315247, "learning_rate": 2.3438744436636214e-05, "loss": 0.0151, "step": 21129 }, { "epoch": 14.848910751932538, "grad_norm": 0.12500065565109253, "learning_rate": 2.3438275942843758e-05, "loss": 0.026, "step": 21130 }, { "epoch": 14.849613492621224, "grad_norm": 0.13316094875335693, "learning_rate": 2.3437807449051302e-05, "loss": 0.0259, "step": 21131 }, { "epoch": 14.85031623330991, "grad_norm": 0.0587179958820343, "learning_rate": 2.3437338955258846e-05, "loss": 0.0061, "step": 21132 }, { "epoch": 14.851018973998595, "grad_norm": 0.11196952313184738, "learning_rate": 2.3436870461466386e-05, "loss": 0.0227, "step": 21133 }, { "epoch": 14.85172171468728, "grad_norm": 0.14687903225421906, "learning_rate": 2.343640196767393e-05, "loss": 0.0102, "step": 21134 }, { "epoch": 14.852424455375965, "grad_norm": 0.10216313600540161, "learning_rate": 2.343593347388147e-05, "loss": 0.0237, "step": 21135 }, { "epoch": 14.853127196064651, "grad_norm": 0.15513435006141663, "learning_rate": 2.3435464980089014e-05, "loss": 0.0132, "step": 21136 }, { "epoch": 14.853829936753337, "grad_norm": 0.18776024878025055, "learning_rate": 2.3434996486296558e-05, "loss": 0.0234, "step": 21137 }, { "epoch": 14.854532677442023, "grad_norm": 0.1888570785522461, "learning_rate": 2.3434527992504098e-05, "loss": 0.0383, "step": 21138 }, { "epoch": 14.85523541813071, "grad_norm": 0.16054046154022217, "learning_rate": 2.343405949871164e-05, "loss": 0.0213, "step": 21139 }, { "epoch": 14.855938158819395, "grad_norm": 0.17934241890907288, "learning_rate": 2.3433591004919185e-05, "loss": 0.0211, "step": 21140 }, { "epoch": 14.856640899508081, "grad_norm": 0.3022003471851349, "learning_rate": 2.343312251112673e-05, "loss": 0.054, "step": 21141 }, { "epoch": 14.857343640196767, "grad_norm": 0.26040294766426086, "learning_rate": 2.343265401733427e-05, "loss": 0.0677, "step": 21142 }, { "epoch": 14.858046380885453, "grad_norm": 0.2597082257270813, "learning_rate": 2.3432185523541813e-05, "loss": 0.0693, "step": 21143 }, { "epoch": 14.858749121574139, "grad_norm": 0.636838436126709, "learning_rate": 2.3431717029749357e-05, "loss": 0.0921, "step": 21144 }, { "epoch": 14.859451862262825, "grad_norm": 1.3161766529083252, "learning_rate": 2.34312485359569e-05, "loss": 0.1301, "step": 21145 }, { "epoch": 14.86015460295151, "grad_norm": 0.6286442875862122, "learning_rate": 2.343078004216444e-05, "loss": 0.1479, "step": 21146 }, { "epoch": 14.860857343640197, "grad_norm": 0.6028013229370117, "learning_rate": 2.3430311548371985e-05, "loss": 0.157, "step": 21147 }, { "epoch": 14.861560084328882, "grad_norm": 0.14149339497089386, "learning_rate": 2.342984305457953e-05, "loss": 0.0627, "step": 21148 }, { "epoch": 14.862262825017568, "grad_norm": 0.09850180894136429, "learning_rate": 2.3429374560787072e-05, "loss": 0.0278, "step": 21149 }, { "epoch": 14.862965565706254, "grad_norm": 0.49988022446632385, "learning_rate": 2.3428906066994616e-05, "loss": 0.0259, "step": 21150 }, { "epoch": 14.86366830639494, "grad_norm": 0.13989828526973724, "learning_rate": 2.3428437573202156e-05, "loss": 0.0153, "step": 21151 }, { "epoch": 14.864371047083626, "grad_norm": 0.16777290403842926, "learning_rate": 2.3427969079409696e-05, "loss": 0.0203, "step": 21152 }, { "epoch": 14.865073787772312, "grad_norm": 0.06469134241342545, "learning_rate": 2.342750058561724e-05, "loss": 0.0087, "step": 21153 }, { "epoch": 14.865776528460998, "grad_norm": 0.15686282515525818, "learning_rate": 2.3427032091824784e-05, "loss": 0.0186, "step": 21154 }, { "epoch": 14.866479269149684, "grad_norm": 0.10542580485343933, "learning_rate": 2.3426563598032324e-05, "loss": 0.0208, "step": 21155 }, { "epoch": 14.86718200983837, "grad_norm": 0.100316621363163, "learning_rate": 2.3426095104239868e-05, "loss": 0.0152, "step": 21156 }, { "epoch": 14.867884750527056, "grad_norm": 0.23767943680286407, "learning_rate": 2.342562661044741e-05, "loss": 0.0086, "step": 21157 }, { "epoch": 14.868587491215742, "grad_norm": 0.14520379900932312, "learning_rate": 2.3425158116654955e-05, "loss": 0.0155, "step": 21158 }, { "epoch": 14.869290231904428, "grad_norm": 0.20017528533935547, "learning_rate": 2.3424689622862496e-05, "loss": 0.0129, "step": 21159 }, { "epoch": 14.869992972593113, "grad_norm": 0.11504537612199783, "learning_rate": 2.342422112907004e-05, "loss": 0.0314, "step": 21160 }, { "epoch": 14.8706957132818, "grad_norm": 0.4215624928474426, "learning_rate": 2.3423752635277583e-05, "loss": 0.015, "step": 21161 }, { "epoch": 14.871398453970485, "grad_norm": 0.19228626787662506, "learning_rate": 2.3423284141485127e-05, "loss": 0.0221, "step": 21162 }, { "epoch": 14.872101194659171, "grad_norm": 0.13876476883888245, "learning_rate": 2.342281564769267e-05, "loss": 0.0193, "step": 21163 }, { "epoch": 14.872803935347857, "grad_norm": 0.13587011396884918, "learning_rate": 2.342234715390021e-05, "loss": 0.0279, "step": 21164 }, { "epoch": 14.873506676036543, "grad_norm": 0.2572200298309326, "learning_rate": 2.3421878660107755e-05, "loss": 0.0422, "step": 21165 }, { "epoch": 14.874209416725229, "grad_norm": 2.590928316116333, "learning_rate": 2.34214101663153e-05, "loss": 0.0433, "step": 21166 }, { "epoch": 14.874912157413915, "grad_norm": 0.3162250518798828, "learning_rate": 2.3420941672522842e-05, "loss": 0.0677, "step": 21167 }, { "epoch": 14.8756148981026, "grad_norm": 1.0938782691955566, "learning_rate": 2.3420473178730382e-05, "loss": 0.0787, "step": 21168 }, { "epoch": 14.876317638791287, "grad_norm": 0.2404192090034485, "learning_rate": 2.3420004684937926e-05, "loss": 0.0924, "step": 21169 }, { "epoch": 14.877020379479973, "grad_norm": 0.5327011346817017, "learning_rate": 2.3419536191145467e-05, "loss": 0.1188, "step": 21170 }, { "epoch": 14.877723120168657, "grad_norm": 0.7279034852981567, "learning_rate": 2.341906769735301e-05, "loss": 0.1506, "step": 21171 }, { "epoch": 14.878425860857345, "grad_norm": 0.6607451438903809, "learning_rate": 2.341859920356055e-05, "loss": 0.1789, "step": 21172 }, { "epoch": 14.879128601546029, "grad_norm": 0.1911916434764862, "learning_rate": 2.3418130709768094e-05, "loss": 0.0687, "step": 21173 }, { "epoch": 14.879831342234715, "grad_norm": 0.12006958574056625, "learning_rate": 2.3417662215975638e-05, "loss": 0.025, "step": 21174 }, { "epoch": 14.8805340829234, "grad_norm": 0.23602336645126343, "learning_rate": 2.3417193722183182e-05, "loss": 0.0327, "step": 21175 }, { "epoch": 14.881236823612086, "grad_norm": 0.16522617638111115, "learning_rate": 2.3416725228390726e-05, "loss": 0.0186, "step": 21176 }, { "epoch": 14.881939564300772, "grad_norm": 0.07957033067941666, "learning_rate": 2.3416256734598266e-05, "loss": 0.0149, "step": 21177 }, { "epoch": 14.882642304989458, "grad_norm": 0.058902762830257416, "learning_rate": 2.341578824080581e-05, "loss": 0.006, "step": 21178 }, { "epoch": 14.883345045678144, "grad_norm": 0.11915824562311172, "learning_rate": 2.3415319747013353e-05, "loss": 0.0138, "step": 21179 }, { "epoch": 14.88404778636683, "grad_norm": 0.15453852713108063, "learning_rate": 2.3414851253220897e-05, "loss": 0.02, "step": 21180 }, { "epoch": 14.884750527055516, "grad_norm": 0.27807047963142395, "learning_rate": 2.3414382759428437e-05, "loss": 0.0318, "step": 21181 }, { "epoch": 14.885453267744202, "grad_norm": 0.2538786232471466, "learning_rate": 2.341391426563598e-05, "loss": 0.0097, "step": 21182 }, { "epoch": 14.886156008432888, "grad_norm": 0.21029353141784668, "learning_rate": 2.3413445771843525e-05, "loss": 0.028, "step": 21183 }, { "epoch": 14.886858749121574, "grad_norm": 0.09748807549476624, "learning_rate": 2.341297727805107e-05, "loss": 0.014, "step": 21184 }, { "epoch": 14.88756148981026, "grad_norm": 0.10498391091823578, "learning_rate": 2.341250878425861e-05, "loss": 0.0314, "step": 21185 }, { "epoch": 14.888264230498946, "grad_norm": 0.17629267275333405, "learning_rate": 2.3412040290466153e-05, "loss": 0.0136, "step": 21186 }, { "epoch": 14.888966971187632, "grad_norm": 0.25594353675842285, "learning_rate": 2.3411571796673693e-05, "loss": 0.0198, "step": 21187 }, { "epoch": 14.889669711876317, "grad_norm": 0.25345784425735474, "learning_rate": 2.3411103302881237e-05, "loss": 0.0215, "step": 21188 }, { "epoch": 14.890372452565003, "grad_norm": 0.15018440783023834, "learning_rate": 2.341063480908878e-05, "loss": 0.0164, "step": 21189 }, { "epoch": 14.89107519325369, "grad_norm": 0.6539618968963623, "learning_rate": 2.341016631529632e-05, "loss": 0.0388, "step": 21190 }, { "epoch": 14.891777933942375, "grad_norm": 0.1669747233390808, "learning_rate": 2.3409697821503864e-05, "loss": 0.0279, "step": 21191 }, { "epoch": 14.892480674631061, "grad_norm": 0.22288890182971954, "learning_rate": 2.3409229327711408e-05, "loss": 0.0467, "step": 21192 }, { "epoch": 14.893183415319747, "grad_norm": 0.5578683018684387, "learning_rate": 2.3408760833918952e-05, "loss": 0.0481, "step": 21193 }, { "epoch": 14.893886156008433, "grad_norm": 0.6365357041358948, "learning_rate": 2.3408292340126492e-05, "loss": 0.0978, "step": 21194 }, { "epoch": 14.894588896697119, "grad_norm": 1.3378297090530396, "learning_rate": 2.3407823846334036e-05, "loss": 0.1062, "step": 21195 }, { "epoch": 14.895291637385805, "grad_norm": 0.5618146657943726, "learning_rate": 2.340735535254158e-05, "loss": 0.1493, "step": 21196 }, { "epoch": 14.89599437807449, "grad_norm": 0.8435612320899963, "learning_rate": 2.3406886858749123e-05, "loss": 0.1709, "step": 21197 }, { "epoch": 14.896697118763177, "grad_norm": 0.4224904775619507, "learning_rate": 2.3406418364956664e-05, "loss": 0.0536, "step": 21198 }, { "epoch": 14.897399859451863, "grad_norm": 0.154163658618927, "learning_rate": 2.3405949871164207e-05, "loss": 0.0144, "step": 21199 }, { "epoch": 14.898102600140549, "grad_norm": 0.3537770211696625, "learning_rate": 2.340548137737175e-05, "loss": 0.0221, "step": 21200 }, { "epoch": 14.898805340829234, "grad_norm": 0.1412203311920166, "learning_rate": 2.3405012883579295e-05, "loss": 0.0135, "step": 21201 }, { "epoch": 14.89950808151792, "grad_norm": 0.11218206584453583, "learning_rate": 2.340454438978684e-05, "loss": 0.0098, "step": 21202 }, { "epoch": 14.900210822206606, "grad_norm": 0.0584990456700325, "learning_rate": 2.340407589599438e-05, "loss": 0.0109, "step": 21203 }, { "epoch": 14.900913562895292, "grad_norm": 0.050117913633584976, "learning_rate": 2.3403607402201923e-05, "loss": 0.0071, "step": 21204 }, { "epoch": 14.901616303583978, "grad_norm": 1.12409508228302, "learning_rate": 2.3403138908409463e-05, "loss": 0.0161, "step": 21205 }, { "epoch": 14.902319044272664, "grad_norm": 0.09039735049009323, "learning_rate": 2.3402670414617007e-05, "loss": 0.0146, "step": 21206 }, { "epoch": 14.90302178496135, "grad_norm": 0.0784100741147995, "learning_rate": 2.3402201920824547e-05, "loss": 0.0118, "step": 21207 }, { "epoch": 14.903724525650036, "grad_norm": 0.3009507954120636, "learning_rate": 2.340173342703209e-05, "loss": 0.0192, "step": 21208 }, { "epoch": 14.904427266338722, "grad_norm": 0.09787777811288834, "learning_rate": 2.3401264933239635e-05, "loss": 0.0131, "step": 21209 }, { "epoch": 14.905130007027406, "grad_norm": 0.18203452229499817, "learning_rate": 2.3400796439447178e-05, "loss": 0.0274, "step": 21210 }, { "epoch": 14.905832747716094, "grad_norm": 0.0978253185749054, "learning_rate": 2.3400327945654722e-05, "loss": 0.0151, "step": 21211 }, { "epoch": 14.906535488404778, "grad_norm": 0.18249835073947906, "learning_rate": 2.3399859451862262e-05, "loss": 0.0305, "step": 21212 }, { "epoch": 14.907238229093464, "grad_norm": 0.11915526539087296, "learning_rate": 2.3399390958069806e-05, "loss": 0.0247, "step": 21213 }, { "epoch": 14.90794096978215, "grad_norm": 0.09473472088575363, "learning_rate": 2.339892246427735e-05, "loss": 0.0167, "step": 21214 }, { "epoch": 14.908643710470836, "grad_norm": 0.19870047271251678, "learning_rate": 2.3398453970484894e-05, "loss": 0.0305, "step": 21215 }, { "epoch": 14.909346451159522, "grad_norm": 0.2613554298877716, "learning_rate": 2.3397985476692434e-05, "loss": 0.0393, "step": 21216 }, { "epoch": 14.910049191848207, "grad_norm": 0.23200690746307373, "learning_rate": 2.3397516982899978e-05, "loss": 0.0467, "step": 21217 }, { "epoch": 14.910751932536893, "grad_norm": 0.21306033432483673, "learning_rate": 2.339704848910752e-05, "loss": 0.0449, "step": 21218 }, { "epoch": 14.91145467322558, "grad_norm": 0.41770845651626587, "learning_rate": 2.3396579995315065e-05, "loss": 0.1031, "step": 21219 }, { "epoch": 14.912157413914265, "grad_norm": 0.6150475144386292, "learning_rate": 2.3396111501522605e-05, "loss": 0.1209, "step": 21220 }, { "epoch": 14.912860154602951, "grad_norm": 0.42898502945899963, "learning_rate": 2.339564300773015e-05, "loss": 0.1225, "step": 21221 }, { "epoch": 14.913562895291637, "grad_norm": 1.6386820077896118, "learning_rate": 2.339517451393769e-05, "loss": 0.1717, "step": 21222 }, { "epoch": 14.914265635980323, "grad_norm": 0.1408357173204422, "learning_rate": 2.3394706020145233e-05, "loss": 0.0516, "step": 21223 }, { "epoch": 14.914968376669009, "grad_norm": 0.11196797341108322, "learning_rate": 2.3394237526352777e-05, "loss": 0.0212, "step": 21224 }, { "epoch": 14.915671117357695, "grad_norm": 0.26186588406562805, "learning_rate": 2.3393769032560317e-05, "loss": 0.0167, "step": 21225 }, { "epoch": 14.91637385804638, "grad_norm": 0.10426562279462814, "learning_rate": 2.339330053876786e-05, "loss": 0.008, "step": 21226 }, { "epoch": 14.917076598735067, "grad_norm": 0.2087331861257553, "learning_rate": 2.3392832044975405e-05, "loss": 0.0187, "step": 21227 }, { "epoch": 14.917779339423753, "grad_norm": 0.08263377100229263, "learning_rate": 2.339236355118295e-05, "loss": 0.0101, "step": 21228 }, { "epoch": 14.918482080112438, "grad_norm": 0.39656344056129456, "learning_rate": 2.339189505739049e-05, "loss": 0.0161, "step": 21229 }, { "epoch": 14.919184820801124, "grad_norm": 0.13796497881412506, "learning_rate": 2.3391426563598032e-05, "loss": 0.0139, "step": 21230 }, { "epoch": 14.91988756148981, "grad_norm": 0.10657881200313568, "learning_rate": 2.3390958069805576e-05, "loss": 0.0176, "step": 21231 }, { "epoch": 14.920590302178496, "grad_norm": 0.16817383468151093, "learning_rate": 2.339048957601312e-05, "loss": 0.0128, "step": 21232 }, { "epoch": 14.921293042867182, "grad_norm": 0.16161766648292542, "learning_rate": 2.339002108222066e-05, "loss": 0.0205, "step": 21233 }, { "epoch": 14.921995783555868, "grad_norm": 0.1116509959101677, "learning_rate": 2.3389552588428204e-05, "loss": 0.0096, "step": 21234 }, { "epoch": 14.922698524244554, "grad_norm": 0.14166216552257538, "learning_rate": 2.3389084094635748e-05, "loss": 0.0191, "step": 21235 }, { "epoch": 14.92340126493324, "grad_norm": 0.10427606105804443, "learning_rate": 2.338861560084329e-05, "loss": 0.0075, "step": 21236 }, { "epoch": 14.924104005621926, "grad_norm": 1.014168381690979, "learning_rate": 2.3388147107050835e-05, "loss": 0.0354, "step": 21237 }, { "epoch": 14.924806746310612, "grad_norm": 0.16320079565048218, "learning_rate": 2.3387678613258375e-05, "loss": 0.0261, "step": 21238 }, { "epoch": 