diff --git "a/checkpoint-1225/trainer_state.json" "b/checkpoint-1225/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-1225/trainer_state.json" @@ -0,0 +1,8608 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.981670061099797, + "eval_steps": 1.0, + "global_step": 1225, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.004073319755600814, + "grad_norm": 34.300819396972656, + "learning_rate": 8.130081300813008e-09, + "loss": 1.59619802236557, + "step": 1 + }, + { + "epoch": 0.008146639511201629, + "grad_norm": 30.720197677612305, + "learning_rate": 1.6260162601626016e-08, + "loss": 1.468272864818573, + "step": 2 + }, + { + "epoch": 0.012219959266802444, + "grad_norm": 30.16754722595215, + "learning_rate": 2.4390243902439023e-08, + "loss": 1.3843095302581787, + "step": 3 + }, + { + "epoch": 0.016293279022403257, + "grad_norm": 38.58047103881836, + "learning_rate": 3.252032520325203e-08, + "loss": 1.7031245231628418, + "step": 4 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 30.89760971069336, + "learning_rate": 4.065040650406504e-08, + "loss": 1.4844104647636414, + "step": 5 + }, + { + "epoch": 0.024439918533604887, + "grad_norm": 34.434993743896484, + "learning_rate": 4.878048780487805e-08, + "loss": 1.574910283088684, + "step": 6 + }, + { + "epoch": 0.028513238289205704, + "grad_norm": 32.540470123291016, + "learning_rate": 5.6910569105691055e-08, + "loss": 1.4606674909591675, + "step": 7 + }, + { + "epoch": 0.032586558044806514, + "grad_norm": 36.41299819946289, + "learning_rate": 6.504065040650406e-08, + "loss": 1.553576111793518, + "step": 8 + }, + { + "epoch": 0.03665987780040733, + "grad_norm": 34.50511932373047, + "learning_rate": 7.317073170731706e-08, + "loss": 1.3344553709030151, + "step": 9 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 27.898704528808594, + "learning_rate": 8.130081300813008e-08, + "loss": 1.3406395316123962, + "step": 10 + }, + { + "epoch": 0.04480651731160896, + "grad_norm": 29.29271125793457, + "learning_rate": 8.943089430894309e-08, + "loss": 1.4415303468704224, + "step": 11 + }, + { + "epoch": 0.048879837067209775, + "grad_norm": 28.2354736328125, + "learning_rate": 9.75609756097561e-08, + "loss": 1.2696096301078796, + "step": 12 + }, + { + "epoch": 0.05295315682281059, + "grad_norm": 35.44163131713867, + "learning_rate": 1.0569105691056911e-07, + "loss": 1.598312497138977, + "step": 13 + }, + { + "epoch": 0.05702647657841141, + "grad_norm": 26.94402313232422, + "learning_rate": 1.1382113821138211e-07, + "loss": 1.3497812747955322, + "step": 14 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 37.78248977661133, + "learning_rate": 1.219512195121951e-07, + "loss": 1.5689660906791687, + "step": 15 + }, + { + "epoch": 0.06517311608961303, + "grad_norm": 31.73078155517578, + "learning_rate": 1.3008130081300813e-07, + "loss": 1.525648295879364, + "step": 16 + }, + { + "epoch": 0.06924643584521385, + "grad_norm": 27.77250862121582, + "learning_rate": 1.3821138211382114e-07, + "loss": 1.304672360420227, + "step": 17 + }, + { + "epoch": 0.07331975560081466, + "grad_norm": 28.092498779296875, + "learning_rate": 1.4634146341463413e-07, + "loss": 1.346445381641388, + "step": 18 + }, + { + "epoch": 0.07739307535641547, + "grad_norm": 30.995866775512695, + "learning_rate": 1.5447154471544717e-07, + "loss": 1.447025179862976, + "step": 19 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 28.858421325683594, + "learning_rate": 1.6260162601626016e-07, + "loss": 1.3801668882369995, + "step": 20 + }, + { + "epoch": 0.0855397148676171, + "grad_norm": 31.91228485107422, + "learning_rate": 1.7073170731707317e-07, + "loss": 1.4577875137329102, + "step": 21 + }, + { + "epoch": 0.08961303462321792, + "grad_norm": 31.215259552001953, + "learning_rate": 1.7886178861788619e-07, + "loss": 1.4091373682022095, + "step": 22 + }, + { + "epoch": 0.09368635437881874, + "grad_norm": 30.24734115600586, + "learning_rate": 1.8699186991869917e-07, + "loss": 1.4649581909179688, + "step": 23 + }, + { + "epoch": 0.09775967413441955, + "grad_norm": 31.560291290283203, + "learning_rate": 1.951219512195122e-07, + "loss": 1.5308585166931152, + "step": 24 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 27.27391242980957, + "learning_rate": 2.032520325203252e-07, + "loss": 1.5144553780555725, + "step": 25 + }, + { + "epoch": 0.10590631364562118, + "grad_norm": 29.813785552978516, + "learning_rate": 2.1138211382113822e-07, + "loss": 1.519466757774353, + "step": 26 + }, + { + "epoch": 0.109979633401222, + "grad_norm": 24.201751708984375, + "learning_rate": 2.195121951219512e-07, + "loss": 1.3116011023521423, + "step": 27 + }, + { + "epoch": 0.11405295315682282, + "grad_norm": 27.95865249633789, + "learning_rate": 2.2764227642276422e-07, + "loss": 1.4637184143066406, + "step": 28 + }, + { + "epoch": 0.11812627291242363, + "grad_norm": 26.65915870666504, + "learning_rate": 2.3577235772357723e-07, + "loss": 1.4885194301605225, + "step": 29 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 27.386289596557617, + "learning_rate": 2.439024390243902e-07, + "loss": 1.3836334347724915, + "step": 30 + }, + { + "epoch": 0.12627291242362526, + "grad_norm": 25.87419319152832, + "learning_rate": 2.520325203252032e-07, + "loss": 1.3642336130142212, + "step": 31 + }, + { + "epoch": 0.13034623217922606, + "grad_norm": 26.620105743408203, + "learning_rate": 2.6016260162601625e-07, + "loss": 1.3461121916770935, + "step": 32 + }, + { + "epoch": 0.13441955193482688, + "grad_norm": 22.665058135986328, + "learning_rate": 2.682926829268293e-07, + "loss": 1.2577590942382812, + "step": 33 + }, + { + "epoch": 0.1384928716904277, + "grad_norm": 23.679920196533203, + "learning_rate": 2.764227642276423e-07, + "loss": 1.2572017908096313, + "step": 34 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 25.136371612548828, + "learning_rate": 2.8455284552845527e-07, + "loss": 1.2670851349830627, + "step": 35 + }, + { + "epoch": 0.14663951120162932, + "grad_norm": 21.567337036132812, + "learning_rate": 2.9268292682926825e-07, + "loss": 1.242683231830597, + "step": 36 + }, + { + "epoch": 0.15071283095723015, + "grad_norm": 20.61647605895996, + "learning_rate": 3.008130081300813e-07, + "loss": 1.279579222202301, + "step": 37 + }, + { + "epoch": 0.15478615071283094, + "grad_norm": 20.656513214111328, + "learning_rate": 3.0894308943089434e-07, + "loss": 1.2040475606918335, + "step": 38 + }, + { + "epoch": 0.15885947046843177, + "grad_norm": 22.86530876159668, + "learning_rate": 3.170731707317073e-07, + "loss": 1.2522715330123901, + "step": 39 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 20.22757911682129, + "learning_rate": 3.252032520325203e-07, + "loss": 1.2012774348258972, + "step": 40 + }, + { + "epoch": 0.1670061099796334, + "grad_norm": 23.09739875793457, + "learning_rate": 3.333333333333333e-07, + "loss": 1.2088268399238586, + "step": 41 + }, + { + "epoch": 0.1710794297352342, + "grad_norm": 22.845685958862305, + "learning_rate": 3.4146341463414634e-07, + "loss": 1.0982880592346191, + "step": 42 + }, + { + "epoch": 0.17515274949083504, + "grad_norm": 19.80814552307129, + "learning_rate": 3.4959349593495933e-07, + "loss": 1.1271469593048096, + "step": 43 + }, + { + "epoch": 0.17922606924643583, + "grad_norm": 20.553686141967773, + "learning_rate": 3.5772357723577237e-07, + "loss": 1.0008204579353333, + "step": 44 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 16.66282844543457, + "learning_rate": 3.6585365853658536e-07, + "loss": 0.9251897931098938, + "step": 45 + }, + { + "epoch": 0.18737270875763748, + "grad_norm": 15.797308921813965, + "learning_rate": 3.7398373983739835e-07, + "loss": 1.0191328525543213, + "step": 46 + }, + { + "epoch": 0.19144602851323828, + "grad_norm": 13.579208374023438, + "learning_rate": 3.821138211382114e-07, + "loss": 0.774791806936264, + "step": 47 + }, + { + "epoch": 0.1955193482688391, + "grad_norm": 14.556002616882324, + "learning_rate": 3.902439024390244e-07, + "loss": 1.0026790797710419, + "step": 48 + }, + { + "epoch": 0.19959266802443992, + "grad_norm": 14.489509582519531, + "learning_rate": 3.9837398373983736e-07, + "loss": 0.9430837631225586, + "step": 49 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 12.495223999023438, + "learning_rate": 4.065040650406504e-07, + "loss": 0.8999880254268646, + "step": 50 + }, + { + "epoch": 0.20773930753564154, + "grad_norm": 11.441575050354004, + "learning_rate": 4.146341463414634e-07, + "loss": 0.8320233225822449, + "step": 51 + }, + { + "epoch": 0.21181262729124237, + "grad_norm": 10.894216537475586, + "learning_rate": 4.2276422764227643e-07, + "loss": 0.8139239549636841, + "step": 52 + }, + { + "epoch": 0.2158859470468432, + "grad_norm": 10.404220581054688, + "learning_rate": 4.308943089430894e-07, + "loss": 0.8323288261890411, + "step": 53 + }, + { + "epoch": 0.219959266802444, + "grad_norm": 10.463072776794434, + "learning_rate": 4.390243902439024e-07, + "loss": 0.882573276758194, + "step": 54 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 10.669075012207031, + "learning_rate": 4.471544715447154e-07, + "loss": 0.749780923128128, + "step": 55 + }, + { + "epoch": 0.22810590631364563, + "grad_norm": 10.453638076782227, + "learning_rate": 4.5528455284552844e-07, + "loss": 0.7727148830890656, + "step": 56 + }, + { + "epoch": 0.23217922606924643, + "grad_norm": 11.427080154418945, + "learning_rate": 4.634146341463415e-07, + "loss": 0.8585084676742554, + "step": 57 + }, + { + "epoch": 0.23625254582484725, + "grad_norm": 8.558117866516113, + "learning_rate": 4.7154471544715447e-07, + "loss": 0.7314337491989136, + "step": 58 + }, + { + "epoch": 0.24032586558044808, + "grad_norm": 9.031648635864258, + "learning_rate": 4.796747967479675e-07, + "loss": 0.701579749584198, + "step": 59 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 8.817708969116211, + "learning_rate": 4.878048780487804e-07, + "loss": 0.7815204560756683, + "step": 60 + }, + { + "epoch": 0.2484725050916497, + "grad_norm": 8.00804615020752, + "learning_rate": 4.959349593495934e-07, + "loss": 0.655106246471405, + "step": 61 + }, + { + "epoch": 0.2525458248472505, + "grad_norm": 6.538842678070068, + "learning_rate": 5.040650406504064e-07, + "loss": 0.6697916388511658, + "step": 62 + }, + { + "epoch": 0.25661914460285135, + "grad_norm": 7.5446553230285645, + "learning_rate": 5.121951219512195e-07, + "loss": 0.7426944077014923, + "step": 63 + }, + { + "epoch": 0.2606924643584521, + "grad_norm": 6.402474403381348, + "learning_rate": 5.203252032520325e-07, + "loss": 0.6401277780532837, + "step": 64 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 7.257569313049316, + "learning_rate": 5.284552845528455e-07, + "loss": 0.6731106042861938, + "step": 65 + }, + { + "epoch": 0.26883910386965376, + "grad_norm": 6.263636589050293, + "learning_rate": 5.365853658536586e-07, + "loss": 0.5806022882461548, + "step": 66 + }, + { + "epoch": 0.2729124236252546, + "grad_norm": 5.273800849914551, + "learning_rate": 5.447154471544715e-07, + "loss": 0.5338439792394638, + "step": 67 + }, + { + "epoch": 0.2769857433808554, + "grad_norm": 5.2786149978637695, + "learning_rate": 5.528455284552846e-07, + "loss": 0.5390533208847046, + "step": 68 + }, + { + "epoch": 0.28105906313645623, + "grad_norm": 4.901702404022217, + "learning_rate": 5.609756097560975e-07, + "loss": 0.5899032056331635, + "step": 69 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 4.853933811187744, + "learning_rate": 5.691056910569105e-07, + "loss": 0.5600310862064362, + "step": 70 + }, + { + "epoch": 0.2892057026476578, + "grad_norm": 4.680273532867432, + "learning_rate": 5.772357723577236e-07, + "loss": 0.5319355428218842, + "step": 71 + }, + { + "epoch": 0.29327902240325865, + "grad_norm": 3.7406885623931885, + "learning_rate": 5.853658536585365e-07, + "loss": 0.508156955242157, + "step": 72 + }, + { + "epoch": 0.2973523421588595, + "grad_norm": 4.389779567718506, + "learning_rate": 5.934959349593496e-07, + "loss": 0.49855048954486847, + "step": 73 + }, + { + "epoch": 0.3014256619144603, + "grad_norm": 4.23866081237793, + "learning_rate": 6.016260162601626e-07, + "loss": 0.5242476612329483, + "step": 74 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 4.1824951171875, + "learning_rate": 6.097560975609756e-07, + "loss": 0.532037615776062, + "step": 75 + }, + { + "epoch": 0.3095723014256619, + "grad_norm": 3.7223150730133057, + "learning_rate": 6.178861788617887e-07, + "loss": 0.46959882974624634, + "step": 76 + }, + { + "epoch": 0.3136456211812627, + "grad_norm": 3.545388698577881, + "learning_rate": 6.260162601626016e-07, + "loss": 0.4825982600450516, + "step": 77 + }, + { + "epoch": 0.31771894093686354, + "grad_norm": 3.6351099014282227, + "learning_rate": 6.341463414634146e-07, + "loss": 0.5095209777355194, + "step": 78 + }, + { + "epoch": 0.32179226069246436, + "grad_norm": 3.243072271347046, + "learning_rate": 6.422764227642276e-07, + "loss": 0.4842926263809204, + "step": 79 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 3.5646300315856934, + "learning_rate": 6.504065040650406e-07, + "loss": 0.4908552020788193, + "step": 80 + }, + { + "epoch": 0.329938900203666, + "grad_norm": 3.5380759239196777, + "learning_rate": 6.585365853658536e-07, + "loss": 0.4536065459251404, + "step": 81 + }, + { + "epoch": 0.3340122199592668, + "grad_norm": 3.128525495529175, + "learning_rate": 6.666666666666666e-07, + "loss": 0.47657161951065063, + "step": 82 + }, + { + "epoch": 0.3380855397148676, + "grad_norm": 3.3621485233306885, + "learning_rate": 6.747967479674797e-07, + "loss": 0.43791596591472626, + "step": 83 + }, + { + "epoch": 0.3421588594704684, + "grad_norm": 3.39066219329834, + "learning_rate": 6.829268292682927e-07, + "loss": 0.42947711050510406, + "step": 84 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 3.7795698642730713, + "learning_rate": 6.910569105691057e-07, + "loss": 0.4219910502433777, + "step": 85 + }, + { + "epoch": 0.35030549898167007, + "grad_norm": 3.633206367492676, + "learning_rate": 6.991869918699187e-07, + "loss": 0.4253977984189987, + "step": 86 + }, + { + "epoch": 0.3543788187372709, + "grad_norm": 3.6160175800323486, + "learning_rate": 7.073170731707316e-07, + "loss": 0.449339896440506, + "step": 87 + }, + { + "epoch": 0.35845213849287166, + "grad_norm": 3.30557918548584, + "learning_rate": 7.154471544715447e-07, + "loss": 0.45001736283302307, + "step": 88 + }, + { + "epoch": 0.3625254582484725, + "grad_norm": 3.1727640628814697, + "learning_rate": 7.235772357723577e-07, + "loss": 0.4165496975183487, + "step": 89 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 3.073976516723633, + "learning_rate": 7.317073170731707e-07, + "loss": 0.4443822205066681, + "step": 90 + }, + { + "epoch": 0.37067209775967414, + "grad_norm": 3.129105567932129, + "learning_rate": 7.398373983739837e-07, + "loss": 0.4265598952770233, + "step": 91 + }, + { + "epoch": 0.37474541751527496, + "grad_norm": 3.1485190391540527, + "learning_rate": 7.479674796747967e-07, + "loss": 0.3882734924554825, + "step": 92 + }, + { + "epoch": 0.3788187372708758, + "grad_norm": 3.1610565185546875, + "learning_rate": 7.560975609756097e-07, + "loss": 0.37010858952999115, + "step": 93 + }, + { + "epoch": 0.38289205702647655, + "grad_norm": 3.039264440536499, + "learning_rate": 7.642276422764228e-07, + "loss": 0.400989294052124, + "step": 94 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 2.9321980476379395, + "learning_rate": 7.723577235772358e-07, + "loss": 0.3771343380212784, + "step": 95 + }, + { + "epoch": 0.3910386965376782, + "grad_norm": 2.807072162628174, + "learning_rate": 7.804878048780488e-07, + "loss": 0.4001482129096985, + "step": 96 + }, + { + "epoch": 0.395112016293279, + "grad_norm": 2.8286941051483154, + "learning_rate": 7.886178861788617e-07, + "loss": 0.4234430640935898, + "step": 97 + }, + { + "epoch": 0.39918533604887985, + "grad_norm": 2.9245986938476562, + "learning_rate": 7.967479674796747e-07, + "loss": 0.3854667395353317, + "step": 98 + }, + { + "epoch": 0.40325865580448067, + "grad_norm": 3.015875816345215, + "learning_rate": 8.048780487804878e-07, + "loss": 0.38027653098106384, + "step": 99 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 2.907216787338257, + "learning_rate": 8.130081300813008e-07, + "loss": 0.34937676787376404, + "step": 100 + }, + { + "epoch": 0.41140529531568226, + "grad_norm": 3.131850004196167, + "learning_rate": 8.211382113821138e-07, + "loss": 0.4414845108985901, + "step": 101 + }, + { + "epoch": 0.4154786150712831, + "grad_norm": 2.9019775390625, + "learning_rate": 8.292682926829268e-07, + "loss": 0.3990558981895447, + "step": 102 + }, + { + "epoch": 0.4195519348268839, + "grad_norm": 2.9362523555755615, + "learning_rate": 8.373983739837398e-07, + "loss": 0.41413092613220215, + "step": 103 + }, + { + "epoch": 0.42362525458248473, + "grad_norm": 3.0895473957061768, + "learning_rate": 8.455284552845529e-07, + "loss": 0.3904542028903961, + "step": 104 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 2.9235992431640625, + "learning_rate": 8.536585365853657e-07, + "loss": 0.3995140939950943, + "step": 105 + }, + { + "epoch": 0.4317718940936864, + "grad_norm": 2.919102668762207, + "learning_rate": 8.617886178861788e-07, + "loss": 0.32857778668403625, + "step": 106 + }, + { + "epoch": 0.43584521384928715, + "grad_norm": 2.831698417663574, + "learning_rate": 8.699186991869918e-07, + "loss": 0.3507983237504959, + "step": 107 + }, + { + "epoch": 0.439918533604888, + "grad_norm": 2.952693223953247, + "learning_rate": 8.780487804878048e-07, + "loss": 0.37046514451503754, + "step": 108 + }, + { + "epoch": 0.4439918533604888, + "grad_norm": 3.315002679824829, + "learning_rate": 8.861788617886179e-07, + "loss": 0.391086682677269, + "step": 109 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 2.7241294384002686, + "learning_rate": 8.943089430894308e-07, + "loss": 0.3864188492298126, + "step": 110 + }, + { + "epoch": 0.45213849287169044, + "grad_norm": 2.782064199447632, + "learning_rate": 9.024390243902439e-07, + "loss": 0.38219109177589417, + "step": 111 + }, + { + "epoch": 0.45621181262729127, + "grad_norm": 4.001572132110596, + "learning_rate": 9.105691056910569e-07, + "loss": 0.3784598410129547, + "step": 112 + }, + { + "epoch": 0.46028513238289204, + "grad_norm": 2.607434034347534, + "learning_rate": 9.186991869918699e-07, + "loss": 0.3763512521982193, + "step": 113 + }, + { + "epoch": 0.46435845213849286, + "grad_norm": 2.97188138961792, + "learning_rate": 9.26829268292683e-07, + "loss": 0.36788034439086914, + "step": 114 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 2.9631524085998535, + "learning_rate": 9.349593495934958e-07, + "loss": 0.3696867823600769, + "step": 115 + }, + { + "epoch": 0.4725050916496945, + "grad_norm": 2.5895049571990967, + "learning_rate": 9.430894308943089e-07, + "loss": 0.3349902927875519, + "step": 116 + }, + { + "epoch": 0.47657841140529533, + "grad_norm": 2.600832462310791, + "learning_rate": 9.512195121951218e-07, + "loss": 0.34966227412223816, + "step": 117 + }, + { + "epoch": 0.48065173116089616, + "grad_norm": 3.0639443397521973, + "learning_rate": 9.59349593495935e-07, + "loss": 0.38310858607292175, + "step": 118 + }, + { + "epoch": 0.4847250509164969, + "grad_norm": 2.6944706439971924, + "learning_rate": 9.67479674796748e-07, + "loss": 0.3360476493835449, + "step": 119 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 2.8398237228393555, + "learning_rate": 9.756097560975609e-07, + "loss": 0.39176714420318604, + "step": 120 + }, + { + "epoch": 0.49287169042769857, + "grad_norm": 2.8028745651245117, + "learning_rate": 9.83739837398374e-07, + "loss": 0.37909021973609924, + "step": 121 + }, + { + "epoch": 0.4969450101832994, + "grad_norm": 2.6169185638427734, + "learning_rate": 9.918699186991869e-07, + "loss": 0.37069061398506165, + "step": 122 + }, + { + "epoch": 0.5010183299389002, + "grad_norm": 2.572046995162964, + "learning_rate": 1e-06, + "loss": 0.3428824096918106, + "step": 123 + }, + { + "epoch": 0.505091649694501, + "grad_norm": 2.7804417610168457, + "learning_rate": 9.999979682219186e-07, + "loss": 0.3680119812488556, + "step": 124 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 2.5910799503326416, + "learning_rate": 9.999918729041868e-07, + "loss": 0.33467385172843933, + "step": 125 + }, + { + "epoch": 0.5132382892057027, + "grad_norm": 2.8417587280273438, + "learning_rate": 9.999817140963419e-07, + "loss": 0.35100705921649933, + "step": 126 + }, + { + "epoch": 0.5173116089613035, + "grad_norm": 2.905728340148926, + "learning_rate": 9.999674918809457e-07, + "loss": 0.32811686396598816, + "step": 127 + }, + { + "epoch": 0.5213849287169042, + "grad_norm": 2.5878095626831055, + "learning_rate": 9.99949206373584e-07, + "loss": 0.32490645349025726, + "step": 128 + }, + { + "epoch": 0.5254582484725051, + "grad_norm": 2.9762229919433594, + "learning_rate": 9.999268577228648e-07, + "loss": 0.3934018760919571, + "step": 129 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 2.792989492416382, + "learning_rate": 9.99900446110418e-07, + "loss": 0.3315049111843109, + "step": 130 + }, + { + "epoch": 0.5336048879837068, + "grad_norm": 2.6891062259674072, + "learning_rate": 9.998699717508945e-07, + "loss": 0.3097301423549652, + "step": 131 + }, + { + "epoch": 0.5376782077393075, + "grad_norm": 2.92191481590271, + "learning_rate": 9.99835434891962e-07, + "loss": 0.34749817848205566, + "step": 132 + }, + { + "epoch": 0.5417515274949084, + "grad_norm": 2.980543851852417, + "learning_rate": 9.99796835814306e-07, + "loss": 0.3367327153682709, + "step": 133 + }, + { + "epoch": 0.5458248472505092, + "grad_norm": 2.50433611869812, + "learning_rate": 9.99754174831625e-07, + "loss": 0.3090934008359909, + "step": 134 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 2.869647979736328, + "learning_rate": 9.9970745229063e-07, + "loss": 0.35603591799736023, + "step": 135 + }, + { + "epoch": 0.5539714867617108, + "grad_norm": 2.6435837745666504, + "learning_rate": 9.9965666857104e-07, + "loss": 0.3288918733596802, + "step": 136 + }, + { + "epoch": 0.5580448065173116, + "grad_norm": 2.7970142364501953, + "learning_rate": 9.996018240855806e-07, + "loss": 0.3878723680973053, + "step": 137 + }, + { + "epoch": 0.5621181262729125, + "grad_norm": 2.593043327331543, + "learning_rate": 9.995429192799788e-07, + "loss": 0.3534126281738281, + "step": 138 + }, + { + "epoch": 0.5661914460285132, + "grad_norm": 2.8867013454437256, + "learning_rate": 9.994799546329602e-07, + "loss": 0.38061630725860596, + "step": 139 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 2.589017152786255, + "learning_rate": 9.994129306562458e-07, + "loss": 0.37725748121738434, + "step": 140 + }, + { + "epoch": 0.5743380855397149, + "grad_norm": 2.369696617126465, + "learning_rate": 9.993418478945472e-07, + "loss": 0.32034583389759064, + "step": 141 + }, + { + "epoch": 0.5784114052953157, + "grad_norm": 2.6410069465637207, + "learning_rate": 9.992667069255618e-07, + "loss": 0.36017628014087677, + "step": 142 + }, + { + "epoch": 0.5824847250509165, + "grad_norm": 2.597259283065796, + "learning_rate": 9.991875083599688e-07, + "loss": 0.32577911019325256, + "step": 143 + }, + { + "epoch": 0.5865580448065173, + "grad_norm": 2.761859655380249, + "learning_rate": 9.991042528414237e-07, + "loss": 0.33353830873966217, + "step": 144 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 2.7634713649749756, + "learning_rate": 9.990169410465536e-07, + "loss": 0.33604632318019867, + "step": 145 + }, + { + "epoch": 0.594704684317719, + "grad_norm": 2.820897340774536, + "learning_rate": 9.98925573684951e-07, + "loss": 0.3069554716348648, + "step": 146 + }, + { + "epoch": 0.5987780040733197, + "grad_norm": 2.856700897216797, + "learning_rate": 9.98830151499169e-07, + "loss": 0.33896636962890625, + "step": 147 + }, + { + "epoch": 0.6028513238289206, + "grad_norm": 2.9203782081604004, + "learning_rate": 9.987306752647142e-07, + "loss": 0.35070909559726715, + "step": 148 + }, + { + "epoch": 0.6069246435845214, + "grad_norm": 2.679352283477783, + "learning_rate": 9.986271457900414e-07, + "loss": 0.3325359970331192, + "step": 149 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 2.4953606128692627, + "learning_rate": 9.98519563916546e-07, + "loss": 0.32330869138240814, + "step": 150 + }, + { + "epoch": 0.615071283095723, + "grad_norm": 2.618744134902954, + "learning_rate": 9.98407930518558e-07, + "loss": 0.33912393450737, + "step": 151 + }, + { + "epoch": 0.6191446028513238, + "grad_norm": 2.6512296199798584, + "learning_rate": 9.982922465033348e-07, + "loss": 0.3045920431613922, + "step": 152 + }, + { + "epoch": 0.6232179226069247, + "grad_norm": 