| { | |
| "best_global_step": 900, | |
| "best_metric": 0.4358259439468384, | |
| "best_model_checkpoint": "./cv_jd_finetuned_model/checkpoint-900", | |
| "epoch": 2.5003474635163307, | |
| "eval_steps": 300, | |
| "global_step": 900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002779708130646282, | |
| "grad_norm": 0.47256213426589966, | |
| "learning_rate": 0.0, | |
| "loss": 1.7714, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005559416261292564, | |
| "grad_norm": 0.43564292788505554, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.7393, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008339124391938846, | |
| "grad_norm": 0.4533561170101166, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.7359, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.011118832522585128, | |
| "grad_norm": 0.4747747778892517, | |
| "learning_rate": 6e-06, | |
| "loss": 1.79, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01389854065323141, | |
| "grad_norm": 0.4859880208969116, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.7613, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01667824878387769, | |
| "grad_norm": 0.46851226687431335, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7779, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.019457956914523976, | |
| "grad_norm": 0.468657910823822, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.7623, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.022237665045170257, | |
| "grad_norm": 0.468654602766037, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.801, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02501737317581654, | |
| "grad_norm": 0.46270835399627686, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.6855, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.02779708130646282, | |
| "grad_norm": 0.4720575511455536, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.7408, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030576789437109102, | |
| "grad_norm": 0.4369213283061981, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7412, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03335649756775538, | |
| "grad_norm": 0.4069555401802063, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.6908, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03613620569840167, | |
| "grad_norm": 0.44282251596450806, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.7716, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03891591382904795, | |
| "grad_norm": 0.4090988039970398, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.66, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04169562195969423, | |
| "grad_norm": 0.44722557067871094, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.7268, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04447533009034051, | |
| "grad_norm": 0.423772931098938, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7199, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0472550382209868, | |
| "grad_norm": 0.3943701386451721, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.7025, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.05003474635163308, | |
| "grad_norm": 0.3880269527435303, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.6418, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05281445448227936, | |
| "grad_norm": 0.3725886046886444, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.6165, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05559416261292564, | |
| "grad_norm": 0.3829341530799866, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.6348, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05837387074357193, | |
| "grad_norm": 0.38574135303497314, | |
| "learning_rate": 4e-05, | |
| "loss": 1.6028, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.061153578874218205, | |
| "grad_norm": 0.34429848194122314, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.6449, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06393328700486449, | |
| "grad_norm": 0.33892038464546204, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.627, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.06671299513551077, | |
| "grad_norm": 0.31090790033340454, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.5466, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06949270326615706, | |
| "grad_norm": 0.3035236895084381, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.5175, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07227241139680333, | |
| "grad_norm": 0.2932650148868561, | |
| "learning_rate": 5e-05, | |
| "loss": 1.4668, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07505211952744961, | |
| "grad_norm": 0.27232736349105835, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.5586, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0778318276580959, | |
| "grad_norm": 0.25519007444381714, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.5489, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.08061153578874218, | |
| "grad_norm": 0.23789168894290924, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.5095, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08339124391938846, | |
| "grad_norm": 0.2281704694032669, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.5199, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08617095205003475, | |
| "grad_norm": 0.23827943205833435, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3646, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08895066018068103, | |
| "grad_norm": 0.24246342480182648, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.4594, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0917303683113273, | |
| "grad_norm": 0.2456520050764084, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.4243, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0945100764419736, | |
| "grad_norm": 0.2509920001029968, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.3821, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09728978457261987, | |
| "grad_norm": 0.2679535448551178, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 1.3136, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.10006949270326616, | |
| "grad_norm": 0.2893657684326172, | |
| "learning_rate": 7e-05, | |
| "loss": 1.3514, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.10284920083391244, | |
| "grad_norm": 0.31409314274787903, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.3186, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.10562890896455872, | |
| "grad_norm": 0.29165610671043396, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.2345, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10840861709520501, | |
| "grad_norm": 0.31779947876930237, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.2938, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.11118832522585129, | |
| "grad_norm": 0.298592209815979, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 1.2937, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11396803335649756, | |
| "grad_norm": 0.30116093158721924, | |
| "learning_rate": 8e-05, | |
| "loss": 1.2387, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11674774148714386, | |
| "grad_norm": 0.28846633434295654, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.2688, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11952744961779013, | |
| "grad_norm": 0.27421048283576965, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.3615, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.12230715774843641, | |
| "grad_norm": 0.25272664427757263, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.2837, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1250868658790827, | |
| "grad_norm": 0.28714367747306824, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.1749, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12786657400972898, | |
| "grad_norm": 0.27853626012802124, | |
| "learning_rate": 9e-05, | |
| "loss": 1.2653, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.13064628214037527, | |
| "grad_norm": 0.29935234785079956, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.0846, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.13342599027102153, | |
| "grad_norm": 0.2770155072212219, | |
| "learning_rate": 9.4e-05, | |
| "loss": 1.1273, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.13620569840166782, | |
| "grad_norm": 0.29866236448287964, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.1003, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13898540653231412, | |
| "grad_norm": 0.3060087561607361, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.0825, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14176511466296038, | |
| "grad_norm": 0.2942337095737457, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1054, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.14454482279360667, | |
| "grad_norm": 0.2934916019439697, | |
| "learning_rate": 0.00010200000000000001, | |
| "loss": 1.0711, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.14732453092425296, | |
| "grad_norm": 0.33025407791137695, | |
| "learning_rate": 0.00010400000000000001, | |
| "loss": 0.9641, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.15010423905489922, | |
| "grad_norm": 0.3368203341960907, | |
| "learning_rate": 0.00010600000000000002, | |
| "loss": 0.9126, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.15288394718554552, | |
| "grad_norm": 0.3417721092700958, | |
| "learning_rate": 0.00010800000000000001, | |
| "loss": 1.0195, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1556636553161918, | |
| "grad_norm": 0.33795756101608276, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 0.8796, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.15844336344683807, | |
| "grad_norm": 0.341162770986557, | |
| "learning_rate": 0.00011200000000000001, | |
| "loss": 0.9361, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.16122307157748436, | |
| "grad_norm": 0.3660062849521637, | |
| "learning_rate": 0.00011399999999999999, | |
| "loss": 0.8179, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.16400277970813065, | |
| "grad_norm": 0.3034508526325226, | |
| "learning_rate": 0.000116, | |
| "loss": 0.8709, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.16678248783877692, | |
| "grad_norm": 0.31419962644577026, | |
| "learning_rate": 0.000118, | |
| "loss": 0.8925, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1695621959694232, | |
| "grad_norm": 0.32202863693237305, | |
| "learning_rate": 0.00012, | |
| "loss": 0.9444, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1723419041000695, | |
| "grad_norm": 0.29973354935646057, | |
| "learning_rate": 0.000122, | |
| "loss": 0.9334, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.17512161223071576, | |
| "grad_norm": 0.2913822531700134, | |
| "learning_rate": 0.000124, | |
| "loss": 0.8633, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.17790132036136205, | |
| "grad_norm": 0.3123023509979248, | |
| "learning_rate": 0.000126, | |
| "loss": 0.8474, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.18068102849200834, | |
| "grad_norm": 0.36691340804100037, | |
| "learning_rate": 0.00012800000000000002, | |
| "loss": 0.8491, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1834607366226546, | |
| "grad_norm": 0.2867358922958374, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 0.8812, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1862404447533009, | |
| "grad_norm": 0.33814680576324463, | |
| "learning_rate": 0.000132, | |
| "loss": 0.7977, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1890201528839472, | |
| "grad_norm": 0.2705424129962921, | |
| "learning_rate": 0.000134, | |
| "loss": 0.8189, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.19179986101459348, | |
| "grad_norm": 0.27234122157096863, | |
| "learning_rate": 0.00013600000000000003, | |
| "loss": 0.8661, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.19457956914523974, | |
| "grad_norm": 0.3268398642539978, | |
| "learning_rate": 0.000138, | |
| "loss": 0.7359, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.19735927727588604, | |
| "grad_norm": 0.29902949929237366, | |
| "learning_rate": 0.00014, | |
| "loss": 0.7943, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.20013898540653233, | |
| "grad_norm": 0.31011486053466797, | |
| "learning_rate": 0.000142, | |
| "loss": 1.0169, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2029186935371786, | |
| "grad_norm": 0.28908616304397583, | |
| "learning_rate": 0.000144, | |
| "loss": 0.7617, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.20569840166782488, | |
| "grad_norm": 0.31617382168769836, | |
| "learning_rate": 0.000146, | |
| "loss": 0.7512, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.20847810979847117, | |
| "grad_norm": 0.32012176513671875, | |
| "learning_rate": 0.000148, | |
| "loss": 0.838, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.21125781792911744, | |
| "grad_norm": 0.32813915610313416, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.8314, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.21403752605976373, | |
| "grad_norm": 0.311210572719574, | |
| "learning_rate": 0.000152, | |
| "loss": 0.7498, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.21681723419041002, | |
| "grad_norm": 0.2896203100681305, | |
| "learning_rate": 0.000154, | |
| "loss": 0.8062, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.21959694232105628, | |
| "grad_norm": 0.30190443992614746, | |
| "learning_rate": 0.00015600000000000002, | |
| "loss": 0.8198, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.22237665045170257, | |
| "grad_norm": 0.31988754868507385, | |
| "learning_rate": 0.00015800000000000002, | |
| "loss": 0.8781, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.22515635858234886, | |
| "grad_norm": 0.29290610551834106, | |
| "learning_rate": 0.00016, | |
| "loss": 0.7494, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.22793606671299513, | |
| "grad_norm": 0.29982951283454895, | |
| "learning_rate": 0.000162, | |
| "loss": 0.6786, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.23071577484364142, | |
| "grad_norm": 0.3004741966724396, | |
| "learning_rate": 0.000164, | |
| "loss": 0.6723, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2334954829742877, | |
| "grad_norm": 0.30331557989120483, | |
| "learning_rate": 0.000166, | |
| "loss": 0.6686, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.23627519110493397, | |
| "grad_norm": 0.320073664188385, | |
| "learning_rate": 0.000168, | |
| "loss": 0.7309, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.23905489923558026, | |