14.925509486999298, "grad_norm": 0.07527431845664978, "learning_rate": 2.3387210119465916e-05, "loss": 0.0083, "step": 21239 }, { "epoch": 14.926212227687984, "grad_norm": 0.17330114543437958, "learning_rate": 2.338674162567346e-05, "loss": 0.037, "step": 21240 }, { "epoch": 14.92691496837667, "grad_norm": 0.1640315055847168, "learning_rate": 2.3386273131881003e-05, "loss": 0.029, "step": 21241 }, { "epoch": 14.927617709065355, "grad_norm": 0.15778714418411255, "learning_rate": 2.3385804638088544e-05, "loss": 0.0473, "step": 21242 }, { "epoch": 14.928320449754041, "grad_norm": 0.3004351556301117, "learning_rate": 2.3385336144296087e-05, "loss": 0.0754, "step": 21243 }, { "epoch": 14.929023190442727, "grad_norm": 0.40213730931282043, "learning_rate": 2.338486765050363e-05, "loss": 0.1004, "step": 21244 }, { "epoch": 14.929725931131413, "grad_norm": 1.8132965564727783, "learning_rate": 2.3384399156711175e-05, "loss": 0.1648, "step": 21245 }, { "epoch": 14.9304286718201, "grad_norm": 0.9101672768592834, "learning_rate": 2.3383930662918715e-05, "loss": 0.1419, "step": 21246 }, { "epoch": 14.931131412508785, "grad_norm": 2.3852758407592773, "learning_rate": 2.338346216912626e-05, "loss": 0.1809, "step": 21247 }, { "epoch": 14.931834153197471, "grad_norm": 0.22216899693012238, "learning_rate": 2.3382993675333803e-05, "loss": 0.0594, "step": 21248 }, { "epoch": 14.932536893886155, "grad_norm": 0.20465916395187378, "learning_rate": 2.3382525181541346e-05, "loss": 0.0264, "step": 21249 }, { "epoch": 14.933239634574843, "grad_norm": 0.0942501425743103, "learning_rate": 2.338205668774889e-05, "loss": 0.0148, "step": 21250 }, { "epoch": 14.933942375263527, "grad_norm": 0.22503767907619476, "learning_rate": 2.338158819395643e-05, "loss": 0.0075, "step": 21251 }, { "epoch": 14.934645115952213, "grad_norm": 0.10419277846813202, "learning_rate": 2.3381119700163974e-05, "loss": 0.0148, "step": 21252 }, { "epoch": 14.935347856640899, "grad_norm": 0.08547132462263107, "learning_rate": 2.3380651206371518e-05, "loss": 0.0118, "step": 21253 }, { "epoch": 14.936050597329585, "grad_norm": 0.18587636947631836, "learning_rate": 2.338018271257906e-05, "loss": 0.0136, "step": 21254 }, { "epoch": 14.93675333801827, "grad_norm": 0.45772814750671387, "learning_rate": 2.3379714218786602e-05, "loss": 0.0308, "step": 21255 }, { "epoch": 14.937456078706957, "grad_norm": 0.17482474446296692, "learning_rate": 2.3379245724994146e-05, "loss": 0.0225, "step": 21256 }, { "epoch": 14.938158819395642, "grad_norm": 0.22656968235969543, "learning_rate": 2.3378777231201686e-05, "loss": 0.0362, "step": 21257 }, { "epoch": 14.938861560084328, "grad_norm": 0.15608638525009155, "learning_rate": 2.337830873740923e-05, "loss": 0.0198, "step": 21258 }, { "epoch": 14.939564300773014, "grad_norm": 0.36339348554611206, "learning_rate": 2.337784024361677e-05, "loss": 0.0168, "step": 21259 }, { "epoch": 14.9402670414617, "grad_norm": 0.36284616589546204, "learning_rate": 2.3377371749824314e-05, "loss": 0.0346, "step": 21260 }, { "epoch": 14.940969782150386, "grad_norm": 0.16834726929664612, "learning_rate": 2.3376903256031857e-05, "loss": 0.0127, "step": 21261 }, { "epoch": 14.941672522839072, "grad_norm": 0.15202881395816803, "learning_rate": 2.33764347622394e-05, "loss": 0.028, "step": 21262 }, { "epoch": 14.942375263527758, "grad_norm": 0.1434553861618042, "learning_rate": 2.3375966268446945e-05, "loss": 0.0297, "step": 21263 }, { "epoch": 14.943078004216444, "grad_norm": 0.11815634369850159, "learning_rate": 2.3375497774654485e-05, "loss": 0.0239, "step": 21264 }, { "epoch": 14.94378074490513, "grad_norm": 0.09478671103715897, "learning_rate": 2.337502928086203e-05, "loss": 0.0182, "step": 21265 }, { "epoch": 14.944483485593816, "grad_norm": 0.18342123925685883, "learning_rate": 2.3374560787069573e-05, "loss": 0.0377, "step": 21266 }, { "epoch": 14.945186226282502, "grad_norm": 0.34598255157470703, "learning_rate": 2.3374092293277116e-05, "loss": 0.0547, "step": 21267 }, { "epoch": 14.945888966971188, "grad_norm": 0.45726412534713745, "learning_rate": 2.3373623799484657e-05, "loss": 0.0645, "step": 21268 }, { "epoch": 14.946591707659874, "grad_norm": 0.30162525177001953, "learning_rate": 2.33731553056922e-05, "loss": 0.1003, "step": 21269 }, { "epoch": 14.94729444834856, "grad_norm": 0.46047621965408325, "learning_rate": 2.3372686811899744e-05, "loss": 0.0947, "step": 21270 }, { "epoch": 14.947997189037245, "grad_norm": 0.9451343417167664, "learning_rate": 2.3372218318107288e-05, "loss": 0.1258, "step": 21271 }, { "epoch": 14.948699929725931, "grad_norm": 0.9672704339027405, "learning_rate": 2.3371749824314828e-05, "loss": 0.2013, "step": 21272 }, { "epoch": 14.949402670414617, "grad_norm": 0.24199660122394562, "learning_rate": 2.3371281330522372e-05, "loss": 0.0694, "step": 21273 }, { "epoch": 14.950105411103303, "grad_norm": 0.11047857999801636, "learning_rate": 2.3370812836729912e-05, "loss": 0.0243, "step": 21274 }, { "epoch": 14.950808151791989, "grad_norm": 0.26453647017478943, "learning_rate": 2.3370344342937456e-05, "loss": 0.0216, "step": 21275 }, { "epoch": 14.951510892480675, "grad_norm": 0.15014316141605377, "learning_rate": 2.3369875849145e-05, "loss": 0.0172, "step": 21276 }, { "epoch": 14.952213633169361, "grad_norm": 0.08932997286319733, "learning_rate": 2.336940735535254e-05, "loss": 0.0069, "step": 21277 }, { "epoch": 14.952916373858047, "grad_norm": 0.2798391282558441, "learning_rate": 2.3368938861560084e-05, "loss": 0.0125, "step": 21278 }, { "epoch": 14.953619114546733, "grad_norm": 0.06787561625242233, "learning_rate": 2.3368470367767628e-05, "loss": 0.0092, "step": 21279 }, { "epoch": 14.954321855235419, "grad_norm": 0.29812926054000854, "learning_rate": 2.336800187397517e-05, "loss": 0.0091, "step": 21280 }, { "epoch": 14.955024595924105, "grad_norm": 0.08436181396245956, "learning_rate": 2.336753338018271e-05, "loss": 0.0196, "step": 21281 }, { "epoch": 14.95572733661279, "grad_norm": 0.13862614333629608, "learning_rate": 2.3367064886390255e-05, "loss": 0.012, "step": 21282 }, { "epoch": 14.956430077301476, "grad_norm": 0.1763206571340561, "learning_rate": 2.33665963925978e-05, "loss": 0.0145, "step": 21283 }, { "epoch": 14.957132817990162, "grad_norm": 0.11885496973991394, "learning_rate": 2.3366127898805343e-05, "loss": 0.0104, "step": 21284 }, { "epoch": 14.957835558678848, "grad_norm": 0.1843503713607788, "learning_rate": 2.3365659405012883e-05, "loss": 0.033, "step": 21285 }, { "epoch": 14.958538299367534, "grad_norm": 0.06385057419538498, "learning_rate": 2.3365190911220427e-05, "loss": 0.0108, "step": 21286 }, { "epoch": 14.95924104005622, "grad_norm": 0.15571105480194092, "learning_rate": 2.336472241742797e-05, "loss": 0.0186, "step": 21287 }, { "epoch": 14.959943780744904, "grad_norm": 0.22347265481948853, "learning_rate": 2.3364253923635514e-05, "loss": 0.0228, "step": 21288 }, { "epoch": 14.96064652143359, "grad_norm": 0.1288778930902481, "learning_rate": 2.3363785429843058e-05, "loss": 0.0174, "step": 21289 }, { "epoch": 14.961349262122276, "grad_norm": 0.1530647873878479, "learning_rate": 2.33633169360506e-05, "loss": 0.0233, "step": 21290 }, { "epoch": 14.962052002810962, "grad_norm": 0.24141591787338257, "learning_rate": 2.3362848442258142e-05, "loss": 0.0358, "step": 21291 }, { "epoch": 14.962754743499648, "grad_norm": 0.19297240674495697, "learning_rate": 2.3362379948465682e-05, "loss": 0.0437, "step": 21292 }, { "epoch": 14.963457484188334, "grad_norm": 0.6179813146591187, "learning_rate": 2.3361911454673226e-05, "loss": 0.0671, "step": 21293 }, { "epoch": 14.96416022487702, "grad_norm": 0.3621077835559845, "learning_rate": 2.3361442960880767e-05, "loss": 0.0838, "step": 21294 }, { "epoch": 14.964862965565706, "grad_norm": 0.7144325375556946, "learning_rate": 2.336097446708831e-05, "loss": 0.13, "step": 21295 }, { "epoch": 14.965565706254392, "grad_norm": 0.7686638236045837, "learning_rate": 2.3360505973295854e-05, "loss": 0.138, "step": 21296 }, { "epoch": 14.966268446943078, "grad_norm": 0.6728037595748901, "learning_rate": 2.3360037479503398e-05, "loss": 0.166, "step": 21297 }, { "epoch": 14.966971187631763, "grad_norm": 0.3854246437549591, "learning_rate": 2.3359568985710938e-05, "loss": 0.0596, "step": 21298 }, { "epoch": 14.96767392832045, "grad_norm": 0.1381918340921402, "learning_rate": 2.3359100491918482e-05, "loss": 0.0283, "step": 21299 }, { "epoch": 14.968376669009135, "grad_norm": 0.25217798352241516, "learning_rate": 2.3358631998126025e-05, "loss": 0.0404, "step": 21300 }, { "epoch": 14.969079409697821, "grad_norm": 0.09687114506959915, "learning_rate": 2.335816350433357e-05, "loss": 0.0165, "step": 21301 }, { "epoch": 14.969782150386507, "grad_norm": 0.15817615389823914, "learning_rate": 2.3357695010541113e-05, "loss": 0.0131, "step": 21302 }, { "epoch": 14.970484891075193, "grad_norm": 0.19363591074943542, "learning_rate": 2.3357226516748653e-05, "loss": 0.0064, "step": 21303 }, { "epoch": 14.971187631763879, "grad_norm": 0.08810415118932724, "learning_rate": 2.3356758022956197e-05, "loss": 0.014, "step": 21304 }, { "epoch": 14.971890372452565, "grad_norm": 0.11354938894510269, "learning_rate": 2.335628952916374e-05, "loss": 0.0162, "step": 21305 }, { "epoch": 14.97259311314125, "grad_norm": 0.280113160610199, "learning_rate": 2.3355821035371284e-05, "loss": 0.0149, "step": 21306 }, { "epoch": 14.973295853829937, "grad_norm": 0.07785142958164215, "learning_rate": 2.3355352541578825e-05, "loss": 0.0126, "step": 21307 }, { "epoch": 14.973998594518623, "grad_norm": 0.09521792829036713, "learning_rate": 2.335488404778637e-05, "loss": 0.0223, "step": 21308 }, { "epoch": 14.974701335207309, "grad_norm": 0.09719119220972061, "learning_rate": 2.335441555399391e-05, "loss": 0.006, "step": 21309 }, { "epoch": 14.975404075895995, "grad_norm": 0.09672295302152634, "learning_rate": 2.3353947060201453e-05, "loss": 0.0204, "step": 21310 }, { "epoch": 14.97610681658468, "grad_norm": 0.14041493833065033, "learning_rate": 2.3353478566408993e-05, "loss": 0.0215, "step": 21311 }, { "epoch": 14.976809557273366, "grad_norm": 0.2625330686569214, "learning_rate": 2.3353010072616537e-05, "loss": 0.0226, "step": 21312 }, { "epoch": 14.977512297962052, "grad_norm": 0.12415438145399094, "learning_rate": 2.335254157882408e-05, "loss": 0.0213, "step": 21313 }, { "epoch": 14.978215038650738, "grad_norm": 0.09674621373414993, "learning_rate": 2.3352073085031624e-05, "loss": 0.0119, "step": 21314 }, { "epoch": 14.978917779339424, "grad_norm": 0.5172224640846252, "learning_rate": 2.3351604591239168e-05, "loss": 0.0281, "step": 21315 }, { "epoch": 14.97962052002811, "grad_norm": 0.18537850677967072, "learning_rate": 2.3351136097446708e-05, "loss": 0.042, "step": 21316 }, { "epoch": 14.980323260716796, "grad_norm": 0.2455025017261505, "learning_rate": 2.3350667603654252e-05, "loss": 0.0479, "step": 21317 }, { "epoch": 14.981026001405482, "grad_norm": 0.2533080279827118, "learning_rate": 2.3350199109861796e-05, "loss": 0.057, "step": 21318 }, { "epoch": 14.981728742094168, "grad_norm": 0.3606742024421692, "learning_rate": 2.334973061606934e-05, "loss": 0.0944, "step": 21319 }, { "epoch": 14.982431482782854, "grad_norm": 0.517499566078186, "learning_rate": 2.334926212227688e-05, "loss": 0.1299, "step": 21320 }, { "epoch": 14.98313422347154, "grad_norm": 0.8153020739555359, "learning_rate": 2.3348793628484423e-05, "loss": 0.1466, "step": 21321 }, { "epoch": 14.983836964160226, "grad_norm": 2.955909252166748, "learning_rate": 2.3348325134691967e-05, "loss": 0.1464, "step": 21322 }, { "epoch": 14.984539704848912, "grad_norm": 0.2706073820590973, "learning_rate": 2.334785664089951e-05, "loss": 0.0598, "step": 21323 }, { "epoch": 14.985242445537597, "grad_norm": 0.19709594547748566, "learning_rate": 2.334738814710705e-05, "loss": 0.0234, "step": 21324 }, { "epoch": 14.985945186226282, "grad_norm": 0.07450412213802338, "learning_rate": 2.3346919653314595e-05, "loss": 0.0191, "step": 21325 }, { "epoch": 14.98664792691497, "grad_norm": 0.08589545637369156, "learning_rate": 2.3346451159522135e-05, "loss": 0.0178, "step": 21326 }, { "epoch": 14.987350667603653, "grad_norm": 0.12933945655822754, "learning_rate": 2.334598266572968e-05, "loss": 0.0142, "step": 21327 }, { "epoch": 14.98805340829234, "grad_norm": 0.12637610733509064, "learning_rate": 2.3345514171937223e-05, "loss": 0.0176, "step": 21328 }, { "epoch": 14.988756148981025, "grad_norm": 0.21057139337062836, "learning_rate": 2.3345045678144763e-05, "loss": 0.0154, "step": 21329 }, { "epoch": 14.989458889669711, "grad_norm": 0.04662243276834488, "learning_rate": 2.3344577184352307e-05, "loss": 0.0116, "step": 21330 }, { "epoch": 14.990161630358397, "grad_norm": 0.10764539986848831, "learning_rate": 2.334410869055985e-05, "loss": 0.0279, "step": 21331 }, { "epoch": 14.990864371047083, "grad_norm": 0.15580172836780548, "learning_rate": 2.3343640196767394e-05, "loss": 0.021, "step": 21332 }, { "epoch": 14.991567111735769, "grad_norm": 0.09377681463956833, "learning_rate": 2.3343171702974935e-05, "loss": 0.0151, "step": 21333 }, { "epoch": 14.992269852424455, "grad_norm": 0.1081911027431488, "learning_rate": 2.3342703209182478e-05, "loss": 0.0189, "step": 21334 }, { "epoch": 14.99297259311314, "grad_norm": 0.10240375250577927, "learning_rate": 2.3342234715390022e-05, "loss": 0.012, "step": 21335 }, { "epoch": 14.993675333801827, "grad_norm": 0.20249904692173004, "learning_rate": 2.3341766221597566e-05, "loss": 0.0323, "step": 21336 }, { "epoch": 14.994378074490513, "grad_norm": 0.12292983382940292, "learning_rate": 2.3341297727805106e-05, "loss": 0.0118, "step": 21337 }, { "epoch": 14.995080815179199, "grad_norm": 0.1173211857676506, "learning_rate": 2.334082923401265e-05, "loss": 0.0258, "step": 21338 }, { "epoch": 14.995783555867884, "grad_norm": 0.1145574152469635, "learning_rate": 2.3340360740220193e-05, "loss": 0.0288, "step": 21339 }, { "epoch": 14.99648629655657, "grad_norm": 0.1451703906059265, "learning_rate": 2.3339892246427737e-05, "loss": 0.0369, "step": 21340 }, { "epoch": 14.997189037245256, "grad_norm": 0.2745712399482727, "learning_rate": 2.333942375263528e-05, "loss": 0.0914, "step": 21341 }, { "epoch": 14.997891777933942, "grad_norm": 0.369119256734848, "learning_rate": 2.333895525884282e-05, "loss": 0.0963, "step": 21342 }, { "epoch": 14.998594518622628, "grad_norm": 0.5468865633010864, "learning_rate": 2.3338486765050365e-05, "loss": 0.1468, "step": 21343 }, { "epoch": 14.999297259311314, "grad_norm": 1.3729636669158936, "learning_rate": 2.3338018271257905e-05, "loss": 0.145, "step": 21344 }, { "epoch": 15.0, "grad_norm": 0.7078835964202881, "learning_rate": 2.333754977746545e-05, "loss": 0.0967, "step": 21345 }, { "epoch": 15.000702740688686, "grad_norm": 0.15166796743869781, "learning_rate": 2.333708128367299e-05, "loss": 0.052, "step": 21346 }, { "epoch": 15.001405481377372, "grad_norm": 0.13112609088420868, "learning_rate": 2.3336612789880533e-05, "loss": 0.0186, "step": 21347 }, { "epoch": 15.002108222066058, "grad_norm": 0.11010733246803284, "learning_rate": 2.3336144296088077e-05, "loss": 0.0126, "step": 