2.7606050968170166, + "learning_rate": 9.981725128110532e-07, + "loss": 0.32916732132434845, + "step": 153 + }, + { + "epoch": 0.6272912423625254, + "grad_norm": 2.95037841796875, + "learning_rate": 9.980487304148024e-07, + "loss": 0.36757831275463104, + "step": 154 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 2.890489339828491, + "learning_rate": 9.97920900320576e-07, + "loss": 0.36117151379585266, + "step": 155 + }, + { + "epoch": 0.6354378818737271, + "grad_norm": 2.7488858699798584, + "learning_rate": 9.97789023567263e-07, + "loss": 0.35026322305202484, + "step": 156 + }, + { + "epoch": 0.639511201629328, + "grad_norm": 2.5479671955108643, + "learning_rate": 9.976531012266413e-07, + "loss": 0.308156818151474, + "step": 157 + }, + { + "epoch": 0.6435845213849287, + "grad_norm": 2.717344045639038, + "learning_rate": 9.975131344033664e-07, + "loss": 0.29827529191970825, + "step": 158 + }, + { + "epoch": 0.6476578411405295, + "grad_norm": 2.569551467895508, + "learning_rate": 9.973691242349648e-07, + "loss": 0.3232528269290924, + "step": 159 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 3.0013420581817627, + "learning_rate": 9.972210718918233e-07, + "loss": 0.3270832598209381, + "step": 160 + }, + { + "epoch": 0.6558044806517311, + "grad_norm": 2.7339162826538086, + "learning_rate": 9.970689785771798e-07, + "loss": 0.3668155074119568, + "step": 161 + }, + { + "epoch": 0.659877800407332, + "grad_norm": 2.6689724922180176, + "learning_rate": 9.969128455271137e-07, + "loss": 0.32853490114212036, + "step": 162 + }, + { + "epoch": 0.6639511201629328, + "grad_norm": 3.042081117630005, + "learning_rate": 9.967526740105358e-07, + "loss": 0.3487651199102402, + "step": 163 + }, + { + "epoch": 0.6680244399185336, + "grad_norm": 2.4641284942626953, + "learning_rate": 9.965884653291783e-07, + "loss": 0.35704147815704346, + "step": 164 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 2.6836225986480713, + "learning_rate": 9.964202208175833e-07, + "loss": 0.33587950468063354, + "step": 165 + }, + { + "epoch": 0.6761710794297352, + "grad_norm": 2.2905988693237305, + "learning_rate": 9.962479418430932e-07, + "loss": 0.3061918318271637, + "step": 166 + }, + { + "epoch": 0.6802443991853361, + "grad_norm": 2.4772934913635254, + "learning_rate": 9.960716298058381e-07, + "loss": 0.2896444499492645, + "step": 167 + }, + { + "epoch": 0.6843177189409368, + "grad_norm": 2.6987321376800537, + "learning_rate": 9.958912861387258e-07, + "loss": 0.3374595195055008, + "step": 168 + }, + { + "epoch": 0.6883910386965377, + "grad_norm": 2.6165449619293213, + "learning_rate": 9.9570691230743e-07, + "loss": 0.33027225732803345, + "step": 169 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 3.1326680183410645, + "learning_rate": 9.955185098103771e-07, + "loss": 0.3138381540775299, + "step": 170 + }, + { + "epoch": 0.6965376782077393, + "grad_norm": 2.5313732624053955, + "learning_rate": 9.953260801787356e-07, + "loss": 0.31824737787246704, + "step": 171 + }, + { + "epoch": 0.7006109979633401, + "grad_norm": 2.529325008392334, + "learning_rate": 9.951296249764025e-07, + "loss": 0.298155277967453, + "step": 172 + }, + { + "epoch": 0.7046843177189409, + "grad_norm": 2.6821744441986084, + "learning_rate": 9.949291457999916e-07, + "loss": 0.33296874165534973, + "step": 173 + }, + { + "epoch": 0.7087576374745418, + "grad_norm": 2.588157892227173, + "learning_rate": 9.947246442788193e-07, + "loss": 0.31226691603660583, + "step": 174 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 2.7822420597076416, + "learning_rate": 9.945161220748927e-07, + "loss": 0.322743222117424, + "step": 175 + }, + { + "epoch": 0.7169042769857433, + "grad_norm": 2.379702091217041, + "learning_rate": 9.943035808828953e-07, + "loss": 0.3056500107049942, + "step": 176 + }, + { + "epoch": 0.7209775967413442, + "grad_norm": 2.4450721740722656, + "learning_rate": 9.94087022430173e-07, + "loss": 0.3037564754486084, + "step": 177 + }, + { + "epoch": 0.725050916496945, + "grad_norm": 2.5885887145996094, + "learning_rate": 9.938664484767205e-07, + "loss": 0.327587828040123, + "step": 178 + }, + { + "epoch": 0.7291242362525459, + "grad_norm": 2.613290309906006, + "learning_rate": 9.936418608151675e-07, + "loss": 0.33323927223682404, + "step": 179 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 2.6541707515716553, + "learning_rate": 9.93413261270763e-07, + "loss": 0.3316569924354553, + "step": 180 + }, + { + "epoch": 0.7372708757637475, + "grad_norm": 2.646383047103882, + "learning_rate": 9.931806517013612e-07, + "loss": 0.35486292839050293, + "step": 181 + }, + { + "epoch": 0.7413441955193483, + "grad_norm": 2.5270328521728516, + "learning_rate": 9.92944033997406e-07, + "loss": 0.3157142102718353, + "step": 182 + }, + { + "epoch": 0.745417515274949, + "grad_norm": 2.5851869583129883, + "learning_rate": 9.927034100819163e-07, + "loss": 0.3013855814933777, + "step": 183 + }, + { + "epoch": 0.7494908350305499, + "grad_norm": 2.75219988822937, + "learning_rate": 9.924587819104695e-07, + "loss": 0.3420049250125885, + "step": 184 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 2.436596632003784, + "learning_rate": 9.922101514711865e-07, + "loss": 0.3062688261270523, + "step": 185 + }, + { + "epoch": 0.7576374745417516, + "grad_norm": 2.9479236602783203, + "learning_rate": 9.919575207847145e-07, + "loss": 0.31793762743473053, + "step": 186 + }, + { + "epoch": 0.7617107942973523, + "grad_norm": 2.5482208728790283, + "learning_rate": 9.917008919042116e-07, + "loss": 0.3306496888399124, + "step": 187 + }, + { + "epoch": 0.7657841140529531, + "grad_norm": 2.609839677810669, + "learning_rate": 9.914402669153295e-07, + "loss": 0.29324449598789215, + "step": 188 + }, + { + "epoch": 0.769857433808554, + "grad_norm": 2.5740039348602295, + "learning_rate": 9.91175647936197e-07, + "loss": 0.3193310797214508, + "step": 189 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 2.3878629207611084, + "learning_rate": 9.909070371174019e-07, + "loss": 0.3040658235549927, + "step": 190 + }, + { + "epoch": 0.7780040733197556, + "grad_norm": 2.755152463912964, + "learning_rate": 9.906344366419746e-07, + "loss": 0.33930477499961853, + "step": 191 + }, + { + "epoch": 0.7820773930753564, + "grad_norm": 2.58367657661438, + "learning_rate": 9.9035784872537e-07, + "loss": 0.3244568109512329, + "step": 192 + }, + { + "epoch": 0.7861507128309573, + "grad_norm": 2.350712537765503, + "learning_rate": 9.90077275615449e-07, + "loss": 0.2779058516025543, + "step": 193 + }, + { + "epoch": 0.790224032586558, + "grad_norm": 2.7418465614318848, + "learning_rate": 9.897927195924608e-07, + "loss": 0.32641272246837616, + "step": 194 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 2.516510009765625, + "learning_rate": 9.895041829690238e-07, + "loss": 0.3083319664001465, + "step": 195 + }, + { + "epoch": 0.7983706720977597, + "grad_norm": 2.7772316932678223, + "learning_rate": 9.892116680901084e-07, + "loss": 0.30357757210731506, + "step": 196 + }, + { + "epoch": 0.8024439918533605, + "grad_norm": 2.5389041900634766, + "learning_rate": 9.88915177333015e-07, + "loss": 0.30694054067134857, + "step": 197 + }, + { + "epoch": 0.8065173116089613, + "grad_norm": 2.7129383087158203, + "learning_rate": 9.886147131073579e-07, + "loss": 0.3402569591999054, + "step": 198 + }, + { + "epoch": 0.8105906313645621, + "grad_norm": 2.654186248779297, + "learning_rate": 9.883102778550434e-07, + "loss": 0.3343619704246521, + "step": 199 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 2.380168914794922, + "learning_rate": 9.880018740502508e-07, + "loss": 0.3020651191473007, + "step": 200 + }, + { + "epoch": 0.8187372708757638, + "grad_norm": 2.771951198577881, + "learning_rate": 9.876895041994127e-07, + "loss": 0.30565840005874634, + "step": 201 + }, + { + "epoch": 0.8228105906313645, + "grad_norm": 2.4966540336608887, + "learning_rate": 9.873731708411939e-07, + "loss": 0.3085058331489563, + "step": 202 + }, + { + "epoch": 0.8268839103869654, + "grad_norm": 2.5919551849365234, + "learning_rate": 9.870528765464711e-07, + "loss": 0.34540820121765137, + "step": 203 + }, + { + "epoch": 0.8309572301425662, + "grad_norm": 3.0668885707855225, + "learning_rate": 9.867286239183122e-07, + "loss": 0.3307037353515625, + "step": 204 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 2.4281554222106934, + "learning_rate": 9.864004155919544e-07, + "loss": 0.28929875791072845, + "step": 205 + }, + { + "epoch": 0.8391038696537678, + "grad_norm": 2.5561623573303223, + "learning_rate": 9.860682542347838e-07, + "loss": 0.3272414803504944, + "step": 206 + }, + { + "epoch": 0.8431771894093686, + "grad_norm": 2.824591636657715, + "learning_rate": 9.85732142546313e-07, + "loss": 0.3192295432090759, + "step": 207 + }, + { + "epoch": 0.8472505091649695, + "grad_norm": 2.643718719482422, + "learning_rate": 9.853920832581597e-07, + "loss": 0.31284041702747345, + "step": 208 + }, + { + "epoch": 0.8513238289205702, + "grad_norm": 2.6777195930480957, + "learning_rate": 9.850480791340236e-07, + "loss": 0.3136574327945709, + "step": 209 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 2.5229766368865967, + "learning_rate": 9.847001329696652e-07, + "loss": 0.3047819435596466, + "step": 210 + }, + { + "epoch": 0.8594704684317719, + "grad_norm": 2.659447431564331, + "learning_rate": 9.843482475928818e-07, + "loss": 0.3642407953739166, + "step": 211 + }, + { + "epoch": 0.8635437881873728, + "grad_norm": 2.697049379348755, + "learning_rate": 9.839924258634853e-07, + "loss": 0.3134022653102875, + "step": 212 + }, + { + "epoch": 0.8676171079429735, + "grad_norm": 2.629868745803833, + "learning_rate": 9.83632670673279e-07, + "loss": 0.306331992149353, + "step": 213 + }, + { + "epoch": 0.8716904276985743, + "grad_norm": 2.4997003078460693, + "learning_rate": 9.832689849460339e-07, + "loss": 0.3142865002155304, + "step": 214 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 2.826869010925293, + "learning_rate": 9.829013716374647e-07, + "loss": 0.2904099076986313, + "step": 215 + }, + { + "epoch": 0.879837067209776, + "grad_norm": 2.6697499752044678, + "learning_rate": 9.825298337352058e-07, + "loss": 0.29838354885578156, + "step": 216 + }, + { + "epoch": 0.8839103869653768, + "grad_norm": 2.5330023765563965, + "learning_rate": 9.821543742587876e-07, + "loss": 0.3052047789096832, + "step": 217 + }, + { + "epoch": 0.8879837067209776, + "grad_norm": 2.806683301925659, + "learning_rate": 9.817749962596114e-07, + "loss": 0.3121778964996338, + "step": 218 + }, + { + "epoch": 0.8920570264765784, + "grad_norm": 2.718122720718384, + "learning_rate": 9.81391702820925e-07, + "loss": 0.32955022156238556, + "step": 219 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 2.346466541290283, + "learning_rate": 9.81004497057797e-07, + "loss": 0.291049063205719, + "step": 220 + }, + { + "epoch": 0.90020366598778, + "grad_norm": 2.4048361778259277, + "learning_rate": 9.806133821170924e-07, + "loss": 0.30249159038066864, + "step": 221 + }, + { + "epoch": 0.9042769857433809, + "grad_norm": 2.681546688079834, + "learning_rate": 9.80218361177446e-07, + "loss": 0.362154021859169, + "step": 222 + }, + { + "epoch": 0.9083503054989817, + "grad_norm": 2.792266368865967, + "learning_rate": 9.798194374492375e-07, + "loss": 0.28344525396823883, + "step": 223 + }, + { + "epoch": 0.9124236252545825, + "grad_norm": 2.507050037384033, + "learning_rate": 9.794166141745646e-07, + "loss": 0.2935172915458679, + "step": 224 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 2.7160379886627197, + "learning_rate": 9.790098946272177e-07, + "loss": 0.3005199581384659, + "step": 225 + }, + { + "epoch": 0.9205702647657841, + "grad_norm": 2.666494131088257, + "learning_rate": 9.785992821126518e-07, + "loss": 0.30710943043231964, + "step": 226 + }, + { + "epoch": 0.924643584521385, + "grad_norm": 2.699313163757324, + "learning_rate": 9.781847799679615e-07, + "loss": 0.3164513558149338, + "step": 227 + }, + { + "epoch": 0.9287169042769857, + "grad_norm": 2.49406099319458, + "learning_rate": 9.777663915618517e-07, + "loss": 0.3061770647764206, + "step": 228 + }, + { + "epoch": 0.9327902240325866, + "grad_norm": 2.552093029022217, + "learning_rate": 9.773441202946121e-07, + "loss": 0.2973909080028534, + "step": 229 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 2.5773231983184814, + "learning_rate": 9.76917969598089e-07, + "loss": 0.31120532751083374, + "step": 230 + }, + { + "epoch": 0.9409368635437881, + "grad_norm": 2.653515100479126, + "learning_rate": 9.76487942935657e-07, + "loss": 0.3365926146507263, + "step": 231 + }, + { + "epoch": 0.945010183299389, + "grad_norm": 2.670433282852173, + "learning_rate": 9.760540438021907e-07, + "loss": 0.3196941614151001, + "step": 232 + }, + { + "epoch": 0.9490835030549898, + "grad_norm": 2.892035961151123, + "learning_rate": 9.756162757240373e-07, + "loss": 0.33982205390930176, + "step": 233 + }, + { + "epoch": 0.9531568228105907, + "grad_norm": 2.5157856941223145, + "learning_rate": 9.751746422589872e-07, + "loss": 0.2537951096892357, + "step": 234 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 2.6808388233184814, + "learning_rate": 9.747291469962452e-07, + "loss": 0.2846526652574539, + "step": 235 + }, + { + "epoch": 0.9613034623217923, + "grad_norm": 2.451559066772461, + "learning_rate": 9.742797935564011e-07, + "loss": 0.29611095786094666, + "step": 236 + }, + { + "epoch": 0.9653767820773931, + "grad_norm": 2.7313358783721924, + "learning_rate": 9.738265855914012e-07, + "loss": 0.3275996297597885, + "step": 237 + }, + { + "epoch": 0.9694501018329938, + "grad_norm": 2.5593299865722656, + "learning_rate": 9.733695267845171e-07, + "loss": 0.2993656247854233, + "step": 238 + }, + { + "epoch": 0.9735234215885947, + "grad_norm": 2.6013288497924805, + "learning_rate": 9.729086208503173e-07, + "loss": 0.31615155935287476, + "step": 239 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 2.5403575897216797, + "learning_rate": 9.72443871534636e-07, + "loss": 0.2843424677848816, + "step": 240 + }, + { + "epoch": 0.9816700610997964, + "grad_norm": 2.4495410919189453, + "learning_rate": 9.719752826145432e-07, + "loss": 0.2987358868122101, + "step": 241 + }, + { + "epoch": 0.9857433808553971, + "grad_norm": 2.719775676727295, + "learning_rate": 9.715028578983136e-07, + "loss": 0.34320636093616486, + "step": 242 + }, + { + "epoch": 0.9898167006109979, + "grad_norm": 2.7152929306030273, + "learning_rate": 9.71026601225396e-07, + "loss": 0.2937510758638382, + "step": 243 + }, + { + "epoch": 0.9938900203665988, + "grad_norm": 2.4305663108825684, + "learning_rate": 9.705465164663817e-07, + "loss": 0.29807206988334656, + "step": 244 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 2.322704792022705, + "learning_rate": 9.700626075229738e-07, + "loss": 0.3189048618078232, + "step": 245 + }, + { + "epoch": 1.0, + "grad_norm": 2.322704792022705, + "learning_rate": 9.695748783279544e-07, + "loss": 0.3195984363555908, + "step": 246 + }, + { + "epoch": 1.0040733197556009, + "grad_norm": 2.505608081817627, + "learning_rate": 9.690833328451532e-07, + "loss": 0.24639207869768143, + "step": 247 + }, + { + "epoch": 1.0081466395112015, + "grad_norm": 2.5328471660614014, + "learning_rate": 9.68587975069416e-07, + "loss": 0.3204229325056076, + "step": 248 + }, + { + "epoch": 1.0122199592668024, + "grad_norm": 2.462449073791504, + "learning_rate": 9.680888090265713e-07, + "loss": 0.29177573323249817, + "step": 249 + }, + { + "epoch": 1.0162932790224033, + "grad_norm": 2.4300286769866943, + "learning_rate": 9.67585838773397e-07, + "loss": 0.2859930992126465, + "step": 250 + }, + { + "epoch": 1.0203665987780042, + "grad_norm": 2.4073538780212402, + "learning_rate": 9.67079068397589e-07, + "loss": 0.253083311021328, + "step": 251 + }, + { + "epoch": 1.0244399185336048, + "grad_norm": 2.4259159564971924, + "learning_rate": 9.66568502017727e-07, + "loss": 0.27018117904663086, + "step": 252 + }, + { + "epoch": 1.0285132382892057, + "grad_norm": 2.490452766418457, + "learning_rate": 9.660541437832416e-07, + "loss": 0.270308181643486, + "step": 253 + }, + { + "epoch": 1.0325865580448066, + "grad_norm": 2.2988967895507812, + "learning_rate": 9.655359978743798e-07, + "loss": 0.2865062654018402, + "step": 254 + }, + { + "epoch": 1.0366598778004072, + "grad_norm": 2.3249478340148926, + "learning_rate": 9.650140685021716e-07, + "loss": 0.30297039449214935, + "step": 255 + }, + { + "epoch": 1.0407331975560081, + "grad_norm": 2.530956506729126, + "learning_rate": 9.644883599083957e-07, + "loss": 0.2875078618526459, + "step": 256 + }, + { + "epoch": 1.044806517311609, + "grad_norm": 2.462820053100586, + "learning_rate": 9.639588763655448e-07, + "loss": 0.3308262676000595, + "step": 257 + }, + { + "epoch": 1.0488798370672097, + "grad_norm": 2.733877658843994, + "learning_rate": 9.634256221767912e-07, + "loss": 0.3172256797552109, + "step": 258 + }, + { + "epoch": 1.0529531568228105, + "grad_norm": 2.5849740505218506, + "learning_rate": 9.628886016759515e-07, + "loss": 0.24582470953464508, + "step": 259 + }, + { + "epoch": 1.0570264765784114, + "grad_norm": 2.495885133743286, + "learning_rate": 9.623478192274517e-07, + "loss": 0.29177717864513397, + "step": 260 + }, + { + "epoch": 1.0610997963340123, + "grad_norm": 2.7560508251190186, + "learning_rate": 9.618032792262914e-07, + "loss": 0.2618248611688614, + "step": 261 + }, + { + "epoch": 1.065173116089613, + "grad_norm": 2.5949246883392334, + "learning_rate": 9.612549860980087e-07, + "loss": 0.29752814769744873, + "step": 262 + }, + { + "epoch": 1.0692464358452138, + "grad_norm": 2.5723235607147217, + "learning_rate": 9.60702944298643e-07, + "loss": 0.283334344625473, + "step": 263 + }, + { + "epoch": 1.0733197556008147, + "grad_norm": 2.5463922023773193, + "learning_rate": 9.601471583147002e-07, + "loss": 0.29830360412597656, + "step": 264 + }, + { + "epoch": 1.0773930753564154, + "grad_norm": 2.4415295124053955, + "learning_rate": 9.595876326631153e-07, + "loss": 0.278356209397316, + "step": 265 + }, + { + "epoch": 1.0814663951120163, + "grad_norm": 2.4167826175689697, + "learning_rate": 9.590243718912164e-07, + "loss": 0.2482159435749054, + "step": 266 + }, + { + "epoch": 1.0855397148676171, + "grad_norm": 2.4946138858795166, + "learning_rate": 9.584573805766867e-07, + "loss": 0.2892283797264099, + "step": 267 + }, + { + "epoch": 1.089613034623218, + "grad_norm": 2.585139036178589, + "learning_rate": 9.578866633275286e-07, + "loss": 0.28597134351730347, + "step": 268 + }, + { + "epoch": 1.0936863543788187, + "grad_norm": 2.5749940872192383, + "learning_rate": 9.573122247820253e-07, + "loss": 0.24586065858602524, + "step": 269 + }, + { + "epoch": 1.0977596741344195, + "grad_norm": 2.6099436283111572, + "learning_rate": 9.567340696087032e-07, + "loss": 0.2968917638063431, + "step": 270 + }, + { + "epoch": 1.1018329938900204, + "grad_norm": 2.5843560695648193, + "learning_rate": 9.561522025062946e-07, + "loss": 0.3214084059000015, + "step": 271 + }, + { + "epoch": 1.105906313645621, + "grad_norm": 2.7599520683288574, + "learning_rate": 9.555666282036984e-07, + "loss": 0.2851483225822449, + "step": 272 + }, + { + "epoch": 1.109979633401222, + "grad_norm": 2.340637683868408, + "learning_rate": 9.549773514599428e-07, + "loss": 0.2864815294742584, + "step": 273 + }, + { + "epoch": 1.1140529531568228, + "grad_norm": 2.682030200958252, + "learning_rate": 9.543843770641463e-07, + "loss": 0.2536205053329468, + "step": 274 + }, + { + "epoch": 1.1181262729124237, + "grad_norm": 2.3765580654144287, + "learning_rate": 9.537877098354784e-07, + "loss": 0.24970313906669617, + "step": 275 + }, + { + "epoch": 1.1221995926680244, + "grad_norm": 2.5046305656433105, + "learning_rate": 9.531873546231208e-07, + "loss": 0.2925817370414734, + "step": 276 + }, + { + "epoch": 1.1262729124236253, + "grad_norm": 2.487609386444092, + "learning_rate": 9.525833163062273e-07, + "loss": 0.28858567774295807, + "step": 277 + }, + { + "epoch": 1.1303462321792261, + "grad_norm": 2.5617923736572266, + "learning_rate": 9.519755997938856e-07, + "loss": 0.29943670332431793, + "step": 278 + }, + { + "epoch": 1.1344195519348268, + "grad_norm": 2.5144989490509033, + "learning_rate": 9.51364210025076e-07, + "loss": 0.26054797321558, + "step": 279 + }, + { + "epoch": 1.1384928716904277, + "grad_norm": 2.393367290496826, + "learning_rate": 9.507491519686315e-07, + "loss": 0.2766056954860687, + "step": 280 + }, + { + "epoch": 1.1425661914460286, + "grad_norm": 2.5914998054504395, + "learning_rate": 9.501304306231983e-07, + "loss": 0.2934035658836365, + "step": 281 + }, + { + "epoch": 1.1466395112016294, + "grad_norm": 2.6555488109588623, + "learning_rate": 9.495080510171942e-07, + "loss": 0.27582375705242157, + "step": 282 + }, + { + "epoch": 1.15071283095723, + "grad_norm": 2.4485878944396973, + "learning_rate": 9.488820182087682e-07, + "loss": 0.27037250250577927, + "step": 283 + }, + { + "epoch": 1.154786150712831, + "grad_norm": 2.5796926021575928, + "learning_rate": 9.482523372857592e-07, + "loss": 0.25297392159700394, + "step": 284 + }, + { + "epoch": 1.1588594704684319, + "grad_norm": 2.4900271892547607, + "learning_rate": 9.476190133656548e-07, + "loss": 0.2928699851036072, + "step": 285 + }, + { + "epoch": 1.1629327902240325, + "grad_norm": 2.378876209259033, + "learning_rate": 9.469820515955497e-07, + "loss": 0.3021889925003052, + "step": 286 + }, + { + "epoch": 1.1670061099796334, + "grad_norm": 2.377063751220703, + "learning_rate": 9.463414571521036e-07, + "loss": 0.26095758378505707, + "step": 287 + }, + { + "epoch": 1.1710794297352343, + "grad_norm": 2.347393274307251, + "learning_rate": 9.456972352414997e-07, + "loss": 0.24494879692792892, + "step": 288 + }, + { + "epoch": 1.175152749490835, + "grad_norm": 2.3959014415740967, + "learning_rate": 9.450493910994017e-07, + "loss": 0.28228186070919037, + "step": 289 + }, + { + "epoch": 1.1792260692464358, + "grad_norm": 2.5725162029266357, + "learning_rate": 9.443979299909117e-07, + "loss": 0.3038964122533798, + "step": 290 + }, + { + "epoch": 1.1832993890020367, + "grad_norm": 3.050133228302002, + "learning_rate": 9.437428572105274e-07, + "loss": 0.28559453785419464, + "step": 291 + }, + { + "epoch": 1.1873727087576376, + "grad_norm": 2.4305083751678467, + "learning_rate": 9.430841780820988e-07, + "loss": 0.26872922480106354, + "step": 292 + }, + { + "epoch": 1.1914460285132382, + "grad_norm": 2.3637914657592773, + "learning_rate": 9.424218979587852e-07, + "loss": 0.2783532440662384, + "step": 293 + }, + { + "epoch": 1.195519348268839, + "grad_norm": 2.4774560928344727, + "learning_rate": 9.417560222230114e-07, + "loss": 0.2842411994934082, + "step": 294 + }, + { + "epoch": 1.19959266802444, + "grad_norm": 2.5469460487365723, + "learning_rate": 9.410865562864245e-07, + "loss": 0.2584770694375038, + "step": 295 + }, + { + "epoch": 1.2036659877800409, + "grad_norm": 2.3821113109588623, + "learning_rate": 9.404135055898495e-07, + "loss": 0.29300089180469513, + "step": 296 + }, + { + "epoch": 1.2077393075356415, + "grad_norm": 2.7043542861938477, + "learning_rate": 9.397368756032444e-07, + "loss": 0.2970714569091797, + "step": 297 + }, + { + "epoch": 1.2118126272912424, + "grad_norm": 2.353179693222046, + "learning_rate": 9.390566718256578e-07, + "loss": 0.28464287519454956, + "step": 298 + }, + { + "epoch": 1.2158859470468433, + "grad_norm": 2.615365743637085, + "learning_rate": 9.383728997851819e-07, + "loss": 0.2843187004327774, + "step": 299 + }, + { + "epoch": 1.219959266802444, + "grad_norm": 2.526761054992676, + "learning_rate": 9.376855650389089e-07, + "loss": 0.29214321076869965, + "step": 300 + }, + { + "epoch": 1.2240325865580448, + "grad_norm": 2.416611909866333, + "learning_rate": 9.369946731728854e-07, + "loss": 0.2746543139219284, + "step": 301 + }, + { + "epoch": 1.2281059063136457, + "grad_norm": 2.6475841999053955, + "learning_rate": 9.363002298020673e-07, + "loss": 0.32700057327747345, + "step": 302 + }, + { + "epoch": 1.2321792260692463, + "grad_norm": 2.7061877250671387, + "learning_rate": 9.356022405702739e-07, + "loss": 0.26850761473178864, + "step": 303 + }, + { + "epoch": 1.2362525458248472, + "grad_norm": 2.578963279724121, + "learning_rate": 9.349007111501419e-07, + "loss": 0.2758927643299103, + "step": 304 + }, + { + "epoch": 1.240325865580448, + "grad_norm": 2.476339340209961, + "learning_rate": 9.341956472430801e-07, + "loss": 0.3046765774488449, + "step": 305 + }, + { + "epoch": 1.2443991853360488, + "grad_norm": 2.594074249267578, + "learning_rate": 9.334870545792217e-07, + "loss": 0.30949874222278595, + "step": 306 + }, + { + "epoch": 