| "grad_norm": 0.37164461612701416, | |
| "learning_rate": 0.00017, | |
| "loss": 0.883, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.24183460736622656, | |
| "grad_norm": 0.3153933584690094, | |
| "learning_rate": 0.000172, | |
| "loss": 0.7538, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.24461431549687282, | |
| "grad_norm": 0.37620604038238525, | |
| "learning_rate": 0.000174, | |
| "loss": 0.7233, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2473940236275191, | |
| "grad_norm": 0.3310216963291168, | |
| "learning_rate": 0.00017600000000000002, | |
| "loss": 0.668, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2501737317581654, | |
| "grad_norm": 0.3648437261581421, | |
| "learning_rate": 0.00017800000000000002, | |
| "loss": 0.7219, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2529534398888117, | |
| "grad_norm": 0.33160319924354553, | |
| "learning_rate": 0.00018, | |
| "loss": 0.6968, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.25573314801945796, | |
| "grad_norm": 0.3842083513736725, | |
| "learning_rate": 0.000182, | |
| "loss": 0.7334, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2585128561501042, | |
| "grad_norm": 0.34885042905807495, | |
| "learning_rate": 0.00018400000000000003, | |
| "loss": 0.6107, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.26129256428075054, | |
| "grad_norm": 0.4350070059299469, | |
| "learning_rate": 0.00018600000000000002, | |
| "loss": 0.6964, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2640722724113968, | |
| "grad_norm": 0.3995032012462616, | |
| "learning_rate": 0.000188, | |
| "loss": 0.7111, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.26685198054204307, | |
| "grad_norm": 0.4035239517688751, | |
| "learning_rate": 0.00019, | |
| "loss": 0.6004, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.2696316886726894, | |
| "grad_norm": 0.39563143253326416, | |
| "learning_rate": 0.000192, | |
| "loss": 0.7174, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.27241139680333565, | |
| "grad_norm": 0.4484492540359497, | |
| "learning_rate": 0.000194, | |
| "loss": 0.732, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2751911049339819, | |
| "grad_norm": 0.4170342683792114, | |
| "learning_rate": 0.000196, | |
| "loss": 0.6415, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.27797081306462823, | |
| "grad_norm": 0.36684471368789673, | |
| "learning_rate": 0.00019800000000000002, | |
| "loss": 0.7303, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2807505211952745, | |
| "grad_norm": 0.417539119720459, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7282, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.28353022932592076, | |
| "grad_norm": 0.3748982846736908, | |
| "learning_rate": 0.00019979591836734694, | |
| "loss": 0.7453, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2863099374565671, | |
| "grad_norm": 0.3669414222240448, | |
| "learning_rate": 0.0001995918367346939, | |
| "loss": 0.7206, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.28908964558721334, | |
| "grad_norm": 0.43745529651641846, | |
| "learning_rate": 0.00019938775510204082, | |
| "loss": 0.7008, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.2918693537178596, | |
| "grad_norm": 0.4588426351547241, | |
| "learning_rate": 0.00019918367346938775, | |
| "loss": 0.6981, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2946490618485059, | |
| "grad_norm": 0.46913576126098633, | |
| "learning_rate": 0.0001989795918367347, | |
| "loss": 0.7276, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2974287699791522, | |
| "grad_norm": 0.38426539301872253, | |
| "learning_rate": 0.00019877551020408164, | |
| "loss": 0.6409, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.30020847810979845, | |
| "grad_norm": 0.4170157313346863, | |
| "learning_rate": 0.0001985714285714286, | |
| "loss": 0.6113, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.30298818624044477, | |
| "grad_norm": 0.4131574332714081, | |
| "learning_rate": 0.00019836734693877553, | |
| "loss": 0.5925, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.30576789437109103, | |
| "grad_norm": 0.4434458017349243, | |
| "learning_rate": 0.00019816326530612246, | |
| "loss": 0.7052, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3085476025017373, | |
| "grad_norm": 0.45088446140289307, | |
| "learning_rate": 0.00019795918367346938, | |
| "loss": 0.5575, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3113273106323836, | |
| "grad_norm": 0.48712992668151855, | |
| "learning_rate": 0.00019775510204081634, | |
| "loss": 0.6706, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3141070187630299, | |
| "grad_norm": 0.44475221633911133, | |
| "learning_rate": 0.00019755102040816327, | |
| "loss": 0.6569, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.31688672689367614, | |
| "grad_norm": 0.4516242742538452, | |
| "learning_rate": 0.0001973469387755102, | |
| "loss": 0.5981, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.31966643502432246, | |
| "grad_norm": 0.4031848907470703, | |
| "learning_rate": 0.00019714285714285716, | |
| "loss": 0.7407, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3224461431549687, | |
| "grad_norm": 0.5948619842529297, | |
| "learning_rate": 0.00019693877551020409, | |
| "loss": 0.5176, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.325225851285615, | |
| "grad_norm": 0.42586585879325867, | |
| "learning_rate": 0.00019673469387755104, | |
| "loss": 0.5305, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3280055594162613, | |
| "grad_norm": 1.2202147245407104, | |
| "learning_rate": 0.00019653061224489797, | |
| "loss": 0.7352, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.33078526754690757, | |
| "grad_norm": 0.4561997354030609, | |
| "learning_rate": 0.0001963265306122449, | |
| "loss": 0.4617, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.33356497567755383, | |
| "grad_norm": 0.38021618127822876, | |
| "learning_rate": 0.00019612244897959183, | |
| "loss": 0.5754, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.33634468380820015, | |
| "grad_norm": 0.4197412431240082, | |
| "learning_rate": 0.0001959183673469388, | |
| "loss": 0.6426, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3391243919388464, | |
| "grad_norm": 0.39121460914611816, | |
| "learning_rate": 0.00019571428571428572, | |
| "loss": 0.6242, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3419041000694927, | |
| "grad_norm": 0.413143515586853, | |
| "learning_rate": 0.00019551020408163265, | |
| "loss": 0.6013, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.344683808200139, | |
| "grad_norm": 0.4672967195510864, | |
| "learning_rate": 0.0001953061224489796, | |
| "loss": 0.5439, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.34746351633078526, | |
| "grad_norm": 0.4725366532802582, | |
| "learning_rate": 0.00019510204081632656, | |
| "loss": 0.6042, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3502432244614315, | |
| "grad_norm": 0.483952134847641, | |
| "learning_rate": 0.0001948979591836735, | |
| "loss": 0.4544, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.35302293259207784, | |
| "grad_norm": 0.39228469133377075, | |
| "learning_rate": 0.00019469387755102042, | |
| "loss": 0.6016, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3558026407227241, | |
| "grad_norm": 0.4152607023715973, | |
| "learning_rate": 0.00019448979591836735, | |
| "loss": 0.4931, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.35858234885337037, | |
| "grad_norm": 0.402338445186615, | |
| "learning_rate": 0.0001942857142857143, | |
| "loss": 0.525, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3613620569840167, | |
| "grad_norm": 0.42365092039108276, | |
| "learning_rate": 0.00019408163265306123, | |
| "loss": 0.6608, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36414176511466295, | |
| "grad_norm": 0.4249265491962433, | |
| "learning_rate": 0.00019387755102040816, | |
| "loss": 0.5565, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3669214732453092, | |
| "grad_norm": 0.6368371248245239, | |
| "learning_rate": 0.0001936734693877551, | |
| "loss": 0.4547, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.36970118137595553, | |
| "grad_norm": 0.37348538637161255, | |
| "learning_rate": 0.00019346938775510205, | |
| "loss": 0.4073, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3724808895066018, | |
| "grad_norm": 0.3562554717063904, | |
| "learning_rate": 0.000193265306122449, | |
| "loss": 0.4968, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3752605976372481, | |
| "grad_norm": 0.42278632521629333, | |
| "learning_rate": 0.00019306122448979593, | |
| "loss": 0.5421, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3780403057678944, | |
| "grad_norm": 0.47804152965545654, | |
| "learning_rate": 0.00019285714285714286, | |
| "loss": 0.6516, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.38082001389854064, | |
| "grad_norm": 0.4298154413700104, | |
| "learning_rate": 0.0001926530612244898, | |
| "loss": 0.4563, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.38359972202918696, | |
| "grad_norm": 0.37862634658813477, | |
| "learning_rate": 0.00019244897959183675, | |
| "loss": 0.5742, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3863794301598332, | |
| "grad_norm": 0.5051096081733704, | |
| "learning_rate": 0.00019224489795918368, | |
| "loss": 0.4427, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3891591382904795, | |
| "grad_norm": 0.4495854079723358, | |
| "learning_rate": 0.0001920408163265306, | |
| "loss": 0.5915, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3919388464211258, | |
| "grad_norm": 0.535527765750885, | |
| "learning_rate": 0.00019183673469387756, | |
| "loss": 0.4995, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.39471855455177207, | |
| "grad_norm": 0.4394996464252472, | |
| "learning_rate": 0.00019163265306122452, | |
| "loss": 0.4898, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.39749826268241834, | |
| "grad_norm": 0.3254806697368622, | |
| "learning_rate": 0.00019142857142857145, | |
| "loss": 0.4628, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.40027797081306465, | |
| "grad_norm": 0.4018654525279999, | |
| "learning_rate": 0.00019122448979591838, | |
| "loss": 0.571, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4030576789437109, | |
| "grad_norm": 0.4496287703514099, | |
| "learning_rate": 0.0001910204081632653, | |
| "loss": 0.4502, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4058373870743572, | |
| "grad_norm": 0.3519289195537567, | |
| "learning_rate": 0.00019081632653061227, | |
| "loss": 0.5355, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.4086170952050035, | |
| "grad_norm": 0.44861000776290894, | |
| "learning_rate": 0.0001906122448979592, | |
| "loss": 0.5026, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.41139680333564976, | |
| "grad_norm": 0.4265352487564087, | |
| "learning_rate": 0.00019040816326530612, | |
| "loss": 0.5444, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.414176511466296, | |
| "grad_norm": 0.3963538706302643, | |
| "learning_rate": 0.00019020408163265305, | |
| "loss": 0.5061, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.41695621959694235, | |
| "grad_norm": 0.4118248522281647, | |
| "learning_rate": 0.00019, | |
| "loss": 0.4564, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4197359277275886, | |
| "grad_norm": 0.3554701507091522, | |
| "learning_rate": 0.00018979591836734697, | |
| "loss": 0.5173, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.4225156358582349, | |
| "grad_norm": 0.3614487648010254, | |
| "learning_rate": 0.0001895918367346939, | |
| "loss": 0.5661, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.4252953439888812, | |
| "grad_norm": 0.3155677318572998, | |
| "learning_rate": 0.00018938775510204083, | |
| "loss": 0.5626, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.42807505211952745, | |
| "grad_norm": 0.4833175241947174, | |
| "learning_rate": 0.00018918367346938776, | |
| "loss": 0.6273, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.4308547602501737, | |
| "grad_norm": 0.37056633830070496, | |
| "learning_rate": 0.0001889795918367347, | |
| "loss": 0.5984, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.43363446838082004, | |
| "grad_norm": 0.36759206652641296, | |
| "learning_rate": 0.00018877551020408164, | |
| "loss": 0.5022, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4364141765114663, | |
| "grad_norm": 0.4722707271575928, | |
| "learning_rate": 0.00018857142857142857, | |
| "loss": 0.5679, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.43919388464211256, | |
| "grad_norm": 0.5502737760543823, | |
| "learning_rate": 0.0001883673469387755, | |
| "loss": 0.5484, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4419735927727589, | |
| "grad_norm": 0.41092830896377563, | |
| "learning_rate": 0.00018816326530612246, | |
| "loss": 0.5346, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.44475330090340515, | |
| "grad_norm": 0.35556185245513916, | |
| "learning_rate": 0.0001879591836734694, | |
| "loss": 0.5094, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4475330090340514, | |
| "grad_norm": 0.4490595757961273, | |
| "learning_rate": 0.00018775510204081634, | |
| "loss": 0.5659, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.45031271716469773, | |
| "grad_norm": 0.4202437102794647, | |
| "learning_rate": 0.00018755102040816327, | |
| "loss": 0.5904, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.453092425295344, | |
| "grad_norm": 0.39136362075805664, | |
| "learning_rate": 0.00018734693877551023, | |
| "loss": 0.5214, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.45587213342599026, | |
| "grad_norm": 0.3859161138534546, | |
| "learning_rate": 0.00018714285714285716, | |
| "loss": 0.5011, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4586518415566366, | |
| "grad_norm": 0.5123438835144043, | |
| "learning_rate": 0.0001869387755102041, | |
| "loss": 0.6053, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.46143154968728284, | |
| "grad_norm": 0.466137170791626, | |
| "learning_rate": 0.00018673469387755102, | |
| "loss": 0.4499, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4642112578179291, | |
| "grad_norm": 0.4053160548210144, | |
| "learning_rate": 0.00018653061224489797, | |
| "loss": 0.5452, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.4669909659485754, | |
| "grad_norm": 0.3608758747577667, | |
| "learning_rate": 0.0001863265306122449, | |