21348 }, { "epoch": 15.002810962754744, "grad_norm": 0.15039335191249847, "learning_rate": 2.333567580229562e-05, "loss": 0.0165, "step": 21349 }, { "epoch": 15.00351370344343, "grad_norm": 0.08839618414640427, "learning_rate": 2.333520730850316e-05, "loss": 0.0191, "step": 21350 }, { "epoch": 15.004216444132116, "grad_norm": 0.19546255469322205, "learning_rate": 2.3334738814710705e-05, "loss": 0.0165, "step": 21351 }, { "epoch": 15.004919184820801, "grad_norm": 0.18803487718105316, "learning_rate": 2.333427032091825e-05, "loss": 0.0189, "step": 21352 }, { "epoch": 15.005621925509487, "grad_norm": 0.14011935889720917, "learning_rate": 2.3333801827125792e-05, "loss": 0.0144, "step": 21353 }, { "epoch": 15.006324666198173, "grad_norm": 0.14018836617469788, "learning_rate": 2.3333333333333336e-05, "loss": 0.0154, "step": 21354 }, { "epoch": 15.00702740688686, "grad_norm": 0.08457675576210022, "learning_rate": 2.3332864839540876e-05, "loss": 0.0105, "step": 21355 }, { "epoch": 15.007730147575545, "grad_norm": 0.13076360523700714, "learning_rate": 2.333239634574842e-05, "loss": 0.0299, "step": 21356 }, { "epoch": 15.008432888264231, "grad_norm": 0.09661714732646942, "learning_rate": 2.3331927851955964e-05, "loss": 0.0097, "step": 21357 }, { "epoch": 15.009135628952917, "grad_norm": 0.19590125977993011, "learning_rate": 2.3331459358163507e-05, "loss": 0.0232, "step": 21358 }, { "epoch": 15.009838369641603, "grad_norm": 0.18672466278076172, "learning_rate": 2.3330990864371048e-05, "loss": 0.0128, "step": 21359 }, { "epoch": 15.010541110330289, "grad_norm": 0.14945532381534576, "learning_rate": 2.333052237057859e-05, "loss": 0.0261, "step": 21360 }, { "epoch": 15.011243851018975, "grad_norm": 0.22535322606563568, "learning_rate": 2.3330053876786132e-05, "loss": 0.0415, "step": 21361 }, { "epoch": 15.01194659170766, "grad_norm": 0.19441373646259308, "learning_rate": 2.3329585382993675e-05, "loss": 0.0193, "step": 21362 }, { "epoch": 15.012649332396347, "grad_norm": 0.1575942188501358, "learning_rate": 2.3329116889201216e-05, "loss": 0.032, "step": 21363 }, { "epoch": 15.013352073085033, "grad_norm": 0.17281021177768707, "learning_rate": 2.332864839540876e-05, "loss": 0.0208, "step": 21364 }, { "epoch": 15.014054813773717, "grad_norm": 0.20601357519626617, "learning_rate": 2.3328179901616303e-05, "loss": 0.051, "step": 21365 }, { "epoch": 15.014757554462403, "grad_norm": 0.3470443785190582, "learning_rate": 2.3327711407823847e-05, "loss": 0.0538, "step": 21366 }, { "epoch": 15.015460295151088, "grad_norm": 0.5059208869934082, "learning_rate": 2.332724291403139e-05, "loss": 0.0861, "step": 21367 }, { "epoch": 15.016163035839774, "grad_norm": 0.34871307015419006, "learning_rate": 2.332677442023893e-05, "loss": 0.1069, "step": 21368 }, { "epoch": 15.01686577652846, "grad_norm": 0.4948224425315857, "learning_rate": 2.3326305926446475e-05, "loss": 0.1556, "step": 21369 }, { "epoch": 15.017568517217146, "grad_norm": 1.5629593133926392, "learning_rate": 2.332583743265402e-05, "loss": 0.1388, "step": 21370 }, { "epoch": 15.018271257905832, "grad_norm": 0.16666138172149658, "learning_rate": 2.3325368938861562e-05, "loss": 0.0578, "step": 21371 }, { "epoch": 15.018973998594518, "grad_norm": 0.09819424897432327, "learning_rate": 2.3324900445069103e-05, "loss": 0.0137, "step": 21372 }, { "epoch": 15.019676739283204, "grad_norm": 0.1430559903383255, "learning_rate": 2.3324431951276646e-05, "loss": 0.0178, "step": 21373 }, { "epoch": 15.02037947997189, "grad_norm": 0.12395504862070084, "learning_rate": 2.332396345748419e-05, "loss": 0.0194, "step": 21374 }, { "epoch": 15.021082220660576, "grad_norm": 0.08698045462369919, "learning_rate": 2.3323494963691734e-05, "loss": 0.0162, "step": 21375 }, { "epoch": 15.021784961349262, "grad_norm": 0.11473765969276428, "learning_rate": 2.3323026469899274e-05, "loss": 0.0084, "step": 21376 }, { "epoch": 15.022487702037948, "grad_norm": 0.08243764936923981, "learning_rate": 2.3322557976106818e-05, "loss": 0.0052, "step": 21377 }, { "epoch": 15.023190442726634, "grad_norm": 0.10806979238986969, "learning_rate": 2.332208948231436e-05, "loss": 0.02, "step": 21378 }, { "epoch": 15.02389318341532, "grad_norm": 0.13874685764312744, "learning_rate": 2.3321620988521902e-05, "loss": 0.0172, "step": 21379 }, { "epoch": 15.024595924104005, "grad_norm": 0.05944996699690819, "learning_rate": 2.3321152494729446e-05, "loss": 0.0058, "step": 21380 }, { "epoch": 15.025298664792691, "grad_norm": 0.1626148372888565, "learning_rate": 2.3320684000936986e-05, "loss": 0.0221, "step": 21381 }, { "epoch": 15.026001405481377, "grad_norm": 0.22879494726657867, "learning_rate": 2.332021550714453e-05, "loss": 0.018, "step": 21382 }, { "epoch": 15.026704146170063, "grad_norm": 0.23681865632534027, "learning_rate": 2.3319747013352073e-05, "loss": 0.0372, "step": 21383 }, { "epoch": 15.02740688685875, "grad_norm": 0.14035138487815857, "learning_rate": 2.3319278519559617e-05, "loss": 0.0122, "step": 21384 }, { "epoch": 15.028109627547435, "grad_norm": 0.2654002606868744, "learning_rate": 2.3318810025767157e-05, "loss": 0.0272, "step": 21385 }, { "epoch": 15.028812368236121, "grad_norm": 0.4843831956386566, "learning_rate": 2.33183415319747e-05, "loss": 0.0292, "step": 21386 }, { "epoch": 15.029515108924807, "grad_norm": 0.07801718264818192, "learning_rate": 2.3317873038182245e-05, "loss": 0.0121, "step": 21387 }, { "epoch": 15.030217849613493, "grad_norm": 0.2921644449234009, "learning_rate": 2.331740454438979e-05, "loss": 0.029, "step": 21388 }, { "epoch": 15.030920590302179, "grad_norm": 0.2160821259021759, "learning_rate": 2.331693605059733e-05, "loss": 0.0511, "step": 21389 }, { "epoch": 15.031623330990865, "grad_norm": 0.17102566361427307, "learning_rate": 2.3316467556804873e-05, "loss": 0.0337, "step": 21390 }, { "epoch": 15.03232607167955, "grad_norm": 0.4597422778606415, "learning_rate": 2.3315999063012416e-05, "loss": 0.0567, "step": 21391 }, { "epoch": 15.033028812368237, "grad_norm": 0.2588365375995636, "learning_rate": 2.331553056921996e-05, "loss": 0.0872, "step": 21392 }, { "epoch": 15.033731553056922, "grad_norm": 0.4592514634132385, "learning_rate": 2.3315062075427504e-05, "loss": 0.108, "step": 21393 }, { "epoch": 15.034434293745608, "grad_norm": 0.7431992292404175, "learning_rate": 2.3314593581635044e-05, "loss": 0.1137, "step": 21394 }, { "epoch": 15.035137034434294, "grad_norm": 3.4883618354797363, "learning_rate": 2.3314125087842588e-05, "loss": 0.1435, "step": 21395 }, { "epoch": 15.03583977512298, "grad_norm": 0.32274129986763, "learning_rate": 2.3313656594050128e-05, "loss": 0.0638, "step": 21396 }, { "epoch": 15.036542515811666, "grad_norm": 0.11711686104536057, "learning_rate": 2.3313188100257672e-05, "loss": 0.0223, "step": 21397 }, { "epoch": 15.037245256500352, "grad_norm": 0.0729406327009201, "learning_rate": 2.3312719606465212e-05, "loss": 0.009, "step": 21398 }, { "epoch": 15.037947997189038, "grad_norm": 0.13344696164131165, "learning_rate": 2.3312251112672756e-05, "loss": 0.016, "step": 21399 }, { "epoch": 15.038650737877724, "grad_norm": 0.16641081869602203, "learning_rate": 2.33117826188803e-05, "loss": 0.0177, "step": 21400 }, { "epoch": 15.03935347856641, "grad_norm": 0.0978194996714592, "learning_rate": 2.3311314125087843e-05, "loss": 0.0116, "step": 21401 }, { "epoch": 15.040056219255096, "grad_norm": 0.12238694727420807, "learning_rate": 2.3310845631295384e-05, "loss": 0.0115, "step": 21402 }, { "epoch": 15.04075895994378, "grad_norm": 0.29003581404685974, "learning_rate": 2.3310377137502928e-05, "loss": 0.017, "step": 21403 }, { "epoch": 15.041461700632466, "grad_norm": 0.15961426496505737, "learning_rate": 2.330990864371047e-05, "loss": 0.0202, "step": 21404 }, { "epoch": 15.042164441321152, "grad_norm": 0.13834892213344574, "learning_rate": 2.3309440149918015e-05, "loss": 0.0112, "step": 21405 }, { "epoch": 15.042867182009838, "grad_norm": 0.10107987374067307, "learning_rate": 2.330897165612556e-05, "loss": 0.0124, "step": 21406 }, { "epoch": 15.043569922698524, "grad_norm": 0.12156203389167786, "learning_rate": 2.33085031623331e-05, "loss": 0.0088, "step": 21407 }, { "epoch": 15.04427266338721, "grad_norm": 0.10431396961212158, "learning_rate": 2.3308034668540643e-05, "loss": 0.019, "step": 21408 }, { "epoch": 15.044975404075895, "grad_norm": 0.11279425770044327, "learning_rate": 2.3307566174748186e-05, "loss": 0.0095, "step": 21409 }, { "epoch": 15.045678144764581, "grad_norm": 0.11470678448677063, "learning_rate": 2.330709768095573e-05, "loss": 0.0248, "step": 21410 }, { "epoch": 15.046380885453267, "grad_norm": 0.1932850182056427, "learning_rate": 2.330662918716327e-05, "loss": 0.0403, "step": 21411 }, { "epoch": 15.047083626141953, "grad_norm": 0.3505580723285675, "learning_rate": 2.3306160693370814e-05, "loss": 0.0126, "step": 21412 }, { "epoch": 15.047786366830639, "grad_norm": 0.15583984553813934, "learning_rate": 2.3305692199578358e-05, "loss": 0.041, "step": 21413 }, { "epoch": 15.048489107519325, "grad_norm": 0.15890412032604218, "learning_rate": 2.33052237057859e-05, "loss": 0.0247, "step": 21414 }, { "epoch": 15.049191848208011, "grad_norm": 0.38312625885009766, "learning_rate": 2.3304755211993442e-05, "loss": 0.0352, "step": 21415 }, { "epoch": 15.049894588896697, "grad_norm": 0.39897117018699646, "learning_rate": 2.3304286718200982e-05, "loss": 0.0723, "step": 21416 }, { "epoch": 15.050597329585383, "grad_norm": 0.9401147365570068, "learning_rate": 2.3303818224408526e-05, "loss": 0.0719, "step": 21417 }, { "epoch": 15.051300070274069, "grad_norm": 0.7511645555496216, "learning_rate": 2.330334973061607e-05, "loss": 0.1076, "step": 21418 }, { "epoch": 15.052002810962755, "grad_norm": 0.8409445881843567, "learning_rate": 2.3302881236823614e-05, "loss": 0.1505, "step": 21419 }, { "epoch": 15.05270555165144, "grad_norm": 0.7458555698394775, "learning_rate": 2.3302412743031154e-05, "loss": 0.1555, "step": 21420 }, { "epoch": 15.053408292340126, "grad_norm": 0.7637832164764404, "learning_rate": 2.3301944249238698e-05, "loss": 0.0501, "step": 21421 }, { "epoch": 15.054111033028812, "grad_norm": 0.13730329275131226, "learning_rate": 2.330147575544624e-05, "loss": 0.0274, "step": 21422 }, { "epoch": 15.054813773717498, "grad_norm": 0.2807796001434326, "learning_rate": 2.3301007261653785e-05, "loss": 0.0138, "step": 21423 }, { "epoch": 15.055516514406184, "grad_norm": 0.08113466203212738, "learning_rate": 2.3300538767861325e-05, "loss": 0.0153, "step": 21424 }, { "epoch": 15.05621925509487, "grad_norm": 0.091338150203228, "learning_rate": 2.330007027406887e-05, "loss": 0.0187, "step": 21425 }, { "epoch": 15.056921995783556, "grad_norm": 0.07952583581209183, "learning_rate": 2.3299601780276413e-05, "loss": 0.0097, "step": 21426 }, { "epoch": 15.057624736472242, "grad_norm": 0.11088517308235168, "learning_rate": 2.3299133286483957e-05, "loss": 0.0131, "step": 21427 }, { "epoch": 15.058327477160928, "grad_norm": 0.20074106752872467, "learning_rate": 2.32986647926915e-05, "loss": 0.0154, "step": 21428 }, { "epoch": 15.059030217849614, "grad_norm": 0.22237232327461243, "learning_rate": 2.329819629889904e-05, "loss": 0.0281, "step": 21429 }, { "epoch": 15.0597329585383, "grad_norm": 0.09689104557037354, "learning_rate": 2.3297727805106584e-05, "loss": 0.0091, "step": 21430 }, { "epoch": 15.060435699226986, "grad_norm": 0.12896430492401123, "learning_rate": 2.3297259311314125e-05, "loss": 0.027, "step": 21431 }, { "epoch": 15.061138439915672, "grad_norm": 0.07867872714996338, "learning_rate": 2.329679081752167e-05, "loss": 0.0135, "step": 21432 }, { "epoch": 15.061841180604358, "grad_norm": 0.39933788776397705, "learning_rate": 2.329632232372921e-05, "loss": 0.0303, "step": 21433 }, { "epoch": 15.062543921293043, "grad_norm": 0.05474699288606644, "learning_rate": 2.3295853829936753e-05, "loss": 0.0098, "step": 21434 }, { "epoch": 15.06324666198173, "grad_norm": 0.31956642866134644, "learning_rate": 2.3295385336144296e-05, "loss": 0.0178, "step": 21435 }, { "epoch": 15.063949402670415, "grad_norm": 0.15361586213111877, "learning_rate": 2.329491684235184e-05, "loss": 0.0367, "step": 21436 }, { "epoch": 15.064652143359101, "grad_norm": 0.17898225784301758, "learning_rate": 2.329444834855938e-05, "loss": 0.0354, "step": 21437 }, { "epoch": 15.065354884047787, "grad_norm": 0.3848382234573364, "learning_rate": 2.3293979854766924e-05, "loss": 0.0267, "step": 21438 }, { "epoch": 15.066057624736473, "grad_norm": 0.32911840081214905, "learning_rate": 2.3293511360974468e-05, "loss": 0.0428, "step": 21439 }, { "epoch": 15.066760365425159, "grad_norm": 0.5581242442131042, "learning_rate": 2.329304286718201e-05, "loss": 0.0391, "step": 21440 }, { "epoch": 15.067463106113845, "grad_norm": 0.32216110825538635, "learning_rate": 2.3292574373389555e-05, "loss": 0.0589, "step": 21441 }, { "epoch": 15.068165846802529, "grad_norm": 0.3984926640987396, "learning_rate": 2.3292105879597096e-05, "loss": 0.0861, "step": 21442 }, { "epoch": 15.068868587491215, "grad_norm": 0.41239699721336365, "learning_rate": 2.329163738580464e-05, "loss": 0.1147, "step": 21443 }, { "epoch": 15.0695713281799, "grad_norm": 0.6747090220451355, "learning_rate": 2.3291168892012183e-05, "loss": 0.1323, "step": 21444 }, { "epoch": 15.070274068868587, "grad_norm": 1.1165107488632202, "learning_rate": 2.3290700398219727e-05, "loss": 0.1943, "step": 21445 }, { "epoch": 15.070976809557273, "grad_norm": 0.3865831792354584, "learning_rate": 2.3290231904427267e-05, "loss": 0.0696, "step": 21446 }, { "epoch": 15.071679550245959, "grad_norm": 0.16389185190200806, "learning_rate": 2.328976341063481e-05, "loss": 0.0215, "step": 21447 }, { "epoch": 15.072382290934645, "grad_norm": 0.09378615766763687, "learning_rate": 2.328929491684235e-05, "loss": 0.0176, "step": 21448 }, { "epoch": 15.07308503162333, "grad_norm": 0.1497759073972702, "learning_rate": 2.3288826423049895e-05, "loss": 0.0123, "step": 21449 }, { "epoch": 15.073787772312016, "grad_norm": 0.09253119677305222, "learning_rate": 2.3288357929257435e-05, "loss": 0.0133, "step": 21450 }, { "epoch": 15.074490513000702, "grad_norm": 0.056711964309215546, "learning_rate": 2.328788943546498e-05, "loss": 0.006, "step": 21451 }, { "epoch": 15.075193253689388, "grad_norm": 0.08459241688251495, "learning_rate": 2.3287420941672523e-05, "loss": 0.0119, "step": 21452 }, { "epoch": 15.075895994378074, "grad_norm": 0.12373089790344238, "learning_rate": 2.3286952447880066e-05, "loss": 0.0178, "step": 21453 }, { "epoch": 15.07659873506676, "grad_norm": 0.13752150535583496, "learning_rate": 2.328648395408761e-05, "loss": 0.0182, "step": 21454 }, { "epoch": 15.077301475755446, "grad_norm": 0.1000804677605629, "learning_rate": 2.328601546029515e-05, "loss": 0.0111, "step": 21455 }, { "epoch": 15.078004216444132, "grad_norm": 0.07762254774570465, "learning_rate": 2.3285546966502694e-05, "loss": 0.011, "step": 21456 }, { "epoch": 15.078706957132818, "grad_norm": 0.09472926706075668, "learning_rate": 2.3285078472710238e-05, "loss": 0.0099, "step": 21457 }, { "epoch": 15.079409697821504, "grad_norm": 0.09841303527355194, "learning_rate": 