1.2484725050916496, + "grad_norm": 2.311086416244507, + "learning_rate": 9.32774938917379e-07, + "loss": 0.27652885019779205, + "step": 307 + }, + { + "epoch": 1.2525458248472505, + "grad_norm": 2.5869719982147217, + "learning_rate": 9.320593060449958e-07, + "loss": 0.2891390472650528, + "step": 308 + }, + { + "epoch": 1.2566191446028514, + "grad_norm": 2.7591726779937744, + "learning_rate": 9.313401617781012e-07, + "loss": 0.2366793006658554, + "step": 309 + }, + { + "epoch": 1.260692464358452, + "grad_norm": 2.394120693206787, + "learning_rate": 9.306175119612612e-07, + "loss": 0.2750696837902069, + "step": 310 + }, + { + "epoch": 1.264765784114053, + "grad_norm": 2.542830228805542, + "learning_rate": 9.29891362467532e-07, + "loss": 0.29537804424762726, + "step": 311 + }, + { + "epoch": 1.2688391038696538, + "grad_norm": 2.63010573387146, + "learning_rate": 9.291617191984123e-07, + "loss": 0.28230538964271545, + "step": 312 + }, + { + "epoch": 1.2729124236252547, + "grad_norm": 2.4835424423217773, + "learning_rate": 9.284285880837946e-07, + "loss": 0.2597920596599579, + "step": 313 + }, + { + "epoch": 1.2769857433808554, + "grad_norm": 2.4826974868774414, + "learning_rate": 9.276919750819181e-07, + "loss": 0.28388310968875885, + "step": 314 + }, + { + "epoch": 1.2810590631364562, + "grad_norm": 2.495941400527954, + "learning_rate": 9.269518861793193e-07, + "loss": 0.2510681226849556, + "step": 315 + }, + { + "epoch": 1.2851323828920571, + "grad_norm": 2.3886849880218506, + "learning_rate": 9.262083273907837e-07, + "loss": 0.28582368791103363, + "step": 316 + }, + { + "epoch": 1.2892057026476578, + "grad_norm": 2.443054437637329, + "learning_rate": 9.254613047592974e-07, + "loss": 0.2807822525501251, + "step": 317 + }, + { + "epoch": 1.2932790224032586, + "grad_norm": 2.2685866355895996, + "learning_rate": 9.247108243559971e-07, + "loss": 0.2636859193444252, + "step": 318 + }, + { + "epoch": 1.2973523421588595, + "grad_norm": 2.4459731578826904, + "learning_rate": 9.239568922801212e-07, + "loss": 0.28120698034763336, + "step": 319 + }, + { + "epoch": 1.3014256619144602, + "grad_norm": 2.4667344093322754, + "learning_rate": 9.231995146589605e-07, + "loss": 0.27484674006700516, + "step": 320 + }, + { + "epoch": 1.305498981670061, + "grad_norm": 2.432385206222534, + "learning_rate": 9.22438697647808e-07, + "loss": 0.3200181722640991, + "step": 321 + }, + { + "epoch": 1.309572301425662, + "grad_norm": 2.594827175140381, + "learning_rate": 9.21674447429909e-07, + "loss": 0.26146167516708374, + "step": 322 + }, + { + "epoch": 1.3136456211812626, + "grad_norm": 2.6781699657440186, + "learning_rate": 9.209067702164108e-07, + "loss": 0.30198925733566284, + "step": 323 + }, + { + "epoch": 1.3177189409368635, + "grad_norm": 2.382455348968506, + "learning_rate": 9.201356722463124e-07, + "loss": 0.2656544893980026, + "step": 324 + }, + { + "epoch": 1.3217922606924644, + "grad_norm": 2.3920562267303467, + "learning_rate": 9.193611597864137e-07, + "loss": 0.30542662739753723, + "step": 325 + }, + { + "epoch": 1.3258655804480652, + "grad_norm": 2.5210139751434326, + "learning_rate": 9.185832391312642e-07, + "loss": 0.33505627512931824, + "step": 326 + }, + { + "epoch": 1.3299389002036661, + "grad_norm": 2.530137777328491, + "learning_rate": 9.178019166031128e-07, + "loss": 0.28073740005493164, + "step": 327 + }, + { + "epoch": 1.3340122199592668, + "grad_norm": 2.6757373809814453, + "learning_rate": 9.170171985518551e-07, + "loss": 0.2617020010948181, + "step": 328 + }, + { + "epoch": 1.3380855397148677, + "grad_norm": 2.737130880355835, + "learning_rate": 9.162290913549831e-07, + "loss": 0.31688614189624786, + "step": 329 + }, + { + "epoch": 1.3421588594704685, + "grad_norm": 2.4106035232543945, + "learning_rate": 9.154376014175325e-07, + "loss": 0.25045838952064514, + "step": 330 + }, + { + "epoch": 1.3462321792260692, + "grad_norm": 2.5006155967712402, + "learning_rate": 9.146427351720307e-07, + "loss": 0.2350526750087738, + "step": 331 + }, + { + "epoch": 1.35030549898167, + "grad_norm": 2.443948268890381, + "learning_rate": 9.138444990784453e-07, + "loss": 0.3193846642971039, + "step": 332 + }, + { + "epoch": 1.354378818737271, + "grad_norm": 2.5271644592285156, + "learning_rate": 9.130428996241304e-07, + "loss": 0.2833334505558014, + "step": 333 + }, + { + "epoch": 1.3584521384928716, + "grad_norm": 2.3211746215820312, + "learning_rate": 9.122379433237748e-07, + "loss": 0.27067267149686813, + "step": 334 + }, + { + "epoch": 1.3625254582484725, + "grad_norm": 2.5455574989318848, + "learning_rate": 9.11429636719349e-07, + "loss": 0.3177233636379242, + "step": 335 + }, + { + "epoch": 1.3665987780040734, + "grad_norm": 2.9286653995513916, + "learning_rate": 9.106179863800513e-07, + "loss": 0.30554522573947906, + "step": 336 + }, + { + "epoch": 1.370672097759674, + "grad_norm": 2.452594757080078, + "learning_rate": 9.098029989022557e-07, + "loss": 0.29210953414440155, + "step": 337 + }, + { + "epoch": 1.374745417515275, + "grad_norm": 2.5022428035736084, + "learning_rate": 9.089846809094563e-07, + "loss": 0.25522106885910034, + "step": 338 + }, + { + "epoch": 1.3788187372708758, + "grad_norm": 2.6464457511901855, + "learning_rate": 9.081630390522157e-07, + "loss": 0.287350669503212, + "step": 339 + }, + { + "epoch": 1.3828920570264764, + "grad_norm": 2.3021206855773926, + "learning_rate": 9.073380800081096e-07, + "loss": 0.2737245708703995, + "step": 340 + }, + { + "epoch": 1.3869653767820773, + "grad_norm": 2.5468103885650635, + "learning_rate": 9.065098104816726e-07, + "loss": 0.26481032371520996, + "step": 341 + }, + { + "epoch": 1.3910386965376782, + "grad_norm": 2.3847126960754395, + "learning_rate": 9.056782372043445e-07, + "loss": 0.270496666431427, + "step": 342 + }, + { + "epoch": 1.395112016293279, + "grad_norm": 2.586550235748291, + "learning_rate": 9.048433669344144e-07, + "loss": 0.32853636145591736, + "step": 343 + }, + { + "epoch": 1.39918533604888, + "grad_norm": 2.5621798038482666, + "learning_rate": 9.04005206456967e-07, + "loss": 0.2707676887512207, + "step": 344 + }, + { + "epoch": 1.4032586558044806, + "grad_norm": 2.405062198638916, + "learning_rate": 9.031637625838264e-07, + "loss": 0.285464346408844, + "step": 345 + }, + { + "epoch": 1.4073319755600815, + "grad_norm": 2.382589817047119, + "learning_rate": 9.023190421535016e-07, + "loss": 0.2704601585865021, + "step": 346 + }, + { + "epoch": 1.4114052953156824, + "grad_norm": 2.2916622161865234, + "learning_rate": 9.014710520311306e-07, + "loss": 0.2510330229997635, + "step": 347 + }, + { + "epoch": 1.415478615071283, + "grad_norm": 2.3999521732330322, + "learning_rate": 9.006197991084241e-07, + "loss": 0.2853076159954071, + "step": 348 + }, + { + "epoch": 1.419551934826884, + "grad_norm": 2.6907191276550293, + "learning_rate": 8.997652903036104e-07, + "loss": 0.27804453670978546, + "step": 349 + }, + { + "epoch": 1.4236252545824848, + "grad_norm": 2.3953359127044678, + "learning_rate": 8.989075325613784e-07, + "loss": 0.24225353449583054, + "step": 350 + }, + { + "epoch": 1.4276985743380854, + "grad_norm": 2.5458950996398926, + "learning_rate": 8.980465328528218e-07, + "loss": 0.29082323610782623, + "step": 351 + }, + { + "epoch": 1.4317718940936863, + "grad_norm": 2.570434093475342, + "learning_rate": 8.971822981753818e-07, + "loss": 0.2637802064418793, + "step": 352 + }, + { + "epoch": 1.4358452138492872, + "grad_norm": 2.608654260635376, + "learning_rate": 8.963148355527908e-07, + "loss": 0.267157182097435, + "step": 353 + }, + { + "epoch": 1.4399185336048879, + "grad_norm": 2.505756139755249, + "learning_rate": 8.954441520350149e-07, + "loss": 0.31443025171756744, + "step": 354 + }, + { + "epoch": 1.4439918533604887, + "grad_norm": 2.8767051696777344, + "learning_rate": 8.945702546981968e-07, + "loss": 0.28173527121543884, + "step": 355 + }, + { + "epoch": 1.4480651731160896, + "grad_norm": 2.5839850902557373, + "learning_rate": 8.936931506445984e-07, + "loss": 0.2735389843583107, + "step": 356 + }, + { + "epoch": 1.4521384928716905, + "grad_norm": 2.380516529083252, + "learning_rate": 8.928128470025429e-07, + "loss": 0.2868722528219223, + "step": 357 + }, + { + "epoch": 1.4562118126272914, + "grad_norm": 2.373098134994507, + "learning_rate": 8.919293509263567e-07, + "loss": 0.266360841691494, + "step": 358 + }, + { + "epoch": 1.460285132382892, + "grad_norm": 2.440216541290283, + "learning_rate": 8.910426695963118e-07, + "loss": 0.26618482172489166, + "step": 359 + }, + { + "epoch": 1.464358452138493, + "grad_norm": 2.374094247817993, + "learning_rate": 8.901528102185669e-07, + "loss": 0.27816291898489, + "step": 360 + }, + { + "epoch": 1.4684317718940938, + "grad_norm": 2.3657944202423096, + "learning_rate": 8.892597800251093e-07, + "loss": 0.25230376422405243, + "step": 361 + }, + { + "epoch": 1.4725050916496945, + "grad_norm": 2.3009097576141357, + "learning_rate": 8.883635862736956e-07, + "loss": 0.26984117925167084, + "step": 362 + }, + { + "epoch": 1.4765784114052953, + "grad_norm": 2.9458165168762207, + "learning_rate": 8.874642362477929e-07, + "loss": 0.29643698036670685, + "step": 363 + }, + { + "epoch": 1.4806517311608962, + "grad_norm": 2.8554880619049072, + "learning_rate": 8.865617372565199e-07, + "loss": 0.256381556391716, + "step": 364 + }, + { + "epoch": 1.4847250509164969, + "grad_norm": 2.5279104709625244, + "learning_rate": 8.856560966345876e-07, + "loss": 0.24138055741786957, + "step": 365 + }, + { + "epoch": 1.4887983706720977, + "grad_norm": 2.4175782203674316, + "learning_rate": 8.847473217422388e-07, + "loss": 0.27801232039928436, + "step": 366 + }, + { + "epoch": 1.4928716904276986, + "grad_norm": 2.7168712615966797, + "learning_rate": 8.838354199651891e-07, + "loss": 0.28772565722465515, + "step": 367 + }, + { + "epoch": 1.4969450101832993, + "grad_norm": 2.5258262157440186, + "learning_rate": 8.829203987145669e-07, + "loss": 0.2848212271928787, + "step": 368 + }, + { + "epoch": 1.5010183299389002, + "grad_norm": 2.424790859222412, + "learning_rate": 8.820022654268525e-07, + "loss": 0.2653035372495651, + "step": 369 + }, + { + "epoch": 1.505091649694501, + "grad_norm": 2.416537284851074, + "learning_rate": 8.810810275638182e-07, + "loss": 0.27215129137039185, + "step": 370 + }, + { + "epoch": 1.5091649694501017, + "grad_norm": 2.5008385181427, + "learning_rate": 8.801566926124677e-07, + "loss": 0.2523413300514221, + "step": 371 + }, + { + "epoch": 1.5132382892057028, + "grad_norm": 2.5779659748077393, + "learning_rate": 8.79229268084975e-07, + "loss": 0.2584775537252426, + "step": 372 + }, + { + "epoch": 1.5173116089613035, + "grad_norm": 2.631009340286255, + "learning_rate": 8.782987615186234e-07, + "loss": 0.27790168672800064, + "step": 373 + }, + { + "epoch": 1.5213849287169041, + "grad_norm": 2.4295825958251953, + "learning_rate": 8.773651804757443e-07, + "loss": 0.26932528614997864, + "step": 374 + }, + { + "epoch": 1.5254582484725052, + "grad_norm": 2.6135988235473633, + "learning_rate": 8.764285325436555e-07, + "loss": 0.28256936371326447, + "step": 375 + }, + { + "epoch": 1.5295315682281059, + "grad_norm": 2.5016775131225586, + "learning_rate": 8.754888253346002e-07, + "loss": 0.28559644520282745, + "step": 376 + }, + { + "epoch": 1.5336048879837068, + "grad_norm": 2.5780680179595947, + "learning_rate": 8.745460664856844e-07, + "loss": 0.2705298289656639, + "step": 377 + }, + { + "epoch": 1.5376782077393076, + "grad_norm": 2.7102303504943848, + "learning_rate": 8.736002636588151e-07, + "loss": 0.27591292560100555, + "step": 378 + }, + { + "epoch": 1.5417515274949083, + "grad_norm": 2.520955801010132, + "learning_rate": 8.72651424540638e-07, + "loss": 0.2621312141418457, + "step": 379 + }, + { + "epoch": 1.5458248472505092, + "grad_norm": 3.2938437461853027, + "learning_rate": 8.716995568424754e-07, + "loss": 0.2692745327949524, + "step": 380 + }, + { + "epoch": 1.54989816700611, + "grad_norm": 2.4736831188201904, + "learning_rate": 8.70744668300263e-07, + "loss": 0.24611817300319672, + "step": 381 + }, + { + "epoch": 1.5539714867617107, + "grad_norm": 2.3731565475463867, + "learning_rate": 8.697867666744871e-07, + "loss": 0.2787970006465912, + "step": 382 + }, + { + "epoch": 1.5580448065173116, + "grad_norm": 2.5973782539367676, + "learning_rate": 8.688258597501219e-07, + "loss": 0.3056941330432892, + "step": 383 + }, + { + "epoch": 1.5621181262729125, + "grad_norm": 2.490133285522461, + "learning_rate": 8.678619553365658e-07, + "loss": 0.28115689754486084, + "step": 384 + }, + { + "epoch": 1.5661914460285131, + "grad_norm": 2.7315447330474854, + "learning_rate": 8.668950612675783e-07, + "loss": 0.2873499393463135, + "step": 385 + }, + { + "epoch": 1.570264765784114, + "grad_norm": 2.518087387084961, + "learning_rate": 8.659251854012161e-07, + "loss": 0.28173641860485077, + "step": 386 + }, + { + "epoch": 1.5743380855397149, + "grad_norm": 2.214815378189087, + "learning_rate": 8.649523356197695e-07, + "loss": 0.2376307100057602, + "step": 387 + }, + { + "epoch": 1.5784114052953155, + "grad_norm": 2.2534103393554688, + "learning_rate": 8.639765198296977e-07, + "loss": 0.2530301362276077, + "step": 388 + }, + { + "epoch": 1.5824847250509166, + "grad_norm": 2.5998353958129883, + "learning_rate": 8.629977459615654e-07, + "loss": 0.26012372225522995, + "step": 389 + }, + { + "epoch": 1.5865580448065173, + "grad_norm": 2.4811596870422363, + "learning_rate": 8.620160219699777e-07, + "loss": 0.26469268649816513, + "step": 390 + }, + { + "epoch": 1.5906313645621182, + "grad_norm": 2.4212377071380615, + "learning_rate": 8.610313558335156e-07, + "loss": 0.2749471068382263, + "step": 391 + }, + { + "epoch": 1.594704684317719, + "grad_norm": 2.3209803104400635, + "learning_rate": 8.600437555546716e-07, + "loss": 0.2666809409856796, + "step": 392 + }, + { + "epoch": 1.5987780040733197, + "grad_norm": 2.6138010025024414, + "learning_rate": 8.59053229159784e-07, + "loss": 0.2631463557481766, + "step": 393 + }, + { + "epoch": 1.6028513238289206, + "grad_norm": 2.4465487003326416, + "learning_rate": 8.580597846989721e-07, + "loss": 0.2840096950531006, + "step": 394 + }, + { + "epoch": 1.6069246435845215, + "grad_norm": 2.781353235244751, + "learning_rate": 8.570634302460706e-07, + "loss": 0.26131805777549744, + "step": 395 + }, + { + "epoch": 1.6109979633401221, + "grad_norm": 2.4190244674682617, + "learning_rate": 8.560641738985641e-07, + "loss": 0.28362704813480377, + "step": 396 + }, + { + "epoch": 1.615071283095723, + "grad_norm": 2.461918354034424, + "learning_rate": 8.550620237775213e-07, + "loss": 0.2682417184114456, + "step": 397 + }, + { + "epoch": 1.6191446028513239, + "grad_norm": 2.4705612659454346, + "learning_rate": 8.540569880275286e-07, + "loss": 0.2832501232624054, + "step": 398 + }, + { + "epoch": 1.6232179226069245, + "grad_norm": 2.401193380355835, + "learning_rate": 8.530490748166244e-07, + "loss": 0.2732377126812935, + "step": 399 + }, + { + "epoch": 1.6272912423625254, + "grad_norm": 2.7205650806427, + "learning_rate": 8.520382923362328e-07, + "loss": 0.2597970813512802, + "step": 400 + }, + { + "epoch": 1.6313645621181263, + "grad_norm": 2.6480019092559814, + "learning_rate": 8.510246488010964e-07, + "loss": 0.2697395384311676, + "step": 401 + }, + { + "epoch": 1.635437881873727, + "grad_norm": 2.5444276332855225, + "learning_rate": 8.5000815244921e-07, + "loss": 0.26779498159885406, + "step": 402 + }, + { + "epoch": 1.639511201629328, + "grad_norm": 2.3247053623199463, + "learning_rate": 8.489888115417537e-07, + "loss": 0.2565397247672081, + "step": 403 + }, + { + "epoch": 1.6435845213849287, + "grad_norm": 2.8525049686431885, + "learning_rate": 8.479666343630256e-07, + "loss": 0.29496023058891296, + "step": 404 + }, + { + "epoch": 1.6476578411405294, + "grad_norm": 2.5987389087677, + "learning_rate": 8.469416292203746e-07, + "loss": 0.24137818068265915, + "step": 405 + }, + { + "epoch": 1.6517311608961305, + "grad_norm": 2.4682445526123047, + "learning_rate": 8.459138044441323e-07, + "loss": 0.3068127781152725, + "step": 406 + }, + { + "epoch": 1.6558044806517311, + "grad_norm": 2.43457293510437, + "learning_rate": 8.448831683875464e-07, + "loss": 0.2586989253759384, + "step": 407 + }, + { + "epoch": 1.659877800407332, + "grad_norm": 2.446392297744751, + "learning_rate": 8.438497294267116e-07, + "loss": 0.2845039367675781, + "step": 408 + }, + { + "epoch": 1.663951120162933, + "grad_norm": 2.4405159950256348, + "learning_rate": 8.428134959605027e-07, + "loss": 0.2668181359767914, + "step": 409 + }, + { + "epoch": 1.6680244399185336, + "grad_norm": 2.398472309112549, + "learning_rate": 8.417744764105053e-07, + "loss": 0.294752761721611, + "step": 410 + }, + { + "epoch": 1.6720977596741344, + "grad_norm": 2.485675096511841, + "learning_rate": 8.407326792209481e-07, + "loss": 0.27970798313617706, + "step": 411 + }, + { + "epoch": 1.6761710794297353, + "grad_norm": 2.326519250869751, + "learning_rate": 8.396881128586338e-07, + "loss": 0.27439168095588684, + "step": 412 + }, + { + "epoch": 1.680244399185336, + "grad_norm": 2.5076746940612793, + "learning_rate": 8.386407858128706e-07, + "loss": 0.24190914630889893, + "step": 413 + }, + { + "epoch": 1.6843177189409368, + "grad_norm": 2.418656349182129, + "learning_rate": 8.375907065954028e-07, + "loss": 0.302260160446167, + "step": 414 + }, + { + "epoch": 1.6883910386965377, + "grad_norm": 2.431206226348877, + "learning_rate": 8.365378837403427e-07, + "loss": 0.24484054744243622, + "step": 415 + }, + { + "epoch": 1.6924643584521384, + "grad_norm": 2.1974987983703613, + "learning_rate": 8.354823258040995e-07, + "loss": 0.23830442130565643, + "step": 416 + }, + { + "epoch": 1.6965376782077393, + "grad_norm": 2.531301498413086, + "learning_rate": 8.344240413653111e-07, + "loss": 0.23077362775802612, + "step": 417 + }, + { + "epoch": 1.7006109979633401, + "grad_norm": 2.3643598556518555, + "learning_rate": 8.333630390247741e-07, + "loss": 0.2691231817007065, + "step": 418 + }, + { + "epoch": 1.7046843177189408, + "grad_norm": 2.429917335510254, + "learning_rate": 8.322993274053738e-07, + "loss": 0.26382092386484146, + "step": 419 + }, + { + "epoch": 1.708757637474542, + "grad_norm": 2.3184382915496826, + "learning_rate": 8.312329151520139e-07, + "loss": 0.24146142601966858, + "step": 420 + }, + { + "epoch": 1.7128309572301426, + "grad_norm": 2.3974194526672363, + "learning_rate": 8.301638109315465e-07, + "loss": 0.28932487964630127, + "step": 421 + }, + { + "epoch": 1.7169042769857432, + "grad_norm": 2.445939064025879, + "learning_rate": 8.290920234327019e-07, + "loss": 0.2816842496395111, + "step": 422 + }, + { + "epoch": 1.7209775967413443, + "grad_norm": 2.517059087753296, + "learning_rate": 8.280175613660175e-07, + "loss": 0.28957797586917877, + "step": 423 + }, + { + "epoch": 1.725050916496945, + "grad_norm": 2.7085039615631104, + "learning_rate": 8.269404334637666e-07, + "loss": 0.2929697036743164, + "step": 424 + }, + { + "epoch": 1.7291242362525459, + "grad_norm": 2.3027985095977783, + "learning_rate": 8.258606484798896e-07, + "loss": 0.2602064907550812, + "step": 425 + }, + { + "epoch": 1.7331975560081467, + "grad_norm": 2.4261462688446045, + "learning_rate": 8.247782151899195e-07, + "loss": 0.2826506048440933, + "step": 426 + }, + { + "epoch": 1.7372708757637474, + "grad_norm": 2.6706831455230713, + "learning_rate": 8.236931423909138e-07, + "loss": 0.25307597219944, + "step": 427 + }, + { + "epoch": 1.7413441955193483, + "grad_norm": 2.6151740550994873, + "learning_rate": 8.226054389013808e-07, + "loss": 0.31587836146354675, + "step": 428 + }, + { + "epoch": 1.7454175152749491, + "grad_norm": 2.3904056549072266, + "learning_rate": 8.215151135612088e-07, + "loss": 0.2650100588798523, + "step": 429 + }, + { + "epoch": 1.7494908350305498, + "grad_norm": 2.48195219039917, + "learning_rate": 8.204221752315948e-07, + "loss": 0.24214383959770203, + "step": 430 + }, + { + "epoch": 1.7535641547861507, + "grad_norm": 2.28658390045166, + "learning_rate": 8.193266327949708e-07, + "loss": 0.26451194286346436, + "step": 431 + }, + { + "epoch": 1.7576374745417516, + "grad_norm": 2.365696430206299, + "learning_rate": 8.182284951549335e-07, + "loss": 0.26112615317106247, + "step": 432 + }, + { + "epoch": 1.7617107942973522, + "grad_norm": 2.467344284057617, + "learning_rate": 8.171277712361708e-07, + "loss": 0.2644128352403641, + "step": 433 + }, + { + "epoch": 1.765784114052953, + "grad_norm": 2.6279022693634033, + "learning_rate": 8.160244699843899e-07, + "loss": 0.26811327785253525, + "step": 434 + }, + { + "epoch": 1.769857433808554, + "grad_norm": 2.196678638458252, + "learning_rate": 8.149186003662437e-07, + "loss": 0.2764963060617447, + "step": 435 + }, + { + "epoch": 1.7739307535641546, + "grad_norm": 2.582368850708008, + "learning_rate": 8.138101713692587e-07, + "loss": 0.24663043022155762, + "step": 436 + }, + { + "epoch": 1.7780040733197557, + "grad_norm": 2.4632701873779297, + "learning_rate": 8.12699192001762e-07, + "loss": 0.24030117690563202, + "step": 437 + }, + { + "epoch": 1.7820773930753564, + "grad_norm": 2.549862861633301, + "learning_rate": 8.115856712928077e-07, + "loss": 0.2775857746601105, + "step": 438 + }, + { + "epoch": 1.7861507128309573, + "grad_norm": 2.460521936416626, + "learning_rate": 8.104696182921039e-07, + "loss": 0.24370034784078598, + "step": 439 + }, + { + "epoch": 1.7902240325865582, + "grad_norm": 2.489588975906372, + "learning_rate": 8.093510420699386e-07, + "loss": 0.2672967165708542, + "step": 440 + }, + { + "epoch": 1.7942973523421588, + "grad_norm": 2.442493438720703, + "learning_rate": 8.082299517171061e-07, + "loss": 0.2774495333433151, + "step": 441 + }, + { + "epoch": 1.7983706720977597, + "grad_norm": 2.6533353328704834, + "learning_rate": 8.071063563448339e-07, + "loss": 0.27777694165706635, + "step": 442 + }, + { + "epoch": 1.8024439918533606, + "grad_norm": 2.546030282974243, + "learning_rate": 8.059802650847077e-07, + "loss": 0.2475699484348297, + "step": 443 + }, + { + "epoch": 1.8065173116089612, + "grad_norm": 2.603074789047241, + "learning_rate": 8.048516870885977e-07, + "loss": 0.2610132396221161, + "step": 444 + }, + { + "epoch": 1.810590631364562, + "grad_norm": 2.5384109020233154, + "learning_rate": 8.037206315285841e-07, + "loss": 0.26541996002197266, + "step": 445 + }, + { + "epoch": 1.814663951120163, + "grad_norm": 2.5843305587768555, + "learning_rate": 8.025871075968826e-07, + "loss": 0.26443643867969513, + "step": 446 + }, + { + "epoch": 1.8187372708757636, + "grad_norm": 2.428213119506836, + "learning_rate": 8.014511245057691e-07, + "loss": 0.27873335778713226, + "step": 447 + }, + { + "epoch": 1.8228105906313645, + "grad_norm": 2.622931480407715, + "learning_rate": 8.003126914875063e-07, + "loss": 0.2784544825553894, + "step": 448 + }, + { + "epoch": 1.8268839103869654, + "grad_norm": 2.4025309085845947, + "learning_rate": 7.991718177942666e-07, + "loss": 0.22745566070079803, + "step": 449 + }, + { + "epoch": 1.830957230142566, + "grad_norm": 2.426877498626709, + "learning_rate": 7.980285126980591e-07, + "loss": 0.2900083065032959, + "step": 450 + }, + { + "epoch": 1.8350305498981672, + "grad_norm": 2.506085157394409, + "learning_rate": 7.968827854906528e-07, + "loss": 0.3019551932811737, + "step": 451 + }, + { + "epoch": 1.8391038696537678, + "grad_norm": 2.428899049758911, + "learning_rate": 7.95734645483501e-07, + "loss": 0.2602303624153137, + "step": 452 + }, + { + "epoch": 1.8431771894093685, + "grad_norm": 2.7357733249664307, + "learning_rate": 7.945841020076669e-07, + "loss": 0.2674318701028824, + "step": 453 + }, + { + "epoch": 1.8472505091649696, + "grad_norm": 2.492051124572754, + "learning_rate": 7.934311644137463e-07, + "loss": 0.29544439911842346, + "step": 454 + }, + { + "epoch": 1.8513238289205702, + "grad_norm": 2.751485586166382, + "learning_rate": 7.922758420717928e-07, + "loss": 0.2730901688337326, + "step": 455 + }, + { + "epoch": 1.8553971486761711, + "grad_norm": 2.469752550125122, + "learning_rate": 7.911181443712407e-07, + "loss": 0.28135350346565247, + "step": 456 + }, + { + "epoch": 1.859470468431772, + "grad_norm": 2.5486271381378174, + "learning_rate": 7.89958080720829e-07, + "loss": 0.2647327482700348, + "step": 457 + }, + { + "epoch": 1.8635437881873727, + "grad_norm": 2.328063726425171, + "learning_rate": 7.887956605485258e-07, + "loss": 0.29177480936050415, + "step": 458 + }, + { + "epoch": 1.8676171079429735, + "grad_norm": 2.3431217670440674, + "learning_rate": 7.876308933014502e-07, + "loss": 0.2461807057261467, + "step": 459 + }, + { + "epoch": 1.8716904276985744, + "grad_norm": 2.346773862838745, + "learning_rate": 7.864637884457961e-07, + "loss": 