| "loss": 0.4941, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4697706740792217, | |
| "grad_norm": 0.36142799258232117, | |
| "learning_rate": 0.00018612244897959183, | |
| "loss": 0.5056, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.47255038220986795, | |
| "grad_norm": 0.2707204818725586, | |
| "learning_rate": 0.0001859183673469388, | |
| "loss": 0.3958, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47533009034051427, | |
| "grad_norm": 0.3678928017616272, | |
| "learning_rate": 0.00018571428571428572, | |
| "loss": 0.47, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.47810979847116053, | |
| "grad_norm": 0.37397581338882446, | |
| "learning_rate": 0.00018551020408163267, | |
| "loss": 0.5774, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4808895066018068, | |
| "grad_norm": 0.3246667981147766, | |
| "learning_rate": 0.0001853061224489796, | |
| "loss": 0.4607, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.4836692147324531, | |
| "grad_norm": 0.34554117918014526, | |
| "learning_rate": 0.00018510204081632653, | |
| "loss": 0.3862, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.4864489228630994, | |
| "grad_norm": 0.356503963470459, | |
| "learning_rate": 0.00018489795918367346, | |
| "loss": 0.4969, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.48922863099374564, | |
| "grad_norm": 0.3777051866054535, | |
| "learning_rate": 0.00018469387755102042, | |
| "loss": 0.4337, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.49200833912439196, | |
| "grad_norm": 0.3148108422756195, | |
| "learning_rate": 0.00018448979591836735, | |
| "loss": 0.4244, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4947880472550382, | |
| "grad_norm": 0.49584245681762695, | |
| "learning_rate": 0.00018428571428571428, | |
| "loss": 0.5775, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4975677553856845, | |
| "grad_norm": 0.34690654277801514, | |
| "learning_rate": 0.00018408163265306123, | |
| "loss": 0.577, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.5003474635163307, | |
| "grad_norm": 0.36074724793434143, | |
| "learning_rate": 0.0001838775510204082, | |
| "loss": 0.441, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5031271716469771, | |
| "grad_norm": 0.3747076392173767, | |
| "learning_rate": 0.00018367346938775512, | |
| "loss": 0.5022, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.5059068797776234, | |
| "grad_norm": 0.41926515102386475, | |
| "learning_rate": 0.00018346938775510205, | |
| "loss": 0.4178, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.5086865879082696, | |
| "grad_norm": 0.32807132601737976, | |
| "learning_rate": 0.00018326530612244898, | |
| "loss": 0.4493, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.5114662960389159, | |
| "grad_norm": 0.3409689962863922, | |
| "learning_rate": 0.00018306122448979593, | |
| "loss": 0.557, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.5142460041695622, | |
| "grad_norm": 0.4119493365287781, | |
| "learning_rate": 0.00018285714285714286, | |
| "loss": 0.4969, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5170257123002084, | |
| "grad_norm": 0.2936202585697174, | |
| "learning_rate": 0.0001826530612244898, | |
| "loss": 0.4222, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.5198054204308548, | |
| "grad_norm": 0.4165465235710144, | |
| "learning_rate": 0.00018244897959183672, | |
| "loss": 0.5563, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.5225851285615011, | |
| "grad_norm": 0.34087347984313965, | |
| "learning_rate": 0.00018224489795918368, | |
| "loss": 0.5508, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5253648366921473, | |
| "grad_norm": 0.39741185307502747, | |
| "learning_rate": 0.00018204081632653064, | |
| "loss": 0.4635, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5281445448227936, | |
| "grad_norm": 0.373943954706192, | |
| "learning_rate": 0.00018183673469387757, | |
| "loss": 0.4667, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5309242529534399, | |
| "grad_norm": 0.3398171067237854, | |
| "learning_rate": 0.0001816326530612245, | |
| "loss": 0.488, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5337039610840861, | |
| "grad_norm": 0.5641401410102844, | |
| "learning_rate": 0.00018142857142857142, | |
| "loss": 0.377, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5364836692147325, | |
| "grad_norm": 0.47961774468421936, | |
| "learning_rate": 0.00018122448979591838, | |
| "loss": 0.5046, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5392633773453788, | |
| "grad_norm": 0.4699658155441284, | |
| "learning_rate": 0.0001810204081632653, | |
| "loss": 0.4936, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.542043085476025, | |
| "grad_norm": 0.2884581983089447, | |
| "learning_rate": 0.00018081632653061224, | |
| "loss": 0.4059, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5448227936066713, | |
| "grad_norm": 0.4616682827472687, | |
| "learning_rate": 0.00018061224489795917, | |
| "loss": 0.553, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5476025017373176, | |
| "grad_norm": 0.35249197483062744, | |
| "learning_rate": 0.00018040816326530615, | |
| "loss": 0.4637, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5503822098679638, | |
| "grad_norm": 0.4296030104160309, | |
| "learning_rate": 0.00018020408163265308, | |
| "loss": 0.5117, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5531619179986101, | |
| "grad_norm": 0.3835342228412628, | |
| "learning_rate": 0.00018, | |
| "loss": 0.5288, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5559416261292565, | |
| "grad_norm": 0.3516342043876648, | |
| "learning_rate": 0.00017979591836734694, | |
| "loss": 0.5322, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5587213342599027, | |
| "grad_norm": 0.4156709909439087, | |
| "learning_rate": 0.0001795918367346939, | |
| "loss": 0.4886, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.561501042390549, | |
| "grad_norm": 0.32229727506637573, | |
| "learning_rate": 0.00017938775510204083, | |
| "loss": 0.4377, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5642807505211953, | |
| "grad_norm": 0.384962260723114, | |
| "learning_rate": 0.00017918367346938776, | |
| "loss": 0.5345, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.5670604586518415, | |
| "grad_norm": 0.41784927248954773, | |
| "learning_rate": 0.00017897959183673469, | |
| "loss": 0.5386, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5698401667824878, | |
| "grad_norm": 0.46640586853027344, | |
| "learning_rate": 0.00017877551020408164, | |
| "loss": 0.4544, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5726198749131342, | |
| "grad_norm": 0.34132063388824463, | |
| "learning_rate": 0.0001785714285714286, | |
| "loss": 0.454, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5753995830437804, | |
| "grad_norm": 0.461137592792511, | |
| "learning_rate": 0.00017836734693877553, | |
| "loss": 0.478, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5781792911744267, | |
| "grad_norm": 0.5866886377334595, | |
| "learning_rate": 0.00017816326530612246, | |
| "loss": 0.546, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.580958999305073, | |
| "grad_norm": 0.3185846209526062, | |
| "learning_rate": 0.0001779591836734694, | |
| "loss": 0.5385, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5837387074357192, | |
| "grad_norm": 0.43802475929260254, | |
| "learning_rate": 0.00017775510204081634, | |
| "loss": 0.554, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5865184155663655, | |
| "grad_norm": 0.2952940762042999, | |
| "learning_rate": 0.00017755102040816327, | |
| "loss": 0.4358, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5892981236970118, | |
| "grad_norm": 0.32370179891586304, | |
| "learning_rate": 0.0001773469387755102, | |
| "loss": 0.4418, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.592077831827658, | |
| "grad_norm": 0.38936758041381836, | |
| "learning_rate": 0.00017714285714285713, | |
| "loss": 0.409, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5948575399583044, | |
| "grad_norm": 0.3157341182231903, | |
| "learning_rate": 0.0001769387755102041, | |
| "loss": 0.4019, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5976372480889507, | |
| "grad_norm": 0.39846348762512207, | |
| "learning_rate": 0.00017673469387755104, | |
| "loss": 0.4804, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.6004169562195969, | |
| "grad_norm": 0.3177434802055359, | |
| "learning_rate": 0.00017653061224489797, | |
| "loss": 0.4092, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.6031966643502432, | |
| "grad_norm": 0.39317747950553894, | |
| "learning_rate": 0.0001763265306122449, | |
| "loss": 0.5265, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.6059763724808895, | |
| "grad_norm": 0.26835039258003235, | |
| "learning_rate": 0.00017612244897959186, | |
| "loss": 0.3989, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.6087560806115357, | |
| "grad_norm": 0.4008518159389496, | |
| "learning_rate": 0.0001759183673469388, | |
| "loss": 0.498, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.6115357887421821, | |
| "grad_norm": 0.31940510869026184, | |
| "learning_rate": 0.00017571428571428572, | |
| "loss": 0.4493, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6143154968728284, | |
| "grad_norm": 0.43578922748565674, | |
| "learning_rate": 0.00017551020408163265, | |
| "loss": 0.4742, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.6170952050034746, | |
| "grad_norm": 0.3231724500656128, | |
| "learning_rate": 0.0001753061224489796, | |
| "loss": 0.5019, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.6198749131341209, | |
| "grad_norm": 0.30763351917266846, | |
| "learning_rate": 0.00017510204081632653, | |
| "loss": 0.6015, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.6226546212647672, | |
| "grad_norm": 0.32532060146331787, | |
| "learning_rate": 0.0001748979591836735, | |
| "loss": 0.3062, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.6254343293954134, | |
| "grad_norm": 0.39833390712738037, | |
| "learning_rate": 0.00017469387755102042, | |
| "loss": 0.3969, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6282140375260598, | |
| "grad_norm": 0.411516010761261, | |
| "learning_rate": 0.00017448979591836735, | |
| "loss": 0.4763, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6309937456567061, | |
| "grad_norm": 0.40843451023101807, | |
| "learning_rate": 0.0001742857142857143, | |
| "loss": 0.4998, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.6337734537873523, | |
| "grad_norm": 0.39122384786605835, | |
| "learning_rate": 0.00017408163265306123, | |
| "loss": 0.4365, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6365531619179986, | |
| "grad_norm": 0.37201231718063354, | |
| "learning_rate": 0.00017387755102040816, | |
| "loss": 0.4427, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.6393328700486449, | |
| "grad_norm": 0.4359400272369385, | |
| "learning_rate": 0.0001736734693877551, | |
| "loss": 0.4686, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6421125781792911, | |
| "grad_norm": 0.4254358112812042, | |
| "learning_rate": 0.00017346938775510205, | |
| "loss": 0.4274, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6448922863099374, | |
| "grad_norm": 0.383859783411026, | |
| "learning_rate": 0.00017326530612244898, | |
| "loss": 0.3683, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.6476719944405838, | |
| "grad_norm": 0.41463732719421387, | |
| "learning_rate": 0.00017306122448979594, | |
| "loss": 0.4606, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.65045170257123, | |
| "grad_norm": 0.32726776599884033, | |
| "learning_rate": 0.00017285714285714287, | |
| "loss": 0.3883, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.6532314107018763, | |
| "grad_norm": 0.4143235683441162, | |
| "learning_rate": 0.00017265306122448982, | |
| "loss": 0.5929, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6560111188325226, | |
| "grad_norm": 0.419869601726532, | |
| "learning_rate": 0.00017244897959183675, | |
| "loss": 0.527, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6587908269631688, | |
| "grad_norm": 0.4330616295337677, | |
| "learning_rate": 0.00017224489795918368, | |
| "loss": 0.523, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6615705350938151, | |
| "grad_norm": 0.42590218782424927, | |
| "learning_rate": 0.0001720408163265306, | |
| "loss": 0.4894, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6643502432244615, | |
| "grad_norm": 0.5940903425216675, | |
| "learning_rate": 0.00017183673469387757, | |
| "loss": 0.458, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.6671299513551077, | |
| "grad_norm": 0.3962993025779724, | |
| "learning_rate": 0.0001716326530612245, | |
| "loss": 0.4934, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.669909659485754, | |
| "grad_norm": 0.3732195496559143, | |
| "learning_rate": 0.00017142857142857143, | |
| "loss": 0.4779, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6726893676164003, | |
| "grad_norm": 0.39993205666542053, | |
| "learning_rate": 0.00017122448979591838, | |
| "loss": 0.3835, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.6754690757470465, | |
| "grad_norm": 0.37663185596466064, | |
| "learning_rate": 0.0001710204081632653, | |
| "loss": 0.3876, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6782487838776928, | |
| "grad_norm": 0.33526360988616943, | |
| "learning_rate": 0.00017081632653061227, | |
| "loss": 0.4415, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.6810284920083391, | |
| "grad_norm": 0.4333009421825409, | |
| "learning_rate": 0.0001706122448979592, | |
| "loss": 0.4897, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6838082001389854, | |
| "grad_norm": 0.4983868896961212, | |
| "learning_rate": 0.00017040816326530613, | |
| "loss": 0.4477, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6865879082696317, | |
| "grad_norm": 0.39352232217788696, | |
| "learning_rate": 0.00017020408163265306, | |
| "loss": 0.4524, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.689367616400278, | |
| "grad_norm": 0.40973153710365295, | |
| "learning_rate": 0.00017, | |
| "loss": 0.4884, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6921473245309242, | |
| "grad_norm": 0.33771470189094543, | |
| "learning_rate": 0.00016979591836734694, | |
| "loss": 0.4938, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6949270326615705, | |
| "grad_norm": 0.34451448917388916, | |
| "learning_rate": 0.00016959183673469387, | |
| "loss": 0.3722, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6977067407922168, | |