2.328460997891778e-05, "loss": 0.0183, "step": 21458 }, { "epoch": 15.08011243851019, "grad_norm": 0.0818975493311882, "learning_rate": 2.3284141485125322e-05, "loss": 0.0112, "step": 21459 }, { "epoch": 15.080815179198876, "grad_norm": 0.24776853621006012, "learning_rate": 2.3283672991332866e-05, "loss": 0.0441, "step": 21460 }, { "epoch": 15.081517919887562, "grad_norm": 0.20144544541835785, "learning_rate": 2.328320449754041e-05, "loss": 0.0223, "step": 21461 }, { "epoch": 15.082220660576247, "grad_norm": 0.11976409703493118, "learning_rate": 2.3282736003747953e-05, "loss": 0.023, "step": 21462 }, { "epoch": 15.082923401264933, "grad_norm": 0.24033500254154205, "learning_rate": 2.3282267509955493e-05, "loss": 0.0303, "step": 21463 }, { "epoch": 15.08362614195362, "grad_norm": 0.1386251002550125, "learning_rate": 2.3281799016163037e-05, "loss": 0.0298, "step": 21464 }, { "epoch": 15.084328882642305, "grad_norm": 0.27976375818252563, "learning_rate": 2.328133052237058e-05, "loss": 0.0449, "step": 21465 }, { "epoch": 15.085031623330991, "grad_norm": 0.20138852298259735, "learning_rate": 2.328086202857812e-05, "loss": 0.0611, "step": 21466 }, { "epoch": 15.085734364019677, "grad_norm": 0.7203060388565063, "learning_rate": 2.3280393534785665e-05, "loss": 0.0923, "step": 21467 }, { "epoch": 15.086437104708363, "grad_norm": 0.3743031620979309, "learning_rate": 2.3279925040993205e-05, "loss": 0.1162, "step": 21468 }, { "epoch": 15.087139845397049, "grad_norm": 0.5660213232040405, "learning_rate": 2.327945654720075e-05, "loss": 0.1468, "step": 21469 }, { "epoch": 15.087842586085735, "grad_norm": 0.815915048122406, "learning_rate": 2.3278988053408293e-05, "loss": 0.129, "step": 21470 }, { "epoch": 15.08854532677442, "grad_norm": 0.8404501080513, "learning_rate": 2.3278519559615836e-05, "loss": 0.057, "step": 21471 }, { "epoch": 15.089248067463107, "grad_norm": 0.11834297329187393, "learning_rate": 2.3278051065823377e-05, "loss": 0.0155, "step": 21472 }, { "epoch": 15.089950808151793, "grad_norm": 0.172062486410141, "learning_rate": 2.327758257203092e-05, "loss": 0.0204, "step": 21473 }, { "epoch": 15.090653548840478, "grad_norm": 0.1066298559308052, "learning_rate": 2.3277114078238464e-05, "loss": 0.0233, "step": 21474 }, { "epoch": 15.091356289529164, "grad_norm": 0.0754081979393959, "learning_rate": 2.3276645584446008e-05, "loss": 0.0083, "step": 21475 }, { "epoch": 15.09205903021785, "grad_norm": 0.13239847123622894, "learning_rate": 2.327617709065355e-05, "loss": 0.0136, "step": 21476 }, { "epoch": 15.092761770906536, "grad_norm": 0.08864174038171768, "learning_rate": 2.3275708596861092e-05, "loss": 0.0117, "step": 21477 }, { "epoch": 15.093464511595222, "grad_norm": 0.08746007829904556, "learning_rate": 2.3275240103068636e-05, "loss": 0.0153, "step": 21478 }, { "epoch": 15.094167252283908, "grad_norm": 0.11294658482074738, "learning_rate": 2.327477160927618e-05, "loss": 0.0184, "step": 21479 }, { "epoch": 15.094869992972592, "grad_norm": 0.07291919738054276, "learning_rate": 2.3274303115483723e-05, "loss": 0.0103, "step": 21480 }, { "epoch": 15.095572733661278, "grad_norm": 0.08463703840970993, "learning_rate": 2.3273834621691264e-05, "loss": 0.0123, "step": 21481 }, { "epoch": 15.096275474349964, "grad_norm": 0.0991198942065239, "learning_rate": 2.3273366127898807e-05, "loss": 0.0123, "step": 21482 }, { "epoch": 15.09697821503865, "grad_norm": 0.1256425678730011, "learning_rate": 2.3272897634106348e-05, "loss": 0.0229, "step": 21483 }, { "epoch": 15.097680955727336, "grad_norm": 0.17452116310596466, "learning_rate": 2.327242914031389e-05, "loss": 0.0257, "step": 21484 }, { "epoch": 15.098383696416022, "grad_norm": 0.17911387979984283, "learning_rate": 2.327196064652143e-05, "loss": 0.0129, "step": 21485 }, { "epoch": 15.099086437104708, "grad_norm": 0.42399969696998596, "learning_rate": 2.3271492152728975e-05, "loss": 0.0304, "step": 21486 }, { "epoch": 15.099789177793394, "grad_norm": 0.6263750195503235, "learning_rate": 2.327102365893652e-05, "loss": 0.0208, "step": 21487 }, { "epoch": 15.10049191848208, "grad_norm": 0.15291152894496918, "learning_rate": 2.3270555165144063e-05, "loss": 0.0362, "step": 21488 }, { "epoch": 15.101194659170766, "grad_norm": 0.17511257529258728, "learning_rate": 2.3270086671351603e-05, "loss": 0.0299, "step": 21489 }, { "epoch": 15.101897399859451, "grad_norm": 0.2936996519565582, "learning_rate": 2.3269618177559147e-05, "loss": 0.0309, "step": 21490 }, { "epoch": 15.102600140548137, "grad_norm": 0.2754088044166565, "learning_rate": 2.326914968376669e-05, "loss": 0.0634, "step": 21491 }, { "epoch": 15.103302881236823, "grad_norm": 0.2650032341480255, "learning_rate": 2.3268681189974234e-05, "loss": 0.0669, "step": 21492 }, { "epoch": 15.10400562192551, "grad_norm": 0.9718489646911621, "learning_rate": 2.3268212696181778e-05, "loss": 0.1025, "step": 21493 }, { "epoch": 15.104708362614195, "grad_norm": 0.6888934969902039, "learning_rate": 2.326774420238932e-05, "loss": 0.1396, "step": 21494 }, { "epoch": 15.105411103302881, "grad_norm": 1.6626263856887817, "learning_rate": 2.3267275708596862e-05, "loss": 0.1261, "step": 21495 }, { "epoch": 15.106113843991567, "grad_norm": 0.40347790718078613, "learning_rate": 2.3266807214804406e-05, "loss": 0.0654, "step": 21496 }, { "epoch": 15.106816584680253, "grad_norm": 0.40710437297821045, "learning_rate": 2.326633872101195e-05, "loss": 0.0273, "step": 21497 }, { "epoch": 15.107519325368939, "grad_norm": 0.09888014942407608, "learning_rate": 2.326587022721949e-05, "loss": 0.0174, "step": 21498 }, { "epoch": 15.108222066057625, "grad_norm": 0.10706045478582382, "learning_rate": 2.3265401733427034e-05, "loss": 0.0119, "step": 21499 }, { "epoch": 15.10892480674631, "grad_norm": 0.19103100895881653, "learning_rate": 2.3264933239634577e-05, "loss": 0.0169, "step": 21500 }, { "epoch": 15.109627547434997, "grad_norm": 0.13055525720119476, "learning_rate": 2.3264464745842118e-05, "loss": 0.0067, "step": 21501 }, { "epoch": 15.110330288123683, "grad_norm": 0.13193637132644653, "learning_rate": 2.3263996252049658e-05, "loss": 0.0143, "step": 21502 }, { "epoch": 15.111033028812368, "grad_norm": 0.129832923412323, "learning_rate": 2.3263527758257202e-05, "loss": 0.0103, "step": 21503 }, { "epoch": 15.111735769501054, "grad_norm": 0.1555197685956955, "learning_rate": 2.3263059264464746e-05, "loss": 0.0175, "step": 21504 }, { "epoch": 15.11243851018974, "grad_norm": 0.16370657086372375, "learning_rate": 2.326259077067229e-05, "loss": 0.0118, "step": 21505 }, { "epoch": 15.113141250878426, "grad_norm": 0.12395352870225906, "learning_rate": 2.3262122276879833e-05, "loss": 0.0294, "step": 21506 }, { "epoch": 15.113843991567112, "grad_norm": 0.04780793562531471, "learning_rate": 2.3261653783087373e-05, "loss": 0.0035, "step": 21507 }, { "epoch": 15.114546732255798, "grad_norm": 0.1119619607925415, "learning_rate": 2.3261185289294917e-05, "loss": 0.0215, "step": 21508 }, { "epoch": 15.115249472944484, "grad_norm": 0.11849185824394226, "learning_rate": 2.326071679550246e-05, "loss": 0.0148, "step": 21509 }, { "epoch": 15.11595221363317, "grad_norm": 0.16335169970989227, "learning_rate": 2.3260248301710004e-05, "loss": 0.0313, "step": 21510 }, { "epoch": 15.116654954321856, "grad_norm": 0.1640399694442749, "learning_rate": 2.3259779807917545e-05, "loss": 0.0215, "step": 21511 }, { "epoch": 15.117357695010542, "grad_norm": 0.09636647999286652, "learning_rate": 2.325931131412509e-05, "loss": 0.0111, "step": 21512 }, { "epoch": 15.118060435699228, "grad_norm": 0.20282751321792603, "learning_rate": 2.3258842820332632e-05, "loss": 0.0269, "step": 21513 }, { "epoch": 15.118763176387914, "grad_norm": 0.15278440713882446, "learning_rate": 2.3258374326540176e-05, "loss": 0.029, "step": 21514 }, { "epoch": 15.1194659170766, "grad_norm": 0.30187249183654785, "learning_rate": 2.3257905832747716e-05, "loss": 0.0426, "step": 21515 }, { "epoch": 15.120168657765285, "grad_norm": 0.19861505925655365, "learning_rate": 2.325743733895526e-05, "loss": 0.0396, "step": 21516 }, { "epoch": 15.120871398453971, "grad_norm": 0.41197669506073, "learning_rate": 2.3256968845162804e-05, "loss": 0.0938, "step": 21517 }, { "epoch": 15.121574139142655, "grad_norm": 0.43255814909935, "learning_rate": 2.3256500351370344e-05, "loss": 0.1154, "step": 21518 }, { "epoch": 15.122276879831341, "grad_norm": 1.095147728919983, "learning_rate": 2.3256031857577888e-05, "loss": 0.1492, "step": 21519 }, { "epoch": 15.122979620520027, "grad_norm": 1.269806146621704, "learning_rate": 2.3255563363785428e-05, "loss": 0.1587, "step": 21520 }, { "epoch": 15.123682361208713, "grad_norm": 0.30052581429481506, "learning_rate": 2.3255094869992972e-05, "loss": 0.0592, "step": 21521 }, { "epoch": 15.1243851018974, "grad_norm": 0.08366592228412628, "learning_rate": 2.3254626376200516e-05, "loss": 0.0208, "step": 21522 }, { "epoch": 15.125087842586085, "grad_norm": 0.24969689548015594, "learning_rate": 2.325415788240806e-05, "loss": 0.0205, "step": 21523 }, { "epoch": 15.125790583274771, "grad_norm": 0.163262277841568, "learning_rate": 2.32536893886156e-05, "loss": 0.0169, "step": 21524 }, { "epoch": 15.126493323963457, "grad_norm": 0.09602135419845581, "learning_rate": 2.3253220894823143e-05, "loss": 0.0191, "step": 21525 }, { "epoch": 15.127196064652143, "grad_norm": 0.030786456540226936, "learning_rate": 2.3252752401030687e-05, "loss": 0.0038, "step": 21526 }, { "epoch": 15.127898805340829, "grad_norm": 0.09351031482219696, "learning_rate": 2.325228390723823e-05, "loss": 0.0172, "step": 21527 }, { "epoch": 15.128601546029515, "grad_norm": 0.12539096176624298, "learning_rate": 2.325181541344577e-05, "loss": 0.0196, "step": 21528 }, { "epoch": 15.1293042867182, "grad_norm": 0.1758139580488205, "learning_rate": 2.3251346919653315e-05, "loss": 0.0205, "step": 21529 }, { "epoch": 15.130007027406887, "grad_norm": 0.10765409469604492, "learning_rate": 2.325087842586086e-05, "loss": 0.0104, "step": 21530 }, { "epoch": 15.130709768095572, "grad_norm": 0.12094926834106445, "learning_rate": 2.3250409932068402e-05, "loss": 0.0171, "step": 21531 }, { "epoch": 15.131412508784258, "grad_norm": 0.07725992053747177, "learning_rate": 2.3249941438275946e-05, "loss": 0.0095, "step": 21532 }, { "epoch": 15.132115249472944, "grad_norm": 0.16105611622333527, "learning_rate": 2.3249472944483486e-05, "loss": 0.0417, "step": 21533 }, { "epoch": 15.13281799016163, "grad_norm": 0.1831364929676056, "learning_rate": 2.324900445069103e-05, "loss": 0.0099, "step": 21534 }, { "epoch": 15.133520730850316, "grad_norm": 0.1299780011177063, "learning_rate": 2.324853595689857e-05, "loss": 0.0314, "step": 21535 }, { "epoch": 15.134223471539002, "grad_norm": 0.16461409628391266, "learning_rate": 2.3248067463106114e-05, "loss": 0.0229, "step": 21536 }, { "epoch": 15.134926212227688, "grad_norm": 0.13501587510108948, "learning_rate": 2.3247598969313655e-05, "loss": 0.0149, "step": 21537 }, { "epoch": 15.135628952916374, "grad_norm": 0.4618009924888611, "learning_rate": 2.3247130475521198e-05, "loss": 0.0288, "step": 21538 }, { "epoch": 15.13633169360506, "grad_norm": 0.17414191365242004, "learning_rate": 2.3246661981728742e-05, "loss": 0.026, "step": 21539 }, { "epoch": 15.137034434293746, "grad_norm": 0.1726926565170288, "learning_rate": 2.3246193487936286e-05, "loss": 0.0314, "step": 21540 }, { "epoch": 15.137737174982432, "grad_norm": 0.24039261043071747, "learning_rate": 2.3245724994143826e-05, "loss": 0.0552, "step": 21541 }, { "epoch": 15.138439915671118, "grad_norm": 0.5757699608802795, "learning_rate": 2.324525650035137e-05, "loss": 0.0672, "step": 21542 }, { "epoch": 15.139142656359803, "grad_norm": 0.5431905388832092, "learning_rate": 2.3244788006558914e-05, "loss": 0.1552, "step": 21543 }, { "epoch": 15.13984539704849, "grad_norm": 0.8465901613235474, "learning_rate": 2.3244319512766457e-05, "loss": 0.1432, "step": 21544 }, { "epoch": 15.140548137737175, "grad_norm": 1.1333945989608765, "learning_rate": 2.3243851018974e-05, "loss": 0.184, "step": 21545 }, { "epoch": 15.141250878425861, "grad_norm": 0.369538277387619, "learning_rate": 2.324338252518154e-05, "loss": 0.0664, "step": 21546 }, { "epoch": 15.141953619114547, "grad_norm": 0.8605588674545288, "learning_rate": 2.3242914031389085e-05, "loss": 0.0245, "step": 21547 }, { "epoch": 15.142656359803233, "grad_norm": 0.1114891991019249, "learning_rate": 2.324244553759663e-05, "loss": 0.019, "step": 21548 }, { "epoch": 15.143359100491919, "grad_norm": 0.10555834323167801, "learning_rate": 2.3241977043804172e-05, "loss": 0.0134, "step": 21549 }, { "epoch": 15.144061841180605, "grad_norm": 0.09834729135036469, "learning_rate": 2.3241508550011713e-05, "loss": 0.0155, "step": 21550 }, { "epoch": 15.14476458186929, "grad_norm": 0.09473413974046707, "learning_rate": 2.3241040056219257e-05, "loss": 0.0078, "step": 21551 }, { "epoch": 15.145467322557977, "grad_norm": 0.09712626039981842, "learning_rate": 2.32405715624268e-05, "loss": 0.0112, "step": 21552 }, { "epoch": 15.146170063246663, "grad_norm": 0.13294638693332672, "learning_rate": 2.324010306863434e-05, "loss": 0.0193, "step": 21553 }, { "epoch": 15.146872803935349, "grad_norm": 0.13041922450065613, "learning_rate": 2.323963457484188e-05, "loss": 0.0145, "step": 21554 }, { "epoch": 15.147575544624035, "grad_norm": 0.20395272970199585, "learning_rate": 2.3239166081049425e-05, "loss": 0.0083, "step": 21555 }, { "epoch": 15.14827828531272, "grad_norm": 0.11278433352708817, "learning_rate": 2.323869758725697e-05, "loss": 0.024, "step": 21556 }, { "epoch": 15.148981026001405, "grad_norm": 0.07525956630706787, "learning_rate": 2.3238229093464512e-05, "loss": 0.0083, "step": 21557 }, { "epoch": 15.14968376669009, "grad_norm": 0.31611818075180054, "learning_rate": 2.3237760599672056e-05, "loss": 0.0231, "step": 21558 }, { "epoch": 15.150386507378776, "grad_norm": 0.20117537677288055, "learning_rate": 2.3237292105879596e-05, "loss": 0.0108, "step": 21559 }, { "epoch": 15.151089248067462, "grad_norm": 0.17169468104839325, "learning_rate": 2.323682361208714e-05, "loss": 0.0236, "step": 21560 }, { "epoch": 15.151791988756148, "grad_norm": 0.2352253645658493, "learning_rate": 2.3236355118294684e-05, "loss": 0.0262, "step": 21561 }, { "epoch": 15.152494729444834, "grad_norm": 0.13565514981746674, "learning_rate": 2.3235886624502227e-05, "loss": 0.0165, "step": 21562 }, { "epoch": 15.15319747013352, "grad_norm": 0.14066866040229797, "learning_rate": 2.3235418130709768e-05, "loss": 0.018, "step": 21563 }, { "epoch": 15.153900210822206, "grad_norm": 0.21165908873081207, "learning_rate": 2.323494963691731e-05, "loss": 0.0433, "step": 21564 }, { "epoch": 15.154602951510892, "grad_norm": 0.3088429868221283, "learning_rate": 2.3234481143124855e-05, "loss": 0.0373, "step": 21565 }, { "epoch": 15.155305692199578, "grad_norm": 0.3620601296424866, "learning_rate": 2.32340126493324e-05, "loss": 0.0611, "step": 21566 }, { "epoch": 15.156008432888264, "grad_norm": 0.6555548310279846, "learning_rate": 2.323354415553994e-05, "loss": 0.0969, "step": 21567 }, { "epoch": 15.15671117357695, "grad_norm": 