0.2737877368927002, + "step": 460 + }, + { + "epoch": 1.875763747454175, + "grad_norm": 2.6286261081695557, + "learning_rate": 7.852943554667561e-07, + "loss": 0.29079559445381165, + "step": 461 + }, + { + "epoch": 1.879837067209776, + "grad_norm": 2.3207788467407227, + "learning_rate": 7.841226038684434e-07, + "loss": 0.261492520570755, + "step": 462 + }, + { + "epoch": 1.8839103869653768, + "grad_norm": 2.4097938537597656, + "learning_rate": 7.829485431738148e-07, + "loss": 0.2779378667473793, + "step": 463 + }, + { + "epoch": 1.8879837067209775, + "grad_norm": 2.580665349960327, + "learning_rate": 7.817721829245935e-07, + "loss": 0.26885151863098145, + "step": 464 + }, + { + "epoch": 1.8920570264765784, + "grad_norm": 2.5190136432647705, + "learning_rate": 7.805935326811912e-07, + "loss": 0.2348337173461914, + "step": 465 + }, + { + "epoch": 1.8961303462321792, + "grad_norm": 2.6823647022247314, + "learning_rate": 7.794126020226309e-07, + "loss": 0.31435824930667877, + "step": 466 + }, + { + "epoch": 1.90020366598778, + "grad_norm": 2.678208827972412, + "learning_rate": 7.782294005464686e-07, + "loss": 0.2894388735294342, + "step": 467 + }, + { + "epoch": 1.904276985743381, + "grad_norm": 2.519721269607544, + "learning_rate": 7.770439378687161e-07, + "loss": 0.3156396448612213, + "step": 468 + }, + { + "epoch": 1.9083503054989817, + "grad_norm": 2.6619021892547607, + "learning_rate": 7.758562236237614e-07, + "loss": 0.2675836831331253, + "step": 469 + }, + { + "epoch": 1.9124236252545825, + "grad_norm": 2.2732584476470947, + "learning_rate": 7.746662674642912e-07, + "loss": 0.21649424731731415, + "step": 470 + }, + { + "epoch": 1.9164969450101834, + "grad_norm": 2.302424669265747, + "learning_rate": 7.734740790612136e-07, + "loss": 0.2436254546046257, + "step": 471 + }, + { + "epoch": 1.920570264765784, + "grad_norm": 2.3057138919830322, + "learning_rate": 7.722796681035769e-07, + "loss": 0.25994570553302765, + "step": 472 + }, + { + "epoch": 1.924643584521385, + "grad_norm": 2.327444076538086, + "learning_rate": 7.710830442984937e-07, + "loss": 0.2572202906012535, + "step": 473 + }, + { + "epoch": 1.9287169042769858, + "grad_norm": 2.3709921836853027, + "learning_rate": 7.698842173710599e-07, + "loss": 0.2600872740149498, + "step": 474 + }, + { + "epoch": 1.9327902240325865, + "grad_norm": 2.463008165359497, + "learning_rate": 7.686831970642766e-07, + "loss": 0.2489926964044571, + "step": 475 + }, + { + "epoch": 1.9368635437881874, + "grad_norm": 2.4627251625061035, + "learning_rate": 7.674799931389708e-07, + "loss": 0.2903194725513458, + "step": 476 + }, + { + "epoch": 1.9409368635437882, + "grad_norm": 2.9676828384399414, + "learning_rate": 7.662746153737156e-07, + "loss": 0.2674560844898224, + "step": 477 + }, + { + "epoch": 1.945010183299389, + "grad_norm": 2.448298692703247, + "learning_rate": 7.65067073564752e-07, + "loss": 0.3171275407075882, + "step": 478 + }, + { + "epoch": 1.9490835030549898, + "grad_norm": 2.524946451187134, + "learning_rate": 7.638573775259077e-07, + "loss": 0.2637060284614563, + "step": 479 + }, + { + "epoch": 1.9531568228105907, + "grad_norm": 2.4001801013946533, + "learning_rate": 7.62645537088518e-07, + "loss": 0.23155802488327026, + "step": 480 + }, + { + "epoch": 1.9572301425661913, + "grad_norm": 2.287005662918091, + "learning_rate": 7.614315621013469e-07, + "loss": 0.27207742631435394, + "step": 481 + }, + { + "epoch": 1.9613034623217924, + "grad_norm": 2.453338623046875, + "learning_rate": 7.60215462430505e-07, + "loss": 0.2895518094301224, + "step": 482 + }, + { + "epoch": 1.965376782077393, + "grad_norm": 2.5567643642425537, + "learning_rate": 7.58997247959371e-07, + "loss": 0.2574731409549713, + "step": 483 + }, + { + "epoch": 1.9694501018329937, + "grad_norm": 2.365522623062134, + "learning_rate": 7.577769285885108e-07, + "loss": 0.22622792422771454, + "step": 484 + }, + { + "epoch": 1.9735234215885948, + "grad_norm": 2.2915074825286865, + "learning_rate": 7.56554514235597e-07, + "loss": 0.26031681150197983, + "step": 485 + }, + { + "epoch": 1.9775967413441955, + "grad_norm": 2.651094436645508, + "learning_rate": 7.553300148353284e-07, + "loss": 0.2904250845313072, + "step": 486 + }, + { + "epoch": 1.9816700610997964, + "grad_norm": 2.416658639907837, + "learning_rate": 7.541034403393489e-07, + "loss": 0.2949088215827942, + "step": 487 + }, + { + "epoch": 1.9857433808553973, + "grad_norm": 2.4140050411224365, + "learning_rate": 7.528748007161676e-07, + "loss": 0.2634105682373047, + "step": 488 + }, + { + "epoch": 1.989816700610998, + "grad_norm": 2.684359550476074, + "learning_rate": 7.516441059510764e-07, + "loss": 0.27155014872550964, + "step": 489 + }, + { + "epoch": 1.9938900203665988, + "grad_norm": 2.4172725677490234, + "learning_rate": 7.5041136604607e-07, + "loss": 0.25667136907577515, + "step": 490 + }, + { + "epoch": 1.9979633401221997, + "grad_norm": 2.3175907135009766, + "learning_rate": 7.491765910197643e-07, + "loss": 0.2724708914756775, + "step": 491 + }, + { + "epoch": 2.0, + "grad_norm": 2.5458946228027344, + "learning_rate": 7.479397909073143e-07, + "loss": 0.2475520372390747, + "step": 492 + }, + { + "epoch": 2.0040733197556007, + "grad_norm": 2.372309446334839, + "learning_rate": 7.467009757603336e-07, + "loss": 0.25162914395332336, + "step": 493 + }, + { + "epoch": 2.0081466395112018, + "grad_norm": 2.3240795135498047, + "learning_rate": 7.454601556468121e-07, + "loss": 0.24550767987966537, + "step": 494 + }, + { + "epoch": 2.0122199592668024, + "grad_norm": 2.3425838947296143, + "learning_rate": 7.442173406510341e-07, + "loss": 0.21918359398841858, + "step": 495 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 2.2755539417266846, + "learning_rate": 7.429725408734968e-07, + "loss": 0.24949443340301514, + "step": 496 + }, + { + "epoch": 2.020366598778004, + "grad_norm": 2.490926742553711, + "learning_rate": 7.417257664308276e-07, + "loss": 0.2353430986404419, + "step": 497 + }, + { + "epoch": 2.024439918533605, + "grad_norm": 2.390965461730957, + "learning_rate": 7.404770274557028e-07, + "loss": 0.25412893295288086, + "step": 498 + }, + { + "epoch": 2.0285132382892055, + "grad_norm": 2.11749005317688, + "learning_rate": 7.392263340967641e-07, + "loss": 0.21416524052619934, + "step": 499 + }, + { + "epoch": 2.0325865580448066, + "grad_norm": 2.317534923553467, + "learning_rate": 7.379736965185368e-07, + "loss": 0.24323680251836777, + "step": 500 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 2.29274845123291, + "learning_rate": 7.367191249013472e-07, + "loss": 0.2301274538040161, + "step": 501 + }, + { + "epoch": 2.0407331975560083, + "grad_norm": 2.2650434970855713, + "learning_rate": 7.354626294412402e-07, + "loss": 0.1820373833179474, + "step": 502 + }, + { + "epoch": 2.044806517311609, + "grad_norm": 2.6822898387908936, + "learning_rate": 7.342042203498951e-07, + "loss": 0.24766084551811218, + "step": 503 + }, + { + "epoch": 2.0488798370672097, + "grad_norm": 2.583789348602295, + "learning_rate": 7.329439078545438e-07, + "loss": 0.24215728789567947, + "step": 504 + }, + { + "epoch": 2.0529531568228108, + "grad_norm": 2.484274387359619, + "learning_rate": 7.316817021978883e-07, + "loss": 0.23763683438301086, + "step": 505 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 2.4976799488067627, + "learning_rate": 7.304176136380149e-07, + "loss": 0.2570403814315796, + "step": 506 + }, + { + "epoch": 2.061099796334012, + "grad_norm": 2.323218584060669, + "learning_rate": 7.291516524483136e-07, + "loss": 0.21409639716148376, + "step": 507 + }, + { + "epoch": 2.065173116089613, + "grad_norm": 2.562451124191284, + "learning_rate": 7.278838289173933e-07, + "loss": 0.2541456073522568, + "step": 508 + }, + { + "epoch": 2.069246435845214, + "grad_norm": 2.538037061691284, + "learning_rate": 7.266141533489983e-07, + "loss": 0.2855856567621231, + "step": 509 + }, + { + "epoch": 2.0733197556008145, + "grad_norm": 2.348076343536377, + "learning_rate": 7.253426360619242e-07, + "loss": 0.2544526904821396, + "step": 510 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 2.2017970085144043, + "learning_rate": 7.240692873899351e-07, + "loss": 0.21077334135770798, + "step": 511 + }, + { + "epoch": 2.0814663951120163, + "grad_norm": 2.3848230838775635, + "learning_rate": 7.227941176816787e-07, + "loss": 0.2178598716855049, + "step": 512 + }, + { + "epoch": 2.085539714867617, + "grad_norm": 2.4316163063049316, + "learning_rate": 7.215171373006024e-07, + "loss": 0.21304114907979965, + "step": 513 + }, + { + "epoch": 2.089613034623218, + "grad_norm": 2.342197895050049, + "learning_rate": 7.202383566248692e-07, + "loss": 0.2186000794172287, + "step": 514 + }, + { + "epoch": 2.0936863543788187, + "grad_norm": 2.3286263942718506, + "learning_rate": 7.189577860472731e-07, + "loss": 0.25711505115032196, + "step": 515 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 2.196580410003662, + "learning_rate": 7.176754359751555e-07, + "loss": 0.20909912884235382, + "step": 516 + }, + { + "epoch": 2.1018329938900204, + "grad_norm": 2.197080373764038, + "learning_rate": 7.163913168303191e-07, + "loss": 0.23562318086624146, + "step": 517 + }, + { + "epoch": 2.105906313645621, + "grad_norm": 2.5636260509490967, + "learning_rate": 7.151054390489444e-07, + "loss": 0.25697334110736847, + "step": 518 + }, + { + "epoch": 2.109979633401222, + "grad_norm": 2.6412978172302246, + "learning_rate": 7.138178130815047e-07, + "loss": 0.24819861352443695, + "step": 519 + }, + { + "epoch": 2.114052953156823, + "grad_norm": 2.4628520011901855, + "learning_rate": 7.125284493926809e-07, + "loss": 0.20018430054187775, + "step": 520 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 2.6035919189453125, + "learning_rate": 7.112373584612763e-07, + "loss": 0.2615906372666359, + "step": 521 + }, + { + "epoch": 2.1221995926680246, + "grad_norm": 2.499178409576416, + "learning_rate": 7.099445507801323e-07, + "loss": 0.25786396861076355, + "step": 522 + }, + { + "epoch": 2.1262729124236253, + "grad_norm": 2.356799364089966, + "learning_rate": 7.086500368560419e-07, + "loss": 0.24771813303232193, + "step": 523 + }, + { + "epoch": 2.130346232179226, + "grad_norm": 2.3499882221221924, + "learning_rate": 7.073538272096651e-07, + "loss": 0.2333463951945305, + "step": 524 + }, + { + "epoch": 2.134419551934827, + "grad_norm": 2.328732490539551, + "learning_rate": 7.060559323754435e-07, + "loss": 0.20822811126708984, + "step": 525 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 2.24107027053833, + "learning_rate": 7.047563629015141e-07, + "loss": 0.2279072254896164, + "step": 526 + }, + { + "epoch": 2.1425661914460283, + "grad_norm": 2.3738133907318115, + "learning_rate": 7.03455129349624e-07, + "loss": 0.23648229241371155, + "step": 527 + }, + { + "epoch": 2.1466395112016294, + "grad_norm": 2.2748498916625977, + "learning_rate": 7.021522422950443e-07, + "loss": 0.21738301217556, + "step": 528 + }, + { + "epoch": 2.15071283095723, + "grad_norm": 2.4731128215789795, + "learning_rate": 7.008477123264847e-07, + "loss": 0.23567791283130646, + "step": 529 + }, + { + "epoch": 2.1547861507128308, + "grad_norm": 2.5964107513427734, + "learning_rate": 6.995415500460067e-07, + "loss": 0.25774678587913513, + "step": 530 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 2.7168128490448, + "learning_rate": 6.982337660689377e-07, + "loss": 0.25715047866106033, + "step": 531 + }, + { + "epoch": 2.1629327902240325, + "grad_norm": 2.4718027114868164, + "learning_rate": 6.96924371023785e-07, + "loss": 0.21807514131069183, + "step": 532 + }, + { + "epoch": 2.167006109979633, + "grad_norm": 2.3053858280181885, + "learning_rate": 6.956133755521496e-07, + "loss": 0.22069019079208374, + "step": 533 + }, + { + "epoch": 2.1710794297352343, + "grad_norm": 2.3815200328826904, + "learning_rate": 6.943007903086387e-07, + "loss": 0.2234855741262436, + "step": 534 + }, + { + "epoch": 2.175152749490835, + "grad_norm": 2.331794261932373, + "learning_rate": 6.929866259607797e-07, + "loss": 0.2392452359199524, + "step": 535 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 2.5421111583709717, + "learning_rate": 6.916708931889344e-07, + "loss": 0.2579016238451004, + "step": 536 + }, + { + "epoch": 2.1832993890020367, + "grad_norm": 2.235800266265869, + "learning_rate": 6.903536026862104e-07, + "loss": 0.21762673556804657, + "step": 537 + }, + { + "epoch": 2.1873727087576373, + "grad_norm": 2.3409221172332764, + "learning_rate": 6.890347651583759e-07, + "loss": 0.21134832501411438, + "step": 538 + }, + { + "epoch": 2.1914460285132384, + "grad_norm": 2.29571795463562, + "learning_rate": 6.877143913237713e-07, + "loss": 0.22032570838928223, + "step": 539 + }, + { + "epoch": 2.195519348268839, + "grad_norm": 2.3287696838378906, + "learning_rate": 6.863924919132236e-07, + "loss": 0.22613044828176498, + "step": 540 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 2.2647016048431396, + "learning_rate": 6.850690776699573e-07, + "loss": 0.20803789049386978, + "step": 541 + }, + { + "epoch": 2.203665987780041, + "grad_norm": 2.5180113315582275, + "learning_rate": 6.837441593495086e-07, + "loss": 0.25074785202741623, + "step": 542 + }, + { + "epoch": 2.2077393075356415, + "grad_norm": 2.4430267810821533, + "learning_rate": 6.824177477196377e-07, + "loss": 0.23671245574951172, + "step": 543 + }, + { + "epoch": 2.211812627291242, + "grad_norm": 2.4262306690216064, + "learning_rate": 6.810898535602411e-07, + "loss": 0.23878173530101776, + "step": 544 + }, + { + "epoch": 2.2158859470468433, + "grad_norm": 2.2304294109344482, + "learning_rate": 6.797604876632632e-07, + "loss": 0.2229127734899521, + "step": 545 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 2.443239450454712, + "learning_rate": 6.784296608326107e-07, + "loss": 0.22968536615371704, + "step": 546 + }, + { + "epoch": 2.224032586558045, + "grad_norm": 2.5297110080718994, + "learning_rate": 6.770973838840622e-07, + "loss": 0.2511328458786011, + "step": 547 + }, + { + "epoch": 2.2281059063136457, + "grad_norm": 2.648350238800049, + "learning_rate": 6.757636676451823e-07, + "loss": 0.2674560844898224, + "step": 548 + }, + { + "epoch": 2.2321792260692463, + "grad_norm": 2.1822292804718018, + "learning_rate": 6.744285229552327e-07, + "loss": 0.19531694799661636, + "step": 549 + }, + { + "epoch": 2.2362525458248474, + "grad_norm": 2.4335954189300537, + "learning_rate": 6.730919606650841e-07, + "loss": 0.23324161767959595, + "step": 550 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 2.5245561599731445, + "learning_rate": 6.717539916371288e-07, + "loss": 0.22070679813623428, + "step": 551 + }, + { + "epoch": 2.2443991853360488, + "grad_norm": 2.4031879901885986, + "learning_rate": 6.704146267451908e-07, + "loss": 0.20751216262578964, + "step": 552 + }, + { + "epoch": 2.24847250509165, + "grad_norm": 2.4731009006500244, + "learning_rate": 6.690738768744394e-07, + "loss": 0.23825813084840775, + "step": 553 + }, + { + "epoch": 2.2525458248472505, + "grad_norm": 2.62092924118042, + "learning_rate": 6.677317529212993e-07, + "loss": 0.23570290952920914, + "step": 554 + }, + { + "epoch": 2.256619144602851, + "grad_norm": 2.4778685569763184, + "learning_rate": 6.663882657933626e-07, + "loss": 0.22070936858654022, + "step": 555 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 2.3577542304992676, + "learning_rate": 6.650434264093e-07, + "loss": 0.2068188190460205, + "step": 556 + }, + { + "epoch": 2.264765784114053, + "grad_norm": 2.421339988708496, + "learning_rate": 6.636972456987725e-07, + "loss": 0.25278185307979584, + "step": 557 + }, + { + "epoch": 2.2688391038696536, + "grad_norm": 2.4010040760040283, + "learning_rate": 6.623497346023417e-07, + "loss": 0.24000639468431473, + "step": 558 + }, + { + "epoch": 2.2729124236252547, + "grad_norm": 2.54424786567688, + "learning_rate": 6.610009040713818e-07, + "loss": 0.27193646132946014, + "step": 559 + }, + { + "epoch": 2.2769857433808554, + "grad_norm": 2.5175139904022217, + "learning_rate": 6.596507650679899e-07, + "loss": 0.23332953453063965, + "step": 560 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 2.4466259479522705, + "learning_rate": 6.582993285648976e-07, + "loss": 0.2428213581442833, + "step": 561 + }, + { + "epoch": 2.285132382892057, + "grad_norm": 2.4644064903259277, + "learning_rate": 6.569466055453807e-07, + "loss": 0.21874462068080902, + "step": 562 + }, + { + "epoch": 2.2892057026476578, + "grad_norm": 2.250108480453491, + "learning_rate": 6.555926070031716e-07, + "loss": 0.20208529382944107, + "step": 563 + }, + { + "epoch": 2.293279022403259, + "grad_norm": 2.5665433406829834, + "learning_rate": 6.542373439423683e-07, + "loss": 0.22964468598365784, + "step": 564 + }, + { + "epoch": 2.2973523421588595, + "grad_norm": 2.6082041263580322, + "learning_rate": 6.528808273773459e-07, + "loss": 0.24122490733861923, + "step": 565 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 2.4170379638671875, + "learning_rate": 6.515230683326669e-07, + "loss": 0.21298009902238846, + "step": 566 + }, + { + "epoch": 2.3054989816700613, + "grad_norm": 2.651233196258545, + "learning_rate": 6.501640778429917e-07, + "loss": 0.24708368629217148, + "step": 567 + }, + { + "epoch": 2.309572301425662, + "grad_norm": 2.462790012359619, + "learning_rate": 6.488038669529886e-07, + "loss": 0.2247847020626068, + "step": 568 + }, + { + "epoch": 2.3136456211812626, + "grad_norm": 2.2574470043182373, + "learning_rate": 6.474424467172442e-07, + "loss": 0.21436495333909988, + "step": 569 + }, + { + "epoch": 2.3177189409368637, + "grad_norm": 2.4516212940216064, + "learning_rate": 6.460798282001738e-07, + "loss": 0.2516315132379532, + "step": 570 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 2.514031410217285, + "learning_rate": 6.447160224759311e-07, + "loss": 0.2250988855957985, + "step": 571 + }, + { + "epoch": 2.325865580448065, + "grad_norm": 2.3617303371429443, + "learning_rate": 6.433510406283185e-07, + "loss": 0.23347290605306625, + "step": 572 + }, + { + "epoch": 2.329938900203666, + "grad_norm": 2.392488479614258, + "learning_rate": 6.419848937506964e-07, + "loss": 0.25712814927101135, + "step": 573 + }, + { + "epoch": 2.3340122199592668, + "grad_norm": 2.4351253509521484, + "learning_rate": 6.406175929458944e-07, + "loss": 0.23134000599384308, + "step": 574 + }, + { + "epoch": 2.3380855397148674, + "grad_norm": 2.351980447769165, + "learning_rate": 6.392491493261198e-07, + "loss": 0.24626444280147552, + "step": 575 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 2.2502481937408447, + "learning_rate": 6.37879574012867e-07, + "loss": 0.2008371651172638, + "step": 576 + }, + { + "epoch": 2.346232179226069, + "grad_norm": 2.413408041000366, + "learning_rate": 6.36508878136829e-07, + "loss": 0.24356064200401306, + "step": 577 + }, + { + "epoch": 2.35030549898167, + "grad_norm": 2.358431577682495, + "learning_rate": 6.351370728378049e-07, + "loss": 0.25313031673431396, + "step": 578 + }, + { + "epoch": 2.354378818737271, + "grad_norm": 2.580531358718872, + "learning_rate": 6.337641692646106e-07, + "loss": 0.2431691735982895, + "step": 579 + }, + { + "epoch": 2.3584521384928716, + "grad_norm": 2.4596192836761475, + "learning_rate": 6.323901785749871e-07, + "loss": 0.2600405141711235, + "step": 580 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 2.41133975982666, + "learning_rate": 6.310151119355118e-07, + "loss": 0.1907319650053978, + "step": 581 + }, + { + "epoch": 2.3665987780040734, + "grad_norm": 2.5591516494750977, + "learning_rate": 6.296389805215054e-07, + "loss": 0.2532680034637451, + "step": 582 + }, + { + "epoch": 2.370672097759674, + "grad_norm": 2.362185478210449, + "learning_rate": 6.282617955169425e-07, + "loss": 0.2242288738489151, + "step": 583 + }, + { + "epoch": 2.374745417515275, + "grad_norm": 2.438898801803589, + "learning_rate": 6.268835681143602e-07, + "loss": 0.2321534976363182, + "step": 584 + }, + { + "epoch": 2.378818737270876, + "grad_norm": 2.066927433013916, + "learning_rate": 6.255043095147678e-07, + "loss": 0.2276027500629425, + "step": 585 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 2.347784996032715, + "learning_rate": 6.241240309275545e-07, + "loss": 0.2264810875058174, + "step": 586 + }, + { + "epoch": 2.3869653767820775, + "grad_norm": 2.444033145904541, + "learning_rate": 6.227427435703995e-07, + "loss": 0.2255028337240219, + "step": 587 + }, + { + "epoch": 2.391038696537678, + "grad_norm": 2.2944633960723877, + "learning_rate": 6.213604586691805e-07, + "loss": 0.2329559102654457, + "step": 588 + }, + { + "epoch": 2.395112016293279, + "grad_norm": 2.346848964691162, + "learning_rate": 6.199771874578819e-07, + "loss": 0.22230961173772812, + "step": 589 + }, + { + "epoch": 2.39918533604888, + "grad_norm": 2.4055089950561523, + "learning_rate": 6.185929411785042e-07, + "loss": 0.22255368530750275, + "step": 590 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 2.5555362701416016, + "learning_rate": 6.172077310809724e-07, + "loss": 0.25746987015008926, + "step": 591 + }, + { + "epoch": 2.4073319755600817, + "grad_norm": 2.5292749404907227, + "learning_rate": 6.15821568423045e-07, + "loss": 0.1872299611568451, + "step": 592 + }, + { + "epoch": 2.4114052953156824, + "grad_norm": 2.6285977363586426, + "learning_rate": 6.144344644702211e-07, + "loss": 0.25099916756153107, + "step": 593 + }, + { + "epoch": 2.415478615071283, + "grad_norm": 2.830077886581421, + "learning_rate": 6.130464304956504e-07, + "loss": 0.26932021975517273, + "step": 594 + }, + { + "epoch": 2.4195519348268837, + "grad_norm": 2.4076106548309326, + "learning_rate": 6.116574777800412e-07, + "loss": 0.25139734894037247, + "step": 595 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 2.45658540725708, + "learning_rate": 6.102676176115681e-07, + "loss": 0.2526377737522125, + "step": 596 + }, + { + "epoch": 2.4276985743380854, + "grad_norm": 2.455595016479492, + "learning_rate": 6.088768612857807e-07, + "loss": 0.20270772278308868, + "step": 597 + }, + { + "epoch": 2.4317718940936865, + "grad_norm": 2.5158026218414307, + "learning_rate": 6.074852201055121e-07, + "loss": 0.2484816238284111, + "step": 598 + }, + { + "epoch": 2.435845213849287, + "grad_norm": 2.8085286617279053, + "learning_rate": 6.060927053807863e-07, + "loss": 0.2813955247402191, + "step": 599 + }, + { + "epoch": 2.439918533604888, + "grad_norm": 2.644648313522339, + "learning_rate": 6.046993284287267e-07, + "loss": 0.2188793271780014, + "step": 600 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 2.362947702407837, + "learning_rate": 6.033051005734647e-07, + "loss": 0.2174539864063263, + "step": 601 + }, + { + "epoch": 2.4480651731160896, + "grad_norm": 2.469346284866333, + "learning_rate": 6.019100331460466e-07, + "loss": 0.24562832713127136, + "step": 602 + }, + { + "epoch": 2.4521384928716903, + "grad_norm": 2.338634729385376, + "learning_rate": 6.005141374843419e-07, + "loss": 0.23371918499469757, + "step": 603 + }, + { + "epoch": 2.4562118126272914, + "grad_norm": 2.3929550647735596, + "learning_rate": 5.991174249329514e-07, + "loss": 0.23879797011613846, + "step": 604 + }, + { + "epoch": 2.460285132382892, + "grad_norm": 2.257073163986206, + "learning_rate": 5.977199068431153e-07, + "loss": 0.22032149881124496, + "step": 605 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 2.408695936203003, + "learning_rate": 5.963215945726198e-07, + "loss": 0.2490229532122612, + "step": 606 + }, + { + "epoch": 2.468431771894094, + "grad_norm": 2.3911116123199463, + "learning_rate": 5.949224994857057e-07, + "loss": 0.23558590561151505, + "step": 607 + }, + { + "epoch": 2.4725050916496945, + "grad_norm": 2.351417303085327, + "learning_rate": 5.93522632952976e-07, + "loss": 0.22598809003829956, + "step": 608 + }, + { + "epoch": 2.4765784114052956, + "grad_norm": 2.258190870285034, + "learning_rate": 5.921220063513034e-07, + "loss": 0.17976737767457962, + "step": 609 + }, + { + "epoch": 2.480651731160896, + "grad_norm": 2.431304454803467, + "learning_rate": 5.907206310637375e-07, + "loss": 0.209254652261734, + "step": 610 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 2.3682570457458496, + "learning_rate": 5.893185184794128e-07, + "loss": 0.23237604647874832, + "step": 611 + }, + { + "epoch": 2.4887983706720975, + "grad_norm": 2.473200798034668, + "learning_rate": 5.879156799934554e-07, + "loss": 0.21567458659410477, + "step": 612 + }, + { + "epoch": 2.4928716904276986, + "grad_norm": 2.312518358230591, + "learning_rate": 5.865121270068916e-07, + "loss": 0.20658506453037262, + "step": 613 + }, + { + "epoch": 2.4969450101832993, + "grad_norm": 2.5240719318389893, + "learning_rate": 5.851078709265541e-07, + "loss": 