| "grad_norm": 0.38862481713294983, | |
| "learning_rate": 0.00016938775510204083, | |
| "loss": 0.426, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.700486448922863, | |
| "grad_norm": 0.41708311438560486, | |
| "learning_rate": 0.00016918367346938778, | |
| "loss": 0.5239, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.7032661570535094, | |
| "grad_norm": 0.4317916929721832, | |
| "learning_rate": 0.0001689795918367347, | |
| "loss": 0.4556, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.7060458651841557, | |
| "grad_norm": 0.3338056802749634, | |
| "learning_rate": 0.00016877551020408164, | |
| "loss": 0.3332, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.7088255733148019, | |
| "grad_norm": 0.41748374700546265, | |
| "learning_rate": 0.00016857142857142857, | |
| "loss": 0.4594, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.7116052814454482, | |
| "grad_norm": 0.4123172461986542, | |
| "learning_rate": 0.00016836734693877553, | |
| "loss": 0.4289, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.7143849895760945, | |
| "grad_norm": 0.40377530455589294, | |
| "learning_rate": 0.00016816326530612246, | |
| "loss": 0.5744, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.7171646977067407, | |
| "grad_norm": 0.369667649269104, | |
| "learning_rate": 0.0001679591836734694, | |
| "loss": 0.5848, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.7199444058373871, | |
| "grad_norm": 0.39214134216308594, | |
| "learning_rate": 0.00016775510204081632, | |
| "loss": 0.3939, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.7227241139680334, | |
| "grad_norm": 0.3648947775363922, | |
| "learning_rate": 0.00016755102040816327, | |
| "loss": 0.3941, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7255038220986796, | |
| "grad_norm": 0.3529266119003296, | |
| "learning_rate": 0.00016734693877551023, | |
| "loss": 0.3505, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.7282835302293259, | |
| "grad_norm": 0.3326796889305115, | |
| "learning_rate": 0.00016714285714285716, | |
| "loss": 0.3973, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.7310632383599722, | |
| "grad_norm": 0.37780439853668213, | |
| "learning_rate": 0.0001669387755102041, | |
| "loss": 0.4882, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7338429464906184, | |
| "grad_norm": 0.4995975196361542, | |
| "learning_rate": 0.00016673469387755102, | |
| "loss": 0.4417, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.7366226546212647, | |
| "grad_norm": 0.40474021434783936, | |
| "learning_rate": 0.00016653061224489797, | |
| "loss": 0.4511, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.7394023627519111, | |
| "grad_norm": 0.8704133629798889, | |
| "learning_rate": 0.0001663265306122449, | |
| "loss": 0.6664, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.7421820708825573, | |
| "grad_norm": 0.7991705536842346, | |
| "learning_rate": 0.00016612244897959183, | |
| "loss": 0.5963, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.7449617790132036, | |
| "grad_norm": 0.4240580201148987, | |
| "learning_rate": 0.00016591836734693876, | |
| "loss": 0.4474, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.7477414871438499, | |
| "grad_norm": 0.4676007032394409, | |
| "learning_rate": 0.00016571428571428575, | |
| "loss": 0.5125, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.7505211952744962, | |
| "grad_norm": 0.2894349694252014, | |
| "learning_rate": 0.00016551020408163268, | |
| "loss": 0.4297, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7533009034051424, | |
| "grad_norm": 0.4876716732978821, | |
| "learning_rate": 0.0001653061224489796, | |
| "loss": 0.391, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.7560806115357888, | |
| "grad_norm": 0.37176764011383057, | |
| "learning_rate": 0.00016510204081632653, | |
| "loss": 0.4825, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.7588603196664351, | |
| "grad_norm": 0.49970927834510803, | |
| "learning_rate": 0.0001648979591836735, | |
| "loss": 0.3729, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.7616400277970813, | |
| "grad_norm": 0.3654176890850067, | |
| "learning_rate": 0.00016469387755102042, | |
| "loss": 0.4255, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.7644197359277276, | |
| "grad_norm": 0.44572046399116516, | |
| "learning_rate": 0.00016448979591836735, | |
| "loss": 0.502, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7671994440583739, | |
| "grad_norm": 0.4408852159976959, | |
| "learning_rate": 0.00016428571428571428, | |
| "loss": 0.4563, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7699791521890201, | |
| "grad_norm": 0.36519378423690796, | |
| "learning_rate": 0.00016408163265306124, | |
| "loss": 0.4626, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7727588603196665, | |
| "grad_norm": 0.48397397994995117, | |
| "learning_rate": 0.0001638775510204082, | |
| "loss": 0.4365, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7755385684503128, | |
| "grad_norm": 0.39511287212371826, | |
| "learning_rate": 0.00016367346938775512, | |
| "loss": 0.4178, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.778318276580959, | |
| "grad_norm": 0.5128254890441895, | |
| "learning_rate": 0.00016346938775510205, | |
| "loss": 0.3998, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7810979847116053, | |
| "grad_norm": 0.5283316969871521, | |
| "learning_rate": 0.00016326530612244898, | |
| "loss": 0.3428, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.7838776928422516, | |
| "grad_norm": 0.4386744201183319, | |
| "learning_rate": 0.00016306122448979594, | |
| "loss": 0.3976, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7866574009728978, | |
| "grad_norm": 0.5863499641418457, | |
| "learning_rate": 0.00016285714285714287, | |
| "loss": 0.5529, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.7894371091035441, | |
| "grad_norm": 0.27297189831733704, | |
| "learning_rate": 0.0001626530612244898, | |
| "loss": 0.3282, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7922168172341905, | |
| "grad_norm": 0.29970136284828186, | |
| "learning_rate": 0.00016244897959183672, | |
| "loss": 0.365, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7949965253648367, | |
| "grad_norm": 0.3835904598236084, | |
| "learning_rate": 0.00016224489795918368, | |
| "loss": 0.4768, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.797776233495483, | |
| "grad_norm": 0.4071420729160309, | |
| "learning_rate": 0.0001620408163265306, | |
| "loss": 0.5426, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.8005559416261293, | |
| "grad_norm": 0.28784051537513733, | |
| "learning_rate": 0.00016183673469387757, | |
| "loss": 0.4108, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.8033356497567755, | |
| "grad_norm": 0.4735048711299896, | |
| "learning_rate": 0.0001616326530612245, | |
| "loss": 0.5495, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.8061153578874218, | |
| "grad_norm": 0.2624104917049408, | |
| "learning_rate": 0.00016142857142857145, | |
| "loss": 0.4767, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8088950660180682, | |
| "grad_norm": 0.38945189118385315, | |
| "learning_rate": 0.00016122448979591838, | |
| "loss": 0.5055, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.8116747741487144, | |
| "grad_norm": 0.4181615114212036, | |
| "learning_rate": 0.0001610204081632653, | |
| "loss": 0.4497, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.8144544822793607, | |
| "grad_norm": 0.37034186720848083, | |
| "learning_rate": 0.00016081632653061224, | |
| "loss": 0.3471, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.817234190410007, | |
| "grad_norm": 0.3208980858325958, | |
| "learning_rate": 0.0001606122448979592, | |
| "loss": 0.3868, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.8200138985406532, | |
| "grad_norm": 0.4345311224460602, | |
| "learning_rate": 0.00016040816326530613, | |
| "loss": 0.439, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.8227936066712995, | |
| "grad_norm": 0.31438905000686646, | |
| "learning_rate": 0.00016020408163265306, | |
| "loss": 0.4661, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.8255733148019458, | |
| "grad_norm": 0.2713527977466583, | |
| "learning_rate": 0.00016, | |
| "loss": 0.3945, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.828353022932592, | |
| "grad_norm": 0.2937558591365814, | |
| "learning_rate": 0.00015979591836734694, | |
| "loss": 0.3287, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.8311327310632384, | |
| "grad_norm": 0.37041494250297546, | |
| "learning_rate": 0.0001595918367346939, | |
| "loss": 0.5133, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.8339124391938847, | |
| "grad_norm": 0.4814389646053314, | |
| "learning_rate": 0.00015938775510204083, | |
| "loss": 0.498, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8339124391938847, | |
| "eval_loss": 0.4747912585735321, | |
| "eval_runtime": 212.458, | |
| "eval_samples_per_second": 1.694, | |
| "eval_steps_per_second": 1.694, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8366921473245309, | |
| "grad_norm": 0.3283718526363373, | |
| "learning_rate": 0.00015918367346938776, | |
| "loss": 0.4485, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.8394718554551772, | |
| "grad_norm": 0.38611340522766113, | |
| "learning_rate": 0.0001589795918367347, | |
| "loss": 0.4188, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.8422515635858235, | |
| "grad_norm": 0.29100102186203003, | |
| "learning_rate": 0.00015877551020408164, | |
| "loss": 0.3489, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.8450312717164697, | |
| "grad_norm": 0.32593274116516113, | |
| "learning_rate": 0.00015857142857142857, | |
| "loss": 0.391, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.8478109798471161, | |
| "grad_norm": 0.5677832365036011, | |
| "learning_rate": 0.0001583673469387755, | |
| "loss": 0.4837, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8505906879777624, | |
| "grad_norm": 0.334756463766098, | |
| "learning_rate": 0.00015816326530612246, | |
| "loss": 0.3565, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.8533703961084086, | |
| "grad_norm": 0.5696679353713989, | |
| "learning_rate": 0.00015795918367346942, | |
| "loss": 0.4759, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.8561501042390549, | |
| "grad_norm": 0.25399741530418396, | |
| "learning_rate": 0.00015775510204081634, | |
| "loss": 0.4281, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.8589298123697012, | |
| "grad_norm": 0.3591265380382538, | |
| "learning_rate": 0.00015755102040816327, | |
| "loss": 0.4464, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.8617095205003474, | |
| "grad_norm": 0.3444579839706421, | |
| "learning_rate": 0.0001573469387755102, | |
| "loss": 0.4565, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8644892286309938, | |
| "grad_norm": 0.3792060315608978, | |
| "learning_rate": 0.00015714285714285716, | |
| "loss": 0.5026, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.8672689367616401, | |
| "grad_norm": 0.2783966362476349, | |
| "learning_rate": 0.0001569387755102041, | |
| "loss": 0.4912, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.8700486448922863, | |
| "grad_norm": 0.31529495120048523, | |
| "learning_rate": 0.00015673469387755102, | |
| "loss": 0.3993, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8728283530229326, | |
| "grad_norm": 0.3652310073375702, | |
| "learning_rate": 0.00015653061224489795, | |
| "loss": 0.5009, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8756080611535789, | |
| "grad_norm": 0.3820590078830719, | |
| "learning_rate": 0.0001563265306122449, | |
| "loss": 0.3424, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8783877692842251, | |
| "grad_norm": 0.3363693654537201, | |
| "learning_rate": 0.00015612244897959186, | |
| "loss": 0.5251, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8811674774148714, | |
| "grad_norm": 0.31948599219322205, | |
| "learning_rate": 0.0001559183673469388, | |
| "loss": 0.4199, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.8839471855455178, | |
| "grad_norm": 0.4196965992450714, | |
| "learning_rate": 0.00015571428571428572, | |
| "loss": 0.493, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.886726893676164, | |
| "grad_norm": 0.4823121428489685, | |
| "learning_rate": 0.00015551020408163265, | |
| "loss": 0.43, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8895066018068103, | |
| "grad_norm": 0.32050636410713196, | |
| "learning_rate": 0.0001553061224489796, | |
| "loss": 0.4247, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8922863099374566, | |
| "grad_norm": 0.3208867609500885, | |
| "learning_rate": 0.00015510204081632654, | |
| "loss": 0.3224, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8950660180681028, | |
| "grad_norm": 0.36326608061790466, | |
| "learning_rate": 0.00015489795918367346, | |
| "loss": 0.4366, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8978457261987491, | |
| "grad_norm": 0.44612210988998413, | |
| "learning_rate": 0.0001546938775510204, | |
| "loss": 0.4212, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.9006254343293955, | |
| "grad_norm": 0.5350055694580078, | |
| "learning_rate": 0.00015448979591836735, | |
| "loss": 0.42, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.9034051424600417, | |
| "grad_norm": 0.5650726556777954, | |
| "learning_rate": 0.0001542857142857143, | |
| "loss": 0.5001, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.906184850590688, | |
| "grad_norm": 0.2960895001888275, | |
| "learning_rate": 0.00015408163265306124, | |
| "loss": 0.2943, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.9089645587213343, | |
| "grad_norm": 0.4059947729110718, | |
| "learning_rate": 0.00015387755102040817, | |
| "loss": 0.4147, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.9117442668519805, | |
| "grad_norm": 0.31508710980415344, | |
| "learning_rate": 0.00015367346938775512, | |
| "loss": 0.3567, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.9145239749826268, | |
| "grad_norm": 0.30250322818756104, | |
| "learning_rate": 0.00015346938775510205, | |
| "loss": 0.4534, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.9173036831132731, | |
| "grad_norm": 0.45266756415367126, | |
| "learning_rate": 0.00015326530612244898, | |
| "loss": 0.4675, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.9200833912439194, | |
| "grad_norm": 0.2555678188800812, | |
| "learning_rate": 0.0001530612244897959, | |
| "loss": 0.4132, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.9228630993745657, | |