0.6008857488632202, "learning_rate": 2.3233075661747483e-05, "loss": 0.1069, "step": 21568 }, { "epoch": 15.157413914265636, "grad_norm": 1.9007482528686523, "learning_rate": 2.3232607167955027e-05, "loss": 0.1425, "step": 21569 }, { "epoch": 15.158116654954322, "grad_norm": 1.003309965133667, "learning_rate": 2.3232138674162567e-05, "loss": 0.1757, "step": 21570 }, { "epoch": 15.158819395643008, "grad_norm": 0.28109049797058105, "learning_rate": 2.323167018037011e-05, "loss": 0.0482, "step": 21571 }, { "epoch": 15.159522136331693, "grad_norm": 0.10293721407651901, "learning_rate": 2.323120168657765e-05, "loss": 0.018, "step": 21572 }, { "epoch": 15.16022487702038, "grad_norm": 0.17658653855323792, "learning_rate": 2.3230733192785195e-05, "loss": 0.0176, "step": 21573 }, { "epoch": 15.160927617709065, "grad_norm": 0.12768509984016418, "learning_rate": 2.323026469899274e-05, "loss": 0.0184, "step": 21574 }, { "epoch": 15.161630358397751, "grad_norm": 0.10864760726690292, "learning_rate": 2.3229796205200282e-05, "loss": 0.0167, "step": 21575 }, { "epoch": 15.162333099086437, "grad_norm": 0.0871705561876297, "learning_rate": 2.3229327711407823e-05, "loss": 0.0115, "step": 21576 }, { "epoch": 15.163035839775123, "grad_norm": 0.3005056381225586, "learning_rate": 2.3228859217615366e-05, "loss": 0.0115, "step": 21577 }, { "epoch": 15.163738580463809, "grad_norm": 0.22051027417182922, "learning_rate": 2.322839072382291e-05, "loss": 0.0139, "step": 21578 }, { "epoch": 15.164441321152495, "grad_norm": 0.09991585463285446, "learning_rate": 2.3227922230030454e-05, "loss": 0.0132, "step": 21579 }, { "epoch": 15.16514406184118, "grad_norm": 0.12712591886520386, "learning_rate": 2.3227453736237994e-05, "loss": 0.0286, "step": 21580 }, { "epoch": 15.165846802529867, "grad_norm": 0.10928316414356232, "learning_rate": 2.3226985242445538e-05, "loss": 0.0117, "step": 21581 }, { "epoch": 15.166549543218553, "grad_norm": 0.15827211737632751, "learning_rate": 2.322651674865308e-05, "loss": 0.0151, "step": 21582 }, { "epoch": 15.167252283907239, "grad_norm": 0.17446018755435944, "learning_rate": 2.3226048254860625e-05, "loss": 0.0198, "step": 21583 }, { "epoch": 15.167955024595924, "grad_norm": 0.1508466750383377, "learning_rate": 2.322557976106817e-05, "loss": 0.0145, "step": 21584 }, { "epoch": 15.16865776528461, "grad_norm": 0.35076820850372314, "learning_rate": 2.322511126727571e-05, "loss": 0.0308, "step": 21585 }, { "epoch": 15.169360505973296, "grad_norm": 0.13948176801204681, "learning_rate": 2.3224642773483253e-05, "loss": 0.0216, "step": 21586 }, { "epoch": 15.170063246661982, "grad_norm": 0.1065329983830452, "learning_rate": 2.3224174279690797e-05, "loss": 0.0165, "step": 21587 }, { "epoch": 15.170765987350668, "grad_norm": 0.15542830526828766, "learning_rate": 2.3223705785898337e-05, "loss": 0.0262, "step": 21588 }, { "epoch": 15.171468728039354, "grad_norm": 0.2021861970424652, "learning_rate": 2.3223237292105877e-05, "loss": 0.0275, "step": 21589 }, { "epoch": 15.17217146872804, "grad_norm": 0.223501056432724, "learning_rate": 2.322276879831342e-05, "loss": 0.0709, "step": 21590 }, { "epoch": 15.172874209416726, "grad_norm": 0.6481139659881592, "learning_rate": 2.3222300304520965e-05, "loss": 0.0571, "step": 21591 }, { "epoch": 15.173576950105412, "grad_norm": 0.45563140511512756, "learning_rate": 2.322183181072851e-05, "loss": 0.0888, "step": 21592 }, { "epoch": 15.174279690794098, "grad_norm": 1.900197982788086, "learning_rate": 2.322136331693605e-05, "loss": 0.1192, "step": 21593 }, { "epoch": 15.174982431482784, "grad_norm": 0.6614814400672913, "learning_rate": 2.3220894823143593e-05, "loss": 0.165, "step": 21594 }, { "epoch": 15.17568517217147, "grad_norm": 1.2416930198669434, "learning_rate": 2.3220426329351136e-05, "loss": 0.1614, "step": 21595 }, { "epoch": 15.176387912860154, "grad_norm": 0.1982998549938202, "learning_rate": 2.321995783555868e-05, "loss": 0.0575, "step": 21596 }, { "epoch": 15.17709065354884, "grad_norm": 0.2219424843788147, "learning_rate": 2.3219489341766224e-05, "loss": 0.0157, "step": 21597 }, { "epoch": 15.177793394237526, "grad_norm": 0.17364443838596344, "learning_rate": 2.3219020847973764e-05, "loss": 0.0224, "step": 21598 }, { "epoch": 15.178496134926212, "grad_norm": 0.11486875265836716, "learning_rate": 2.3218552354181308e-05, "loss": 0.0104, "step": 21599 }, { "epoch": 15.179198875614897, "grad_norm": 0.15627197921276093, "learning_rate": 2.321808386038885e-05, "loss": 0.0228, "step": 21600 }, { "epoch": 15.179901616303583, "grad_norm": 0.07724685966968536, "learning_rate": 2.3217615366596395e-05, "loss": 0.011, "step": 21601 }, { "epoch": 15.18060435699227, "grad_norm": 0.309604287147522, "learning_rate": 2.3217146872803936e-05, "loss": 0.0135, "step": 21602 }, { "epoch": 15.181307097680955, "grad_norm": 0.08446294069290161, "learning_rate": 2.321667837901148e-05, "loss": 0.0144, "step": 21603 }, { "epoch": 15.182009838369641, "grad_norm": 0.12358222901821136, "learning_rate": 2.3216209885219023e-05, "loss": 0.0188, "step": 21604 }, { "epoch": 15.182712579058327, "grad_norm": 0.1968967467546463, "learning_rate": 2.3215741391426563e-05, "loss": 0.0198, "step": 21605 }, { "epoch": 15.183415319747013, "grad_norm": 0.21313712000846863, "learning_rate": 2.3215272897634107e-05, "loss": 0.0221, "step": 21606 }, { "epoch": 15.184118060435699, "grad_norm": 0.19770684838294983, "learning_rate": 2.3214804403841648e-05, "loss": 0.0145, "step": 21607 }, { "epoch": 15.184820801124385, "grad_norm": 0.27363452315330505, "learning_rate": 2.321433591004919e-05, "loss": 0.0236, "step": 21608 }, { "epoch": 15.18552354181307, "grad_norm": 0.11963734775781631, "learning_rate": 2.3213867416256735e-05, "loss": 0.014, "step": 21609 }, { "epoch": 15.186226282501757, "grad_norm": 0.4551324248313904, "learning_rate": 2.321339892246428e-05, "loss": 0.0236, "step": 21610 }, { "epoch": 15.186929023190443, "grad_norm": 0.19681981205940247, "learning_rate": 2.321293042867182e-05, "loss": 0.0269, "step": 21611 }, { "epoch": 15.187631763879128, "grad_norm": 0.10867714881896973, "learning_rate": 2.3212461934879363e-05, "loss": 0.0214, "step": 21612 }, { "epoch": 15.188334504567814, "grad_norm": 0.20089630782604218, "learning_rate": 2.3211993441086907e-05, "loss": 0.0343, "step": 21613 }, { "epoch": 15.1890372452565, "grad_norm": 0.33489537239074707, "learning_rate": 2.321152494729445e-05, "loss": 0.0302, "step": 21614 }, { "epoch": 15.189739985945186, "grad_norm": 0.28877875208854675, "learning_rate": 2.321105645350199e-05, "loss": 0.0362, "step": 21615 }, { "epoch": 15.190442726633872, "grad_norm": 0.48555150628089905, "learning_rate": 2.3210587959709534e-05, "loss": 0.0551, "step": 21616 }, { "epoch": 15.191145467322558, "grad_norm": 0.6092302799224854, "learning_rate": 2.3210119465917078e-05, "loss": 0.0928, "step": 21617 }, { "epoch": 15.191848208011244, "grad_norm": 1.3422744274139404, "learning_rate": 2.3209650972124622e-05, "loss": 0.1491, "step": 21618 }, { "epoch": 15.19255094869993, "grad_norm": 0.558525800704956, "learning_rate": 2.3209182478332162e-05, "loss": 0.1562, "step": 21619 }, { "epoch": 15.193253689388616, "grad_norm": 1.2546230554580688, "learning_rate": 2.3208713984539706e-05, "loss": 0.1184, "step": 21620 }, { "epoch": 15.193956430077302, "grad_norm": 0.2800127863883972, "learning_rate": 2.320824549074725e-05, "loss": 0.0592, "step": 21621 }, { "epoch": 15.194659170765988, "grad_norm": 0.09509279578924179, "learning_rate": 2.320777699695479e-05, "loss": 0.027, "step": 21622 }, { "epoch": 15.195361911454674, "grad_norm": 0.15485301613807678, "learning_rate": 2.3207308503162334e-05, "loss": 0.0229, "step": 21623 }, { "epoch": 15.19606465214336, "grad_norm": 0.1473291963338852, "learning_rate": 2.3206840009369874e-05, "loss": 0.0154, "step": 21624 }, { "epoch": 15.196767392832045, "grad_norm": 0.07789316773414612, "learning_rate": 2.3206371515577418e-05, "loss": 0.0067, "step": 21625 }, { "epoch": 15.197470133520731, "grad_norm": 0.08134112507104874, "learning_rate": 2.320590302178496e-05, "loss": 0.0096, "step": 21626 }, { "epoch": 15.198172874209417, "grad_norm": 0.05447182431817055, "learning_rate": 2.3205434527992505e-05, "loss": 0.0087, "step": 21627 }, { "epoch": 15.198875614898103, "grad_norm": 0.13200819492340088, "learning_rate": 2.3204966034200045e-05, "loss": 0.0148, "step": 21628 }, { "epoch": 15.19957835558679, "grad_norm": 0.09519356489181519, "learning_rate": 2.320449754040759e-05, "loss": 0.0175, "step": 21629 }, { "epoch": 15.200281096275475, "grad_norm": 0.055397920310497284, "learning_rate": 2.3204029046615133e-05, "loss": 0.0072, "step": 21630 }, { "epoch": 15.200983836964161, "grad_norm": 0.2852911949157715, "learning_rate": 2.3203560552822677e-05, "loss": 0.0293, "step": 21631 }, { "epoch": 15.201686577652847, "grad_norm": 0.12462198734283447, "learning_rate": 2.320309205903022e-05, "loss": 0.0096, "step": 21632 }, { "epoch": 15.202389318341533, "grad_norm": 0.19739535450935364, "learning_rate": 2.320262356523776e-05, "loss": 0.0339, "step": 21633 }, { "epoch": 15.203092059030217, "grad_norm": 0.6009523868560791, "learning_rate": 2.3202155071445304e-05, "loss": 0.0123, "step": 21634 }, { "epoch": 15.203794799718903, "grad_norm": 0.3417830467224121, "learning_rate": 2.3201686577652848e-05, "loss": 0.0181, "step": 21635 }, { "epoch": 15.204497540407589, "grad_norm": 0.22764289379119873, "learning_rate": 2.3201218083860392e-05, "loss": 0.0341, "step": 21636 }, { "epoch": 15.205200281096275, "grad_norm": 0.4317256212234497, "learning_rate": 2.3200749590067932e-05, "loss": 0.0195, "step": 21637 }, { "epoch": 15.20590302178496, "grad_norm": 0.16903547942638397, "learning_rate": 2.3200281096275476e-05, "loss": 0.0412, "step": 21638 }, { "epoch": 15.206605762473647, "grad_norm": 0.1468101590871811, "learning_rate": 2.319981260248302e-05, "loss": 0.0292, "step": 21639 }, { "epoch": 15.207308503162333, "grad_norm": 0.2571030259132385, "learning_rate": 2.319934410869056e-05, "loss": 0.0608, "step": 21640 }, { "epoch": 15.208011243851018, "grad_norm": 0.31402361392974854, "learning_rate": 2.31988756148981e-05, "loss": 0.0521, "step": 21641 }, { "epoch": 15.208713984539704, "grad_norm": 0.5026158690452576, "learning_rate": 2.3198407121105644e-05, "loss": 0.0993, "step": 21642 }, { "epoch": 15.20941672522839, "grad_norm": 0.6380947828292847, "learning_rate": 2.3197938627313188e-05, "loss": 0.1054, "step": 21643 }, { "epoch": 15.210119465917076, "grad_norm": 0.7643046379089355, "learning_rate": 2.319747013352073e-05, "loss": 0.14, "step": 21644 }, { "epoch": 15.210822206605762, "grad_norm": 0.6601099967956543, "learning_rate": 2.3197001639728275e-05, "loss": 0.1532, "step": 21645 }, { "epoch": 15.211524947294448, "grad_norm": 0.24277274310588837, "learning_rate": 2.3196533145935816e-05, "loss": 0.0625, "step": 21646 }, { "epoch": 15.212227687983134, "grad_norm": 0.2515694499015808, "learning_rate": 2.319606465214336e-05, "loss": 0.0235, "step": 21647 }, { "epoch": 15.21293042867182, "grad_norm": 0.10247813910245895, "learning_rate": 2.3195596158350903e-05, "loss": 0.0256, "step": 21648 }, { "epoch": 15.213633169360506, "grad_norm": 0.069987952709198, "learning_rate": 2.3195127664558447e-05, "loss": 0.0135, "step": 21649 }, { "epoch": 15.214335910049192, "grad_norm": 0.1302199363708496, "learning_rate": 2.3194659170765987e-05, "loss": 0.0127, "step": 21650 }, { "epoch": 15.215038650737878, "grad_norm": 0.09728232771158218, "learning_rate": 2.319419067697353e-05, "loss": 0.011, "step": 21651 }, { "epoch": 15.215741391426564, "grad_norm": 0.13453538715839386, "learning_rate": 2.3193722183181075e-05, "loss": 0.0123, "step": 21652 }, { "epoch": 15.21644413211525, "grad_norm": 0.2138078510761261, "learning_rate": 2.3193253689388618e-05, "loss": 0.0126, "step": 21653 }, { "epoch": 15.217146872803935, "grad_norm": 0.085929274559021, "learning_rate": 2.319278519559616e-05, "loss": 0.0107, "step": 21654 }, { "epoch": 15.217849613492621, "grad_norm": 0.07917873561382294, "learning_rate": 2.3192316701803702e-05, "loss": 0.0096, "step": 21655 }, { "epoch": 15.218552354181307, "grad_norm": 0.16228553652763367, "learning_rate": 2.3191848208011246e-05, "loss": 0.0228, "step": 21656 }, { "epoch": 15.219255094869993, "grad_norm": 0.10432934761047363, "learning_rate": 2.3191379714218786e-05, "loss": 0.0127, "step": 21657 }, { "epoch": 15.219957835558679, "grad_norm": 0.09799295663833618, "learning_rate": 2.319091122042633e-05, "loss": 0.0146, "step": 21658 }, { "epoch": 15.220660576247365, "grad_norm": 0.06709616631269455, "learning_rate": 2.319044272663387e-05, "loss": 0.0061, "step": 21659 }, { "epoch": 15.221363316936051, "grad_norm": 0.17910785973072052, "learning_rate": 2.3189974232841414e-05, "loss": 0.0389, "step": 21660 }, { "epoch": 15.222066057624737, "grad_norm": 0.9531939625740051, "learning_rate": 2.3189505739048958e-05, "loss": 0.0335, "step": 21661 }, { "epoch": 15.222768798313423, "grad_norm": 0.11458829045295715, "learning_rate": 2.31890372452565e-05, "loss": 0.0163, "step": 21662 }, { "epoch": 15.223471539002109, "grad_norm": 0.15693756937980652, "learning_rate": 2.3188568751464042e-05, "loss": 0.0233, "step": 21663 }, { "epoch": 15.224174279690795, "grad_norm": 0.2647891044616699, "learning_rate": 2.3188100257671586e-05, "loss": 0.0363, "step": 21664 }, { "epoch": 15.22487702037948, "grad_norm": 0.20142436027526855, "learning_rate": 2.318763176387913e-05, "loss": 0.0333, "step": 21665 }, { "epoch": 15.225579761068166, "grad_norm": 0.2603228688240051, "learning_rate": 2.3187163270086673e-05, "loss": 0.0632, "step": 21666 }, { "epoch": 15.226282501756852, "grad_norm": 0.4057525098323822, "learning_rate": 2.3186694776294213e-05, "loss": 0.0915, "step": 21667 }, { "epoch": 15.226985242445538, "grad_norm": 0.983169674873352, "learning_rate": 2.3186226282501757e-05, "loss": 0.1084, "step": 21668 }, { "epoch": 15.227687983134224, "grad_norm": 0.3843376636505127, "learning_rate": 2.31857577887093e-05, "loss": 0.1235, "step": 21669 }, { "epoch": 15.22839072382291, "grad_norm": 1.2051626443862915, "learning_rate": 2.3185289294916845e-05, "loss": 0.1453, "step": 21670 }, { "epoch": 15.229093464511596, "grad_norm": 0.1974441409111023, "learning_rate": 2.318482080112439e-05, "loss": 0.0561, "step": 21671 }, { "epoch": 15.22979620520028, "grad_norm": 0.3388596773147583, "learning_rate": 2.318435230733193e-05, "loss": 0.0232, "step": 21672 }, { "epoch": 15.230498945888966, "grad_norm": 0.18382832407951355, "learning_rate": 2.3183883813539472e-05, "loss": 0.0209, "step": 21673 }, { "epoch": 15.231201686577652, "grad_norm": 0.16397608816623688, "learning_rate": 2.3183415319747016e-05, "loss": 0.015, "step": 21674 }, { "epoch": 15.231904427266338, "grad_norm": 0.09822817891836166, "learning_rate": 2.3182946825954556e-05, "loss": 0.0155, "step": 21675 }, { "epoch": 15.232607167955024, "grad_norm": 0.11578644067049026, "learning_rate": 2.3182478332162097e-05, "loss": 0.0154, "step": 21676 }, { "epoch": 15.23330990864371, "grad_norm": 0.1256403625011444, "learning_rate": 2.318200983836964e-05, "loss": 0.0105, "step": 21677 }, { "epoch": 