0.2315971404314041, + "step": 614 + }, + { + "epoch": 2.5010183299389004, + "grad_norm": 2.444183349609375, + "learning_rate": 5.837029231649898e-07, + "loss": 0.248056061565876, + "step": 615 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 2.928251028060913, + "learning_rate": 5.82297295140367e-07, + "loss": 0.190412700176239, + "step": 616 + }, + { + "epoch": 2.5091649694501017, + "grad_norm": 2.3546931743621826, + "learning_rate": 5.808909982763825e-07, + "loss": 0.2290133684873581, + "step": 617 + }, + { + "epoch": 2.513238289205703, + "grad_norm": 2.4045937061309814, + "learning_rate": 5.794840440021686e-07, + "loss": 0.20809811353683472, + "step": 618 + }, + { + "epoch": 2.5173116089613035, + "grad_norm": 2.304481029510498, + "learning_rate": 5.780764437522012e-07, + "loss": 0.21659423410892487, + "step": 619 + }, + { + "epoch": 2.521384928716904, + "grad_norm": 2.5403997898101807, + "learning_rate": 5.766682089662054e-07, + "loss": 0.24090662598609924, + "step": 620 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 2.365463972091675, + "learning_rate": 5.752593510890635e-07, + "loss": 0.22003071755170822, + "step": 621 + }, + { + "epoch": 2.529531568228106, + "grad_norm": 2.4152722358703613, + "learning_rate": 5.738498815707219e-07, + "loss": 0.22138181328773499, + "step": 622 + }, + { + "epoch": 2.5336048879837065, + "grad_norm": 2.3184590339660645, + "learning_rate": 5.724398118660973e-07, + "loss": 0.22900409996509552, + "step": 623 + }, + { + "epoch": 2.5376782077393076, + "grad_norm": 2.4730820655822754, + "learning_rate": 5.710291534349849e-07, + "loss": 0.2507135570049286, + "step": 624 + }, + { + "epoch": 2.5417515274949083, + "grad_norm": 2.771104097366333, + "learning_rate": 5.696179177419642e-07, + "loss": 0.2613483816385269, + "step": 625 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 2.334442377090454, + "learning_rate": 5.682061162563061e-07, + "loss": 0.19708166271448135, + "step": 626 + }, + { + "epoch": 2.54989816700611, + "grad_norm": 2.361771821975708, + "learning_rate": 5.667937604518798e-07, + "loss": 0.2150387316942215, + "step": 627 + }, + { + "epoch": 2.5539714867617107, + "grad_norm": 2.5752477645874023, + "learning_rate": 5.653808618070597e-07, + "loss": 0.25150199234485626, + "step": 628 + }, + { + "epoch": 2.5580448065173114, + "grad_norm": 2.685345411300659, + "learning_rate": 5.639674318046317e-07, + "loss": 0.22753620892763138, + "step": 629 + }, + { + "epoch": 2.5621181262729125, + "grad_norm": 2.3737359046936035, + "learning_rate": 5.625534819317004e-07, + "loss": 0.2346680760383606, + "step": 630 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 2.5365428924560547, + "learning_rate": 5.61139023679595e-07, + "loss": 0.27340593934059143, + "step": 631 + }, + { + "epoch": 2.5702647657841142, + "grad_norm": 2.5359578132629395, + "learning_rate": 5.597240685437765e-07, + "loss": 0.21642443537712097, + "step": 632 + }, + { + "epoch": 2.574338085539715, + "grad_norm": 2.651430130004883, + "learning_rate": 5.583086280237446e-07, + "loss": 0.2263542339205742, + "step": 633 + }, + { + "epoch": 2.5784114052953155, + "grad_norm": 2.7270631790161133, + "learning_rate": 5.568927136229432e-07, + "loss": 0.2491724044084549, + "step": 634 + }, + { + "epoch": 2.5824847250509166, + "grad_norm": 2.3526268005371094, + "learning_rate": 5.554763368486674e-07, + "loss": 0.21600161492824554, + "step": 635 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 2.307650327682495, + "learning_rate": 5.540595092119708e-07, + "loss": 0.19877354055643082, + "step": 636 + }, + { + "epoch": 2.5906313645621184, + "grad_norm": 2.3778960704803467, + "learning_rate": 5.526422422275707e-07, + "loss": 0.2133459597826004, + "step": 637 + }, + { + "epoch": 2.594704684317719, + "grad_norm": 2.5685219764709473, + "learning_rate": 5.512245474137546e-07, + "loss": 0.2637478709220886, + "step": 638 + }, + { + "epoch": 2.5987780040733197, + "grad_norm": 2.3442726135253906, + "learning_rate": 5.498064362922882e-07, + "loss": 0.2268994301557541, + "step": 639 + }, + { + "epoch": 2.6028513238289204, + "grad_norm": 2.780430793762207, + "learning_rate": 5.483879203883194e-07, + "loss": 0.24991512298583984, + "step": 640 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 2.5000061988830566, + "learning_rate": 5.469690112302863e-07, + "loss": 0.2421465590596199, + "step": 641 + }, + { + "epoch": 2.610997963340122, + "grad_norm": 2.5583932399749756, + "learning_rate": 5.455497203498231e-07, + "loss": 0.22834140062332153, + "step": 642 + }, + { + "epoch": 2.6150712830957232, + "grad_norm": 2.4214282035827637, + "learning_rate": 5.441300592816662e-07, + "loss": 0.24269834905862808, + "step": 643 + }, + { + "epoch": 2.619144602851324, + "grad_norm": 2.352479934692383, + "learning_rate": 5.427100395635607e-07, + "loss": 0.23605701327323914, + "step": 644 + }, + { + "epoch": 2.6232179226069245, + "grad_norm": 2.3162059783935547, + "learning_rate": 5.412896727361662e-07, + "loss": 0.2196669727563858, + "step": 645 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 2.438443183898926, + "learning_rate": 5.398689703429634e-07, + "loss": 0.23577384650707245, + "step": 646 + }, + { + "epoch": 2.6313645621181263, + "grad_norm": 2.6072158813476562, + "learning_rate": 5.384479439301605e-07, + "loss": 0.22815095633268356, + "step": 647 + }, + { + "epoch": 2.635437881873727, + "grad_norm": 2.495229959487915, + "learning_rate": 5.370266050465983e-07, + "loss": 0.21826496720314026, + "step": 648 + }, + { + "epoch": 2.639511201629328, + "grad_norm": 2.4141361713409424, + "learning_rate": 5.356049652436579e-07, + "loss": 0.2289327010512352, + "step": 649 + }, + { + "epoch": 2.6435845213849287, + "grad_norm": 2.461887836456299, + "learning_rate": 5.341830360751658e-07, + "loss": 0.24295459687709808, + "step": 650 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 2.4014673233032227, + "learning_rate": 5.327608290972998e-07, + "loss": 0.23736917972564697, + "step": 651 + }, + { + "epoch": 2.6517311608961305, + "grad_norm": 2.3798553943634033, + "learning_rate": 5.313383558684957e-07, + "loss": 0.21394247561693192, + "step": 652 + }, + { + "epoch": 2.655804480651731, + "grad_norm": 2.433729648590088, + "learning_rate": 5.299156279493535e-07, + "loss": 0.22641988098621368, + "step": 653 + }, + { + "epoch": 2.6598778004073322, + "grad_norm": 2.801403284072876, + "learning_rate": 5.284926569025428e-07, + "loss": 0.19061069190502167, + "step": 654 + }, + { + "epoch": 2.663951120162933, + "grad_norm": 2.345667839050293, + "learning_rate": 5.270694542927088e-07, + "loss": 0.2233218401670456, + "step": 655 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 2.7157375812530518, + "learning_rate": 5.256460316863791e-07, + "loss": 0.268241822719574, + "step": 656 + }, + { + "epoch": 2.672097759674134, + "grad_norm": 2.566452980041504, + "learning_rate": 5.242224006518694e-07, + "loss": 0.23765669763088226, + "step": 657 + }, + { + "epoch": 2.6761710794297353, + "grad_norm": 2.395622491836548, + "learning_rate": 5.227985727591888e-07, + "loss": 0.2502652183175087, + "step": 658 + }, + { + "epoch": 2.680244399185336, + "grad_norm": 2.3431246280670166, + "learning_rate": 5.213745595799462e-07, + "loss": 0.2402183562517166, + "step": 659 + }, + { + "epoch": 2.684317718940937, + "grad_norm": 2.374241352081299, + "learning_rate": 5.199503726872573e-07, + "loss": 0.23906593769788742, + "step": 660 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 2.4988749027252197, + "learning_rate": 5.185260236556484e-07, + "loss": 0.23993954807519913, + "step": 661 + }, + { + "epoch": 2.6924643584521384, + "grad_norm": 2.4878809452056885, + "learning_rate": 5.171015240609644e-07, + "loss": 0.23122139275074005, + "step": 662 + }, + { + "epoch": 2.696537678207739, + "grad_norm": 2.6239712238311768, + "learning_rate": 5.156768854802734e-07, + "loss": 0.2619614750146866, + "step": 663 + }, + { + "epoch": 2.70061099796334, + "grad_norm": 3.4436697959899902, + "learning_rate": 5.142521194917733e-07, + "loss": 0.24487808346748352, + "step": 664 + }, + { + "epoch": 2.704684317718941, + "grad_norm": 2.537623643875122, + "learning_rate": 5.128272376746971e-07, + "loss": 0.2131681889295578, + "step": 665 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 2.5551114082336426, + "learning_rate": 5.114022516092194e-07, + "loss": 0.2128780037164688, + "step": 666 + }, + { + "epoch": 2.7128309572301426, + "grad_norm": 2.3989691734313965, + "learning_rate": 5.099771728763623e-07, + "loss": 0.24740803986787796, + "step": 667 + }, + { + "epoch": 2.716904276985743, + "grad_norm": 2.2155869007110596, + "learning_rate": 5.085520130579005e-07, + "loss": 0.19293303787708282, + "step": 668 + }, + { + "epoch": 2.7209775967413443, + "grad_norm": 2.773521661758423, + "learning_rate": 5.07126783736268e-07, + "loss": 0.24105177074670792, + "step": 669 + }, + { + "epoch": 2.725050916496945, + "grad_norm": 2.3562657833099365, + "learning_rate": 5.057014964944634e-07, + "loss": 0.2287985384464264, + "step": 670 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 2.4878883361816406, + "learning_rate": 5.042761629159566e-07, + "loss": 0.2295224368572235, + "step": 671 + }, + { + "epoch": 2.7331975560081467, + "grad_norm": 2.9576029777526855, + "learning_rate": 5.028507945845932e-07, + "loss": 0.24781977385282516, + "step": 672 + }, + { + "epoch": 2.7372708757637474, + "grad_norm": 2.7853589057922363, + "learning_rate": 5.014254030845021e-07, + "loss": 0.2594607025384903, + "step": 673 + }, + { + "epoch": 2.741344195519348, + "grad_norm": 2.313760280609131, + "learning_rate": 5e-07, + "loss": 0.2149473801255226, + "step": 674 + }, + { + "epoch": 2.745417515274949, + "grad_norm": 2.6287894248962402, + "learning_rate": 4.98574596915498e-07, + "loss": 0.2292870730161667, + "step": 675 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 2.3970835208892822, + "learning_rate": 4.971492054154068e-07, + "loss": 0.24747398495674133, + "step": 676 + }, + { + "epoch": 2.753564154786151, + "grad_norm": 2.5344555377960205, + "learning_rate": 4.957238370840436e-07, + "loss": 0.24396724253892899, + "step": 677 + }, + { + "epoch": 2.7576374745417516, + "grad_norm": 2.3144986629486084, + "learning_rate": 4.942985035055366e-07, + "loss": 0.2170693725347519, + "step": 678 + }, + { + "epoch": 2.7617107942973522, + "grad_norm": 2.3190531730651855, + "learning_rate": 4.928732162637321e-07, + "loss": 0.22908472269773483, + "step": 679 + }, + { + "epoch": 2.765784114052953, + "grad_norm": 2.263834238052368, + "learning_rate": 4.914479869420994e-07, + "loss": 0.22652582079172134, + "step": 680 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 2.4412286281585693, + "learning_rate": 4.900228271236377e-07, + "loss": 0.22030826658010483, + "step": 681 + }, + { + "epoch": 2.7739307535641546, + "grad_norm": 2.443682909011841, + "learning_rate": 4.885977483907804e-07, + "loss": 0.22983287274837494, + "step": 682 + }, + { + "epoch": 2.7780040733197557, + "grad_norm": 2.451335906982422, + "learning_rate": 4.871727623253028e-07, + "loss": 0.22219268232584, + "step": 683 + }, + { + "epoch": 2.7820773930753564, + "grad_norm": 2.3448593616485596, + "learning_rate": 4.857478805082267e-07, + "loss": 0.2025885134935379, + "step": 684 + }, + { + "epoch": 2.786150712830957, + "grad_norm": 2.2007482051849365, + "learning_rate": 4.843231145197266e-07, + "loss": 0.2085644006729126, + "step": 685 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 2.331252336502075, + "learning_rate": 4.828984759390356e-07, + "loss": 0.20427662134170532, + "step": 686 + }, + { + "epoch": 2.794297352342159, + "grad_norm": 2.243265151977539, + "learning_rate": 4.814739763443515e-07, + "loss": 0.2013532519340515, + "step": 687 + }, + { + "epoch": 2.79837067209776, + "grad_norm": 2.581249475479126, + "learning_rate": 4.800496273127429e-07, + "loss": 0.2316228449344635, + "step": 688 + }, + { + "epoch": 2.8024439918533606, + "grad_norm": 2.5763373374938965, + "learning_rate": 4.786254404200538e-07, + "loss": 0.2278730794787407, + "step": 689 + }, + { + "epoch": 2.8065173116089612, + "grad_norm": 2.4369475841522217, + "learning_rate": 4.772014272408114e-07, + "loss": 0.22459274530410767, + "step": 690 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 2.458749294281006, + "learning_rate": 4.757775993481306e-07, + "loss": 0.21142005175352097, + "step": 691 + }, + { + "epoch": 2.814663951120163, + "grad_norm": 2.3797214031219482, + "learning_rate": 4.743539683136209e-07, + "loss": 0.23049592226743698, + "step": 692 + }, + { + "epoch": 2.8187372708757636, + "grad_norm": 2.5822553634643555, + "learning_rate": 4.7293054570729126e-07, + "loss": 0.2521442621946335, + "step": 693 + }, + { + "epoch": 2.8228105906313647, + "grad_norm": 2.489661931991577, + "learning_rate": 4.715073430974573e-07, + "loss": 0.24500177055597305, + "step": 694 + }, + { + "epoch": 2.8268839103869654, + "grad_norm": 2.592890977859497, + "learning_rate": 4.7008437205064634e-07, + "loss": 0.2384454905986786, + "step": 695 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 2.5368080139160156, + "learning_rate": 4.686616441315043e-07, + "loss": 0.25886720418930054, + "step": 696 + }, + { + "epoch": 2.835030549898167, + "grad_norm": 2.4768271446228027, + "learning_rate": 4.672391709027002e-07, + "loss": 0.2593514025211334, + "step": 697 + }, + { + "epoch": 2.839103869653768, + "grad_norm": 2.3749191761016846, + "learning_rate": 4.658169639248342e-07, + "loss": 0.19730783253908157, + "step": 698 + }, + { + "epoch": 2.8431771894093685, + "grad_norm": 2.4880032539367676, + "learning_rate": 4.643950347563421e-07, + "loss": 0.22945521771907806, + "step": 699 + }, + { + "epoch": 2.8472505091649696, + "grad_norm": 2.3967182636260986, + "learning_rate": 4.6297339495340165e-07, + "loss": 0.24117758870124817, + "step": 700 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 2.144585609436035, + "learning_rate": 4.615520560698397e-07, + "loss": 0.181608684360981, + "step": 701 + }, + { + "epoch": 2.855397148676171, + "grad_norm": 2.489408493041992, + "learning_rate": 4.601310296570366e-07, + "loss": 0.27073855698108673, + "step": 702 + }, + { + "epoch": 2.859470468431772, + "grad_norm": 2.4170804023742676, + "learning_rate": 4.5871032726383385e-07, + "loss": 0.20901280641555786, + "step": 703 + }, + { + "epoch": 2.8635437881873727, + "grad_norm": 2.4658868312835693, + "learning_rate": 4.572899604364392e-07, + "loss": 0.22699516266584396, + "step": 704 + }, + { + "epoch": 2.8676171079429738, + "grad_norm": 2.32704758644104, + "learning_rate": 4.5586994071833377e-07, + "loss": 0.2194635197520256, + "step": 705 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 2.540724277496338, + "learning_rate": 4.5445027965017683e-07, + "loss": 0.23322076350450516, + "step": 706 + }, + { + "epoch": 2.875763747454175, + "grad_norm": 2.4795303344726562, + "learning_rate": 4.5303098876971373e-07, + "loss": 0.22777557373046875, + "step": 707 + }, + { + "epoch": 2.8798370672097757, + "grad_norm": 2.5126326084136963, + "learning_rate": 4.516120796116806e-07, + "loss": 0.2254505679011345, + "step": 708 + }, + { + "epoch": 2.883910386965377, + "grad_norm": 2.6765761375427246, + "learning_rate": 4.5019356370771185e-07, + "loss": 0.23741194605827332, + "step": 709 + }, + { + "epoch": 2.8879837067209775, + "grad_norm": 2.261955976486206, + "learning_rate": 4.487754525862453e-07, + "loss": 0.22391848266124725, + "step": 710 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 2.5763208866119385, + "learning_rate": 4.473577577724293e-07, + "loss": 0.2411726713180542, + "step": 711 + }, + { + "epoch": 2.8961303462321792, + "grad_norm": 2.489358425140381, + "learning_rate": 4.459404907880292e-07, + "loss": 0.25273367017507553, + "step": 712 + }, + { + "epoch": 2.90020366598778, + "grad_norm": 2.3726134300231934, + "learning_rate": 4.4452366315133256e-07, + "loss": 0.20239847898483276, + "step": 713 + }, + { + "epoch": 2.904276985743381, + "grad_norm": 2.5708673000335693, + "learning_rate": 4.43107286377057e-07, + "loss": 0.23958415538072586, + "step": 714 + }, + { + "epoch": 2.9083503054989817, + "grad_norm": 2.319916248321533, + "learning_rate": 4.4169137197625537e-07, + "loss": 0.2106548771262169, + "step": 715 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 2.5788509845733643, + "learning_rate": 4.4027593145622357e-07, + "loss": 0.2554449290037155, + "step": 716 + }, + { + "epoch": 2.9164969450101834, + "grad_norm": 2.4445812702178955, + "learning_rate": 4.388609763204051e-07, + "loss": 0.22007200866937637, + "step": 717 + }, + { + "epoch": 2.920570264765784, + "grad_norm": 2.534118175506592, + "learning_rate": 4.3744651806829967e-07, + "loss": 0.24423322826623917, + "step": 718 + }, + { + "epoch": 2.9246435845213847, + "grad_norm": 2.4413700103759766, + "learning_rate": 4.3603256819536817e-07, + "loss": 0.22403902560472488, + "step": 719 + }, + { + "epoch": 2.928716904276986, + "grad_norm": 2.397890567779541, + "learning_rate": 4.3461913819294035e-07, + "loss": 0.2157697230577469, + "step": 720 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 2.469019651412964, + "learning_rate": 4.332062395481203e-07, + "loss": 0.22792110592126846, + "step": 721 + }, + { + "epoch": 2.9368635437881876, + "grad_norm": 2.254711866378784, + "learning_rate": 4.3179388374369396e-07, + "loss": 0.22459496557712555, + "step": 722 + }, + { + "epoch": 2.9409368635437882, + "grad_norm": 2.2734055519104004, + "learning_rate": 4.3038208225803594e-07, + "loss": 0.22033336013555527, + "step": 723 + }, + { + "epoch": 2.945010183299389, + "grad_norm": 2.565237283706665, + "learning_rate": 4.289708465650151e-07, + "loss": 0.24250654131174088, + "step": 724 + }, + { + "epoch": 2.9490835030549896, + "grad_norm": 2.408484935760498, + "learning_rate": 4.275601881339027e-07, + "loss": 0.22946030646562576, + "step": 725 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 2.5278210639953613, + "learning_rate": 4.261501184292782e-07, + "loss": 0.22008834779262543, + "step": 726 + }, + { + "epoch": 2.9572301425661913, + "grad_norm": 2.2692785263061523, + "learning_rate": 4.2474064891093655e-07, + "loss": 0.22004914283752441, + "step": 727 + }, + { + "epoch": 2.9613034623217924, + "grad_norm": 2.8039700984954834, + "learning_rate": 4.2333179103379445e-07, + "loss": 0.22936520725488663, + "step": 728 + }, + { + "epoch": 2.965376782077393, + "grad_norm": 2.5510387420654297, + "learning_rate": 4.2192355624779884e-07, + "loss": 0.2276434227824211, + "step": 729 + }, + { + "epoch": 2.9694501018329937, + "grad_norm": 2.512366771697998, + "learning_rate": 4.205159559978313e-07, + "loss": 0.23083247244358063, + "step": 730 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 2.492103338241577, + "learning_rate": 4.1910900172361763e-07, + "loss": 0.2253756895661354, + "step": 731 + }, + { + "epoch": 2.9775967413441955, + "grad_norm": 2.4701900482177734, + "learning_rate": 4.1770270485963294e-07, + "loss": 0.2475346326828003, + "step": 732 + }, + { + "epoch": 2.9816700610997966, + "grad_norm": 2.8806581497192383, + "learning_rate": 4.162970768350102e-07, + "loss": 0.27113020420074463, + "step": 733 + }, + { + "epoch": 2.9857433808553973, + "grad_norm": 2.374631643295288, + "learning_rate": 4.148921290734459e-07, + "loss": 0.22385699301958084, + "step": 734 + }, + { + "epoch": 2.989816700610998, + "grad_norm": 2.5273098945617676, + "learning_rate": 4.134878729931083e-07, + "loss": 0.2260419949889183, + "step": 735 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 2.4065239429473877, + "learning_rate": 4.120843200065447e-07, + "loss": 0.2051657810807228, + "step": 736 + }, + { + "epoch": 2.9979633401221997, + "grad_norm": 2.419027328491211, + "learning_rate": 4.106814815205873e-07, + "loss": 0.22403547167778015, + "step": 737 + }, + { + "epoch": 3.0, + "grad_norm": 2.419027328491211, + "learning_rate": 4.092793689362625e-07, + "loss": 0.2715803384780884, + "step": 738 + }, + { + "epoch": 3.0040733197556007, + "grad_norm": 2.3535571098327637, + "learning_rate": 4.078779936486965e-07, + "loss": 0.21160051226615906, + "step": 739 + }, + { + "epoch": 3.0081466395112018, + "grad_norm": 2.379765272140503, + "learning_rate": 4.06477367047024e-07, + "loss": 0.22423982620239258, + "step": 740 + }, + { + "epoch": 3.0122199592668024, + "grad_norm": 2.3597190380096436, + "learning_rate": 4.050775005142943e-07, + "loss": 0.22356468439102173, + "step": 741 + }, + { + "epoch": 3.016293279022403, + "grad_norm": 2.3479771614074707, + "learning_rate": 4.036784054273803e-07, + "loss": 0.1949443370103836, + "step": 742 + }, + { + "epoch": 3.020366598778004, + "grad_norm": 2.1719963550567627, + "learning_rate": 4.0228009315688463e-07, + "loss": 0.16644436120986938, + "step": 743 + }, + { + "epoch": 3.024439918533605, + "grad_norm": 2.2947304248809814, + "learning_rate": 4.0088257506704853e-07, + "loss": 0.21931639313697815, + "step": 744 + }, + { + "epoch": 3.0285132382892055, + "grad_norm": 2.383701801300049, + "learning_rate": 3.994858625156582e-07, + "loss": 0.2001979798078537, + "step": 745 + }, + { + "epoch": 3.0325865580448066, + "grad_norm": 2.237454891204834, + "learning_rate": 3.9808996685395344e-07, + "loss": 0.20605531334877014, + "step": 746 + }, + { + "epoch": 3.0366598778004072, + "grad_norm": 2.3682029247283936, + "learning_rate": 3.966948994265354e-07, + "loss": 0.21250516921281815, + "step": 747 + }, + { + "epoch": 3.0407331975560083, + "grad_norm": 2.4874134063720703, + "learning_rate": 3.953006715712733e-07, + "loss": 0.21023060381412506, + "step": 748 + }, + { + "epoch": 3.044806517311609, + "grad_norm": 2.21878719329834, + "learning_rate": 3.939072946192139e-07, + "loss": 0.22278980910778046, + "step": 749 + }, + { + "epoch": 3.0488798370672097, + "grad_norm": 2.3506107330322266, + "learning_rate": 3.9251477989448795e-07, + "loss": 0.2010231390595436, + "step": 750 + }, + { + "epoch": 3.0529531568228108, + "grad_norm": 2.3747055530548096, + "learning_rate": 3.9112313871421937e-07, + "loss": 0.18845809996128082, + "step": 751 + }, + { + "epoch": 3.0570264765784114, + "grad_norm": 2.1489953994750977, + "learning_rate": 3.897323823884318e-07, + "loss": 0.19643110036849976, + "step": 752 + }, + { + "epoch": 3.061099796334012, + "grad_norm": 2.3836452960968018, + "learning_rate": 3.8834252221995877e-07, + "loss": 0.22072193771600723, + "step": 753 + }, + { + "epoch": 3.065173116089613, + "grad_norm": 2.418147563934326, + "learning_rate": 3.8695356950434945e-07, + "loss": 0.2128564417362213, + "step": 754 + }, + { + "epoch": 3.069246435845214, + "grad_norm": 2.341552734375, + "learning_rate": 3.855655355297789e-07, + "loss": 0.21925050020217896, + "step": 755 + }, + { + "epoch": 3.0733197556008145, + "grad_norm": 2.7095723152160645, + "learning_rate": 3.8417843157695497e-07, + "loss": 0.19232793152332306, + "step": 756 + }, + { + "epoch": 3.0773930753564156, + "grad_norm": 2.275331497192383, + "learning_rate": 3.827922689190275e-07, + "loss": 0.203746996819973, + "step": 757 + }, + { + "epoch": 3.0814663951120163, + "grad_norm": 2.3823604583740234, + "learning_rate": 3.8140705882149585e-07, + "loss": 0.20427367091178894, + "step": 758 + }, + { + "epoch": 3.085539714867617, + "grad_norm": 3.103059768676758, + "learning_rate": 3.8002281254211815e-07, + "loss": 0.22299692034721375, + "step": 759 + }, + { + "epoch": 3.089613034623218, + "grad_norm": 2.3718128204345703, + "learning_rate": 3.7863954133081966e-07, + "loss": 0.18740925192832947, + "step": 760 + }, + { + "epoch": 3.0936863543788187, + "grad_norm": 2.494307518005371, + "learning_rate": 3.772572564296004e-07, + "loss": 0.18933183699846268, + "step": 761 + }, + { + "epoch": 3.0977596741344193, + "grad_norm": 2.2291083335876465, + "learning_rate": 3.7587596907244545e-07, + "loss": 0.20075885951519012, + "step": 762 + }, + { + "epoch": 3.1018329938900204, + "grad_norm": 2.42327618598938, + "learning_rate": 3.744956904852321e-07, + "loss": 0.20842499285936356, + "step": 763 + }, + { + "epoch": 3.105906313645621, + "grad_norm": 2.4771366119384766, + "learning_rate": 3.7311643188563967e-07, + "loss": 0.20921579748392105, + "step": 764 + }, + { + "epoch": 3.109979633401222, + "grad_norm": 2.459636688232422, + "learning_rate": 3.717382044830575e-07, + "loss": 0.20464809238910675, + "step": 765 + }, + { + "epoch": 3.114052953156823, + "grad_norm": 2.4493255615234375, + "learning_rate": 3.7036101947849456e-07, + "loss": 0.19754133373498917, + "step": 766 + }, + { + "epoch": 3.1181262729124235, + "grad_norm": 2.426888942718506, + "learning_rate": 3.6898488806448807e-07, + "loss": 0.22133717685937881, + "step": 767 + }, + { + "epoch": 3.1221995926680246, + "grad_norm": 