| "grad_norm": 0.3064277768135071, | |
| "learning_rate": 0.00015285714285714287, | |
| "loss": 0.5, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.925642807505212, | |
| "grad_norm": 0.40955591201782227, | |
| "learning_rate": 0.00015265306122448982, | |
| "loss": 0.4131, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.9284225156358582, | |
| "grad_norm": 0.32479333877563477, | |
| "learning_rate": 0.00015244897959183675, | |
| "loss": 0.4303, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.9312022237665045, | |
| "grad_norm": 0.4044603705406189, | |
| "learning_rate": 0.00015224489795918368, | |
| "loss": 0.4498, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.9339819318971508, | |
| "grad_norm": 0.4078894555568695, | |
| "learning_rate": 0.0001520408163265306, | |
| "loss": 0.4442, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.936761640027797, | |
| "grad_norm": 0.4486389756202698, | |
| "learning_rate": 0.00015183673469387757, | |
| "loss": 0.3902, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.9395413481584434, | |
| "grad_norm": 0.3673665523529053, | |
| "learning_rate": 0.0001516326530612245, | |
| "loss": 0.4158, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.9423210562890897, | |
| "grad_norm": 0.28217577934265137, | |
| "learning_rate": 0.00015142857142857143, | |
| "loss": 0.4327, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.9451007644197359, | |
| "grad_norm": 0.30868950486183167, | |
| "learning_rate": 0.00015122448979591836, | |
| "loss": 0.411, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9478804725503822, | |
| "grad_norm": 0.27955666184425354, | |
| "learning_rate": 0.0001510204081632653, | |
| "loss": 0.4125, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.9506601806810285, | |
| "grad_norm": 0.49935731291770935, | |
| "learning_rate": 0.00015081632653061227, | |
| "loss": 0.369, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.9534398888116747, | |
| "grad_norm": 0.26663827896118164, | |
| "learning_rate": 0.0001506122448979592, | |
| "loss": 0.3864, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.9562195969423211, | |
| "grad_norm": 0.32989761233329773, | |
| "learning_rate": 0.00015040816326530613, | |
| "loss": 0.3859, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.9589993050729674, | |
| "grad_norm": 0.37399861216545105, | |
| "learning_rate": 0.00015020408163265306, | |
| "loss": 0.4813, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9617790132036136, | |
| "grad_norm": 0.3359721302986145, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.4135, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.9645587213342599, | |
| "grad_norm": 0.29389145970344543, | |
| "learning_rate": 0.00014979591836734694, | |
| "loss": 0.407, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.9673384294649062, | |
| "grad_norm": 0.39862900972366333, | |
| "learning_rate": 0.00014959183673469387, | |
| "loss": 0.4659, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.9701181375955524, | |
| "grad_norm": 0.36769983172416687, | |
| "learning_rate": 0.00014938775510204083, | |
| "loss": 0.4076, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.9728978457261988, | |
| "grad_norm": 0.29756975173950195, | |
| "learning_rate": 0.00014918367346938776, | |
| "loss": 0.3922, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9756775538568451, | |
| "grad_norm": 0.4466356337070465, | |
| "learning_rate": 0.00014897959183673472, | |
| "loss": 0.4123, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9784572619874913, | |
| "grad_norm": 0.5409598350524902, | |
| "learning_rate": 0.00014877551020408164, | |
| "loss": 0.2904, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9812369701181376, | |
| "grad_norm": 0.5088945627212524, | |
| "learning_rate": 0.00014857142857142857, | |
| "loss": 0.3555, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.9840166782487839, | |
| "grad_norm": 0.41926005482673645, | |
| "learning_rate": 0.00014836734693877553, | |
| "loss": 0.3168, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9867963863794301, | |
| "grad_norm": 0.3465840220451355, | |
| "learning_rate": 0.00014816326530612246, | |
| "loss": 0.4542, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9895760945100764, | |
| "grad_norm": 0.29874542355537415, | |
| "learning_rate": 0.0001479591836734694, | |
| "loss": 0.4541, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.9923558026407228, | |
| "grad_norm": 0.32591426372528076, | |
| "learning_rate": 0.00014775510204081632, | |
| "loss": 0.3355, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.995135510771369, | |
| "grad_norm": 0.6298475861549377, | |
| "learning_rate": 0.00014755102040816328, | |
| "loss": 0.4996, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9979152189020153, | |
| "grad_norm": 0.6214368939399719, | |
| "learning_rate": 0.0001473469387755102, | |
| "loss": 0.5361, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.29986897110939026, | |
| "learning_rate": 0.00014714285714285716, | |
| "loss": 0.2962, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.0027797081306462, | |
| "grad_norm": 0.27898886799812317, | |
| "learning_rate": 0.0001469387755102041, | |
| "loss": 0.3485, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.0055594162612926, | |
| "grad_norm": 0.3615298569202423, | |
| "learning_rate": 0.00014673469387755102, | |
| "loss": 0.4651, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.0083391243919388, | |
| "grad_norm": 0.31298184394836426, | |
| "learning_rate": 0.00014653061224489798, | |
| "loss": 0.4272, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.011118832522585, | |
| "grad_norm": 0.27693745493888855, | |
| "learning_rate": 0.0001463265306122449, | |
| "loss": 0.36, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.0138985406532315, | |
| "grad_norm": 0.3590083122253418, | |
| "learning_rate": 0.00014612244897959183, | |
| "loss": 0.4548, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.0166782487838777, | |
| "grad_norm": 0.3620007336139679, | |
| "learning_rate": 0.0001459183673469388, | |
| "loss": 0.3749, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.019457956914524, | |
| "grad_norm": 0.46811267733573914, | |
| "learning_rate": 0.00014571428571428572, | |
| "loss": 0.4619, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.0222376650451703, | |
| "grad_norm": 0.3233739137649536, | |
| "learning_rate": 0.00014551020408163265, | |
| "loss": 0.4955, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.0250173731758165, | |
| "grad_norm": 0.3408876061439514, | |
| "learning_rate": 0.0001453061224489796, | |
| "loss": 0.3564, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.0277970813064627, | |
| "grad_norm": 0.33044156432151794, | |
| "learning_rate": 0.00014510204081632654, | |
| "loss": 0.4784, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.0305767894371092, | |
| "grad_norm": 0.4836321175098419, | |
| "learning_rate": 0.0001448979591836735, | |
| "loss": 0.2986, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.0333564975677554, | |
| "grad_norm": 0.3213842213153839, | |
| "learning_rate": 0.00014469387755102042, | |
| "loss": 0.3768, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.0361362056984016, | |
| "grad_norm": 0.4250502288341522, | |
| "learning_rate": 0.00014448979591836735, | |
| "loss": 0.4024, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.038915913829048, | |
| "grad_norm": 0.5815131068229675, | |
| "learning_rate": 0.00014428571428571428, | |
| "loss": 0.4631, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.0416956219596942, | |
| "grad_norm": 0.3646114468574524, | |
| "learning_rate": 0.00014408163265306124, | |
| "loss": 0.4658, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.0444753300903404, | |
| "grad_norm": 0.4233134388923645, | |
| "learning_rate": 0.00014387755102040817, | |
| "loss": 0.4287, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.0472550382209869, | |
| "grad_norm": 0.29038846492767334, | |
| "learning_rate": 0.0001436734693877551, | |
| "loss": 0.3537, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.050034746351633, | |
| "grad_norm": 0.3281858265399933, | |
| "learning_rate": 0.00014346938775510205, | |
| "loss": 0.3594, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.0528144544822793, | |
| "grad_norm": 0.3003385066986084, | |
| "learning_rate": 0.00014326530612244898, | |
| "loss": 0.429, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.0555941626129257, | |
| "grad_norm": 0.42301732301712036, | |
| "learning_rate": 0.00014306122448979594, | |
| "loss": 0.3517, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.058373870743572, | |
| "grad_norm": 0.3951142728328705, | |
| "learning_rate": 0.00014285714285714287, | |
| "loss": 0.3341, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.0611535788742181, | |
| "grad_norm": 0.3241204023361206, | |
| "learning_rate": 0.0001426530612244898, | |
| "loss": 0.4393, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.0639332870048646, | |
| "grad_norm": 0.37029820680618286, | |
| "learning_rate": 0.00014244897959183673, | |
| "loss": 0.2978, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.0667129951355108, | |
| "grad_norm": 0.2742837965488434, | |
| "learning_rate": 0.00014224489795918368, | |
| "loss": 0.3306, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.069492703266157, | |
| "grad_norm": 0.26682788133621216, | |
| "learning_rate": 0.0001420408163265306, | |
| "loss": 0.4009, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.0722724113968034, | |
| "grad_norm": 0.3267010450363159, | |
| "learning_rate": 0.00014183673469387754, | |
| "loss": 0.303, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.0750521195274496, | |
| "grad_norm": 0.3354584276676178, | |
| "learning_rate": 0.0001416326530612245, | |
| "loss": 0.3721, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.0778318276580958, | |
| "grad_norm": 0.33933278918266296, | |
| "learning_rate": 0.00014142857142857145, | |
| "loss": 0.437, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.0806115357887423, | |
| "grad_norm": 0.6241405606269836, | |
| "learning_rate": 0.00014122448979591838, | |
| "loss": 0.4561, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.0833912439193885, | |
| "grad_norm": 0.6265623569488525, | |
| "learning_rate": 0.00014102040816326531, | |
| "loss": 0.5705, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0861709520500347, | |
| "grad_norm": 0.4538445770740509, | |
| "learning_rate": 0.00014081632653061224, | |
| "loss": 0.3084, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.088950660180681, | |
| "grad_norm": 0.2949851453304291, | |
| "learning_rate": 0.0001406122448979592, | |
| "loss": 0.3774, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.0917303683113273, | |
| "grad_norm": 0.25528761744499207, | |
| "learning_rate": 0.00014040816326530613, | |
| "loss": 0.362, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.0945100764419735, | |
| "grad_norm": 0.25837084650993347, | |
| "learning_rate": 0.00014020408163265306, | |
| "loss": 0.341, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.09728978457262, | |
| "grad_norm": 0.3381750285625458, | |
| "learning_rate": 0.00014, | |
| "loss": 0.409, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.1000694927032661, | |
| "grad_norm": 0.296481728553772, | |
| "learning_rate": 0.00013979591836734694, | |
| "loss": 0.3378, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.1028492008339124, | |
| "grad_norm": 0.2741848826408386, | |
| "learning_rate": 0.0001395918367346939, | |
| "loss": 0.3112, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.1056289089645588, | |
| "grad_norm": 0.2974790036678314, | |
| "learning_rate": 0.00013938775510204083, | |
| "loss": 0.434, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.108408617095205, | |
| "grad_norm": 0.5039945840835571, | |
| "learning_rate": 0.00013918367346938776, | |
| "loss": 0.4185, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.1111883252258512, | |
| "grad_norm": 0.3443140387535095, | |
| "learning_rate": 0.0001389795918367347, | |
| "loss": 0.2975, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.1139680333564976, | |
| "grad_norm": 0.38616564869880676, | |
| "learning_rate": 0.00013877551020408165, | |
| "loss": 0.5452, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.1167477414871438, | |
| "grad_norm": 0.34680166840553284, | |
| "learning_rate": 0.00013857142857142857, | |
| "loss": 0.3738, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.11952744961779, | |
| "grad_norm": 0.30565282702445984, | |
| "learning_rate": 0.0001383673469387755, | |
| "loss": 0.4436, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.1223071577484365, | |
| "grad_norm": 0.3291468322277069, | |
| "learning_rate": 0.00013816326530612243, | |
| "loss": 0.369, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.1250868658790827, | |
| "grad_norm": 0.36906489729881287, | |
| "learning_rate": 0.00013795918367346942, | |
| "loss": 0.4367, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.127866574009729, | |
| "grad_norm": 0.2758554220199585, | |
| "learning_rate": 0.00013775510204081635, | |
| "loss": 0.437, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.1306462821403753, | |
| "grad_norm": 0.3159145414829254, | |
| "learning_rate": 0.00013755102040816328, | |
| "loss": 0.4224, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.1334259902710215, | |
| "grad_norm": 0.25819459557533264, | |
| "learning_rate": 0.0001373469387755102, | |
| "loss": 0.331, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.1362056984016677, | |
| "grad_norm": 0.33512744307518005, | |
| "learning_rate": 0.00013714285714285716, | |
| "loss": 0.4196, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.1389854065323142, | |
| "grad_norm": 0.3556046187877655, | |
| "learning_rate": 0.0001369387755102041, | |
| "loss": 0.3745, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.1417651146629604, | |
| "grad_norm": 0.48709914088249207, | |
| "learning_rate": 0.00013673469387755102, | |
| "loss": 0.346, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.1445448227936066, | |
| "grad_norm": 0.3741767704486847, | |
| "learning_rate": 0.00013653061224489795, | |
| "loss": 0.4871, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.147324530924253, | |
| "grad_norm": 0.4699570834636688, | |
| "learning_rate": 0.0001363265306122449, | |
| "loss": 0.3674, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.1501042390548992, | |
| "grad_norm": 0.32232385873794556, | |
| "learning_rate": 0.00013612244897959184, | |