15.234012649332396, "grad_norm": 0.4531199336051941, "learning_rate": 2.3181541344577184e-05, "loss": 0.0218, "step": 21678 }, { "epoch": 15.234715390021082, "grad_norm": 0.0894237831234932, "learning_rate": 2.3181072850784728e-05, "loss": 0.0129, "step": 21679 }, { "epoch": 15.235418130709768, "grad_norm": 0.06709054857492447, "learning_rate": 2.318060435699227e-05, "loss": 0.009, "step": 21680 }, { "epoch": 15.236120871398454, "grad_norm": 0.1397094577550888, "learning_rate": 2.3180135863199812e-05, "loss": 0.0173, "step": 21681 }, { "epoch": 15.23682361208714, "grad_norm": 0.14640754461288452, "learning_rate": 2.3179667369407356e-05, "loss": 0.0071, "step": 21682 }, { "epoch": 15.237526352775825, "grad_norm": 0.6632722020149231, "learning_rate": 2.31791988756149e-05, "loss": 0.0225, "step": 21683 }, { "epoch": 15.238229093464511, "grad_norm": 0.8393206596374512, "learning_rate": 2.3178730381822443e-05, "loss": 0.0188, "step": 21684 }, { "epoch": 15.238931834153197, "grad_norm": 0.14093901216983795, "learning_rate": 2.3178261888029984e-05, "loss": 0.017, "step": 21685 }, { "epoch": 15.239634574841883, "grad_norm": 0.3506373465061188, "learning_rate": 2.3177793394237527e-05, "loss": 0.056, "step": 21686 }, { "epoch": 15.240337315530569, "grad_norm": 0.42569246888160706, "learning_rate": 2.317732490044507e-05, "loss": 0.0176, "step": 21687 }, { "epoch": 15.241040056219255, "grad_norm": 0.20322014391422272, "learning_rate": 2.3176856406652615e-05, "loss": 0.0273, "step": 21688 }, { "epoch": 15.24174279690794, "grad_norm": 0.2504371404647827, "learning_rate": 2.3176387912860155e-05, "loss": 0.0295, "step": 21689 }, { "epoch": 15.242445537596627, "grad_norm": 0.17894677817821503, "learning_rate": 2.31759194190677e-05, "loss": 0.0308, "step": 21690 }, { "epoch": 15.243148278285313, "grad_norm": 0.35292181372642517, "learning_rate": 2.3175450925275243e-05, "loss": 0.0547, "step": 21691 }, { "epoch": 15.243851018973999, "grad_norm": 0.4224717617034912, "learning_rate": 2.3174982431482783e-05, "loss": 0.0988, "step": 21692 }, { "epoch": 15.244553759662685, "grad_norm": 0.6903510093688965, "learning_rate": 2.3174513937690323e-05, "loss": 0.1191, "step": 21693 }, { "epoch": 15.24525650035137, "grad_norm": 1.248191237449646, "learning_rate": 2.3174045443897867e-05, "loss": 0.1471, "step": 21694 }, { "epoch": 15.245959241040056, "grad_norm": 1.1272226572036743, "learning_rate": 2.317357695010541e-05, "loss": 0.1926, "step": 21695 }, { "epoch": 15.246661981728742, "grad_norm": 0.26955708861351013, "learning_rate": 2.3173108456312954e-05, "loss": 0.0538, "step": 21696 }, { "epoch": 15.247364722417428, "grad_norm": 0.1287744641304016, "learning_rate": 2.3172639962520498e-05, "loss": 0.0146, "step": 21697 }, { "epoch": 15.248067463106114, "grad_norm": 0.3875480592250824, "learning_rate": 2.317217146872804e-05, "loss": 0.0189, "step": 21698 }, { "epoch": 15.2487702037948, "grad_norm": 0.16307248175144196, "learning_rate": 2.3171702974935582e-05, "loss": 0.0203, "step": 21699 }, { "epoch": 15.249472944483486, "grad_norm": 0.11123693734407425, "learning_rate": 2.3171234481143126e-05, "loss": 0.0157, "step": 21700 }, { "epoch": 15.250175685172172, "grad_norm": 0.07463769614696503, "learning_rate": 2.317076598735067e-05, "loss": 0.0113, "step": 21701 }, { "epoch": 15.250878425860858, "grad_norm": 0.12810224294662476, "learning_rate": 2.317029749355821e-05, "loss": 0.0161, "step": 21702 }, { "epoch": 15.251581166549544, "grad_norm": 0.30266404151916504, "learning_rate": 2.3169828999765754e-05, "loss": 0.0114, "step": 21703 }, { "epoch": 15.25228390723823, "grad_norm": 0.11024602502584457, "learning_rate": 2.3169360505973297e-05, "loss": 0.0145, "step": 21704 }, { "epoch": 15.252986647926916, "grad_norm": 0.07195290178060532, "learning_rate": 2.316889201218084e-05, "loss": 0.0073, "step": 21705 }, { "epoch": 15.253689388615602, "grad_norm": 0.15843793749809265, "learning_rate": 2.316842351838838e-05, "loss": 0.016, "step": 21706 }, { "epoch": 15.254392129304287, "grad_norm": 0.5237612724304199, "learning_rate": 2.3167955024595925e-05, "loss": 0.006, "step": 21707 }, { "epoch": 15.255094869992973, "grad_norm": 0.3137483596801758, "learning_rate": 2.316748653080347e-05, "loss": 0.0182, "step": 21708 }, { "epoch": 15.25579761068166, "grad_norm": 0.07599838823080063, "learning_rate": 2.3167018037011013e-05, "loss": 0.0111, "step": 21709 }, { "epoch": 15.256500351370345, "grad_norm": 0.1368364840745926, "learning_rate": 2.3166549543218553e-05, "loss": 0.0194, "step": 21710 }, { "epoch": 15.25720309205903, "grad_norm": 0.20405492186546326, "learning_rate": 2.3166081049426093e-05, "loss": 0.0254, "step": 21711 }, { "epoch": 15.257905832747715, "grad_norm": 0.3085726201534271, "learning_rate": 2.3165612555633637e-05, "loss": 0.0129, "step": 21712 }, { "epoch": 15.258608573436401, "grad_norm": 0.2666810154914856, "learning_rate": 2.316514406184118e-05, "loss": 0.0423, "step": 21713 }, { "epoch": 15.259311314125087, "grad_norm": 0.30107414722442627, "learning_rate": 2.3164675568048725e-05, "loss": 0.0472, "step": 21714 }, { "epoch": 15.260014054813773, "grad_norm": 0.16834387183189392, "learning_rate": 2.3164207074256265e-05, "loss": 0.0326, "step": 21715 }, { "epoch": 15.260716795502459, "grad_norm": 1.0347633361816406, "learning_rate": 2.316373858046381e-05, "loss": 0.0726, "step": 21716 }, { "epoch": 15.261419536191145, "grad_norm": 0.48005664348602295, "learning_rate": 2.3163270086671352e-05, "loss": 0.0833, "step": 21717 }, { "epoch": 15.26212227687983, "grad_norm": 0.6334994435310364, "learning_rate": 2.3162801592878896e-05, "loss": 0.1043, "step": 21718 }, { "epoch": 15.262825017568517, "grad_norm": 0.6472219824790955, "learning_rate": 2.3162333099086436e-05, "loss": 0.1724, "step": 21719 }, { "epoch": 15.263527758257203, "grad_norm": 1.0173823833465576, "learning_rate": 2.316186460529398e-05, "loss": 0.1601, "step": 21720 }, { "epoch": 15.264230498945889, "grad_norm": 0.2627590000629425, "learning_rate": 2.3161396111501524e-05, "loss": 0.0606, "step": 21721 }, { "epoch": 15.264933239634574, "grad_norm": 0.09605082124471664, "learning_rate": 2.3160927617709068e-05, "loss": 0.0226, "step": 21722 }, { "epoch": 15.26563598032326, "grad_norm": 0.14151543378829956, "learning_rate": 2.316045912391661e-05, "loss": 0.0126, "step": 21723 }, { "epoch": 15.266338721011946, "grad_norm": 0.11514393240213394, "learning_rate": 2.315999063012415e-05, "loss": 0.015, "step": 21724 }, { "epoch": 15.267041461700632, "grad_norm": 0.08673609793186188, "learning_rate": 2.3159522136331695e-05, "loss": 0.0099, "step": 21725 }, { "epoch": 15.267744202389318, "grad_norm": 0.07441238313913345, "learning_rate": 2.315905364253924e-05, "loss": 0.0086, "step": 21726 }, { "epoch": 15.268446943078004, "grad_norm": 0.4512379467487335, "learning_rate": 2.315858514874678e-05, "loss": 0.0144, "step": 21727 }, { "epoch": 15.26914968376669, "grad_norm": 1.2181978225708008, "learning_rate": 2.315811665495432e-05, "loss": 0.04, "step": 21728 }, { "epoch": 15.269852424455376, "grad_norm": 0.09927879273891449, "learning_rate": 2.3157648161161863e-05, "loss": 0.0225, "step": 21729 }, { "epoch": 15.270555165144062, "grad_norm": 0.049756623804569244, "learning_rate": 2.3157179667369407e-05, "loss": 0.0057, "step": 21730 }, { "epoch": 15.271257905832748, "grad_norm": 0.15216735005378723, "learning_rate": 2.315671117357695e-05, "loss": 0.03, "step": 21731 }, { "epoch": 15.271960646521434, "grad_norm": 0.3315196633338928, "learning_rate": 2.315624267978449e-05, "loss": 0.0141, "step": 21732 }, { "epoch": 15.27266338721012, "grad_norm": 0.2041492909193039, "learning_rate": 2.3155774185992035e-05, "loss": 0.0195, "step": 21733 }, { "epoch": 15.273366127898806, "grad_norm": 0.1501573920249939, "learning_rate": 2.315530569219958e-05, "loss": 0.0216, "step": 21734 }, { "epoch": 15.274068868587491, "grad_norm": 0.1989631950855255, "learning_rate": 2.3154837198407122e-05, "loss": 0.0138, "step": 21735 }, { "epoch": 15.274771609276177, "grad_norm": 0.17962558567523956, "learning_rate": 2.3154368704614666e-05, "loss": 0.029, "step": 21736 }, { "epoch": 15.275474349964863, "grad_norm": 0.190634086728096, "learning_rate": 2.3153900210822206e-05, "loss": 0.0152, "step": 21737 }, { "epoch": 15.27617709065355, "grad_norm": 0.3514901399612427, "learning_rate": 2.315343171702975e-05, "loss": 0.0205, "step": 21738 }, { "epoch": 15.276879831342235, "grad_norm": 0.17024099826812744, "learning_rate": 2.3152963223237294e-05, "loss": 0.0311, "step": 21739 }, { "epoch": 15.277582572030921, "grad_norm": 0.32372644543647766, "learning_rate": 2.3152494729444838e-05, "loss": 0.0504, "step": 21740 }, { "epoch": 15.278285312719607, "grad_norm": 0.3169518709182739, "learning_rate": 2.3152026235652378e-05, "loss": 0.0656, "step": 21741 }, { "epoch": 15.278988053408293, "grad_norm": 0.48059406876564026, "learning_rate": 2.3151557741859922e-05, "loss": 0.0937, "step": 21742 }, { "epoch": 15.279690794096979, "grad_norm": 0.754338800907135, "learning_rate": 2.3151089248067465e-05, "loss": 0.1209, "step": 21743 }, { "epoch": 15.280393534785665, "grad_norm": 0.4888177514076233, "learning_rate": 2.3150620754275006e-05, "loss": 0.1681, "step": 21744 }, { "epoch": 15.28109627547435, "grad_norm": 2.028165817260742, "learning_rate": 2.3150152260482546e-05, "loss": 0.1719, "step": 21745 }, { "epoch": 15.281799016163037, "grad_norm": 0.5441020131111145, "learning_rate": 2.314968376669009e-05, "loss": 0.0547, "step": 21746 }, { "epoch": 15.282501756851723, "grad_norm": 0.1598404049873352, "learning_rate": 2.3149215272897634e-05, "loss": 0.0265, "step": 21747 }, { "epoch": 15.283204497540408, "grad_norm": 0.106757752597332, "learning_rate": 2.3148746779105177e-05, "loss": 0.023, "step": 21748 }, { "epoch": 15.283907238229094, "grad_norm": 0.14884309470653534, "learning_rate": 2.314827828531272e-05, "loss": 0.0133, "step": 21749 }, { "epoch": 15.284609978917779, "grad_norm": 0.14075350761413574, "learning_rate": 2.314780979152026e-05, "loss": 0.022, "step": 21750 }, { "epoch": 15.285312719606464, "grad_norm": 0.07093711197376251, "learning_rate": 2.3147341297727805e-05, "loss": 0.009, "step": 21751 }, { "epoch": 15.28601546029515, "grad_norm": 0.19656312465667725, "learning_rate": 2.314687280393535e-05, "loss": 0.013, "step": 21752 }, { "epoch": 15.286718200983836, "grad_norm": 0.13310998678207397, "learning_rate": 2.3146404310142893e-05, "loss": 0.0147, "step": 21753 }, { "epoch": 15.287420941672522, "grad_norm": 0.18692316114902496, "learning_rate": 2.3145935816350433e-05, "loss": 0.0155, "step": 21754 }, { "epoch": 15.288123682361208, "grad_norm": 0.1290460228919983, "learning_rate": 2.3145467322557977e-05, "loss": 0.0185, "step": 21755 }, { "epoch": 15.288826423049894, "grad_norm": 0.11264680325984955, "learning_rate": 2.314499882876552e-05, "loss": 0.0238, "step": 21756 }, { "epoch": 15.28952916373858, "grad_norm": 0.07430517673492432, "learning_rate": 2.3144530334973064e-05, "loss": 0.0088, "step": 21757 }, { "epoch": 15.290231904427266, "grad_norm": 0.6229949593544006, "learning_rate": 2.3144061841180604e-05, "loss": 0.0257, "step": 21758 }, { "epoch": 15.290934645115952, "grad_norm": 0.12401559948921204, "learning_rate": 2.3143593347388148e-05, "loss": 0.0139, "step": 21759 }, { "epoch": 15.291637385804638, "grad_norm": 0.35859915614128113, "learning_rate": 2.3143124853595692e-05, "loss": 0.0297, "step": 21760 }, { "epoch": 15.292340126493324, "grad_norm": 0.22301959991455078, "learning_rate": 2.3142656359803236e-05, "loss": 0.0423, "step": 21761 }, { "epoch": 15.29304286718201, "grad_norm": 0.11820690333843231, "learning_rate": 2.3142187866010776e-05, "loss": 0.017, "step": 21762 }, { "epoch": 15.293745607870695, "grad_norm": 0.1790880411863327, "learning_rate": 2.3141719372218316e-05, "loss": 0.0338, "step": 21763 }, { "epoch": 15.294448348559381, "grad_norm": 0.13391545414924622, "learning_rate": 2.314125087842586e-05, "loss": 0.024, "step": 21764 }, { "epoch": 15.295151089248067, "grad_norm": 0.6259467005729675, "learning_rate": 2.3140782384633404e-05, "loss": 0.0539, "step": 21765 }, { "epoch": 15.295853829936753, "grad_norm": 0.37925875186920166, "learning_rate": 2.3140313890840947e-05, "loss": 0.0498, "step": 21766 }, { "epoch": 15.29655657062544, "grad_norm": 0.4401091933250427, "learning_rate": 2.3139845397048488e-05, "loss": 0.0815, "step": 21767 }, { "epoch": 15.297259311314125, "grad_norm": 0.765630841255188, "learning_rate": 2.313937690325603e-05, "loss": 0.1116, "step": 21768 }, { "epoch": 15.297962052002811, "grad_norm": 0.5121971964836121, "learning_rate": 2.3138908409463575e-05, "loss": 0.1377, "step": 21769 }, { "epoch": 15.298664792691497, "grad_norm": 1.3123060464859009, "learning_rate": 2.313843991567112e-05, "loss": 0.1554, "step": 21770 }, { "epoch": 15.299367533380183, "grad_norm": 0.4730585217475891, "learning_rate": 2.313797142187866e-05, "loss": 0.0574, "step": 21771 }, { "epoch": 15.300070274068869, "grad_norm": 0.24435622990131378, "learning_rate": 2.3137502928086203e-05, "loss": 0.0318, "step": 21772 }, { "epoch": 15.300773014757555, "grad_norm": 0.2385556995868683, "learning_rate": 2.3137034434293747e-05, "loss": 0.0202, "step": 21773 }, { "epoch": 15.30147575544624, "grad_norm": 0.16415159404277802, "learning_rate": 2.313656594050129e-05, "loss": 0.0283, "step": 21774 }, { "epoch": 15.302178496134927, "grad_norm": 0.18570637702941895, "learning_rate": 2.3136097446708834e-05, "loss": 0.0168, "step": 21775 }, { "epoch": 15.302881236823612, "grad_norm": 0.2897723615169525, "learning_rate": 2.3135628952916374e-05, "loss": 0.0148, "step": 21776 }, { "epoch": 15.303583977512298, "grad_norm": 0.08166272193193436, "learning_rate": 2.3135160459123918e-05, "loss": 0.015, "step": 21777 }, { "epoch": 15.304286718200984, "grad_norm": 0.5202643275260925, "learning_rate": 2.3134691965331462e-05, "loss": 0.0313, "step": 21778 }, { "epoch": 15.30498945888967, "grad_norm": 0.2772580683231354, "learning_rate": 2.3134223471539002e-05, "loss": 0.0234, "step": 21779 }, { "epoch": 15.305692199578356, "grad_norm": 0.5156728625297546, "learning_rate": 2.3133754977746543e-05, "loss": 0.0107, "step": 21780 }, { "epoch": 15.306394940267042, "grad_norm": 0.14529360830783844, "learning_rate": 2.3133286483954086e-05, "loss": 0.0177, "step": 21781 }, { "epoch": 15.307097680955728, "grad_norm": 0.09646078199148178, "learning_rate": 2.313281799016163e-05, "loss": 0.0087, "step": 21782 }, { "epoch": 15.307800421644414, "grad_norm": 0.42645469307899475, "learning_rate": 2.3132349496369174e-05, "loss": 0.0214, "step": 21783 }, { "epoch": 15.3085031623331, "grad_norm": 0.08080544322729111, "learning_rate": 2.3131881002576714e-05, "loss": 0.0133, "step": 21784 }, { "epoch": 15.309205903021786, "grad_norm": 0.08787249028682709, "learning_rate": 2.3131412508784258e-05, "loss": 0.0208, "step": 21785 }, { "epoch": 15.309908643710472, "grad_norm": 0.11150606721639633, "learning_rate": 2.31309440149918e-05, "loss": 0.0189, "step": 21786 }, { "epoch": 15.310611384399156, "grad_norm": 0.11927749961614609, "learning_rate": 2.3130475521199345e-05, "loss": 