2.7352089881896973, + "learning_rate": 3.6760982142501284e-07, + "loss": 0.19161057472229004, + "step": 768 + }, + { + "epoch": 3.1262729124236253, + "grad_norm": 2.271554946899414, + "learning_rate": 3.6623583073538965e-07, + "loss": 0.1958870366215706, + "step": 769 + }, + { + "epoch": 3.130346232179226, + "grad_norm": 2.6821932792663574, + "learning_rate": 3.6486292716219514e-07, + "loss": 0.20984943211078644, + "step": 770 + }, + { + "epoch": 3.134419551934827, + "grad_norm": 2.6320438385009766, + "learning_rate": 3.634911218631711e-07, + "loss": 0.20410407334566116, + "step": 771 + }, + { + "epoch": 3.1384928716904277, + "grad_norm": 2.2215492725372314, + "learning_rate": 3.6212042598713296e-07, + "loss": 0.21011168509721756, + "step": 772 + }, + { + "epoch": 3.1425661914460283, + "grad_norm": 2.485713005065918, + "learning_rate": 3.607508506738803e-07, + "loss": 0.20080970227718353, + "step": 773 + }, + { + "epoch": 3.1466395112016294, + "grad_norm": 2.4904892444610596, + "learning_rate": 3.5938240705410537e-07, + "loss": 0.21116189658641815, + "step": 774 + }, + { + "epoch": 3.15071283095723, + "grad_norm": 2.4921395778656006, + "learning_rate": 3.580151062493036e-07, + "loss": 0.1990169882774353, + "step": 775 + }, + { + "epoch": 3.1547861507128308, + "grad_norm": 2.323054790496826, + "learning_rate": 3.566489593716816e-07, + "loss": 0.19865593314170837, + "step": 776 + }, + { + "epoch": 3.158859470468432, + "grad_norm": 2.533299684524536, + "learning_rate": 3.5528397752406894e-07, + "loss": 0.19558026641607285, + "step": 777 + }, + { + "epoch": 3.1629327902240325, + "grad_norm": 2.2772974967956543, + "learning_rate": 3.5392017179982613e-07, + "loss": 0.18929005414247513, + "step": 778 + }, + { + "epoch": 3.167006109979633, + "grad_norm": 2.433457374572754, + "learning_rate": 3.5255755328275584e-07, + "loss": 0.2181885540485382, + "step": 779 + }, + { + "epoch": 3.1710794297352343, + "grad_norm": 2.350560426712036, + "learning_rate": 3.511961330470115e-07, + "loss": 0.19552721083164215, + "step": 780 + }, + { + "epoch": 3.175152749490835, + "grad_norm": 2.4112911224365234, + "learning_rate": 3.498359221570083e-07, + "loss": 0.20244888216257095, + "step": 781 + }, + { + "epoch": 3.179226069246436, + "grad_norm": 2.4253125190734863, + "learning_rate": 3.484769316673331e-07, + "loss": 0.22686784714460373, + "step": 782 + }, + { + "epoch": 3.1832993890020367, + "grad_norm": 2.535693645477295, + "learning_rate": 3.471191726226541e-07, + "loss": 0.22888771444559097, + "step": 783 + }, + { + "epoch": 3.1873727087576373, + "grad_norm": 2.3272151947021484, + "learning_rate": 3.4576265605763185e-07, + "loss": 0.20377343893051147, + "step": 784 + }, + { + "epoch": 3.1914460285132384, + "grad_norm": 2.6908118724823, + "learning_rate": 3.444073929968284e-07, + "loss": 0.20916848629713058, + "step": 785 + }, + { + "epoch": 3.195519348268839, + "grad_norm": 2.3921995162963867, + "learning_rate": 3.4305339445461923e-07, + "loss": 0.1810828298330307, + "step": 786 + }, + { + "epoch": 3.1995926680244398, + "grad_norm": 2.4223461151123047, + "learning_rate": 3.417006714351024e-07, + "loss": 0.2161688134074211, + "step": 787 + }, + { + "epoch": 3.203665987780041, + "grad_norm": 2.4449656009674072, + "learning_rate": 3.4034923493201007e-07, + "loss": 0.18497437238693237, + "step": 788 + }, + { + "epoch": 3.2077393075356415, + "grad_norm": 2.400811195373535, + "learning_rate": 3.3899909592861816e-07, + "loss": 0.1974419429898262, + "step": 789 + }, + { + "epoch": 3.211812627291242, + "grad_norm": 2.381777763366699, + "learning_rate": 3.3765026539765827e-07, + "loss": 0.18730898946523666, + "step": 790 + }, + { + "epoch": 3.2158859470468433, + "grad_norm": 2.3965096473693848, + "learning_rate": 3.3630275430122747e-07, + "loss": 0.17667900025844574, + "step": 791 + }, + { + "epoch": 3.219959266802444, + "grad_norm": 2.2706997394561768, + "learning_rate": 3.349565735907e-07, + "loss": 0.21489760279655457, + "step": 792 + }, + { + "epoch": 3.224032586558045, + "grad_norm": 2.5842676162719727, + "learning_rate": 3.336117342066375e-07, + "loss": 0.21414875984191895, + "step": 793 + }, + { + "epoch": 3.2281059063136457, + "grad_norm": 2.563056707382202, + "learning_rate": 3.3226824707870073e-07, + "loss": 0.22138936817646027, + "step": 794 + }, + { + "epoch": 3.2321792260692463, + "grad_norm": 2.35392689704895, + "learning_rate": 3.3092612312556075e-07, + "loss": 0.20381484925746918, + "step": 795 + }, + { + "epoch": 3.2362525458248474, + "grad_norm": 2.532590389251709, + "learning_rate": 3.2958537325480924e-07, + "loss": 0.21896713972091675, + "step": 796 + }, + { + "epoch": 3.240325865580448, + "grad_norm": 2.476938247680664, + "learning_rate": 3.282460083628713e-07, + "loss": 0.21506690233945847, + "step": 797 + }, + { + "epoch": 3.2443991853360488, + "grad_norm": 2.544861078262329, + "learning_rate": 3.2690803933491576e-07, + "loss": 0.22808198630809784, + "step": 798 + }, + { + "epoch": 3.24847250509165, + "grad_norm": 2.4569203853607178, + "learning_rate": 3.255714770447674e-07, + "loss": 0.2176097184419632, + "step": 799 + }, + { + "epoch": 3.2525458248472505, + "grad_norm": 2.4378373622894287, + "learning_rate": 3.242363323548177e-07, + "loss": 0.19017792493104935, + "step": 800 + }, + { + "epoch": 3.256619144602851, + "grad_norm": 2.4247865676879883, + "learning_rate": 3.229026161159378e-07, + "loss": 0.2059084177017212, + "step": 801 + }, + { + "epoch": 3.2606924643584523, + "grad_norm": 2.5162672996520996, + "learning_rate": 3.215703391673893e-07, + "loss": 0.19650442153215408, + "step": 802 + }, + { + "epoch": 3.264765784114053, + "grad_norm": 2.570923328399658, + "learning_rate": 3.202395123367367e-07, + "loss": 0.2457878440618515, + "step": 803 + }, + { + "epoch": 3.2688391038696536, + "grad_norm": 2.5794918537139893, + "learning_rate": 3.189101464397591e-07, + "loss": 0.1937229335308075, + "step": 804 + }, + { + "epoch": 3.2729124236252547, + "grad_norm": 2.4221835136413574, + "learning_rate": 3.1758225228036227e-07, + "loss": 0.17538277059793472, + "step": 805 + }, + { + "epoch": 3.2769857433808554, + "grad_norm": 2.418198347091675, + "learning_rate": 3.1625584065049155e-07, + "loss": 0.20809274911880493, + "step": 806 + }, + { + "epoch": 3.281059063136456, + "grad_norm": 2.56083607673645, + "learning_rate": 3.1493092233004277e-07, + "loss": 0.2350323647260666, + "step": 807 + }, + { + "epoch": 3.285132382892057, + "grad_norm": 2.516634702682495, + "learning_rate": 3.136075080867765e-07, + "loss": 0.1945355385541916, + "step": 808 + }, + { + "epoch": 3.2892057026476578, + "grad_norm": 2.5624663829803467, + "learning_rate": 3.1228560867622854e-07, + "loss": 0.21247724443674088, + "step": 809 + }, + { + "epoch": 3.293279022403259, + "grad_norm": 2.6713180541992188, + "learning_rate": 3.1096523484162407e-07, + "loss": 0.20905288308858871, + "step": 810 + }, + { + "epoch": 3.2973523421588595, + "grad_norm": 2.5594656467437744, + "learning_rate": 3.0964639731378947e-07, + "loss": 0.21280484646558762, + "step": 811 + }, + { + "epoch": 3.30142566191446, + "grad_norm": 2.5005033016204834, + "learning_rate": 3.0832910681106565e-07, + "loss": 0.1858508661389351, + "step": 812 + }, + { + "epoch": 3.3054989816700613, + "grad_norm": 2.5059120655059814, + "learning_rate": 3.070133740392202e-07, + "loss": 0.17269013077020645, + "step": 813 + }, + { + "epoch": 3.309572301425662, + "grad_norm": 2.484297752380371, + "learning_rate": 3.0569920969136135e-07, + "loss": 0.21136271953582764, + "step": 814 + }, + { + "epoch": 3.3136456211812626, + "grad_norm": 2.5234310626983643, + "learning_rate": 3.043866244478505e-07, + "loss": 0.18130285292863846, + "step": 815 + }, + { + "epoch": 3.3177189409368637, + "grad_norm": 2.338294267654419, + "learning_rate": 3.0307562897621485e-07, + "loss": 0.20136955380439758, + "step": 816 + }, + { + "epoch": 3.3217922606924644, + "grad_norm": 2.490870714187622, + "learning_rate": 3.0176623393106235e-07, + "loss": 0.18799114972352982, + "step": 817 + }, + { + "epoch": 3.325865580448065, + "grad_norm": 2.756178140640259, + "learning_rate": 3.0045844995399327e-07, + "loss": 0.17684923857450485, + "step": 818 + }, + { + "epoch": 3.329938900203666, + "grad_norm": 2.6265006065368652, + "learning_rate": 2.9915228767351535e-07, + "loss": 0.20372115820646286, + "step": 819 + }, + { + "epoch": 3.3340122199592668, + "grad_norm": 2.3552069664001465, + "learning_rate": 2.978477577049556e-07, + "loss": 0.17303332686424255, + "step": 820 + }, + { + "epoch": 3.3380855397148674, + "grad_norm": 2.3262510299682617, + "learning_rate": 2.965448706503761e-07, + "loss": 0.22534170746803284, + "step": 821 + }, + { + "epoch": 3.3421588594704685, + "grad_norm": 2.619785785675049, + "learning_rate": 2.952436370984859e-07, + "loss": 0.2306053191423416, + "step": 822 + }, + { + "epoch": 3.346232179226069, + "grad_norm": 2.4102282524108887, + "learning_rate": 2.939440676245566e-07, + "loss": 0.2103247046470642, + "step": 823 + }, + { + "epoch": 3.35030549898167, + "grad_norm": 2.646646738052368, + "learning_rate": 2.926461727903349e-07, + "loss": 0.21324314177036285, + "step": 824 + }, + { + "epoch": 3.354378818737271, + "grad_norm": 2.454810619354248, + "learning_rate": 2.9134996314395817e-07, + "loss": 0.1971464827656746, + "step": 825 + }, + { + "epoch": 3.3584521384928716, + "grad_norm": 2.4926371574401855, + "learning_rate": 2.900554492198677e-07, + "loss": 0.19652055203914642, + "step": 826 + }, + { + "epoch": 3.3625254582484727, + "grad_norm": 2.251152992248535, + "learning_rate": 2.887626415387237e-07, + "loss": 0.22214417904615402, + "step": 827 + }, + { + "epoch": 3.3665987780040734, + "grad_norm": 2.5524468421936035, + "learning_rate": 2.8747155060731937e-07, + "loss": 0.21327239274978638, + "step": 828 + }, + { + "epoch": 3.370672097759674, + "grad_norm": 2.429053783416748, + "learning_rate": 2.8618218691849545e-07, + "loss": 0.2205718532204628, + "step": 829 + }, + { + "epoch": 3.374745417515275, + "grad_norm": 2.5766923427581787, + "learning_rate": 2.8489456095105566e-07, + "loss": 0.23655060678720474, + "step": 830 + }, + { + "epoch": 3.378818737270876, + "grad_norm": 2.752376079559326, + "learning_rate": 2.836086831696809e-07, + "loss": 0.24048178642988205, + "step": 831 + }, + { + "epoch": 3.3828920570264764, + "grad_norm": 2.4999852180480957, + "learning_rate": 2.8232456402484463e-07, + "loss": 0.16257788240909576, + "step": 832 + }, + { + "epoch": 3.3869653767820775, + "grad_norm": 2.4269254207611084, + "learning_rate": 2.8104221395272674e-07, + "loss": 0.2149578034877777, + "step": 833 + }, + { + "epoch": 3.391038696537678, + "grad_norm": 2.354278326034546, + "learning_rate": 2.797616433751309e-07, + "loss": 0.19306989759206772, + "step": 834 + }, + { + "epoch": 3.395112016293279, + "grad_norm": 2.495457649230957, + "learning_rate": 2.784828626993976e-07, + "loss": 0.16760513186454773, + "step": 835 + }, + { + "epoch": 3.39918533604888, + "grad_norm": 2.299818754196167, + "learning_rate": 2.772058823183212e-07, + "loss": 0.1885610893368721, + "step": 836 + }, + { + "epoch": 3.4032586558044806, + "grad_norm": 2.441559076309204, + "learning_rate": 2.7593071261006473e-07, + "loss": 0.19247783720493317, + "step": 837 + }, + { + "epoch": 3.4073319755600817, + "grad_norm": 2.5072267055511475, + "learning_rate": 2.746573639380758e-07, + "loss": 0.16321063041687012, + "step": 838 + }, + { + "epoch": 3.4114052953156824, + "grad_norm": 2.3393125534057617, + "learning_rate": 2.7338584665100195e-07, + "loss": 0.1672486811876297, + "step": 839 + }, + { + "epoch": 3.415478615071283, + "grad_norm": 2.5106165409088135, + "learning_rate": 2.7211617108260674e-07, + "loss": 0.20449652522802353, + "step": 840 + }, + { + "epoch": 3.4195519348268837, + "grad_norm": 2.5324196815490723, + "learning_rate": 2.708483475516865e-07, + "loss": 0.2249668836593628, + "step": 841 + }, + { + "epoch": 3.423625254582485, + "grad_norm": 2.581486225128174, + "learning_rate": 2.695823863619853e-07, + "loss": 0.2402847856283188, + "step": 842 + }, + { + "epoch": 3.4276985743380854, + "grad_norm": 2.535104274749756, + "learning_rate": 2.683182978021118e-07, + "loss": 0.1992955058813095, + "step": 843 + }, + { + "epoch": 3.4317718940936865, + "grad_norm": 2.2774689197540283, + "learning_rate": 2.6705609214545585e-07, + "loss": 0.19953173398971558, + "step": 844 + }, + { + "epoch": 3.435845213849287, + "grad_norm": 2.4968085289001465, + "learning_rate": 2.65795779650105e-07, + "loss": 0.19426950812339783, + "step": 845 + }, + { + "epoch": 3.439918533604888, + "grad_norm": 2.459174156188965, + "learning_rate": 2.6453737055875974e-07, + "loss": 0.1926077976822853, + "step": 846 + }, + { + "epoch": 3.443991853360489, + "grad_norm": 2.254603624343872, + "learning_rate": 2.632808750986527e-07, + "loss": 0.1818878874182701, + "step": 847 + }, + { + "epoch": 3.4480651731160896, + "grad_norm": 2.502915382385254, + "learning_rate": 2.620263034814632e-07, + "loss": 0.17121271044015884, + "step": 848 + }, + { + "epoch": 3.4521384928716903, + "grad_norm": 2.478640079498291, + "learning_rate": 2.6077366590323605e-07, + "loss": 0.17577876895666122, + "step": 849 + }, + { + "epoch": 3.4562118126272914, + "grad_norm": 2.799896478652954, + "learning_rate": 2.5952297254429725e-07, + "loss": 0.2371537759900093, + "step": 850 + }, + { + "epoch": 3.460285132382892, + "grad_norm": 2.4503631591796875, + "learning_rate": 2.582742335691722e-07, + "loss": 0.18308546394109726, + "step": 851 + }, + { + "epoch": 3.4643584521384927, + "grad_norm": 2.511253595352173, + "learning_rate": 2.5702745912650327e-07, + "loss": 0.1863907426595688, + "step": 852 + }, + { + "epoch": 3.468431771894094, + "grad_norm": 2.441244602203369, + "learning_rate": 2.5578265934896586e-07, + "loss": 0.17125633358955383, + "step": 853 + }, + { + "epoch": 3.4725050916496945, + "grad_norm": 2.5986480712890625, + "learning_rate": 2.54539844353188e-07, + "loss": 0.20211002230644226, + "step": 854 + }, + { + "epoch": 3.4765784114052956, + "grad_norm": 2.5657708644866943, + "learning_rate": 2.5329902423966636e-07, + "loss": 0.18166958540678024, + "step": 855 + }, + { + "epoch": 3.480651731160896, + "grad_norm": 2.759941577911377, + "learning_rate": 2.5206020909268575e-07, + "loss": 0.2045290172100067, + "step": 856 + }, + { + "epoch": 3.484725050916497, + "grad_norm": 2.3406291007995605, + "learning_rate": 2.508234089802356e-07, + "loss": 0.20272044837474823, + "step": 857 + }, + { + "epoch": 3.4887983706720975, + "grad_norm": 2.5458176136016846, + "learning_rate": 2.4958863395392985e-07, + "loss": 0.20205383747816086, + "step": 858 + }, + { + "epoch": 3.4928716904276986, + "grad_norm": 2.6369478702545166, + "learning_rate": 2.483558940489235e-07, + "loss": 0.21245616674423218, + "step": 859 + }, + { + "epoch": 3.4969450101832993, + "grad_norm": 2.407942533493042, + "learning_rate": 2.4712519928383245e-07, + "loss": 0.20060960948467255, + "step": 860 + }, + { + "epoch": 3.5010183299389004, + "grad_norm": 2.491145133972168, + "learning_rate": 2.45896559660651e-07, + "loss": 0.2001042366027832, + "step": 861 + }, + { + "epoch": 3.505091649694501, + "grad_norm": 2.3934686183929443, + "learning_rate": 2.4466998516467176e-07, + "loss": 0.20040663331747055, + "step": 862 + }, + { + "epoch": 3.5091649694501017, + "grad_norm": 2.350095510482788, + "learning_rate": 2.4344548576440293e-07, + "loss": 0.1889752671122551, + "step": 863 + }, + { + "epoch": 3.513238289205703, + "grad_norm": 2.4073328971862793, + "learning_rate": 2.4222307141148906e-07, + "loss": 0.18654479831457138, + "step": 864 + }, + { + "epoch": 3.5173116089613035, + "grad_norm": 2.4485344886779785, + "learning_rate": 2.4100275204062897e-07, + "loss": 0.21063948422670364, + "step": 865 + }, + { + "epoch": 3.521384928716904, + "grad_norm": 2.4200923442840576, + "learning_rate": 2.397845375694949e-07, + "loss": 0.1860312521457672, + "step": 866 + }, + { + "epoch": 3.525458248472505, + "grad_norm": 2.775789260864258, + "learning_rate": 2.3856843789865303e-07, + "loss": 0.2309775874018669, + "step": 867 + }, + { + "epoch": 3.529531568228106, + "grad_norm": 2.7096245288848877, + "learning_rate": 2.3735446291148176e-07, + "loss": 0.2163269817829132, + "step": 868 + }, + { + "epoch": 3.5336048879837065, + "grad_norm": 2.4550371170043945, + "learning_rate": 2.361426224740924e-07, + "loss": 0.1779681146144867, + "step": 869 + }, + { + "epoch": 3.5376782077393076, + "grad_norm": 2.2727365493774414, + "learning_rate": 2.3493292643524799e-07, + "loss": 0.19602014124393463, + "step": 870 + }, + { + "epoch": 3.5417515274949083, + "grad_norm": 2.445028066635132, + "learning_rate": 2.3372538462628422e-07, + "loss": 0.19011924415826797, + "step": 871 + }, + { + "epoch": 3.5458248472505094, + "grad_norm": 2.537092447280884, + "learning_rate": 2.3252000686102912e-07, + "loss": 0.22465527802705765, + "step": 872 + }, + { + "epoch": 3.54989816700611, + "grad_norm": 2.5350615978240967, + "learning_rate": 2.3131680293572336e-07, + "loss": 0.2048494815826416, + "step": 873 + }, + { + "epoch": 3.5539714867617107, + "grad_norm": 2.435441255569458, + "learning_rate": 2.3011578262894015e-07, + "loss": 0.22786639630794525, + "step": 874 + }, + { + "epoch": 3.5580448065173114, + "grad_norm": 2.409935235977173, + "learning_rate": 2.2891695570150631e-07, + "loss": 0.1882152482867241, + "step": 875 + }, + { + "epoch": 3.5621181262729125, + "grad_norm": 2.6188433170318604, + "learning_rate": 2.2772033189642321e-07, + "loss": 0.21629157662391663, + "step": 876 + }, + { + "epoch": 3.566191446028513, + "grad_norm": 2.4537856578826904, + "learning_rate": 2.2652592093878665e-07, + "loss": 0.19376155734062195, + "step": 877 + }, + { + "epoch": 3.5702647657841142, + "grad_norm": 2.438338279724121, + "learning_rate": 2.2533373253570875e-07, + "loss": 0.22396929562091827, + "step": 878 + }, + { + "epoch": 3.574338085539715, + "grad_norm": 2.4425530433654785, + "learning_rate": 2.2414377637623865e-07, + "loss": 0.1896074339747429, + "step": 879 + }, + { + "epoch": 3.5784114052953155, + "grad_norm": 2.46877384185791, + "learning_rate": 2.2295606213128387e-07, + "loss": 0.1916242465376854, + "step": 880 + }, + { + "epoch": 3.5824847250509166, + "grad_norm": 2.4224820137023926, + "learning_rate": 2.2177059945353115e-07, + "loss": 0.23046725243330002, + "step": 881 + }, + { + "epoch": 3.5865580448065173, + "grad_norm": 2.678439140319824, + "learning_rate": 2.2058739797736914e-07, + "loss": 0.1764805093407631, + "step": 882 + }, + { + "epoch": 3.5906313645621184, + "grad_norm": 2.5862419605255127, + "learning_rate": 2.1940646731880885e-07, + "loss": 0.1832810789346695, + "step": 883 + }, + { + "epoch": 3.594704684317719, + "grad_norm": 2.3499772548675537, + "learning_rate": 2.1822781707540667e-07, + "loss": 0.19466058164834976, + "step": 884 + }, + { + "epoch": 3.5987780040733197, + "grad_norm": 2.416409492492676, + "learning_rate": 2.1705145682618502e-07, + "loss": 0.20160751044750214, + "step": 885 + }, + { + "epoch": 3.6028513238289204, + "grad_norm": 2.7396693229675293, + "learning_rate": 2.1587739613155653e-07, + "loss": 0.2220790833234787, + "step": 886 + }, + { + "epoch": 3.6069246435845215, + "grad_norm": 2.438356876373291, + "learning_rate": 2.1470564453324392e-07, + "loss": 0.1909223347902298, + "step": 887 + }, + { + "epoch": 3.610997963340122, + "grad_norm": 2.637033224105835, + "learning_rate": 2.1353621155420393e-07, + "loss": 0.19567319750785828, + "step": 888 + }, + { + "epoch": 3.6150712830957232, + "grad_norm": 2.6245222091674805, + "learning_rate": 2.1236910669855006e-07, + "loss": 0.19575632363557816, + "step": 889 + }, + { + "epoch": 3.619144602851324, + "grad_norm": 2.6102139949798584, + "learning_rate": 2.112043394514742e-07, + "loss": 0.19621288776397705, + "step": 890 + }, + { + "epoch": 3.6232179226069245, + "grad_norm": 2.5597336292266846, + "learning_rate": 2.100419192791708e-07, + "loss": 0.17981623113155365, + "step": 891 + }, + { + "epoch": 3.627291242362525, + "grad_norm": 2.345719575881958, + "learning_rate": 2.088818556287592e-07, + "loss": 0.18147233873605728, + "step": 892 + }, + { + "epoch": 3.6313645621181263, + "grad_norm": 2.4701764583587646, + "learning_rate": 2.0772415792820713e-07, + "loss": 0.20388461649417877, + "step": 893 + }, + { + "epoch": 3.635437881873727, + "grad_norm": 2.6218173503875732, + "learning_rate": 2.0656883558625348e-07, + "loss": 0.20968149602413177, + "step": 894 + }, + { + "epoch": 3.639511201629328, + "grad_norm": 2.433162212371826, + "learning_rate": 2.054158979923331e-07, + "loss": 0.23324060440063477, + "step": 895 + }, + { + "epoch": 3.6435845213849287, + "grad_norm": 2.5269041061401367, + "learning_rate": 2.042653545164989e-07, + "loss": 0.22314583510160446, + "step": 896 + }, + { + "epoch": 3.6476578411405294, + "grad_norm": 2.772693634033203, + "learning_rate": 2.0311721450934732e-07, + "loss": 0.17379353195428848, + "step": 897 + }, + { + "epoch": 3.6517311608961305, + "grad_norm": 2.4417383670806885, + "learning_rate": 2.0197148730194085e-07, + "loss": 0.196005217730999, + "step": 898 + }, + { + "epoch": 3.655804480651731, + "grad_norm": 2.5393450260162354, + "learning_rate": 2.0082818220573332e-07, + "loss": 0.21743719279766083, + "step": 899 + }, + { + "epoch": 3.6598778004073322, + "grad_norm": 2.476929187774658, + "learning_rate": 1.9968730851249388e-07, + "loss": 0.20515238493680954, + "step": 900 + }, + { + "epoch": 3.663951120162933, + "grad_norm": 2.441044569015503, + "learning_rate": 1.9854887549423082e-07, + "loss": 0.1906992271542549, + "step": 901 + }, + { + "epoch": 3.6680244399185336, + "grad_norm": 2.5373618602752686, + "learning_rate": 1.9741289240311754e-07, + "loss": 0.20473621785640717, + "step": 902 + }, + { + "epoch": 3.672097759674134, + "grad_norm": 2.405190944671631, + "learning_rate": 1.962793684714158e-07, + "loss": 0.21750831604003906, + "step": 903 + }, + { + "epoch": 3.6761710794297353, + "grad_norm": 2.40073299407959, + "learning_rate": 1.9514831291140228e-07, + "loss": 0.19307416677474976, + "step": 904 + }, + { + "epoch": 3.680244399185336, + "grad_norm": 2.7207555770874023, + "learning_rate": 1.940197349152923e-07, + "loss": 0.2060309201478958, + "step": 905 + }, + { + "epoch": 3.684317718940937, + "grad_norm": 2.4056124687194824, + "learning_rate": 1.9289364365516607e-07, + "loss": 0.1817646324634552, + "step": 906 + }, + { + "epoch": 3.6883910386965377, + "grad_norm": 2.5248098373413086, + "learning_rate": 1.9177004828289383e-07, + "loss": 0.20411433279514313, + "step": 907 + }, + { + "epoch": 3.6924643584521384, + "grad_norm": 2.365913152694702, + "learning_rate": 1.9064895793006153e-07, + "loss": 0.1928766593337059, + "step": 908 + }, + { + "epoch": 3.696537678207739, + "grad_norm": 2.5754306316375732, + "learning_rate": 1.8953038170789615e-07, + "loss": 0.21009906381368637, + "step": 909 + }, + { + "epoch": 3.70061099796334, + "grad_norm": 2.641087055206299, + "learning_rate": 1.8841432870719226e-07, + "loss": 0.21582353860139847, + "step": 910 + }, + { + "epoch": 3.704684317718941, + "grad_norm": 2.567742347717285, + "learning_rate": 1.8730080799823815e-07, + "loss": 0.19261349737644196, + "step": 911 + }, + { + "epoch": 3.708757637474542, + "grad_norm": 2.521796703338623, + "learning_rate": 1.861898286307413e-07, + "loss": 0.20253480970859528, + "step": 912 + }, + { + "epoch": 3.7128309572301426, + "grad_norm": 2.504960298538208, + "learning_rate": 1.8508139963375646e-07, + "loss": 0.196384996175766, + "step": 913 + }, + { + "epoch": 3.716904276985743, + "grad_norm": 2.448503017425537, + "learning_rate": 1.8397553001561012e-07, + "loss": 0.22694706171751022, + "step": 914 + }, + { + "epoch": 3.7209775967413443, + "grad_norm": 2.448093891143799, + "learning_rate": 1.8287222876382912e-07, + "loss": 0.18733669072389603, + "step": 915 + }, + { + "epoch": 3.725050916496945, + "grad_norm": 2.4246034622192383, + "learning_rate": 1.8177150484506642e-07, + "loss": 0.19822125136852264, + "step": 916 + }, + { + "epoch": 3.729124236252546, + "grad_norm": 2.5895848274230957, + "learning_rate": 1.806733672050293e-07, + "loss": 0.18903044611215591, + "step": 917 + }, + { + "epoch": 3.7331975560081467, + "grad_norm": 2.5084774494171143, + "learning_rate": 1.7957782476840528e-07, + "loss": 0.2244538515806198, + "step": 918 + }, + { + "epoch": 3.7372708757637474, + "grad_norm": 2.436156749725342, + "learning_rate": 1.78484886438791e-07, + "loss": 0.19061604887247086, + "step": 919 + }, + { + "epoch": 3.741344195519348, + "grad_norm": 2.3762974739074707, + "learning_rate": 1.7739456109861912e-07, + "loss": 0.18180037289857864, + "step": 920 + }, + { + "epoch": 3.745417515274949, + "grad_norm": 