| "loss": 0.3821, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.1528839471855454, | |
| "grad_norm": 0.34662458300590515, | |
| "learning_rate": 0.0001359183673469388, | |
| "loss": 0.3121, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.1556636553161919, | |
| "grad_norm": 0.32288941740989685, | |
| "learning_rate": 0.00013571428571428572, | |
| "loss": 0.3808, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.158443363446838, | |
| "grad_norm": 0.3495519459247589, | |
| "learning_rate": 0.00013551020408163265, | |
| "loss": 0.352, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.1612230715774843, | |
| "grad_norm": 0.3813597559928894, | |
| "learning_rate": 0.0001353061224489796, | |
| "loss": 0.4044, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.1640027797081307, | |
| "grad_norm": 0.2824418246746063, | |
| "learning_rate": 0.00013510204081632654, | |
| "loss": 0.3902, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.166782487838777, | |
| "grad_norm": 0.25736352801322937, | |
| "learning_rate": 0.00013489795918367347, | |
| "loss": 0.3357, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1695621959694231, | |
| "grad_norm": 0.33888882398605347, | |
| "learning_rate": 0.0001346938775510204, | |
| "loss": 0.3691, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.1723419041000696, | |
| "grad_norm": 0.3581472933292389, | |
| "learning_rate": 0.00013448979591836735, | |
| "loss": 0.3576, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.1751216122307158, | |
| "grad_norm": 0.3296295702457428, | |
| "learning_rate": 0.00013428571428571428, | |
| "loss": 0.3832, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.177901320361362, | |
| "grad_norm": 0.33110418915748596, | |
| "learning_rate": 0.00013408163265306124, | |
| "loss": 0.3247, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.1806810284920084, | |
| "grad_norm": 0.3340078592300415, | |
| "learning_rate": 0.00013387755102040817, | |
| "loss": 0.3718, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.1834607366226546, | |
| "grad_norm": 0.3025374114513397, | |
| "learning_rate": 0.00013367346938775512, | |
| "loss": 0.3766, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.1862404447533008, | |
| "grad_norm": 0.2760503590106964, | |
| "learning_rate": 0.00013346938775510205, | |
| "loss": 0.3311, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.1890201528839472, | |
| "grad_norm": 0.27493569254875183, | |
| "learning_rate": 0.00013326530612244898, | |
| "loss": 0.3487, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.1917998610145935, | |
| "grad_norm": 0.27025917172431946, | |
| "learning_rate": 0.0001330612244897959, | |
| "loss": 0.3817, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.1945795691452397, | |
| "grad_norm": 0.6159951686859131, | |
| "learning_rate": 0.00013285714285714287, | |
| "loss": 0.5815, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.197359277275886, | |
| "grad_norm": 0.30732738971710205, | |
| "learning_rate": 0.0001326530612244898, | |
| "loss": 0.3646, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.2001389854065323, | |
| "grad_norm": 0.3065405786037445, | |
| "learning_rate": 0.00013244897959183673, | |
| "loss": 0.4138, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.2029186935371785, | |
| "grad_norm": 0.4519249498844147, | |
| "learning_rate": 0.00013224489795918368, | |
| "loss": 0.5033, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.205698401667825, | |
| "grad_norm": 0.2661092281341553, | |
| "learning_rate": 0.0001320408163265306, | |
| "loss": 0.3642, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.2084781097984711, | |
| "grad_norm": 0.27004894614219666, | |
| "learning_rate": 0.00013183673469387757, | |
| "loss": 0.3814, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.2112578179291174, | |
| "grad_norm": 0.39225614070892334, | |
| "learning_rate": 0.0001316326530612245, | |
| "loss": 0.4773, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.2140375260597638, | |
| "grad_norm": 0.2685422897338867, | |
| "learning_rate": 0.00013142857142857143, | |
| "loss": 0.3679, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.21681723419041, | |
| "grad_norm": 0.32003405690193176, | |
| "learning_rate": 0.00013122448979591836, | |
| "loss": 0.4101, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.2195969423210562, | |
| "grad_norm": 0.2831343114376068, | |
| "learning_rate": 0.00013102040816326531, | |
| "loss": 0.3832, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.2223766504517026, | |
| "grad_norm": 0.327888548374176, | |
| "learning_rate": 0.00013081632653061224, | |
| "loss": 0.427, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.2251563585823488, | |
| "grad_norm": 0.37563320994377136, | |
| "learning_rate": 0.00013061224489795917, | |
| "loss": 0.4982, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.227936066712995, | |
| "grad_norm": 0.374957799911499, | |
| "learning_rate": 0.00013040816326530613, | |
| "loss": 0.3724, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.2307157748436415, | |
| "grad_norm": 0.3069044053554535, | |
| "learning_rate": 0.00013020408163265309, | |
| "loss": 0.4736, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.2334954829742877, | |
| "grad_norm": 0.2836786210536957, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 0.3139, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.236275191104934, | |
| "grad_norm": 0.5021543502807617, | |
| "learning_rate": 0.00012979591836734695, | |
| "loss": 0.4687, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.2390548992355803, | |
| "grad_norm": 0.34825125336647034, | |
| "learning_rate": 0.00012959183673469387, | |
| "loss": 0.4624, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.2418346073662265, | |
| "grad_norm": 0.2713720202445984, | |
| "learning_rate": 0.00012938775510204083, | |
| "loss": 0.3841, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.2446143154968727, | |
| "grad_norm": 0.2995285093784332, | |
| "learning_rate": 0.00012918367346938776, | |
| "loss": 0.4582, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.2473940236275192, | |
| "grad_norm": 0.2539880871772766, | |
| "learning_rate": 0.0001289795918367347, | |
| "loss": 0.3232, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.2501737317581654, | |
| "grad_norm": 0.24956463277339935, | |
| "learning_rate": 0.00012877551020408162, | |
| "loss": 0.3282, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.2529534398888118, | |
| "grad_norm": 0.23669935762882233, | |
| "learning_rate": 0.00012857142857142858, | |
| "loss": 0.3036, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.255733148019458, | |
| "grad_norm": 0.31620267033576965, | |
| "learning_rate": 0.00012836734693877553, | |
| "loss": 0.4203, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.2585128561501042, | |
| "grad_norm": 0.37676119804382324, | |
| "learning_rate": 0.00012816326530612246, | |
| "loss": 0.4163, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.2612925642807506, | |
| "grad_norm": 0.2864663004875183, | |
| "learning_rate": 0.0001279591836734694, | |
| "loss": 0.4024, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.2640722724113969, | |
| "grad_norm": 0.27227628231048584, | |
| "learning_rate": 0.00012775510204081632, | |
| "loss": 0.32, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.266851980542043, | |
| "grad_norm": 0.402497261762619, | |
| "learning_rate": 0.00012755102040816328, | |
| "loss": 0.4626, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.2696316886726895, | |
| "grad_norm": 0.3603473901748657, | |
| "learning_rate": 0.0001273469387755102, | |
| "loss": 0.3218, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.2724113968033357, | |
| "grad_norm": 0.2873530685901642, | |
| "learning_rate": 0.00012714285714285714, | |
| "loss": 0.4026, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.275191104933982, | |
| "grad_norm": 0.29491978883743286, | |
| "learning_rate": 0.00012693877551020406, | |
| "loss": 0.2958, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.2779708130646283, | |
| "grad_norm": 0.3286297917366028, | |
| "learning_rate": 0.00012673469387755105, | |
| "loss": 0.3922, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2807505211952745, | |
| "grad_norm": 0.3373448848724365, | |
| "learning_rate": 0.00012653061224489798, | |
| "loss": 0.4216, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.2835302293259208, | |
| "grad_norm": 0.305073082447052, | |
| "learning_rate": 0.0001263265306122449, | |
| "loss": 0.3676, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.2863099374565672, | |
| "grad_norm": 0.3803780674934387, | |
| "learning_rate": 0.00012612244897959184, | |
| "loss": 0.3714, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.2890896455872134, | |
| "grad_norm": 0.38346725702285767, | |
| "learning_rate": 0.0001259183673469388, | |
| "loss": 0.4816, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.2918693537178596, | |
| "grad_norm": 0.3128606379032135, | |
| "learning_rate": 0.00012571428571428572, | |
| "loss": 0.4051, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.294649061848506, | |
| "grad_norm": 0.2748924493789673, | |
| "learning_rate": 0.00012551020408163265, | |
| "loss": 0.383, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.2974287699791522, | |
| "grad_norm": 0.2946728765964508, | |
| "learning_rate": 0.00012530612244897958, | |
| "loss": 0.3495, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.3002084781097984, | |
| "grad_norm": 0.4133062958717346, | |
| "learning_rate": 0.00012510204081632654, | |
| "loss": 0.4196, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.3029881862404449, | |
| "grad_norm": 0.2911101281642914, | |
| "learning_rate": 0.0001248979591836735, | |
| "loss": 0.3733, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.305767894371091, | |
| "grad_norm": 0.35561490058898926, | |
| "learning_rate": 0.00012469387755102042, | |
| "loss": 0.4773, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.3085476025017373, | |
| "grad_norm": 0.23788857460021973, | |
| "learning_rate": 0.00012448979591836735, | |
| "loss": 0.3111, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.3113273106323837, | |
| "grad_norm": 0.3194721043109894, | |
| "learning_rate": 0.00012428571428571428, | |
| "loss": 0.3641, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.31410701876303, | |
| "grad_norm": 0.30017781257629395, | |
| "learning_rate": 0.00012408163265306124, | |
| "loss": 0.3239, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.3168867268936761, | |
| "grad_norm": 0.2877587676048279, | |
| "learning_rate": 0.00012387755102040817, | |
| "loss": 0.3622, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.3196664350243226, | |
| "grad_norm": 0.3146721422672272, | |
| "learning_rate": 0.0001236734693877551, | |
| "loss": 0.3629, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.3224461431549688, | |
| "grad_norm": 0.28086069226264954, | |
| "learning_rate": 0.00012346938775510203, | |
| "loss": 0.3967, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.325225851285615, | |
| "grad_norm": 0.2777217626571655, | |
| "learning_rate": 0.00012326530612244898, | |
| "loss": 0.3531, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.3280055594162614, | |
| "grad_norm": 0.2867282032966614, | |
| "learning_rate": 0.00012306122448979594, | |
| "loss": 0.293, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.3307852675469076, | |
| "grad_norm": 0.3004007339477539, | |
| "learning_rate": 0.00012285714285714287, | |
| "loss": 0.3577, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.3335649756775538, | |
| "grad_norm": 0.32095468044281006, | |
| "learning_rate": 0.0001226530612244898, | |
| "loss": 0.4233, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.3363446838082003, | |
| "grad_norm": 0.2812075614929199, | |
| "learning_rate": 0.00012244897959183676, | |
| "loss": 0.3405, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.3391243919388465, | |
| "grad_norm": 0.38106903433799744, | |
| "learning_rate": 0.00012224489795918368, | |
| "loss": 0.3918, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.3419041000694927, | |
| "grad_norm": 0.3127501606941223, | |
| "learning_rate": 0.00012204081632653061, | |
| "loss": 0.3959, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.344683808200139, | |
| "grad_norm": 0.31547605991363525, | |
| "learning_rate": 0.00012183673469387756, | |
| "loss": 0.4783, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.3474635163307853, | |
| "grad_norm": 0.3430207669734955, | |
| "learning_rate": 0.00012163265306122449, | |
| "loss": 0.4181, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.3502432244614315, | |
| "grad_norm": 0.30673524737358093, | |
| "learning_rate": 0.00012142857142857143, | |
| "loss": 0.4545, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.353022932592078, | |
| "grad_norm": 0.2744535505771637, | |
| "learning_rate": 0.00012122448979591839, | |
| "loss": 0.3568, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.3558026407227242, | |
| "grad_norm": 0.28088897466659546, | |
| "learning_rate": 0.00012102040816326532, | |
| "loss": 0.2957, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.3585823488533704, | |
| "grad_norm": 0.2807769775390625, | |
| "learning_rate": 0.00012081632653061226, | |
| "loss": 0.316, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.3613620569840168, | |
| "grad_norm": 0.33393171429634094, | |
| "learning_rate": 0.00012061224489795919, | |
| "loss": 0.3952, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.364141765114663, | |
| "grad_norm": 0.270470529794693, | |
| "learning_rate": 0.00012040816326530613, | |
| "loss": 0.3263, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.3669214732453092, | |
| "grad_norm": 0.26981666684150696, | |
| "learning_rate": 0.00012020408163265306, | |
| "loss": 0.3758, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.3697011813759556, | |
| "grad_norm": 0.41900643706321716, | |
| "learning_rate": 0.00012, | |
| "loss": 0.4978, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.3724808895066019, | |
| "grad_norm": 0.30232396721839905, | |
| "learning_rate": 0.00011979591836734693, | |
| "loss": 0.3857, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.375260597637248, | |
| "grad_norm": 0.302409827709198, | |
| "learning_rate": 0.00011959183673469388, | |
| "loss": 0.4455, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.3780403057678945, | |
| "grad_norm": 0.30566728115081787, | |
| "learning_rate": 0.00011938775510204083, | |