0.0192, "step": 21787 }, { "epoch": 15.311314125087842, "grad_norm": 0.10321450978517532, "learning_rate": 2.313000702740689e-05, "loss": 0.0223, "step": 21788 }, { "epoch": 15.312016865776528, "grad_norm": 0.22525596618652344, "learning_rate": 2.312953853361443e-05, "loss": 0.0416, "step": 21789 }, { "epoch": 15.312719606465214, "grad_norm": 0.2755291759967804, "learning_rate": 2.3129070039821973e-05, "loss": 0.0501, "step": 21790 }, { "epoch": 15.3134223471539, "grad_norm": 0.4377020001411438, "learning_rate": 2.3128601546029517e-05, "loss": 0.0603, "step": 21791 }, { "epoch": 15.314125087842585, "grad_norm": 0.35862040519714355, "learning_rate": 2.312813305223706e-05, "loss": 0.0854, "step": 21792 }, { "epoch": 15.314827828531271, "grad_norm": 0.7055469751358032, "learning_rate": 2.31276645584446e-05, "loss": 0.1183, "step": 21793 }, { "epoch": 15.315530569219957, "grad_norm": 0.6183928847312927, "learning_rate": 2.3127196064652145e-05, "loss": 0.1168, "step": 21794 }, { "epoch": 15.316233309908643, "grad_norm": 0.7172726392745972, "learning_rate": 2.312672757085969e-05, "loss": 0.1436, "step": 21795 }, { "epoch": 15.316936050597329, "grad_norm": 0.30919840931892395, "learning_rate": 2.3126259077067232e-05, "loss": 0.0565, "step": 21796 }, { "epoch": 15.317638791286015, "grad_norm": 0.150959774851799, "learning_rate": 2.3125790583274772e-05, "loss": 0.0218, "step": 21797 }, { "epoch": 15.318341531974701, "grad_norm": 0.07667982578277588, "learning_rate": 2.3125322089482313e-05, "loss": 0.012, "step": 21798 }, { "epoch": 15.319044272663387, "grad_norm": 0.1842474490404129, "learning_rate": 2.3124853595689856e-05, "loss": 0.0172, "step": 21799 }, { "epoch": 15.319747013352073, "grad_norm": 0.11105459928512573, "learning_rate": 2.31243851018974e-05, "loss": 0.0115, "step": 21800 }, { "epoch": 15.320449754040759, "grad_norm": 0.07671438902616501, "learning_rate": 2.3123916608104944e-05, "loss": 0.0121, "step": 21801 }, { "epoch": 15.321152494729445, "grad_norm": 0.08158186078071594, "learning_rate": 2.3123448114312484e-05, "loss": 0.0102, "step": 21802 }, { "epoch": 15.32185523541813, "grad_norm": 0.14116643369197845, "learning_rate": 2.3122979620520028e-05, "loss": 0.013, "step": 21803 }, { "epoch": 15.322557976106816, "grad_norm": 0.1148390993475914, "learning_rate": 2.312251112672757e-05, "loss": 0.0164, "step": 21804 }, { "epoch": 15.323260716795502, "grad_norm": 0.11418785899877548, "learning_rate": 2.3122042632935115e-05, "loss": 0.0117, "step": 21805 }, { "epoch": 15.323963457484188, "grad_norm": 0.08434926718473434, "learning_rate": 2.3121574139142656e-05, "loss": 0.0136, "step": 21806 }, { "epoch": 15.324666198172874, "grad_norm": 0.04973762854933739, "learning_rate": 2.31211056453502e-05, "loss": 0.0042, "step": 21807 }, { "epoch": 15.32536893886156, "grad_norm": 0.10402385890483856, "learning_rate": 2.3120637151557743e-05, "loss": 0.0223, "step": 21808 }, { "epoch": 15.326071679550246, "grad_norm": 0.11479059606790543, "learning_rate": 2.3120168657765287e-05, "loss": 0.011, "step": 21809 }, { "epoch": 15.326774420238932, "grad_norm": 0.1380072981119156, "learning_rate": 2.3119700163972827e-05, "loss": 0.0165, "step": 21810 }, { "epoch": 15.327477160927618, "grad_norm": 0.3533885180950165, "learning_rate": 2.311923167018037e-05, "loss": 0.0496, "step": 21811 }, { "epoch": 15.328179901616304, "grad_norm": 0.2327951341867447, "learning_rate": 2.3118763176387915e-05, "loss": 0.0151, "step": 21812 }, { "epoch": 15.32888264230499, "grad_norm": 0.18963269889354706, "learning_rate": 2.311829468259546e-05, "loss": 0.0244, "step": 21813 }, { "epoch": 15.329585382993676, "grad_norm": 0.43363654613494873, "learning_rate": 2.3117826188803e-05, "loss": 0.0599, "step": 21814 }, { "epoch": 15.330288123682362, "grad_norm": 0.18485507369041443, "learning_rate": 2.311735769501054e-05, "loss": 0.0169, "step": 21815 }, { "epoch": 15.330990864371048, "grad_norm": 1.3327730894088745, "learning_rate": 2.3116889201218083e-05, "loss": 0.0551, "step": 21816 }, { "epoch": 15.331693605059733, "grad_norm": 0.32620927691459656, "learning_rate": 2.3116420707425627e-05, "loss": 0.0805, "step": 21817 }, { "epoch": 15.33239634574842, "grad_norm": 0.6911879181861877, "learning_rate": 2.311595221363317e-05, "loss": 0.1091, "step": 21818 }, { "epoch": 15.333099086437105, "grad_norm": 0.6746777892112732, "learning_rate": 2.311548371984071e-05, "loss": 0.148, "step": 21819 }, { "epoch": 15.333801827125791, "grad_norm": 1.2900035381317139, "learning_rate": 2.3115015226048254e-05, "loss": 0.1716, "step": 21820 }, { "epoch": 15.334504567814477, "grad_norm": 0.37755516171455383, "learning_rate": 2.3114546732255798e-05, "loss": 0.0583, "step": 21821 }, { "epoch": 15.335207308503163, "grad_norm": 0.1126868724822998, "learning_rate": 2.3114078238463342e-05, "loss": 0.0222, "step": 21822 }, { "epoch": 15.335910049191849, "grad_norm": 0.15139919519424438, "learning_rate": 2.3113609744670886e-05, "loss": 0.019, "step": 21823 }, { "epoch": 15.336612789880535, "grad_norm": 0.1014680489897728, "learning_rate": 2.3113141250878426e-05, "loss": 0.0214, "step": 21824 }, { "epoch": 15.33731553056922, "grad_norm": 0.11558612436056137, "learning_rate": 2.311267275708597e-05, "loss": 0.0172, "step": 21825 }, { "epoch": 15.338018271257905, "grad_norm": 0.15296390652656555, "learning_rate": 2.3112204263293513e-05, "loss": 0.0198, "step": 21826 }, { "epoch": 15.33872101194659, "grad_norm": 0.09733225405216217, "learning_rate": 2.3111735769501057e-05, "loss": 0.0116, "step": 21827 }, { "epoch": 15.339423752635277, "grad_norm": 0.09520622342824936, "learning_rate": 2.3111267275708597e-05, "loss": 0.0106, "step": 21828 }, { "epoch": 15.340126493323963, "grad_norm": 0.2831893265247345, "learning_rate": 2.311079878191614e-05, "loss": 0.0212, "step": 21829 }, { "epoch": 15.340829234012649, "grad_norm": 0.08693176507949829, "learning_rate": 2.3110330288123685e-05, "loss": 0.0099, "step": 21830 }, { "epoch": 15.341531974701335, "grad_norm": 0.37961649894714355, "learning_rate": 2.3109861794331225e-05, "loss": 0.0271, "step": 21831 }, { "epoch": 15.34223471539002, "grad_norm": 0.2548750042915344, "learning_rate": 2.3109393300538765e-05, "loss": 0.0236, "step": 21832 }, { "epoch": 15.342937456078706, "grad_norm": 0.3251765966415405, "learning_rate": 2.310892480674631e-05, "loss": 0.02, "step": 21833 }, { "epoch": 15.343640196767392, "grad_norm": 0.2870011031627655, "learning_rate": 2.3108456312953853e-05, "loss": 0.0185, "step": 21834 }, { "epoch": 15.344342937456078, "grad_norm": 0.17856313288211823, "learning_rate": 2.3107987819161397e-05, "loss": 0.0284, "step": 21835 }, { "epoch": 15.345045678144764, "grad_norm": 0.1951395869255066, "learning_rate": 2.310751932536894e-05, "loss": 0.0161, "step": 21836 }, { "epoch": 15.34574841883345, "grad_norm": 0.25876304507255554, "learning_rate": 2.310705083157648e-05, "loss": 0.0279, "step": 21837 }, { "epoch": 15.346451159522136, "grad_norm": 0.13440482318401337, "learning_rate": 2.3106582337784024e-05, "loss": 0.0202, "step": 21838 }, { "epoch": 15.347153900210822, "grad_norm": 0.3461669683456421, "learning_rate": 2.3106113843991568e-05, "loss": 0.0578, "step": 21839 }, { "epoch": 15.347856640899508, "grad_norm": 0.14609573781490326, "learning_rate": 2.3105645350199112e-05, "loss": 0.0416, "step": 21840 }, { "epoch": 15.348559381588194, "grad_norm": 0.2450731098651886, "learning_rate": 2.3105176856406652e-05, "loss": 0.0528, "step": 21841 }, { "epoch": 15.34926212227688, "grad_norm": 0.32700279355049133, "learning_rate": 2.3104708362614196e-05, "loss": 0.0976, "step": 21842 }, { "epoch": 15.349964862965566, "grad_norm": 0.7525007128715515, "learning_rate": 2.310423986882174e-05, "loss": 0.1245, "step": 21843 }, { "epoch": 15.350667603654252, "grad_norm": 0.49913549423217773, "learning_rate": 2.3103771375029283e-05, "loss": 0.1278, "step": 21844 }, { "epoch": 15.351370344342937, "grad_norm": 0.7848547101020813, "learning_rate": 2.3103302881236824e-05, "loss": 0.1669, "step": 21845 }, { "epoch": 15.352073085031623, "grad_norm": 0.2111300826072693, "learning_rate": 2.3102834387444367e-05, "loss": 0.0551, "step": 21846 }, { "epoch": 15.35277582572031, "grad_norm": 0.28236740827560425, "learning_rate": 2.310236589365191e-05, "loss": 0.0191, "step": 21847 }, { "epoch": 15.353478566408995, "grad_norm": 0.11234427988529205, "learning_rate": 2.3101897399859455e-05, "loss": 0.0249, "step": 21848 }, { "epoch": 15.354181307097681, "grad_norm": 0.07548581808805466, "learning_rate": 2.3101428906066995e-05, "loss": 0.012, "step": 21849 }, { "epoch": 15.354884047786367, "grad_norm": 0.10848148167133331, "learning_rate": 2.3100960412274536e-05, "loss": 0.0195, "step": 21850 }, { "epoch": 15.355586788475053, "grad_norm": 0.06766823679208755, "learning_rate": 2.310049191848208e-05, "loss": 0.0126, "step": 21851 }, { "epoch": 15.356289529163739, "grad_norm": 0.13438834249973297, "learning_rate": 2.3100023424689623e-05, "loss": 0.0132, "step": 21852 }, { "epoch": 15.356992269852425, "grad_norm": 0.24813343584537506, "learning_rate": 2.3099554930897167e-05, "loss": 0.0213, "step": 21853 }, { "epoch": 15.35769501054111, "grad_norm": 0.07460132241249084, "learning_rate": 2.3099086437104707e-05, "loss": 0.0144, "step": 21854 }, { "epoch": 15.358397751229797, "grad_norm": 0.1708766669034958, "learning_rate": 2.309861794331225e-05, "loss": 0.0146, "step": 21855 }, { "epoch": 15.359100491918483, "grad_norm": 0.11096685379743576, "learning_rate": 2.3098149449519795e-05, "loss": 0.0175, "step": 21856 }, { "epoch": 15.359803232607169, "grad_norm": 0.07947687804698944, "learning_rate": 2.3097680955727338e-05, "loss": 0.0126, "step": 21857 }, { "epoch": 15.360505973295854, "grad_norm": 0.22841444611549377, "learning_rate": 2.309721246193488e-05, "loss": 0.0309, "step": 21858 }, { "epoch": 15.36120871398454, "grad_norm": 0.2127084583044052, "learning_rate": 2.3096743968142422e-05, "loss": 0.0153, "step": 21859 }, { "epoch": 15.361911454673226, "grad_norm": 0.14398299157619476, "learning_rate": 2.3096275474349966e-05, "loss": 0.0214, "step": 21860 }, { "epoch": 15.362614195361912, "grad_norm": 0.21148154139518738, "learning_rate": 2.309580698055751e-05, "loss": 0.0323, "step": 21861 }, { "epoch": 15.363316936050598, "grad_norm": 0.15562686324119568, "learning_rate": 2.3095338486765054e-05, "loss": 0.0161, "step": 21862 }, { "epoch": 15.364019676739284, "grad_norm": 0.15092700719833374, "learning_rate": 2.3094869992972594e-05, "loss": 0.0315, "step": 21863 }, { "epoch": 15.36472241742797, "grad_norm": 0.15922275185585022, "learning_rate": 2.3094401499180138e-05, "loss": 0.0362, "step": 21864 }, { "epoch": 15.365425158116654, "grad_norm": 0.16709758341312408, "learning_rate": 2.309393300538768e-05, "loss": 0.0476, "step": 21865 }, { "epoch": 15.36612789880534, "grad_norm": 0.7113425135612488, "learning_rate": 2.309346451159522e-05, "loss": 0.0941, "step": 21866 }, { "epoch": 15.366830639494026, "grad_norm": 0.47166821360588074, "learning_rate": 2.3092996017802762e-05, "loss": 0.0861, "step": 21867 }, { "epoch": 15.367533380182712, "grad_norm": 0.6670225858688354, "learning_rate": 2.3092527524010306e-05, "loss": 0.1254, "step": 21868 }, { "epoch": 15.368236120871398, "grad_norm": 0.4228339195251465, "learning_rate": 2.309205903021785e-05, "loss": 0.1262, "step": 21869 }, { "epoch": 15.368938861560084, "grad_norm": 1.0864351987838745, "learning_rate": 2.3091590536425393e-05, "loss": 0.1608, "step": 21870 }, { "epoch": 15.36964160224877, "grad_norm": 0.2563804090023041, "learning_rate": 2.3091122042632934e-05, "loss": 0.061, "step": 21871 }, { "epoch": 15.370344342937456, "grad_norm": 0.10993191599845886, "learning_rate": 2.3090653548840477e-05, "loss": 0.0191, "step": 21872 }, { "epoch": 15.371047083626141, "grad_norm": 0.16611982882022858, "learning_rate": 2.309018505504802e-05, "loss": 0.0271, "step": 21873 }, { "epoch": 15.371749824314827, "grad_norm": 0.07901473343372345, "learning_rate": 2.3089716561255565e-05, "loss": 0.0112, "step": 21874 }, { "epoch": 15.372452565003513, "grad_norm": 0.06062311679124832, "learning_rate": 2.308924806746311e-05, "loss": 0.0177, "step": 21875 }, { "epoch": 15.3731553056922, "grad_norm": 0.12998899817466736, "learning_rate": 2.308877957367065e-05, "loss": 0.0193, "step": 21876 }, { "epoch": 15.373858046380885, "grad_norm": 0.09645821154117584, "learning_rate": 2.3088311079878192e-05, "loss": 0.0088, "step": 21877 }, { "epoch": 15.374560787069571, "grad_norm": 0.08370429277420044, "learning_rate": 2.3087842586085736e-05, "loss": 0.0109, "step": 21878 }, { "epoch": 15.375263527758257, "grad_norm": 0.6716512441635132, "learning_rate": 2.308737409229328e-05, "loss": 0.0357, "step": 21879 }, { "epoch": 15.375966268446943, "grad_norm": 0.10828372836112976, "learning_rate": 2.308690559850082e-05, "loss": 0.0086, "step": 21880 }, { "epoch": 15.376669009135629, "grad_norm": 0.12643131613731384, "learning_rate": 2.3086437104708364e-05, "loss": 0.02, "step": 21881 }, { "epoch": 15.377371749824315, "grad_norm": 0.12498686462640762, "learning_rate": 2.3085968610915908e-05, "loss": 0.0169, "step": 21882 }, { "epoch": 15.378074490513, "grad_norm": 0.18074223399162292, "learning_rate": 2.308550011712345e-05, "loss": 0.03, "step": 21883 }, { "epoch": 15.378777231201687, "grad_norm": 0.19757744669914246, "learning_rate": 2.308503162333099e-05, "loss": 0.0171, "step": 21884 }, { "epoch": 15.379479971890373, "grad_norm": 0.13504287600517273, "learning_rate": 2.3084563129538532e-05, "loss": 0.0329, "step": 21885 }, { "epoch": 15.380182712579058, "grad_norm": 0.11474493891000748, "learning_rate": 2.3084094635746076e-05, "loss": 0.0198, "step": 21886 }, { "epoch": 15.380885453267744, "grad_norm": 0.24261574447155, "learning_rate": 2.308362614195362e-05, "loss": 0.0111, "step": 21887 }, { "epoch": 15.38158819395643, "grad_norm": 0.15043988823890686, "learning_rate": 2.3083157648161163e-05, "loss": 0.0242, "step": 21888 }, { "epoch": 15.382290934645116, "grad_norm": 0.23273180425167084, "learning_rate": 2.3082689154368704e-05, "loss": 0.024, "step": 21889 }, { "epoch": 15.382993675333802, "grad_norm": 0.6559271216392517, "learning_rate": 2.3082220660576247e-05, "loss": 0.063, "step": 21890 }, { "epoch": 15.383696416022488, "grad_norm": 0.44204169511795044, "learning_rate": 2.308175216678379e-05, "loss": 0.0686, "step": 21891 }, { "epoch": 15.384399156711174, "grad_norm": 0.5920888185501099, "learning_rate": 2.3081283672991335e-05, "loss": 0.1125, "step": 21892 }, { "epoch": 15.38510189739986, "grad_norm": 0.4228442311286926, "learning_rate": 2.3080815179198875e-05, "loss": 0.1127, "step": 21893 }, { "epoch": 15.385804638088546, "grad_norm": 0.9956089854240417, "learning_rate": 2.308034668540642e-05, "loss": 0.1588, "step": 21894 }, { "epoch": 15.386507378777232, "grad_norm": 0.6832162737846375, "learning_rate": 2.3079878191613963e-05, "loss": 0.169, "step": 21895 }, { "epoch": 15.387210119465918, "grad_norm": 0.6003009080886841, "learning_rate": 2.3079409697821506e-05, "loss": 0.0539, "step": 21896 }, { "epoch": 15.387912860154604, "grad_norm": 0.21226896345615387, "learning_rate": 