2.5922470092773438, + "learning_rate": 1.763068576090862e-07, + "loss": 0.21426790952682495, + "step": 921 + }, + { + "epoch": 3.74949083503055, + "grad_norm": 2.582359790802002, + "learning_rate": 1.7522178481008054e-07, + "loss": 0.20259950309991837, + "step": 922 + }, + { + "epoch": 3.753564154786151, + "grad_norm": 2.3691251277923584, + "learning_rate": 1.7413935152011055e-07, + "loss": 0.21478895843029022, + "step": 923 + }, + { + "epoch": 3.7576374745417516, + "grad_norm": 2.511850595474243, + "learning_rate": 1.7305956653623343e-07, + "loss": 0.20169981569051743, + "step": 924 + }, + { + "epoch": 3.7617107942973522, + "grad_norm": 2.342623472213745, + "learning_rate": 1.719824386339827e-07, + "loss": 0.21613454818725586, + "step": 925 + }, + { + "epoch": 3.765784114052953, + "grad_norm": 2.58245587348938, + "learning_rate": 1.7090797656729804e-07, + "loss": 0.1654214784502983, + "step": 926 + }, + { + "epoch": 3.769857433808554, + "grad_norm": 2.3268702030181885, + "learning_rate": 1.6983618906845332e-07, + "loss": 0.18952950835227966, + "step": 927 + }, + { + "epoch": 3.7739307535641546, + "grad_norm": 2.42199444770813, + "learning_rate": 1.6876708484798608e-07, + "loss": 0.16338826343417168, + "step": 928 + }, + { + "epoch": 3.7780040733197557, + "grad_norm": 2.2870736122131348, + "learning_rate": 1.677006725946261e-07, + "loss": 0.18167713284492493, + "step": 929 + }, + { + "epoch": 3.7820773930753564, + "grad_norm": 2.5234158039093018, + "learning_rate": 1.6663696097522585e-07, + "loss": 0.1719643920660019, + "step": 930 + }, + { + "epoch": 3.786150712830957, + "grad_norm": 2.5721402168273926, + "learning_rate": 1.6557595863468886e-07, + "loss": 0.1708434671163559, + "step": 931 + }, + { + "epoch": 3.790224032586558, + "grad_norm": 2.392375946044922, + "learning_rate": 1.6451767419590062e-07, + "loss": 0.20495689660310745, + "step": 932 + }, + { + "epoch": 3.794297352342159, + "grad_norm": 2.815039873123169, + "learning_rate": 1.6346211625965732e-07, + "loss": 0.1895192414522171, + "step": 933 + }, + { + "epoch": 3.79837067209776, + "grad_norm": 2.6176204681396484, + "learning_rate": 1.6240929340459703e-07, + "loss": 0.19631709158420563, + "step": 934 + }, + { + "epoch": 3.8024439918533606, + "grad_norm": 2.4265332221984863, + "learning_rate": 1.6135921418712955e-07, + "loss": 0.19473101943731308, + "step": 935 + }, + { + "epoch": 3.8065173116089612, + "grad_norm": 2.5783004760742188, + "learning_rate": 1.6031188714136623e-07, + "loss": 0.2178090512752533, + "step": 936 + }, + { + "epoch": 3.810590631364562, + "grad_norm": 2.534191846847534, + "learning_rate": 1.5926732077905203e-07, + "loss": 0.16174981743097305, + "step": 937 + }, + { + "epoch": 3.814663951120163, + "grad_norm": 2.6376049518585205, + "learning_rate": 1.582255235894947e-07, + "loss": 0.22103732079267502, + "step": 938 + }, + { + "epoch": 3.8187372708757636, + "grad_norm": 2.646101474761963, + "learning_rate": 1.571865040394973e-07, + "loss": 0.21458610147237778, + "step": 939 + }, + { + "epoch": 3.8228105906313647, + "grad_norm": 2.335294246673584, + "learning_rate": 1.561502705732883e-07, + "loss": 0.21040001511573792, + "step": 940 + }, + { + "epoch": 3.8268839103869654, + "grad_norm": 2.527137279510498, + "learning_rate": 1.5511683161245365e-07, + "loss": 0.19486284255981445, + "step": 941 + }, + { + "epoch": 3.830957230142566, + "grad_norm": 2.3123862743377686, + "learning_rate": 1.540861955558676e-07, + "loss": 0.18999101221561432, + "step": 942 + }, + { + "epoch": 3.835030549898167, + "grad_norm": 2.6126627922058105, + "learning_rate": 1.5305837077962542e-07, + "loss": 0.20924139767885208, + "step": 943 + }, + { + "epoch": 3.839103869653768, + "grad_norm": 2.4893648624420166, + "learning_rate": 1.5203336563697444e-07, + "loss": 0.18669888377189636, + "step": 944 + }, + { + "epoch": 3.8431771894093685, + "grad_norm": 2.406766414642334, + "learning_rate": 1.5101118845824628e-07, + "loss": 0.18617846816778183, + "step": 945 + }, + { + "epoch": 3.8472505091649696, + "grad_norm": 2.2685155868530273, + "learning_rate": 1.4999184755079004e-07, + "loss": 0.1924063339829445, + "step": 946 + }, + { + "epoch": 3.8513238289205702, + "grad_norm": 2.443790912628174, + "learning_rate": 1.4897535119890364e-07, + "loss": 0.1853998601436615, + "step": 947 + }, + { + "epoch": 3.855397148676171, + "grad_norm": 2.320080041885376, + "learning_rate": 1.4796170766376727e-07, + "loss": 0.21173150092363358, + "step": 948 + }, + { + "epoch": 3.859470468431772, + "grad_norm": 2.5018677711486816, + "learning_rate": 1.4695092518337554e-07, + "loss": 0.1776503399014473, + "step": 949 + }, + { + "epoch": 3.8635437881873727, + "grad_norm": 2.493777275085449, + "learning_rate": 1.459430119724715e-07, + "loss": 0.17813850194215775, + "step": 950 + }, + { + "epoch": 3.8676171079429738, + "grad_norm": 2.4366796016693115, + "learning_rate": 1.4493797622247867e-07, + "loss": 0.19571475684642792, + "step": 951 + }, + { + "epoch": 3.8716904276985744, + "grad_norm": 2.559058427810669, + "learning_rate": 1.439358261014359e-07, + "loss": 0.19421598315238953, + "step": 952 + }, + { + "epoch": 3.875763747454175, + "grad_norm": 2.4444687366485596, + "learning_rate": 1.4293656975392937e-07, + "loss": 0.21629701554775238, + "step": 953 + }, + { + "epoch": 3.8798370672097757, + "grad_norm": 2.388728380203247, + "learning_rate": 1.4194021530102783e-07, + "loss": 0.19527916610240936, + "step": 954 + }, + { + "epoch": 3.883910386965377, + "grad_norm": 2.286078929901123, + "learning_rate": 1.4094677084021588e-07, + "loss": 0.1616881936788559, + "step": 955 + }, + { + "epoch": 3.8879837067209775, + "grad_norm": 2.5002880096435547, + "learning_rate": 1.3995624444532844e-07, + "loss": 0.19413011521100998, + "step": 956 + }, + { + "epoch": 3.8920570264765786, + "grad_norm": 2.6207797527313232, + "learning_rate": 1.3896864416648452e-07, + "loss": 0.15424808859825134, + "step": 957 + }, + { + "epoch": 3.8961303462321792, + "grad_norm": 2.5780515670776367, + "learning_rate": 1.3798397803002237e-07, + "loss": 0.20987361669540405, + "step": 958 + }, + { + "epoch": 3.90020366598778, + "grad_norm": 2.439293622970581, + "learning_rate": 1.370022540384347e-07, + "loss": 0.19203339517116547, + "step": 959 + }, + { + "epoch": 3.904276985743381, + "grad_norm": 2.810973644256592, + "learning_rate": 1.360234801703023e-07, + "loss": 0.25363121181726456, + "step": 960 + }, + { + "epoch": 3.9083503054989817, + "grad_norm": 2.6023387908935547, + "learning_rate": 1.3504766438023042e-07, + "loss": 0.21091164648532867, + "step": 961 + }, + { + "epoch": 3.9124236252545828, + "grad_norm": 2.3644959926605225, + "learning_rate": 1.3407481459878366e-07, + "loss": 0.21235870569944382, + "step": 962 + }, + { + "epoch": 3.9164969450101834, + "grad_norm": 2.3764562606811523, + "learning_rate": 1.3310493873242167e-07, + "loss": 0.1808951571583748, + "step": 963 + }, + { + "epoch": 3.920570264765784, + "grad_norm": 2.3367507457733154, + "learning_rate": 1.321380446634342e-07, + "loss": 0.2204703390598297, + "step": 964 + }, + { + "epoch": 3.9246435845213847, + "grad_norm": 2.595667600631714, + "learning_rate": 1.3117414024987823e-07, + "loss": 0.19948850572109222, + "step": 965 + }, + { + "epoch": 3.928716904276986, + "grad_norm": 2.6602208614349365, + "learning_rate": 1.3021323332551294e-07, + "loss": 0.22697525471448898, + "step": 966 + }, + { + "epoch": 3.9327902240325865, + "grad_norm": 2.483463764190674, + "learning_rate": 1.2925533169973695e-07, + "loss": 0.17191919684410095, + "step": 967 + }, + { + "epoch": 3.9368635437881876, + "grad_norm": 2.4962165355682373, + "learning_rate": 1.283004431575246e-07, + "loss": 0.20009542256593704, + "step": 968 + }, + { + "epoch": 3.9409368635437882, + "grad_norm": 2.5772528648376465, + "learning_rate": 1.273485754593619e-07, + "loss": 0.2056511491537094, + "step": 969 + }, + { + "epoch": 3.945010183299389, + "grad_norm": 2.4633169174194336, + "learning_rate": 1.26399736341185e-07, + "loss": 0.19378910213708878, + "step": 970 + }, + { + "epoch": 3.9490835030549896, + "grad_norm": 2.5064918994903564, + "learning_rate": 1.254539335143156e-07, + "loss": 0.2224956750869751, + "step": 971 + }, + { + "epoch": 3.9531568228105907, + "grad_norm": 2.723224401473999, + "learning_rate": 1.2451117466539985e-07, + "loss": 0.20264600962400436, + "step": 972 + }, + { + "epoch": 3.9572301425661913, + "grad_norm": 2.501723289489746, + "learning_rate": 1.235714674563445e-07, + "loss": 0.17863625288009644, + "step": 973 + }, + { + "epoch": 3.9613034623217924, + "grad_norm": 2.344696521759033, + "learning_rate": 1.226348195242557e-07, + "loss": 0.23713428527116776, + "step": 974 + }, + { + "epoch": 3.965376782077393, + "grad_norm": 2.421300172805786, + "learning_rate": 1.2170123848137648e-07, + "loss": 0.20281652361154556, + "step": 975 + }, + { + "epoch": 3.9694501018329937, + "grad_norm": 2.5802454948425293, + "learning_rate": 1.2077073191502496e-07, + "loss": 0.21466045081615448, + "step": 976 + }, + { + "epoch": 3.973523421588595, + "grad_norm": 2.3439295291900635, + "learning_rate": 1.1984330738753218e-07, + "loss": 0.1579430103302002, + "step": 977 + }, + { + "epoch": 3.9775967413441955, + "grad_norm": 2.491952896118164, + "learning_rate": 1.1891897243618183e-07, + "loss": 0.1682949811220169, + "step": 978 + }, + { + "epoch": 3.9816700610997966, + "grad_norm": 2.3951187133789062, + "learning_rate": 1.1799773457314766e-07, + "loss": 0.20190123468637466, + "step": 979 + }, + { + "epoch": 3.9857433808553973, + "grad_norm": 2.4365179538726807, + "learning_rate": 1.1707960128543314e-07, + "loss": 0.18969932198524475, + "step": 980 + }, + { + "epoch": 3.989816700610998, + "grad_norm": 2.2501721382141113, + "learning_rate": 1.1616458003481084e-07, + "loss": 0.16813133656978607, + "step": 981 + }, + { + "epoch": 3.9938900203665986, + "grad_norm": 2.550407886505127, + "learning_rate": 1.1525267825776114e-07, + "loss": 0.19014200568199158, + "step": 982 + }, + { + "epoch": 3.9979633401221997, + "grad_norm": 2.3324475288391113, + "learning_rate": 1.1434390336541238e-07, + "loss": 0.21611415594816208, + "step": 983 + }, + { + "epoch": 4.0, + "grad_norm": 2.8509156703948975, + "learning_rate": 1.1343826274347995e-07, + "loss": 0.2155466079711914, + "step": 984 + }, + { + "epoch": 4.004073319755601, + "grad_norm": 2.371936321258545, + "learning_rate": 1.125357637522072e-07, + "loss": 0.21221815049648285, + "step": 985 + }, + { + "epoch": 4.008146639511201, + "grad_norm": 2.5109193325042725, + "learning_rate": 1.1163641372630445e-07, + "loss": 0.19040236622095108, + "step": 986 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 2.3751659393310547, + "learning_rate": 1.1074021997489074e-07, + "loss": 0.17234822362661362, + "step": 987 + }, + { + "epoch": 4.0162932790224035, + "grad_norm": 2.3197624683380127, + "learning_rate": 1.0984718978143287e-07, + "loss": 0.1829347461462021, + "step": 988 + }, + { + "epoch": 4.020366598778004, + "grad_norm": 2.230381965637207, + "learning_rate": 1.0895733040368815e-07, + "loss": 0.1833319142460823, + "step": 989 + }, + { + "epoch": 4.024439918533605, + "grad_norm": 2.195852279663086, + "learning_rate": 1.0807064907364321e-07, + "loss": 0.1787407100200653, + "step": 990 + }, + { + "epoch": 4.0285132382892055, + "grad_norm": 2.331566572189331, + "learning_rate": 1.0718715299745717e-07, + "loss": 0.16609113663434982, + "step": 991 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 2.600375175476074, + "learning_rate": 1.0630684935540168e-07, + "loss": 0.2246694266796112, + "step": 992 + }, + { + "epoch": 4.036659877800408, + "grad_norm": 2.4915621280670166, + "learning_rate": 1.0542974530180327e-07, + "loss": 0.1964123621582985, + "step": 993 + }, + { + "epoch": 4.040733197556008, + "grad_norm": 2.434880256652832, + "learning_rate": 1.0455584796498512e-07, + "loss": 0.19002247601747513, + "step": 994 + }, + { + "epoch": 4.044806517311609, + "grad_norm": 2.3584625720977783, + "learning_rate": 1.0368516444720915e-07, + "loss": 0.16979426890611649, + "step": 995 + }, + { + "epoch": 4.04887983706721, + "grad_norm": 2.557974338531494, + "learning_rate": 1.0281770182461813e-07, + "loss": 0.20458902418613434, + "step": 996 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 2.3859493732452393, + "learning_rate": 1.0195346714717812e-07, + "loss": 0.16649136692285538, + "step": 997 + }, + { + "epoch": 4.057026476578411, + "grad_norm": 2.168416738510132, + "learning_rate": 1.0109246743862155e-07, + "loss": 0.15576700866222382, + "step": 998 + }, + { + "epoch": 4.0610997963340125, + "grad_norm": 2.2460920810699463, + "learning_rate": 1.0023470969638953e-07, + "loss": 0.1829686239361763, + "step": 999 + }, + { + "epoch": 4.065173116089613, + "grad_norm": 2.477757215499878, + "learning_rate": 9.938020089157595e-08, + "loss": 0.20117176324129105, + "step": 1000 + }, + { + "epoch": 4.069246435845214, + "grad_norm": 2.3903391361236572, + "learning_rate": 9.85289479688694e-08, + "loss": 0.1714942306280136, + "step": 1001 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 2.4241321086883545, + "learning_rate": 9.768095784649833e-08, + "loss": 0.1976744383573532, + "step": 1002 + }, + { + "epoch": 4.077393075356415, + "grad_norm": 2.3917348384857178, + "learning_rate": 9.68362374161737e-08, + "loss": 0.18097112327814102, + "step": 1003 + }, + { + "epoch": 4.081466395112017, + "grad_norm": 2.710338830947876, + "learning_rate": 9.599479354303308e-08, + "loss": 0.21830029785633087, + "step": 1004 + }, + { + "epoch": 4.085539714867617, + "grad_norm": 2.2726480960845947, + "learning_rate": 9.515663306558568e-08, + "loss": 0.1627790406346321, + "step": 1005 + }, + { + "epoch": 4.089613034623218, + "grad_norm": 2.563021183013916, + "learning_rate": 9.432176279565557e-08, + "loss": 0.17855525016784668, + "step": 1006 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 2.2680087089538574, + "learning_rate": 9.349018951832738e-08, + "loss": 0.17709980905056, + "step": 1007 + }, + { + "epoch": 4.097759674134419, + "grad_norm": 2.483131170272827, + "learning_rate": 9.266191999189044e-08, + "loss": 0.18354866653680801, + "step": 1008 + }, + { + "epoch": 4.10183299389002, + "grad_norm": 2.2514729499816895, + "learning_rate": 9.18369609477842e-08, + "loss": 0.17437510192394257, + "step": 1009 + }, + { + "epoch": 4.1059063136456215, + "grad_norm": 2.2802059650421143, + "learning_rate": 9.10153190905436e-08, + "loss": 0.18543030321598053, + "step": 1010 + }, + { + "epoch": 4.109979633401222, + "grad_norm": 2.3787717819213867, + "learning_rate": 9.019700109774436e-08, + "loss": 0.16922692209482193, + "step": 1011 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 3.4778051376342773, + "learning_rate": 8.938201361994846e-08, + "loss": 0.20204298198223114, + "step": 1012 + }, + { + "epoch": 4.1181262729124235, + "grad_norm": 2.3797998428344727, + "learning_rate": 8.857036328065098e-08, + "loss": 0.19148864597082138, + "step": 1013 + }, + { + "epoch": 4.122199592668024, + "grad_norm": 2.5149481296539307, + "learning_rate": 8.776205667622527e-08, + "loss": 0.18599937111139297, + "step": 1014 + }, + { + "epoch": 4.126272912423625, + "grad_norm": 2.5857813358306885, + "learning_rate": 8.695710037586957e-08, + "loss": 0.22922367602586746, + "step": 1015 + }, + { + "epoch": 4.130346232179226, + "grad_norm": 2.539100408554077, + "learning_rate": 8.615550092155477e-08, + "loss": 0.21377216279506683, + "step": 1016 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 2.3746533393859863, + "learning_rate": 8.535726482796918e-08, + "loss": 0.17763541638851166, + "step": 1017 + }, + { + "epoch": 4.138492871690428, + "grad_norm": 2.46425199508667, + "learning_rate": 8.456239858246755e-08, + "loss": 0.1617206409573555, + "step": 1018 + }, + { + "epoch": 4.142566191446028, + "grad_norm": 2.2839174270629883, + "learning_rate": 8.37709086450168e-08, + "loss": 0.17701925337314606, + "step": 1019 + }, + { + "epoch": 4.146639511201629, + "grad_norm": 2.5324790477752686, + "learning_rate": 8.29828014481449e-08, + "loss": 0.18900563567876816, + "step": 1020 + }, + { + "epoch": 4.1507128309572305, + "grad_norm": 2.781304359436035, + "learning_rate": 8.219808339688722e-08, + "loss": 0.2167806699872017, + "step": 1021 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 2.32759165763855, + "learning_rate": 8.141676086873573e-08, + "loss": 0.15807153284549713, + "step": 1022 + }, + { + "epoch": 4.158859470468432, + "grad_norm": 2.4917571544647217, + "learning_rate": 8.063884021358624e-08, + "loss": 0.19083156436681747, + "step": 1023 + }, + { + "epoch": 4.1629327902240325, + "grad_norm": 2.430983543395996, + "learning_rate": 7.986432775368756e-08, + "loss": 0.1890602707862854, + "step": 1024 + }, + { + "epoch": 4.167006109979633, + "grad_norm": 2.4754414558410645, + "learning_rate": 7.909322978358912e-08, + "loss": 0.1957196220755577, + "step": 1025 + }, + { + "epoch": 4.171079429735234, + "grad_norm": 2.4021310806274414, + "learning_rate": 7.832555257009105e-08, + "loss": 0.1860102415084839, + "step": 1026 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 2.6945855617523193, + "learning_rate": 7.75613023521921e-08, + "loss": 0.207361102104187, + "step": 1027 + }, + { + "epoch": 4.179226069246436, + "grad_norm": 2.431584358215332, + "learning_rate": 7.68004853410395e-08, + "loss": 0.17377741634845734, + "step": 1028 + }, + { + "epoch": 4.183299389002037, + "grad_norm": 2.4929258823394775, + "learning_rate": 7.604310771987882e-08, + "loss": 0.18458444625139236, + "step": 1029 + }, + { + "epoch": 4.187372708757637, + "grad_norm": 2.394940137863159, + "learning_rate": 7.528917564400289e-08, + "loss": 0.16298695653676987, + "step": 1030 + }, + { + "epoch": 4.191446028513238, + "grad_norm": 2.4729163646698, + "learning_rate": 7.45386952407026e-08, + "loss": 0.19453544914722443, + "step": 1031 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 2.406733751296997, + "learning_rate": 7.379167260921621e-08, + "loss": 0.18493016809225082, + "step": 1032 + }, + { + "epoch": 4.19959266802444, + "grad_norm": 2.6580498218536377, + "learning_rate": 7.304811382068077e-08, + "loss": 0.2018485963344574, + "step": 1033 + }, + { + "epoch": 4.203665987780041, + "grad_norm": 2.409303903579712, + "learning_rate": 7.230802491808191e-08, + "loss": 0.18995004892349243, + "step": 1034 + }, + { + "epoch": 4.2077393075356415, + "grad_norm": 2.679262399673462, + "learning_rate": 7.157141191620548e-08, + "loss": 0.18551041930913925, + "step": 1035 + }, + { + "epoch": 4.211812627291242, + "grad_norm": 2.4415478706359863, + "learning_rate": 7.083828080158783e-08, + "loss": 0.19411860406398773, + "step": 1036 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 2.6177053451538086, + "learning_rate": 7.010863753246798e-08, + "loss": 0.18280935287475586, + "step": 1037 + }, + { + "epoch": 4.219959266802444, + "grad_norm": 2.4606175422668457, + "learning_rate": 6.938248803873887e-08, + "loss": 0.17776849120855331, + "step": 1038 + }, + { + "epoch": 4.224032586558045, + "grad_norm": 2.459441900253296, + "learning_rate": 6.86598382218988e-08, + "loss": 0.18201512843370438, + "step": 1039 + }, + { + "epoch": 4.228105906313646, + "grad_norm": 2.5489375591278076, + "learning_rate": 6.794069395500418e-08, + "loss": 0.183207206428051, + "step": 1040 + }, + { + "epoch": 4.232179226069246, + "grad_norm": 2.3916819095611572, + "learning_rate": 6.722506108262111e-08, + "loss": 0.17223259061574936, + "step": 1041 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 2.3839707374572754, + "learning_rate": 6.651294542077846e-08, + "loss": 0.18934150785207748, + "step": 1042 + }, + { + "epoch": 4.240325865580448, + "grad_norm": 2.680551052093506, + "learning_rate": 6.580435275691987e-08, + "loss": 0.19397541880607605, + "step": 1043 + }, + { + "epoch": 4.244399185336049, + "grad_norm": 2.7202107906341553, + "learning_rate": 6.509928884985799e-08, + "loss": 0.20066425204277039, + "step": 1044 + }, + { + "epoch": 4.24847250509165, + "grad_norm": 2.458122730255127, + "learning_rate": 6.439775942972609e-08, + "loss": 0.19551938772201538, + "step": 1045 + }, + { + "epoch": 4.2525458248472505, + "grad_norm": 2.5357983112335205, + "learning_rate": 6.369977019793271e-08, + "loss": 0.20287911593914032, + "step": 1046 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 2.4644596576690674, + "learning_rate": 6.300532682711456e-08, + "loss": 0.20531214773654938, + "step": 1047 + }, + { + "epoch": 4.260692464358452, + "grad_norm": 2.4811034202575684, + "learning_rate": 6.231443496109117e-08, + "loss": 0.16013487428426743, + "step": 1048 + }, + { + "epoch": 4.2647657841140525, + "grad_norm": 2.415850877761841, + "learning_rate": 6.16271002148181e-08, + "loss": 0.18164420872926712, + "step": 1049 + }, + { + "epoch": 4.268839103869654, + "grad_norm": 2.4283230304718018, + "learning_rate": 6.094332817434211e-08, + "loss": 0.1915610432624817, + "step": 1050 + }, + { + "epoch": 4.272912423625255, + "grad_norm": 2.526712656021118, + "learning_rate": 6.026312439675551e-08, + "loss": 0.2085861638188362, + "step": 1051 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 2.4190962314605713, + "learning_rate": 5.9586494410150565e-08, + "loss": 0.17680321633815765, + "step": 1052 + }, + { + "epoch": 4.281059063136456, + "grad_norm": 2.567119598388672, + "learning_rate": 5.891344371357549e-08, + "loss": 0.15278445929288864, + "step": 1053 + }, + { + "epoch": 4.285132382892057, + "grad_norm": 2.6076533794403076, + "learning_rate": 5.824397777698858e-08, + "loss": 0.20571757853031158, + "step": 1054 + }, + { + "epoch": 4.289205702647658, + "grad_norm": 2.405752658843994, + "learning_rate": 5.7578102041214936e-08, + "loss": 0.1562228798866272, + "step": 1055 + }, + { + "epoch": 4.293279022403259, + "grad_norm": 2.524439811706543, + "learning_rate": 5.691582191790123e-08, + "loss": 0.20757433027029037, + "step": 1056 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 2.387105703353882, + "learning_rate": 5.6257142789472576e-08, + "loss": 0.1555587202310562, + "step": 1057 + }, + { + "epoch": 4.30142566191446, + "grad_norm": 2.396641969680786, + "learning_rate": 5.560207000908823e-08, + "loss": 0.18656989187002182, + "step": 1058 + }, + { + "epoch": 4.305498981670061, + "grad_norm": 2.687645196914673, + "learning_rate": 5.4950608900598326e-08, + "loss": 0.21030457317829132, + "step": 1059 + }, + { + "epoch": 4.3095723014256615, + "grad_norm": 2.4285991191864014, + "learning_rate": 5.430276475850026e-08, + "loss": 0.19240695238113403, + "step": 1060 + }, + { + "epoch": 4.313645621181263, + "grad_norm": 2.5999672412872314, + "learning_rate": 5.3658542847896395e-08, + "loss": 0.18602756410837173, + "step": 1061 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 2.3800199031829834, + "learning_rate": 5.30179484044504e-08, + "loss": 0.17461714148521423, + "step": 1062 + }, + { + "epoch": 4.321792260692464, + "grad_norm": 2.4830610752105713, + "learning_rate": 5.2380986634345204e-08, + "loss": 0.210846908390522, + "step": 1063 + }, + { + "epoch": 4.325865580448065, + "grad_norm": 2.3398642539978027, + "learning_rate": 5.1747662714240806e-08, + "loss": 0.18094143271446228, + "step": 1064 + }, + { + "epoch": 4.329938900203666, + "grad_norm": 2.7047762870788574, + "learning_rate": 5.1117981791231726e-08, + "loss": 0.21604597568511963, + "step": 1065 + }, + { + "epoch": 4.334012219959266, + "grad_norm": 2.4406027793884277, + "learning_rate": 5.049194898280579e-08, + "loss": 0.19460473209619522, + "step": 1066 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 2.416376829147339, + "learning_rate": 4.986956937680165e-08, + "loss": 0.18469534814357758, + "step": 1067 + }, + { + "epoch": 4.3421588594704685, + "grad_norm": 2.3338382244110107, + "learning_rate": 4.925084803136853e-08, + "loss": 0.16407328099012375, + "step": 1068 + }, + { + "epoch": 4.346232179226069, + "grad_norm": 2.4940404891967773, + "learning_rate": 4.863578997492407e-08, + "loss": 0.19178733974695206, + "step": 1069 + }, + { + "epoch": 4.35030549898167, + "grad_norm": 2.358466863632202, + "learning_rate": 4.802440020611426e-08, + "loss": 0.18211781978607178, + "step": 1070 + }, + { + "epoch": 4.3543788187372705, + "grad_norm": 2.4675703048706055, + "learning_rate": 4.7416683693772495e-08, + "loss": 0.20454513281583786, + "step": 1071 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 2.691836357116699, + "learning_rate": 4.6812645376879236e-08, + "loss": 0.20274968445301056, + "step": 1072 + }, + { + "epoch": 4.362525458248473, + "grad_norm": 2.635066032409668, + "learning_rate": 4.621229016452155e-08, + "loss": 0.18851862847805023, + "step": 1073 + }, + { + "epoch": 4.366598778004073, + "grad_norm": 