| "loss": 0.3365, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.3808200138985407, | |
| "grad_norm": 0.2815055847167969, | |
| "learning_rate": 0.00011918367346938777, | |
| "loss": 0.395, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.383599722029187, | |
| "grad_norm": 0.27659547328948975, | |
| "learning_rate": 0.0001189795918367347, | |
| "loss": 0.3635, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.3863794301598333, | |
| "grad_norm": 0.2691234052181244, | |
| "learning_rate": 0.00011877551020408165, | |
| "loss": 0.354, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.3891591382904795, | |
| "grad_norm": 0.26110008358955383, | |
| "learning_rate": 0.00011857142857142858, | |
| "loss": 0.3056, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3919388464211258, | |
| "grad_norm": 0.3188588619232178, | |
| "learning_rate": 0.00011836734693877552, | |
| "loss": 0.3483, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.3947185545517722, | |
| "grad_norm": 0.26784244179725647, | |
| "learning_rate": 0.00011816326530612245, | |
| "loss": 0.4165, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.3974982626824184, | |
| "grad_norm": 0.27212825417518616, | |
| "learning_rate": 0.00011795918367346939, | |
| "loss": 0.3358, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.4002779708130646, | |
| "grad_norm": 0.27269822359085083, | |
| "learning_rate": 0.00011775510204081632, | |
| "loss": 0.3894, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.403057678943711, | |
| "grad_norm": 0.33584582805633545, | |
| "learning_rate": 0.00011755102040816328, | |
| "loss": 0.3801, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.4058373870743572, | |
| "grad_norm": 0.34887197613716125, | |
| "learning_rate": 0.00011734693877551022, | |
| "loss": 0.4057, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.4086170952050034, | |
| "grad_norm": 0.2805575728416443, | |
| "learning_rate": 0.00011714285714285715, | |
| "loss": 0.3556, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.4113968033356499, | |
| "grad_norm": 0.3376087546348572, | |
| "learning_rate": 0.00011693877551020409, | |
| "loss": 0.3511, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.414176511466296, | |
| "grad_norm": 0.29646357893943787, | |
| "learning_rate": 0.00011673469387755102, | |
| "loss": 0.3396, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.4169562195969423, | |
| "grad_norm": 0.2987593710422516, | |
| "learning_rate": 0.00011653061224489797, | |
| "loss": 0.3152, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.4197359277275887, | |
| "grad_norm": 0.27871114015579224, | |
| "learning_rate": 0.0001163265306122449, | |
| "loss": 0.3039, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.422515635858235, | |
| "grad_norm": 0.3051932752132416, | |
| "learning_rate": 0.00011612244897959184, | |
| "loss": 0.3619, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.4252953439888811, | |
| "grad_norm": 0.31153935194015503, | |
| "learning_rate": 0.00011591836734693877, | |
| "loss": 0.4433, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.4280750521195276, | |
| "grad_norm": 0.30891695618629456, | |
| "learning_rate": 0.00011571428571428574, | |
| "loss": 0.4737, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.4308547602501738, | |
| "grad_norm": 0.2646159529685974, | |
| "learning_rate": 0.00011551020408163267, | |
| "loss": 0.3155, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.43363446838082, | |
| "grad_norm": 0.31023016571998596, | |
| "learning_rate": 0.00011530612244897961, | |
| "loss": 0.468, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.4364141765114664, | |
| "grad_norm": 0.32484954595565796, | |
| "learning_rate": 0.00011510204081632654, | |
| "loss": 0.4442, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.4391938846421126, | |
| "grad_norm": 0.30216851830482483, | |
| "learning_rate": 0.00011489795918367348, | |
| "loss": 0.3959, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.4419735927727588, | |
| "grad_norm": 0.3013352155685425, | |
| "learning_rate": 0.00011469387755102041, | |
| "loss": 0.3682, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.4447533009034053, | |
| "grad_norm": 0.3074597716331482, | |
| "learning_rate": 0.00011448979591836735, | |
| "loss": 0.3796, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.4475330090340515, | |
| "grad_norm": 0.29347458481788635, | |
| "learning_rate": 0.00011428571428571428, | |
| "loss": 0.387, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.4503127171646977, | |
| "grad_norm": 0.2919844686985016, | |
| "learning_rate": 0.00011408163265306123, | |
| "loss": 0.3413, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.453092425295344, | |
| "grad_norm": 0.274503618478775, | |
| "learning_rate": 0.00011387755102040818, | |
| "loss": 0.3384, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.4558721334259903, | |
| "grad_norm": 0.30001312494277954, | |
| "learning_rate": 0.00011367346938775511, | |
| "loss": 0.3218, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.4586518415566365, | |
| "grad_norm": 0.2913724482059479, | |
| "learning_rate": 0.00011346938775510206, | |
| "loss": 0.3212, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.461431549687283, | |
| "grad_norm": 0.2810291051864624, | |
| "learning_rate": 0.00011326530612244898, | |
| "loss": 0.4783, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.4642112578179292, | |
| "grad_norm": 0.28154751658439636, | |
| "learning_rate": 0.00011306122448979593, | |
| "loss": 0.384, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.4669909659485754, | |
| "grad_norm": 0.3322899639606476, | |
| "learning_rate": 0.00011285714285714286, | |
| "loss": 0.345, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.4697706740792218, | |
| "grad_norm": 0.2776385247707367, | |
| "learning_rate": 0.0001126530612244898, | |
| "loss": 0.3287, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.472550382209868, | |
| "grad_norm": 0.3325115144252777, | |
| "learning_rate": 0.00011244897959183673, | |
| "loss": 0.4717, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.4753300903405142, | |
| "grad_norm": 0.30507102608680725, | |
| "learning_rate": 0.00011224489795918367, | |
| "loss": 0.3773, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.4781097984711606, | |
| "grad_norm": 0.28410202264785767, | |
| "learning_rate": 0.0001120408163265306, | |
| "loss": 0.3844, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.4808895066018068, | |
| "grad_norm": 0.3436387777328491, | |
| "learning_rate": 0.00011183673469387757, | |
| "loss": 0.4171, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.483669214732453, | |
| "grad_norm": 0.2561332881450653, | |
| "learning_rate": 0.0001116326530612245, | |
| "loss": 0.3597, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.4864489228630995, | |
| "grad_norm": 0.31354856491088867, | |
| "learning_rate": 0.00011142857142857144, | |
| "loss": 0.3473, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.4892286309937457, | |
| "grad_norm": 0.26828843355178833, | |
| "learning_rate": 0.00011122448979591837, | |
| "loss": 0.2624, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.492008339124392, | |
| "grad_norm": 0.3968587815761566, | |
| "learning_rate": 0.00011102040816326532, | |
| "loss": 0.387, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.4947880472550383, | |
| "grad_norm": 0.28001683950424194, | |
| "learning_rate": 0.00011081632653061225, | |
| "loss": 0.3883, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.4975677553856845, | |
| "grad_norm": 0.2872998118400574, | |
| "learning_rate": 0.00011061224489795919, | |
| "loss": 0.3847, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.5003474635163307, | |
| "grad_norm": 0.27555397152900696, | |
| "learning_rate": 0.00011040816326530612, | |
| "loss": 0.2961, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.5031271716469772, | |
| "grad_norm": 0.27156969904899597, | |
| "learning_rate": 0.00011020408163265306, | |
| "loss": 0.3722, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.5059068797776234, | |
| "grad_norm": 0.3042210638523102, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 0.4093, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.5086865879082696, | |
| "grad_norm": 0.2673455774784088, | |
| "learning_rate": 0.00010979591836734695, | |
| "loss": 0.3467, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.511466296038916, | |
| "grad_norm": 0.4710778594017029, | |
| "learning_rate": 0.00010959183673469389, | |
| "loss": 0.4335, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.5142460041695622, | |
| "grad_norm": 0.28414055705070496, | |
| "learning_rate": 0.00010938775510204082, | |
| "loss": 0.3329, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.5170257123002084, | |
| "grad_norm": 0.28005218505859375, | |
| "learning_rate": 0.00010918367346938776, | |
| "loss": 0.3585, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.5198054204308549, | |
| "grad_norm": 0.36730489134788513, | |
| "learning_rate": 0.00010897959183673469, | |
| "loss": 0.3679, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.522585128561501, | |
| "grad_norm": 0.2829911410808563, | |
| "learning_rate": 0.00010877551020408163, | |
| "loss": 0.364, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.5253648366921473, | |
| "grad_norm": 0.27309224009513855, | |
| "learning_rate": 0.00010857142857142856, | |
| "loss": 0.413, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.5281445448227937, | |
| "grad_norm": 0.25464004278182983, | |
| "learning_rate": 0.00010836734693877551, | |
| "loss": 0.3196, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.53092425295344, | |
| "grad_norm": 0.2633742094039917, | |
| "learning_rate": 0.00010816326530612246, | |
| "loss": 0.312, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.5337039610840861, | |
| "grad_norm": 0.34395769238471985, | |
| "learning_rate": 0.0001079591836734694, | |
| "loss": 0.3483, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.5364836692147326, | |
| "grad_norm": 0.27649009227752686, | |
| "learning_rate": 0.00010775510204081634, | |
| "loss": 0.4565, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.5392633773453788, | |
| "grad_norm": 0.4239721894264221, | |
| "learning_rate": 0.00010755102040816328, | |
| "loss": 0.4087, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.542043085476025, | |
| "grad_norm": 0.5541340708732605, | |
| "learning_rate": 0.00010734693877551021, | |
| "loss": 0.4309, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.5448227936066714, | |
| "grad_norm": 0.2893712818622589, | |
| "learning_rate": 0.00010714285714285715, | |
| "loss": 0.3447, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.5476025017373176, | |
| "grad_norm": 0.26238903403282166, | |
| "learning_rate": 0.00010693877551020408, | |
| "loss": 0.3256, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.5503822098679638, | |
| "grad_norm": 0.29899922013282776, | |
| "learning_rate": 0.00010673469387755102, | |
| "loss": 0.3142, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.5531619179986103, | |
| "grad_norm": 0.28399357199668884, | |
| "learning_rate": 0.00010653061224489795, | |
| "loss": 0.3314, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.5559416261292565, | |
| "grad_norm": 0.3450354337692261, | |
| "learning_rate": 0.00010632653061224491, | |
| "loss": 0.3733, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.5587213342599027, | |
| "grad_norm": 0.283597856760025, | |
| "learning_rate": 0.00010612244897959185, | |
| "loss": 0.3828, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.561501042390549, | |
| "grad_norm": 0.28581732511520386, | |
| "learning_rate": 0.00010591836734693878, | |
| "loss": 0.3131, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.5642807505211953, | |
| "grad_norm": 0.2778254747390747, | |
| "learning_rate": 0.00010571428571428572, | |
| "loss": 0.4744, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.5670604586518415, | |
| "grad_norm": 0.2736669182777405, | |
| "learning_rate": 0.00010551020408163265, | |
| "loss": 0.4084, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.569840166782488, | |
| "grad_norm": 0.31702694296836853, | |
| "learning_rate": 0.0001053061224489796, | |
| "loss": 0.3877, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.5726198749131342, | |
| "grad_norm": 0.4662436246871948, | |
| "learning_rate": 0.00010510204081632653, | |
| "loss": 0.3619, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.5753995830437804, | |
| "grad_norm": 0.26047173142433167, | |
| "learning_rate": 0.00010489795918367347, | |
| "loss": 0.3031, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.5781792911744268, | |
| "grad_norm": 0.4014153778553009, | |
| "learning_rate": 0.0001046938775510204, | |
| "loss": 0.416, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.580958999305073, | |
| "grad_norm": 0.2595798969268799, | |
| "learning_rate": 0.00010448979591836735, | |
| "loss": 0.3635, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.5837387074357192, | |
| "grad_norm": 0.3327747881412506, | |
| "learning_rate": 0.0001042857142857143, | |
| "loss": 0.5032, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.5865184155663656, | |
| "grad_norm": 0.28941619396209717, | |
| "learning_rate": 0.00010408163265306123, | |
| "loss": 0.4001, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.5892981236970118, | |
| "grad_norm": 0.29737424850463867, | |
| "learning_rate": 0.00010387755102040817, | |
| "loss": 0.4263, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.592077831827658, | |
| "grad_norm": 0.3156750202178955, | |
| "learning_rate": 0.00010367346938775511, | |
| "loss": 0.326, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.5948575399583045, | |
| "grad_norm": 0.33826690912246704, | |
| "learning_rate": 0.00010346938775510204, | |
| "loss": 0.4291, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.5976372480889507, | |
| "grad_norm": 0.2737540900707245, | |
| "learning_rate": 0.00010326530612244899, | |
| "loss": 0.4464, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.600416956219597, | |
| "grad_norm": 0.3012180030345917, | |
| "learning_rate": 0.00010306122448979591, | |
| "loss": 0.4044, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.6031966643502433, | |
| "grad_norm": 0.3628576397895813, | |
| "learning_rate": 0.00010285714285714286, | |
| "loss": 0.4376, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.6059763724808895, | |
| "grad_norm": 0.3489641845226288, | |
| "learning_rate": 0.00010265306122448981, | |