2.3078941204029047e-05, "loss": 0.0322, "step": 21897 }, { "epoch": 15.38861560084329, "grad_norm": 0.09877544641494751, "learning_rate": 2.307847271023659e-05, "loss": 0.0144, "step": 21898 }, { "epoch": 15.389318341531975, "grad_norm": 0.12122132629156113, "learning_rate": 2.3078004216444134e-05, "loss": 0.0179, "step": 21899 }, { "epoch": 15.390021082220661, "grad_norm": 0.2852117717266083, "learning_rate": 2.3077535722651678e-05, "loss": 0.0179, "step": 21900 }, { "epoch": 15.390723822909347, "grad_norm": 0.2256508618593216, "learning_rate": 2.3077067228859218e-05, "loss": 0.0199, "step": 21901 }, { "epoch": 15.391426563598033, "grad_norm": 0.09294512122869492, "learning_rate": 2.307659873506676e-05, "loss": 0.0176, "step": 21902 }, { "epoch": 15.392129304286719, "grad_norm": 0.11309178173542023, "learning_rate": 2.3076130241274302e-05, "loss": 0.022, "step": 21903 }, { "epoch": 15.392832044975403, "grad_norm": 0.12007950991392136, "learning_rate": 2.3075661747481846e-05, "loss": 0.0146, "step": 21904 }, { "epoch": 15.39353478566409, "grad_norm": 0.04905923083424568, "learning_rate": 2.307519325368939e-05, "loss": 0.0083, "step": 21905 }, { "epoch": 15.394237526352775, "grad_norm": 0.19321970641613007, "learning_rate": 2.307472475989693e-05, "loss": 0.0302, "step": 21906 }, { "epoch": 15.394940267041461, "grad_norm": 0.13362325727939606, "learning_rate": 2.3074256266104474e-05, "loss": 0.0151, "step": 21907 }, { "epoch": 15.395643007730147, "grad_norm": 0.14737297594547272, "learning_rate": 2.3073787772312017e-05, "loss": 0.0221, "step": 21908 }, { "epoch": 15.396345748418833, "grad_norm": 0.16647127270698547, "learning_rate": 2.307331927851956e-05, "loss": 0.015, "step": 21909 }, { "epoch": 15.397048489107519, "grad_norm": 0.6267066597938538, "learning_rate": 2.30728507847271e-05, "loss": 0.0183, "step": 21910 }, { "epoch": 15.397751229796205, "grad_norm": 0.24137909710407257, "learning_rate": 2.3072382290934645e-05, "loss": 0.0298, "step": 21911 }, { "epoch": 15.39845397048489, "grad_norm": 0.20750722289085388, "learning_rate": 2.307191379714219e-05, "loss": 0.0355, "step": 21912 }, { "epoch": 15.399156711173577, "grad_norm": 0.08361097425222397, "learning_rate": 2.3071445303349733e-05, "loss": 0.0136, "step": 21913 }, { "epoch": 15.399859451862262, "grad_norm": 0.4337790012359619, "learning_rate": 2.3070976809557276e-05, "loss": 0.0353, "step": 21914 }, { "epoch": 15.400562192550948, "grad_norm": 0.28436046838760376, "learning_rate": 2.3070508315764817e-05, "loss": 0.0425, "step": 21915 }, { "epoch": 15.401264933239634, "grad_norm": 0.5543561577796936, "learning_rate": 2.307003982197236e-05, "loss": 0.0806, "step": 21916 }, { "epoch": 15.40196767392832, "grad_norm": 0.6559550166130066, "learning_rate": 2.3069571328179904e-05, "loss": 0.0947, "step": 21917 }, { "epoch": 15.402670414617006, "grad_norm": 0.4180174469947815, "learning_rate": 2.3069102834387448e-05, "loss": 0.1185, "step": 21918 }, { "epoch": 15.403373155305692, "grad_norm": 1.0030614137649536, "learning_rate": 2.3068634340594985e-05, "loss": 0.124, "step": 21919 }, { "epoch": 15.404075895994378, "grad_norm": 0.7590761780738831, "learning_rate": 2.306816584680253e-05, "loss": 0.1594, "step": 21920 }, { "epoch": 15.404778636683064, "grad_norm": 0.16343903541564941, "learning_rate": 2.3067697353010072e-05, "loss": 0.0535, "step": 21921 }, { "epoch": 15.40548137737175, "grad_norm": 0.14197242259979248, "learning_rate": 2.3067228859217616e-05, "loss": 0.0321, "step": 21922 }, { "epoch": 15.406184118060436, "grad_norm": 0.18964573740959167, "learning_rate": 2.3066760365425156e-05, "loss": 0.0186, "step": 21923 }, { "epoch": 15.406886858749122, "grad_norm": 0.4340343177318573, "learning_rate": 2.30662918716327e-05, "loss": 0.0198, "step": 21924 }, { "epoch": 15.407589599437808, "grad_norm": 0.12274917215108871, "learning_rate": 2.3065823377840244e-05, "loss": 0.0146, "step": 21925 }, { "epoch": 15.408292340126494, "grad_norm": 0.15708012878894806, "learning_rate": 2.3065354884047788e-05, "loss": 0.0105, "step": 21926 }, { "epoch": 15.40899508081518, "grad_norm": 0.15794821083545685, "learning_rate": 2.306488639025533e-05, "loss": 0.0188, "step": 21927 }, { "epoch": 15.409697821503865, "grad_norm": 0.5967220067977905, "learning_rate": 2.306441789646287e-05, "loss": 0.0147, "step": 21928 }, { "epoch": 15.410400562192551, "grad_norm": 0.11677131801843643, "learning_rate": 2.3063949402670415e-05, "loss": 0.0187, "step": 21929 }, { "epoch": 15.411103302881237, "grad_norm": 0.15194295346736908, "learning_rate": 2.306348090887796e-05, "loss": 0.0174, "step": 21930 }, { "epoch": 15.411806043569923, "grad_norm": 0.18143989145755768, "learning_rate": 2.3063012415085503e-05, "loss": 0.0151, "step": 21931 }, { "epoch": 15.412508784258609, "grad_norm": 0.17899581789970398, "learning_rate": 2.3062543921293043e-05, "loss": 0.0179, "step": 21932 }, { "epoch": 15.413211524947295, "grad_norm": 0.17270676791667938, "learning_rate": 2.3062075427500587e-05, "loss": 0.0234, "step": 21933 }, { "epoch": 15.41391426563598, "grad_norm": 0.1757778376340866, "learning_rate": 2.306160693370813e-05, "loss": 0.0198, "step": 21934 }, { "epoch": 15.414617006324667, "grad_norm": 0.16677339375019073, "learning_rate": 2.3061138439915674e-05, "loss": 0.024, "step": 21935 }, { "epoch": 15.415319747013353, "grad_norm": 0.23022471368312836, "learning_rate": 2.306066994612321e-05, "loss": 0.0241, "step": 21936 }, { "epoch": 15.416022487702039, "grad_norm": 0.10927430540323257, "learning_rate": 2.3060201452330755e-05, "loss": 0.0168, "step": 21937 }, { "epoch": 15.416725228390725, "grad_norm": 0.4287322461605072, "learning_rate": 2.30597329585383e-05, "loss": 0.0265, "step": 21938 }, { "epoch": 15.41742796907941, "grad_norm": 0.08970166742801666, "learning_rate": 2.3059264464745842e-05, "loss": 0.0215, "step": 21939 }, { "epoch": 15.418130709768096, "grad_norm": 0.23982776701450348, "learning_rate": 2.3058795970953386e-05, "loss": 0.0339, "step": 21940 }, { "epoch": 15.41883345045678, "grad_norm": 1.1185771226882935, "learning_rate": 2.3058327477160927e-05, "loss": 0.0553, "step": 21941 }, { "epoch": 15.419536191145466, "grad_norm": 0.5837534070014954, "learning_rate": 2.305785898336847e-05, "loss": 0.101, "step": 21942 }, { "epoch": 15.420238931834152, "grad_norm": 0.6146341562271118, "learning_rate": 2.3057390489576014e-05, "loss": 0.1159, "step": 21943 }, { "epoch": 15.420941672522838, "grad_norm": 0.6277382373809814, "learning_rate": 2.3056921995783558e-05, "loss": 0.1445, "step": 21944 }, { "epoch": 15.421644413211524, "grad_norm": 0.48787808418273926, "learning_rate": 2.3056453501991098e-05, "loss": 0.1559, "step": 21945 }, { "epoch": 15.42234715390021, "grad_norm": 0.22822751104831696, "learning_rate": 2.3055985008198642e-05, "loss": 0.0602, "step": 21946 }, { "epoch": 15.423049894588896, "grad_norm": 0.08903741836547852, "learning_rate": 2.3055516514406185e-05, "loss": 0.0154, "step": 21947 }, { "epoch": 15.423752635277582, "grad_norm": 0.12630002200603485, "learning_rate": 2.305504802061373e-05, "loss": 0.0298, "step": 21948 }, { "epoch": 15.424455375966268, "grad_norm": 0.11841589957475662, "learning_rate": 2.305457952682127e-05, "loss": 0.0119, "step": 21949 }, { "epoch": 15.425158116654954, "grad_norm": 0.09199385344982147, "learning_rate": 2.3054111033028813e-05, "loss": 0.0131, "step": 21950 }, { "epoch": 15.42586085734364, "grad_norm": 0.1405792087316513, "learning_rate": 2.3053642539236357e-05, "loss": 0.007, "step": 21951 }, { "epoch": 15.426563598032326, "grad_norm": 0.11485636979341507, "learning_rate": 2.30531740454439e-05, "loss": 0.0141, "step": 21952 }, { "epoch": 15.427266338721012, "grad_norm": 0.09989002346992493, "learning_rate": 2.305270555165144e-05, "loss": 0.0134, "step": 21953 }, { "epoch": 15.427969079409698, "grad_norm": 0.13692782819271088, "learning_rate": 2.305223705785898e-05, "loss": 0.0219, "step": 21954 }, { "epoch": 15.428671820098383, "grad_norm": 0.100208580493927, "learning_rate": 2.3051768564066525e-05, "loss": 0.0107, "step": 21955 }, { "epoch": 15.42937456078707, "grad_norm": 0.26748085021972656, "learning_rate": 2.305130007027407e-05, "loss": 0.0239, "step": 21956 }, { "epoch": 15.430077301475755, "grad_norm": 0.05930578336119652, "learning_rate": 2.3050831576481613e-05, "loss": 0.0073, "step": 21957 }, { "epoch": 15.430780042164441, "grad_norm": 0.1274440437555313, "learning_rate": 2.3050363082689153e-05, "loss": 0.0162, "step": 21958 }, { "epoch": 15.431482782853127, "grad_norm": 0.23153451085090637, "learning_rate": 2.3049894588896697e-05, "loss": 0.0128, "step": 21959 }, { "epoch": 15.432185523541813, "grad_norm": 0.14020098745822906, "learning_rate": 2.304942609510424e-05, "loss": 0.0237, "step": 21960 }, { "epoch": 15.432888264230499, "grad_norm": 0.2064196914434433, "learning_rate": 2.3048957601311784e-05, "loss": 0.0234, "step": 21961 }, { "epoch": 15.433591004919185, "grad_norm": 0.1333177089691162, "learning_rate": 2.3048489107519324e-05, "loss": 0.0264, "step": 21962 }, { "epoch": 15.43429374560787, "grad_norm": 0.36969777941703796, "learning_rate": 2.3048020613726868e-05, "loss": 0.0119, "step": 21963 }, { "epoch": 15.434996486296557, "grad_norm": 0.43483439087867737, "learning_rate": 2.3047552119934412e-05, "loss": 0.0474, "step": 21964 }, { "epoch": 15.435699226985243, "grad_norm": 0.21602274477481842, "learning_rate": 2.3047083626141956e-05, "loss": 0.0604, "step": 21965 }, { "epoch": 15.436401967673929, "grad_norm": 0.24344561994075775, "learning_rate": 2.30466151323495e-05, "loss": 0.0698, "step": 21966 }, { "epoch": 15.437104708362615, "grad_norm": 0.5091447830200195, "learning_rate": 2.304614663855704e-05, "loss": 0.0851, "step": 21967 }, { "epoch": 15.4378074490513, "grad_norm": 0.4680374562740326, "learning_rate": 2.3045678144764583e-05, "loss": 0.15, "step": 21968 }, { "epoch": 15.438510189739986, "grad_norm": 0.5161880850791931, "learning_rate": 2.3045209650972127e-05, "loss": 0.133, "step": 21969 }, { "epoch": 15.439212930428672, "grad_norm": 0.8232362270355225, "learning_rate": 2.304474115717967e-05, "loss": 0.1593, "step": 21970 }, { "epoch": 15.439915671117358, "grad_norm": 0.23246538639068604, "learning_rate": 2.3044272663387208e-05, "loss": 0.0587, "step": 21971 }, { "epoch": 15.440618411806044, "grad_norm": 0.13103187084197998, "learning_rate": 2.304380416959475e-05, "loss": 0.0263, "step": 21972 }, { "epoch": 15.44132115249473, "grad_norm": 0.06038490682840347, "learning_rate": 2.3043335675802295e-05, "loss": 0.0108, "step": 21973 }, { "epoch": 15.442023893183416, "grad_norm": 0.11346457153558731, "learning_rate": 2.304286718200984e-05, "loss": 0.0172, "step": 21974 }, { "epoch": 15.442726633872102, "grad_norm": 0.12011868506669998, "learning_rate": 2.304239868821738e-05, "loss": 0.013, "step": 21975 }, { "epoch": 15.443429374560788, "grad_norm": 0.07600774616003036, "learning_rate": 2.3041930194424923e-05, "loss": 0.0117, "step": 21976 }, { "epoch": 15.444132115249474, "grad_norm": 0.09737632423639297, "learning_rate": 2.3041461700632467e-05, "loss": 0.0201, "step": 21977 }, { "epoch": 15.44483485593816, "grad_norm": 0.19350218772888184, "learning_rate": 2.304099320684001e-05, "loss": 0.0114, "step": 21978 }, { "epoch": 15.445537596626846, "grad_norm": 0.09053528308868408, "learning_rate": 2.3040524713047554e-05, "loss": 0.0174, "step": 21979 }, { "epoch": 15.44624033731553, "grad_norm": 0.40547218918800354, "learning_rate": 2.3040056219255095e-05, "loss": 0.0183, "step": 21980 }, { "epoch": 15.446943078004216, "grad_norm": 0.1316988468170166, "learning_rate": 2.3039587725462638e-05, "loss": 0.0158, "step": 21981 }, { "epoch": 15.447645818692902, "grad_norm": 0.11417914927005768, "learning_rate": 2.3039119231670182e-05, "loss": 0.0084, "step": 21982 }, { "epoch": 15.448348559381587, "grad_norm": 0.4248321056365967, "learning_rate": 2.3038650737877726e-05, "loss": 0.0169, "step": 21983 }, { "epoch": 15.449051300070273, "grad_norm": 0.2595472037792206, "learning_rate": 2.3038182244085266e-05, "loss": 0.0124, "step": 21984 }, { "epoch": 15.44975404075896, "grad_norm": 0.12851634621620178, "learning_rate": 2.303771375029281e-05, "loss": 0.0178, "step": 21985 }, { "epoch": 15.450456781447645, "grad_norm": 0.3344787359237671, "learning_rate": 2.3037245256500353e-05, "loss": 0.0306, "step": 21986 }, { "epoch": 15.451159522136331, "grad_norm": 0.268375962972641, "learning_rate": 2.3036776762707897e-05, "loss": 0.0092, "step": 21987 }, { "epoch": 15.451862262825017, "grad_norm": 0.14249053597450256, "learning_rate": 2.3036308268915434e-05, "loss": 0.029, "step": 21988 }, { "epoch": 15.452565003513703, "grad_norm": 0.18395978212356567, "learning_rate": 2.3035839775122978e-05, "loss": 0.0499, "step": 21989 }, { "epoch": 15.453267744202389, "grad_norm": 0.1935296505689621, "learning_rate": 2.303537128133052e-05, "loss": 0.0382, "step": 21990 }, { "epoch": 15.453970484891075, "grad_norm": 0.33329999446868896, "learning_rate": 2.3034902787538065e-05, "loss": 0.063, "step": 21991 }, { "epoch": 15.45467322557976, "grad_norm": 0.3787148594856262, "learning_rate": 2.303443429374561e-05, "loss": 0.0904, "step": 21992 }, { "epoch": 15.455375966268447, "grad_norm": 0.7759064435958862, "learning_rate": 2.303396579995315e-05, "loss": 0.1196, "step": 21993 }, { "epoch": 15.456078706957133, "grad_norm": 2.3593344688415527, "learning_rate": 2.3033497306160693e-05, "loss": 0.164, "step": 21994 }, { "epoch": 15.456781447645819, "grad_norm": 1.499649167060852, "learning_rate": 2.3033028812368237e-05, "loss": 0.2039, "step": 21995 }, { "epoch": 15.457484188334504, "grad_norm": 0.16385012865066528, "learning_rate": 2.303256031857578e-05, "loss": 0.0603, "step": 21996 }, { "epoch": 15.45818692902319, "grad_norm": 0.20630183815956116, "learning_rate": 2.303209182478332e-05, "loss": 0.0237, "step": 21997 }, { "epoch": 15.458889669711876, "grad_norm": 0.09823406487703323, "learning_rate": 2.3031623330990865e-05, "loss": 0.0217, "step": 21998 }, { "epoch": 15.459592410400562, "grad_norm": 0.14190173149108887, "learning_rate": 2.303115483719841e-05, "loss": 0.0126, "step": 21999 }, { "epoch": 15.460295151089248, "grad_norm": 0.05911704897880554, "learning_rate": 2.3030686343405952e-05, "loss": 0.0085, "step": 22000 }, { "epoch": 15.460295151089248, "eval_cer": 0.192857844260417, "eval_loss": 0.2595928907394409, "eval_runtime": 18.248, "eval_samples_per_second": 248.684, "eval_steps_per_second": 0.822, "eval_wer": 0.34180609480182117, "step": 22000 }, { "epoch": 15.460295151089248, "step": 22000, "total_flos": 8.390221112256642e+20, "train_loss": 0.14257890296495646, "train_runtime": 52320.6877, "train_samples_per_second": 434.961, "train_steps_per_second": 1.36 } ], "logging_steps": 1.0, "max_steps": 71150, "num_input_tokens_seen": 0, "num_train_epochs": 50, "save_steps": 1000, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 5, "early_stopping_threshold": 0.0 }, "attributes": { "early_stopping_patience_counter": 5 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 8.390221112256642e+20, "train_batch_size": 160, "trial_name": null, "trial_params": null }