2.4487643241882324, + "learning_rate": 4.561562293585364e-08, + "loss": 0.1614024117588997, + "step": 1074 + }, + { + "epoch": 4.370672097759674, + "grad_norm": 2.554473638534546, + "learning_rate": 4.5022648540057274e-08, + "loss": 0.18884680420160294, + "step": 1075 + }, + { + "epoch": 4.374745417515275, + "grad_norm": 2.5699715614318848, + "learning_rate": 4.443337179630174e-08, + "loss": 0.18893162161111832, + "step": 1076 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 2.412440061569214, + "learning_rate": 4.384779749370554e-08, + "loss": 0.18390610069036484, + "step": 1077 + }, + { + "epoch": 4.382892057026477, + "grad_norm": 2.4353432655334473, + "learning_rate": 4.3265930391296726e-08, + "loss": 0.19943667203187943, + "step": 1078 + }, + { + "epoch": 4.3869653767820775, + "grad_norm": 2.5288360118865967, + "learning_rate": 4.2687775217974674e-08, + "loss": 0.15438083559274673, + "step": 1079 + }, + { + "epoch": 4.391038696537678, + "grad_norm": 2.6911988258361816, + "learning_rate": 4.2113336672471245e-08, + "loss": 0.1961485669016838, + "step": 1080 + }, + { + "epoch": 4.395112016293279, + "grad_norm": 2.4872078895568848, + "learning_rate": 4.1542619423313276e-08, + "loss": 0.17868036031723022, + "step": 1081 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 2.635779619216919, + "learning_rate": 4.097562810878369e-08, + "loss": 0.18501786887645721, + "step": 1082 + }, + { + "epoch": 4.403258655804481, + "grad_norm": 2.4047963619232178, + "learning_rate": 4.041236733688474e-08, + "loss": 0.1908714473247528, + "step": 1083 + }, + { + "epoch": 4.407331975560082, + "grad_norm": 2.582057476043701, + "learning_rate": 3.985284168529995e-08, + "loss": 0.19209705293178558, + "step": 1084 + }, + { + "epoch": 4.411405295315682, + "grad_norm": 2.320059061050415, + "learning_rate": 3.929705570135711e-08, + "loss": 0.17084672302007675, + "step": 1085 + }, + { + "epoch": 4.415478615071283, + "grad_norm": 2.5907821655273438, + "learning_rate": 3.874501390199148e-08, + "loss": 0.18601538985967636, + "step": 1086 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 2.4668545722961426, + "learning_rate": 3.819672077370856e-08, + "loss": 0.17312505841255188, + "step": 1087 + }, + { + "epoch": 4.423625254582484, + "grad_norm": 2.8270812034606934, + "learning_rate": 3.7652180772548393e-08, + "loss": 0.15925130993127823, + "step": 1088 + }, + { + "epoch": 4.427698574338086, + "grad_norm": 2.352118968963623, + "learning_rate": 3.711139832404853e-08, + "loss": 0.1883508414030075, + "step": 1089 + }, + { + "epoch": 4.4317718940936865, + "grad_norm": 2.426455020904541, + "learning_rate": 3.657437782320888e-08, + "loss": 0.19179560989141464, + "step": 1090 + }, + { + "epoch": 4.435845213849287, + "grad_norm": 2.449763059616089, + "learning_rate": 3.604112363445522e-08, + "loss": 0.17501786351203918, + "step": 1091 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 2.496755361557007, + "learning_rate": 3.551164009160429e-08, + "loss": 0.17811695486307144, + "step": 1092 + }, + { + "epoch": 4.4439918533604885, + "grad_norm": 2.3908495903015137, + "learning_rate": 3.49859314978283e-08, + "loss": 0.16527411341667175, + "step": 1093 + }, + { + "epoch": 4.44806517311609, + "grad_norm": 2.344762086868286, + "learning_rate": 3.446400212562017e-08, + "loss": 0.1888575553894043, + "step": 1094 + }, + { + "epoch": 4.452138492871691, + "grad_norm": 2.396965503692627, + "learning_rate": 3.3945856216758384e-08, + "loss": 0.19079575687646866, + "step": 1095 + }, + { + "epoch": 4.456211812627291, + "grad_norm": 2.5549583435058594, + "learning_rate": 3.343149798227301e-08, + "loss": 0.16446177661418915, + "step": 1096 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 2.4434475898742676, + "learning_rate": 3.2920931602411105e-08, + "loss": 0.20346921682357788, + "step": 1097 + }, + { + "epoch": 4.464358452138493, + "grad_norm": 2.409318208694458, + "learning_rate": 3.241416122660312e-08, + "loss": 0.1957874372601509, + "step": 1098 + }, + { + "epoch": 4.468431771894093, + "grad_norm": 2.5695855617523193, + "learning_rate": 3.191119097342876e-08, + "loss": 0.20461179316043854, + "step": 1099 + }, + { + "epoch": 4.472505091649695, + "grad_norm": 2.499413013458252, + "learning_rate": 3.141202493058376e-08, + "loss": 0.19887447357177734, + "step": 1100 + }, + { + "epoch": 4.4765784114052956, + "grad_norm": 2.511868476867676, + "learning_rate": 3.0916667154846644e-08, + "loss": 0.21206560730934143, + "step": 1101 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 2.3708322048187256, + "learning_rate": 3.042512167204569e-08, + "loss": 0.19482630491256714, + "step": 1102 + }, + { + "epoch": 4.484725050916497, + "grad_norm": 2.418764114379883, + "learning_rate": 2.9937392477026256e-08, + "loss": 0.15218085050582886, + "step": 1103 + }, + { + "epoch": 4.4887983706720975, + "grad_norm": 2.526432752609253, + "learning_rate": 2.9453483533618208e-08, + "loss": 0.19341377913951874, + "step": 1104 + }, + { + "epoch": 4.492871690427698, + "grad_norm": 2.466676950454712, + "learning_rate": 2.8973398774603976e-08, + "loss": 0.20285610854625702, + "step": 1105 + }, + { + "epoch": 4.4969450101833, + "grad_norm": 2.626288890838623, + "learning_rate": 2.84971421016863e-08, + "loss": 0.21343690901994705, + "step": 1106 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 2.4493844509124756, + "learning_rate": 2.8024717385456752e-08, + "loss": 0.1735098510980606, + "step": 1107 + }, + { + "epoch": 4.505091649694501, + "grad_norm": 2.614302158355713, + "learning_rate": 2.755612846536398e-08, + "loss": 0.2130608856678009, + "step": 1108 + }, + { + "epoch": 4.509164969450102, + "grad_norm": 2.4521331787109375, + "learning_rate": 2.7091379149682682e-08, + "loss": 0.16439496725797653, + "step": 1109 + }, + { + "epoch": 4.513238289205702, + "grad_norm": 2.3728718757629395, + "learning_rate": 2.6630473215482895e-08, + "loss": 0.17642652988433838, + "step": 1110 + }, + { + "epoch": 4.517311608961304, + "grad_norm": 2.4868416786193848, + "learning_rate": 2.6173414408598826e-08, + "loss": 0.19394922256469727, + "step": 1111 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 2.190237283706665, + "learning_rate": 2.5720206443598736e-08, + "loss": 0.15093064308166504, + "step": 1112 + }, + { + "epoch": 4.525458248472505, + "grad_norm": 2.2836077213287354, + "learning_rate": 2.52708530037547e-08, + "loss": 0.16533400863409042, + "step": 1113 + }, + { + "epoch": 4.529531568228106, + "grad_norm": 2.499483823776245, + "learning_rate": 2.4825357741012686e-08, + "loss": 0.18762270361185074, + "step": 1114 + }, + { + "epoch": 4.5336048879837065, + "grad_norm": 2.6462838649749756, + "learning_rate": 2.438372427596258e-08, + "loss": 0.1893811672925949, + "step": 1115 + }, + { + "epoch": 4.537678207739307, + "grad_norm": 2.453942060470581, + "learning_rate": 2.394595619780926e-08, + "loss": 0.18495626747608185, + "step": 1116 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 2.721647024154663, + "learning_rate": 2.3512057064343006e-08, + "loss": 0.19172073900699615, + "step": 1117 + }, + { + "epoch": 4.545824847250509, + "grad_norm": 2.3400092124938965, + "learning_rate": 2.3082030401910868e-08, + "loss": 0.18133129179477692, + "step": 1118 + }, + { + "epoch": 4.54989816700611, + "grad_norm": 2.5185601711273193, + "learning_rate": 2.2655879705387683e-08, + "loss": 0.18830690532922745, + "step": 1119 + }, + { + "epoch": 4.553971486761711, + "grad_norm": 2.7922866344451904, + "learning_rate": 2.223360843814831e-08, + "loss": 0.17920264601707458, + "step": 1120 + }, + { + "epoch": 4.558044806517311, + "grad_norm": 2.25423526763916, + "learning_rate": 2.181522003203862e-08, + "loss": 0.1760845184326172, + "step": 1121 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 2.8005192279815674, + "learning_rate": 2.1400717887348096e-08, + "loss": 0.17970797419548035, + "step": 1122 + }, + { + "epoch": 4.566191446028514, + "grad_norm": 2.4870920181274414, + "learning_rate": 2.0990105372782342e-08, + "loss": 0.19520121812820435, + "step": 1123 + }, + { + "epoch": 4.570264765784114, + "grad_norm": 2.4267923831939697, + "learning_rate": 2.058338582543534e-08, + "loss": 0.18133552372455597, + "step": 1124 + }, + { + "epoch": 4.574338085539715, + "grad_norm": 2.4744374752044678, + "learning_rate": 2.0180562550762557e-08, + "loss": 0.20324096083641052, + "step": 1125 + }, + { + "epoch": 4.5784114052953155, + "grad_norm": 2.4335596561431885, + "learning_rate": 1.9781638822553958e-08, + "loss": 0.19217759370803833, + "step": 1126 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 2.4571220874786377, + "learning_rate": 1.9386617882907576e-08, + "loss": 0.1853143498301506, + "step": 1127 + }, + { + "epoch": 4.586558044806518, + "grad_norm": 2.7526497840881348, + "learning_rate": 1.8995502942202866e-08, + "loss": 0.15644675493240356, + "step": 1128 + }, + { + "epoch": 4.590631364562118, + "grad_norm": 2.5981698036193848, + "learning_rate": 1.8608297179074973e-08, + "loss": 0.1661221832036972, + "step": 1129 + }, + { + "epoch": 4.594704684317719, + "grad_norm": 2.5742714405059814, + "learning_rate": 1.8225003740388545e-08, + "loss": 0.1731417402625084, + "step": 1130 + }, + { + "epoch": 4.59877800407332, + "grad_norm": 2.418259620666504, + "learning_rate": 1.7845625741212455e-08, + "loss": 0.17661213874816895, + "step": 1131 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 2.518803358078003, + "learning_rate": 1.7470166264794252e-08, + "loss": 0.18510932475328445, + "step": 1132 + }, + { + "epoch": 4.606924643584521, + "grad_norm": 2.540313482284546, + "learning_rate": 1.709862836253534e-08, + "loss": 0.1894361674785614, + "step": 1133 + }, + { + "epoch": 4.610997963340123, + "grad_norm": 2.5993340015411377, + "learning_rate": 1.6731015053966012e-08, + "loss": 0.17485319077968597, + "step": 1134 + }, + { + "epoch": 4.615071283095723, + "grad_norm": 2.4957704544067383, + "learning_rate": 1.6367329326720856e-08, + "loss": 0.19985325634479523, + "step": 1135 + }, + { + "epoch": 4.619144602851324, + "grad_norm": 2.43721604347229, + "learning_rate": 1.6007574136514712e-08, + "loss": 0.18134716898202896, + "step": 1136 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 2.3459789752960205, + "learning_rate": 1.5651752407118247e-08, + "loss": 0.17994606494903564, + "step": 1137 + }, + { + "epoch": 4.627291242362525, + "grad_norm": 2.4452524185180664, + "learning_rate": 1.5299867030334813e-08, + "loss": 0.2009306699037552, + "step": 1138 + }, + { + "epoch": 4.631364562118126, + "grad_norm": 2.472385883331299, + "learning_rate": 1.4951920865976298e-08, + "loss": 0.18430952727794647, + "step": 1139 + }, + { + "epoch": 4.635437881873727, + "grad_norm": 2.39799165725708, + "learning_rate": 1.4607916741840254e-08, + "loss": 0.16439027339220047, + "step": 1140 + }, + { + "epoch": 4.639511201629328, + "grad_norm": 2.3951544761657715, + "learning_rate": 1.426785745368686e-08, + "loss": 0.17046649008989334, + "step": 1141 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 2.521660566329956, + "learning_rate": 1.3931745765216218e-08, + "loss": 0.19015470147132874, + "step": 1142 + }, + { + "epoch": 4.647657841140529, + "grad_norm": 2.4929487705230713, + "learning_rate": 1.3599584408045705e-08, + "loss": 0.195196270942688, + "step": 1143 + }, + { + "epoch": 4.65173116089613, + "grad_norm": 2.4070982933044434, + "learning_rate": 1.3271376081687934e-08, + "loss": 0.18301567435264587, + "step": 1144 + }, + { + "epoch": 4.655804480651732, + "grad_norm": 2.6266772747039795, + "learning_rate": 1.2947123453528885e-08, + "loss": 0.19102585315704346, + "step": 1145 + }, + { + "epoch": 4.659877800407332, + "grad_norm": 2.549821138381958, + "learning_rate": 1.2626829158806084e-08, + "loss": 0.1963408663868904, + "step": 1146 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 2.394179344177246, + "learning_rate": 1.2310495800587295e-08, + "loss": 0.17852062731981277, + "step": 1147 + }, + { + "epoch": 4.6680244399185336, + "grad_norm": 2.6832633018493652, + "learning_rate": 1.1998125949749194e-08, + "loss": 0.21865876764059067, + "step": 1148 + }, + { + "epoch": 4.672097759674134, + "grad_norm": 2.521888494491577, + "learning_rate": 1.168972214495667e-08, + "loss": 0.2011728659272194, + "step": 1149 + }, + { + "epoch": 4.676171079429735, + "grad_norm": 2.46395206451416, + "learning_rate": 1.1385286892642066e-08, + "loss": 0.2030443474650383, + "step": 1150 + }, + { + "epoch": 4.680244399185336, + "grad_norm": 2.5812623500823975, + "learning_rate": 1.1084822666984905e-08, + "loss": 0.20721124857664108, + "step": 1151 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 2.2974133491516113, + "learning_rate": 1.0788331909891646e-08, + "loss": 0.164021335542202, + "step": 1152 + }, + { + "epoch": 4.688391038696538, + "grad_norm": 2.2822930812835693, + "learning_rate": 1.0495817030976018e-08, + "loss": 0.14888529479503632, + "step": 1153 + }, + { + "epoch": 4.692464358452138, + "grad_norm": 2.5392258167266846, + "learning_rate": 1.0207280407539265e-08, + "loss": 0.1809844821691513, + "step": 1154 + }, + { + "epoch": 4.696537678207739, + "grad_norm": 2.4750709533691406, + "learning_rate": 9.922724384551051e-09, + "loss": 0.1907663643360138, + "step": 1155 + }, + { + "epoch": 4.70061099796334, + "grad_norm": 2.3656814098358154, + "learning_rate": 9.642151274630028e-09, + "loss": 0.1763923540711403, + "step": 1156 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 2.6838109493255615, + "learning_rate": 9.365563358025353e-09, + "loss": 0.1965761035680771, + "step": 1157 + }, + { + "epoch": 4.708757637474542, + "grad_norm": 2.723710775375366, + "learning_rate": 9.092962882598144e-09, + "loss": 0.20456959307193756, + "step": 1158 + }, + { + "epoch": 4.712830957230143, + "grad_norm": 2.6089608669281006, + "learning_rate": 8.824352063803053e-09, + "loss": 0.19157623499631882, + "step": 1159 + }, + { + "epoch": 4.716904276985743, + "grad_norm": 2.320207357406616, + "learning_rate": 8.559733084670396e-09, + "loss": 0.1886928454041481, + "step": 1160 + }, + { + "epoch": 4.720977596741344, + "grad_norm": 2.386166572570801, + "learning_rate": 8.29910809578832e-09, + "loss": 0.16740842908620834, + "step": 1161 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 2.4986507892608643, + "learning_rate": 8.042479215285503e-09, + "loss": 0.18630076944828033, + "step": 1162 + }, + { + "epoch": 4.729124236252546, + "grad_norm": 2.5233304500579834, + "learning_rate": 7.789848528813537e-09, + "loss": 0.17323897778987885, + "step": 1163 + }, + { + "epoch": 4.733197556008147, + "grad_norm": 2.336219072341919, + "learning_rate": 7.541218089530454e-09, + "loss": 0.18500008434057236, + "step": 1164 + }, + { + "epoch": 4.737270875763747, + "grad_norm": 2.426069736480713, + "learning_rate": 7.296589918083684e-09, + "loss": 0.1783820241689682, + "step": 1165 + }, + { + "epoch": 4.741344195519348, + "grad_norm": 2.6765623092651367, + "learning_rate": 7.055966002594005e-09, + "loss": 0.207184799015522, + "step": 1166 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 2.5222582817077637, + "learning_rate": 6.819348298638839e-09, + "loss": 0.2139280065894127, + "step": 1167 + }, + { + "epoch": 4.74949083503055, + "grad_norm": 2.328120470046997, + "learning_rate": 6.5867387292369295e-09, + "loss": 0.15476347506046295, + "step": 1168 + }, + { + "epoch": 4.753564154786151, + "grad_norm": 2.431166887283325, + "learning_rate": 6.358139184832412e-09, + "loss": 0.19098040461540222, + "step": 1169 + }, + { + "epoch": 4.757637474541752, + "grad_norm": 2.378601551055908, + "learning_rate": 6.1335515232793786e-09, + "loss": 0.1613384410738945, + "step": 1170 + }, + { + "epoch": 4.761710794297352, + "grad_norm": 2.562450647354126, + "learning_rate": 5.91297756982706e-09, + "loss": 0.18494479358196259, + "step": 1171 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 2.483607292175293, + "learning_rate": 5.696419117104667e-09, + "loss": 0.18615225702524185, + "step": 1172 + }, + { + "epoch": 4.7698574338085535, + "grad_norm": 2.571909189224243, + "learning_rate": 5.483877925107127e-09, + "loss": 0.15820663422346115, + "step": 1173 + }, + { + "epoch": 4.773930753564155, + "grad_norm": 2.3289737701416016, + "learning_rate": 5.275355721180541e-09, + "loss": 0.17181526124477386, + "step": 1174 + }, + { + "epoch": 4.778004073319756, + "grad_norm": 2.635626792907715, + "learning_rate": 5.070854200008356e-09, + "loss": 0.18911509215831757, + "step": 1175 + }, + { + "epoch": 4.782077393075356, + "grad_norm": 2.4061386585235596, + "learning_rate": 4.870375023597384e-09, + "loss": 0.19500445574522018, + "step": 1176 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 2.6663994789123535, + "learning_rate": 4.67391982126436e-09, + "loss": 0.18807929754257202, + "step": 1177 + }, + { + "epoch": 4.790224032586558, + "grad_norm": 2.541975259780884, + "learning_rate": 4.481490189622794e-09, + "loss": 0.2035481184720993, + "step": 1178 + }, + { + "epoch": 4.794297352342159, + "grad_norm": 2.6013271808624268, + "learning_rate": 4.293087692569974e-09, + "loss": 0.1858099028468132, + "step": 1179 + }, + { + "epoch": 4.79837067209776, + "grad_norm": 2.4530036449432373, + "learning_rate": 4.108713861274038e-09, + "loss": 0.1730939969420433, + "step": 1180 + }, + { + "epoch": 4.802443991853361, + "grad_norm": 2.538501501083374, + "learning_rate": 3.928370194161923e-09, + "loss": 0.19924624264240265, + "step": 1181 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 2.5811445713043213, + "learning_rate": 3.7520581569068785e-09, + "loss": 0.170890673995018, + "step": 1182 + }, + { + "epoch": 4.810590631364562, + "grad_norm": 2.467304229736328, + "learning_rate": 3.579779182416587e-09, + "loss": 0.19038555771112442, + "step": 1183 + }, + { + "epoch": 4.814663951120163, + "grad_norm": 2.435349702835083, + "learning_rate": 3.4115346708216695e-09, + "loss": 0.19310712069272995, + "step": 1184 + }, + { + "epoch": 4.818737270875764, + "grad_norm": 2.652028799057007, + "learning_rate": 3.247325989464089e-09, + "loss": 0.18610060214996338, + "step": 1185 + }, + { + "epoch": 4.822810590631365, + "grad_norm": 2.5963809490203857, + "learning_rate": 3.0871544728863217e-09, + "loss": 0.18380168080329895, + "step": 1186 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 2.5820107460021973, + "learning_rate": 2.9310214228202014e-09, + "loss": 0.1849624440073967, + "step": 1187 + }, + { + "epoch": 4.830957230142566, + "grad_norm": 2.3870606422424316, + "learning_rate": 2.778928108176648e-09, + "loss": 0.1531737670302391, + "step": 1188 + }, + { + "epoch": 4.835030549898167, + "grad_norm": 2.368928909301758, + "learning_rate": 2.630875765035068e-09, + "loss": 0.16189177334308624, + "step": 1189 + }, + { + "epoch": 4.839103869653767, + "grad_norm": 2.522529363632202, + "learning_rate": 2.4868655966334696e-09, + "loss": 0.18868660926818848, + "step": 1190 + }, + { + "epoch": 4.843177189409369, + "grad_norm": 2.583470582962036, + "learning_rate": 2.346898773358752e-09, + "loss": 0.21254148334264755, + "step": 1191 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 2.4336440563201904, + "learning_rate": 2.2109764327368774e-09, + "loss": 0.17807240039110184, + "step": 1192 + }, + { + "epoch": 4.85132382892057, + "grad_norm": 2.539942741394043, + "learning_rate": 2.079099679424101e-09, + "loss": 0.167090006172657, + "step": 1193 + }, + { + "epoch": 4.855397148676171, + "grad_norm": 2.442253351211548, + "learning_rate": 1.9512695851975346e-09, + "loss": 0.16879206895828247, + "step": 1194 + }, + { + "epoch": 4.8594704684317716, + "grad_norm": 2.496445894241333, + "learning_rate": 1.827487188946708e-09, + "loss": 0.1911175176501274, + "step": 1195 + }, + { + "epoch": 4.863543788187373, + "grad_norm": 2.352762222290039, + "learning_rate": 1.7077534966650765e-09, + "loss": 0.1814107596874237, + "step": 1196 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 2.6556501388549805, + "learning_rate": 1.59206948144186e-09, + "loss": 0.16374102234840393, + "step": 1197 + }, + { + "epoch": 4.871690427698574, + "grad_norm": 2.3610963821411133, + "learning_rate": 1.4804360834539398e-09, + "loss": 0.1801617071032524, + "step": 1198 + }, + { + "epoch": 4.875763747454175, + "grad_norm": 2.3272058963775635, + "learning_rate": 1.372854209958585e-09, + "loss": 0.15738121420145035, + "step": 1199 + }, + { + "epoch": 4.879837067209776, + "grad_norm": 2.4722721576690674, + "learning_rate": 1.2693247352856817e-09, + "loss": 0.17832408845424652, + "step": 1200 + }, + { + "epoch": 4.883910386965377, + "grad_norm": 2.7059404850006104, + "learning_rate": 1.169848500830961e-09, + "loss": 0.19288718700408936, + "step": 1201 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 2.5134401321411133, + "learning_rate": 1.0744263150489486e-09, + "loss": 0.16510120779275894, + "step": 1202 + }, + { + "epoch": 4.892057026476579, + "grad_norm": 2.3228554725646973, + "learning_rate": 9.830589534464139e-10, + "loss": 0.1725562885403633, + "step": 1203 + }, + { + "epoch": 4.896130346232179, + "grad_norm": 2.576732635498047, + "learning_rate": 8.957471585762655e-10, + "loss": 0.19249098747968674, + "step": 1204 + }, + { + "epoch": 4.90020366598778, + "grad_norm": 3.2054977416992188, + "learning_rate": 8.124916400311654e-10, + "loss": 0.19399181008338928, + "step": 1205 + }, + { + "epoch": 4.904276985743381, + "grad_norm": 2.5403683185577393, + "learning_rate": 7.332930744380905e-10, + "loss": 0.21494056284427643, + "step": 1206 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 2.5669965744018555, + "learning_rate": 6.581521054526695e-10, + "loss": 0.213642418384552, + "step": 1207 + }, + { + "epoch": 4.912423625254583, + "grad_norm": 2.3600635528564453, + "learning_rate": 5.870693437540764e-10, + "loss": 0.17898286134004593, + "step": 1208 + }, + { + "epoch": 4.916496945010183, + "grad_norm": 2.6826629638671875, + "learning_rate": 5.20045367039812e-10, + "loss": 0.18938665091991425, + "step": 1209 + }, + { + "epoch": 4.920570264765784, + "grad_norm": 2.3554890155792236, + "learning_rate": 4.570807200212634e-10, + "loss": 0.18093843758106232, + "step": 1210 + }, + { + "epoch": 4.924643584521385, + "grad_norm": 2.351396083831787, + "learning_rate": 3.9817591441926313e-10, + "loss": 0.17892272025346756, + "step": 1211 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 2.481386423110962, + "learning_rate": 3.4333142895975886e-10, + "loss": 0.18873242288827896, + "step": 1212 + }, + { + "epoch": 4.932790224032587, + "grad_norm": 2.3141331672668457, + "learning_rate": 2.9254770936998354e-10, + "loss": 0.15687178820371628, + "step": 1213 + }, + { + "epoch": 4.936863543788188, + "grad_norm": 2.5041322708129883, + "learning_rate": 2.458251683750134e-10, + "loss": 0.19991402328014374, + "step": 1214 + }, + { + "epoch": 4.940936863543788, + "grad_norm": 2.6752336025238037, + "learning_rate": 2.0316418569416015e-10, + "loss": 0.2112450748682022, + "step": 1215 + }, + { + "epoch": 4.945010183299389, + "grad_norm": 2.4857139587402344, + "learning_rate": 1.6456510803797286e-10, + "loss": 0.16738834232091904, + "step": 1216 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 2.5286214351654053, + "learning_rate": 1.3002824910551824e-10, + "loss": 0.19572018086910248, + "step": 1217 + }, + { + "epoch": 4.953156822810591, + "grad_norm": 2.585662841796875, + "learning_rate": 9.955388958177158e-11, + "loss": 0.21787291765213013, + "step": 1218 + }, + { + "epoch": 4.957230142566192, + "grad_norm": 2.3523364067077637, + "learning_rate": 7.314227713522969e-11, + "loss": 0.18198108673095703, + "step": 1219 + }, + { + "epoch": 4.961303462321792, + "grad_norm": 2.401686191558838, + "learning_rate": 5.079362641602358e-11, + "loss": 0.1640736535191536, + "step": 1220 + }, + { + "epoch": 4.965376782077393, + "grad_norm": 2.5535383224487305, + "learning_rate": 3.250811905419759e-11, + "loss": 0.20682203024625778, + "step": 1221 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 2.499046564102173, + "learning_rate": 1.8285903658099566e-11, + "loss": 0.16233273595571518, + "step": 1222 + }, + { + "epoch": 4.973523421588594, + "grad_norm": 2.4775757789611816, + "learning_rate": 8.127095813215135e-12, + "loss": 0.16816455125808716, + "step": 1223 + }, + { + "epoch": 4.977596741344195, + "grad_norm": 2.65974760055542, + "learning_rate": 2.0317780814460563e-12, + "loss": 0.20536164939403534, + "step": 1224 + }, + { + "epoch": 4.981670061099797, + "grad_norm": 2.462040901184082, + "learning_rate": 0.0, + "loss": 0.19432584941387177, + "step": 1225 + } + ], + "logging_steps": 1.0, + "max_steps": 1225, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 100.0, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}