| "loss": 0.3054, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.6087560806115357, | |
| "grad_norm": 0.28231537342071533, | |
| "learning_rate": 0.00010244897959183674, | |
| "loss": 0.2566, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.6115357887421822, | |
| "grad_norm": 0.2675357758998871, | |
| "learning_rate": 0.00010224489795918369, | |
| "loss": 0.3917, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.6143154968728284, | |
| "grad_norm": 0.25992777943611145, | |
| "learning_rate": 0.00010204081632653062, | |
| "loss": 0.3511, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.6170952050034746, | |
| "grad_norm": 0.2999131977558136, | |
| "learning_rate": 0.00010183673469387756, | |
| "loss": 0.3135, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.619874913134121, | |
| "grad_norm": 0.2928028404712677, | |
| "learning_rate": 0.00010163265306122449, | |
| "loss": 0.34, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.6226546212647672, | |
| "grad_norm": 0.24898597598075867, | |
| "learning_rate": 0.00010142857142857143, | |
| "loss": 0.3334, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.6254343293954134, | |
| "grad_norm": 0.309908002614975, | |
| "learning_rate": 0.00010122448979591836, | |
| "loss": 0.4202, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.6282140375260599, | |
| "grad_norm": 0.2774292230606079, | |
| "learning_rate": 0.0001010204081632653, | |
| "loss": 0.3986, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.630993745656706, | |
| "grad_norm": 0.4288952648639679, | |
| "learning_rate": 0.00010081632653061226, | |
| "loss": 0.4282, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.6337734537873523, | |
| "grad_norm": 0.2605331540107727, | |
| "learning_rate": 0.00010061224489795919, | |
| "loss": 0.356, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.6365531619179987, | |
| "grad_norm": 0.2896113395690918, | |
| "learning_rate": 0.00010040816326530613, | |
| "loss": 0.3454, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.639332870048645, | |
| "grad_norm": 0.26861974596977234, | |
| "learning_rate": 0.00010020408163265306, | |
| "loss": 0.4541, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.6421125781792911, | |
| "grad_norm": 0.3458561599254608, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3547, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.6448922863099376, | |
| "grad_norm": 0.287142276763916, | |
| "learning_rate": 9.979591836734695e-05, | |
| "loss": 0.3585, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.6476719944405838, | |
| "grad_norm": 0.2605501711368561, | |
| "learning_rate": 9.959183673469388e-05, | |
| "loss": 0.345, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.65045170257123, | |
| "grad_norm": 0.2797659635543823, | |
| "learning_rate": 9.938775510204082e-05, | |
| "loss": 0.3176, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.6532314107018764, | |
| "grad_norm": 0.26743918657302856, | |
| "learning_rate": 9.918367346938776e-05, | |
| "loss": 0.2928, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.6560111188325226, | |
| "grad_norm": 0.2472202330827713, | |
| "learning_rate": 9.897959183673469e-05, | |
| "loss": 0.2373, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.6587908269631688, | |
| "grad_norm": 0.26467829942703247, | |
| "learning_rate": 9.877551020408164e-05, | |
| "loss": 0.3824, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.6615705350938152, | |
| "grad_norm": 0.272390753030777, | |
| "learning_rate": 9.857142857142858e-05, | |
| "loss": 0.322, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.6643502432244615, | |
| "grad_norm": 0.2785828709602356, | |
| "learning_rate": 9.836734693877552e-05, | |
| "loss": 0.3201, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.6671299513551077, | |
| "grad_norm": 0.28154081106185913, | |
| "learning_rate": 9.816326530612245e-05, | |
| "loss": 0.3767, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.6671299513551077, | |
| "eval_loss": 0.4453812837600708, | |
| "eval_runtime": 212.689, | |
| "eval_samples_per_second": 1.693, | |
| "eval_steps_per_second": 1.693, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.669909659485754, | |
| "grad_norm": 0.26373493671417236, | |
| "learning_rate": 9.79591836734694e-05, | |
| "loss": 0.3756, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.6726893676164003, | |
| "grad_norm": 0.31205254793167114, | |
| "learning_rate": 9.775510204081632e-05, | |
| "loss": 0.2962, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.6754690757470465, | |
| "grad_norm": 0.2865038812160492, | |
| "learning_rate": 9.755102040816328e-05, | |
| "loss": 0.345, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.678248783877693, | |
| "grad_norm": 0.2950930893421173, | |
| "learning_rate": 9.734693877551021e-05, | |
| "loss": 0.3678, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.6810284920083391, | |
| "grad_norm": 0.35823482275009155, | |
| "learning_rate": 9.714285714285715e-05, | |
| "loss": 0.4096, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.6838082001389854, | |
| "grad_norm": 0.2647465169429779, | |
| "learning_rate": 9.693877551020408e-05, | |
| "loss": 0.4015, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.6865879082696318, | |
| "grad_norm": 0.30505606532096863, | |
| "learning_rate": 9.673469387755102e-05, | |
| "loss": 0.4116, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.689367616400278, | |
| "grad_norm": 0.3076132833957672, | |
| "learning_rate": 9.653061224489797e-05, | |
| "loss": 0.3867, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.6921473245309242, | |
| "grad_norm": 0.2777102589607239, | |
| "learning_rate": 9.63265306122449e-05, | |
| "loss": 0.3944, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.6949270326615706, | |
| "grad_norm": 0.3163359463214874, | |
| "learning_rate": 9.612244897959184e-05, | |
| "loss": 0.4266, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.6977067407922168, | |
| "grad_norm": 0.3227400779724121, | |
| "learning_rate": 9.591836734693878e-05, | |
| "loss": 0.397, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.700486448922863, | |
| "grad_norm": 0.27346518635749817, | |
| "learning_rate": 9.571428571428573e-05, | |
| "loss": 0.3463, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.7032661570535095, | |
| "grad_norm": 0.3154374957084656, | |
| "learning_rate": 9.551020408163265e-05, | |
| "loss": 0.3921, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.7060458651841557, | |
| "grad_norm": 0.2859618365764618, | |
| "learning_rate": 9.53061224489796e-05, | |
| "loss": 0.3736, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.708825573314802, | |
| "grad_norm": 0.28541162610054016, | |
| "learning_rate": 9.510204081632653e-05, | |
| "loss": 0.3689, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.7116052814454483, | |
| "grad_norm": 0.2870318293571472, | |
| "learning_rate": 9.489795918367348e-05, | |
| "loss": 0.2594, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.7143849895760945, | |
| "grad_norm": 0.2808590531349182, | |
| "learning_rate": 9.469387755102041e-05, | |
| "loss": 0.4471, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.7171646977067407, | |
| "grad_norm": 0.2777983248233795, | |
| "learning_rate": 9.448979591836736e-05, | |
| "loss": 0.304, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.7199444058373872, | |
| "grad_norm": 0.2791382372379303, | |
| "learning_rate": 9.428571428571429e-05, | |
| "loss": 0.3117, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.7227241139680334, | |
| "grad_norm": 0.3301670253276825, | |
| "learning_rate": 9.408163265306123e-05, | |
| "loss": 0.3445, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.7255038220986796, | |
| "grad_norm": 0.2752249240875244, | |
| "learning_rate": 9.387755102040817e-05, | |
| "loss": 0.3807, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.728283530229326, | |
| "grad_norm": 0.32973968982696533, | |
| "learning_rate": 9.367346938775511e-05, | |
| "loss": 0.4078, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.7310632383599722, | |
| "grad_norm": 0.30001533031463623, | |
| "learning_rate": 9.346938775510204e-05, | |
| "loss": 0.4355, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.7338429464906184, | |
| "grad_norm": 0.255593866109848, | |
| "learning_rate": 9.326530612244899e-05, | |
| "loss": 0.3447, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.7366226546212649, | |
| "grad_norm": 0.2974906861782074, | |
| "learning_rate": 9.306122448979592e-05, | |
| "loss": 0.4654, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.739402362751911, | |
| "grad_norm": 0.3247474133968353, | |
| "learning_rate": 9.285714285714286e-05, | |
| "loss": 0.3117, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.7421820708825573, | |
| "grad_norm": 0.27801772952079773, | |
| "learning_rate": 9.26530612244898e-05, | |
| "loss": 0.4573, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.7449617790132037, | |
| "grad_norm": 0.2884966731071472, | |
| "learning_rate": 9.244897959183673e-05, | |
| "loss": 0.3575, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.74774148714385, | |
| "grad_norm": 0.27776768803596497, | |
| "learning_rate": 9.224489795918367e-05, | |
| "loss": 0.4243, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.7505211952744961, | |
| "grad_norm": 0.2852678596973419, | |
| "learning_rate": 9.204081632653062e-05, | |
| "loss": 0.4214, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.7533009034051426, | |
| "grad_norm": 0.7854850888252258, | |
| "learning_rate": 9.183673469387756e-05, | |
| "loss": 0.4479, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.7560806115357888, | |
| "grad_norm": 0.3441762626171112, | |
| "learning_rate": 9.163265306122449e-05, | |
| "loss": 0.3926, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.758860319666435, | |
| "grad_norm": 0.27086740732192993, | |
| "learning_rate": 9.142857142857143e-05, | |
| "loss": 0.3389, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.7616400277970814, | |
| "grad_norm": 0.273234486579895, | |
| "learning_rate": 9.122448979591836e-05, | |
| "loss": 0.3376, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.7644197359277276, | |
| "grad_norm": 0.29030004143714905, | |
| "learning_rate": 9.102040816326532e-05, | |
| "loss": 0.3156, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.7671994440583738, | |
| "grad_norm": 0.2696784436702728, | |
| "learning_rate": 9.081632653061225e-05, | |
| "loss": 0.3006, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.7699791521890202, | |
| "grad_norm": 0.2931046485900879, | |
| "learning_rate": 9.061224489795919e-05, | |
| "loss": 0.3289, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.7727588603196665, | |
| "grad_norm": 0.2624610364437103, | |
| "learning_rate": 9.040816326530612e-05, | |
| "loss": 0.316, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.7755385684503127, | |
| "grad_norm": 0.2946118414402008, | |
| "learning_rate": 9.020408163265308e-05, | |
| "loss": 0.4373, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.778318276580959, | |
| "grad_norm": 0.32654690742492676, | |
| "learning_rate": 9e-05, | |
| "loss": 0.3699, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.7810979847116053, | |
| "grad_norm": 0.2824501693248749, | |
| "learning_rate": 8.979591836734695e-05, | |
| "loss": 0.3846, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.7838776928422515, | |
| "grad_norm": 0.3406214118003845, | |
| "learning_rate": 8.959183673469388e-05, | |
| "loss": 0.3932, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.786657400972898, | |
| "grad_norm": 0.26329147815704346, | |
| "learning_rate": 8.938775510204082e-05, | |
| "loss": 0.3152, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.7894371091035441, | |
| "grad_norm": 0.2712422311306, | |
| "learning_rate": 8.918367346938776e-05, | |
| "loss": 0.3595, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.7922168172341904, | |
| "grad_norm": 0.30618801712989807, | |
| "learning_rate": 8.89795918367347e-05, | |
| "loss": 0.3724, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.7949965253648368, | |
| "grad_norm": 0.2918386161327362, | |
| "learning_rate": 8.877551020408164e-05, | |
| "loss": 0.3537, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.797776233495483, | |
| "grad_norm": 0.27820420265197754, | |
| "learning_rate": 8.857142857142857e-05, | |
| "loss": 0.3404, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.8005559416261292, | |
| "grad_norm": 0.296987920999527, | |
| "learning_rate": 8.836734693877552e-05, | |
| "loss": 0.3644, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.8033356497567756, | |
| "grad_norm": 0.29092004895210266, | |
| "learning_rate": 8.816326530612245e-05, | |
| "loss": 0.428, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.8061153578874218, | |
| "grad_norm": 0.2843736410140991, | |
| "learning_rate": 8.79591836734694e-05, | |
| "loss": 0.3709, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.808895066018068, | |
| "grad_norm": 0.2677766978740692, | |
| "learning_rate": 8.775510204081632e-05, | |
| "loss": 0.2649, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.8116747741487145, | |
| "grad_norm": 0.2586863338947296, | |
| "learning_rate": 8.755102040816327e-05, | |
| "loss": 0.3241, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.8144544822793607, | |
| "grad_norm": 0.32978034019470215, | |
| "learning_rate": 8.734693877551021e-05, | |
| "loss": 0.3787, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.8172341904100069, | |
| "grad_norm": 0.27411404252052307, | |
| "learning_rate": 8.714285714285715e-05, | |
| "loss": 0.3447, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.8200138985406533, | |
| "grad_norm": 0.26756396889686584, | |
| "learning_rate": 8.693877551020408e-05, | |
| "loss": 0.3528, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.8227936066712995, | |
| "grad_norm": 0.30654609203338623, | |
| "learning_rate": 8.673469387755102e-05, | |
| "loss": 0.3805, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.8255733148019457, | |
| "grad_norm": 0.29328277707099915, | |
| "learning_rate": 8.653061224489797e-05, | |
| "loss": 0.3276, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.8283530229325922, | |
| "grad_norm": 0.2501872777938843, | |
| "learning_rate": 8.632653061224491e-05, | |
| "loss": 0.3368, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.8311327310632384, | |
| "grad_norm": 0.3712775707244873, | |
| "learning_rate": 8.612244897959184e-05, | |
| "loss": 0.3671, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.8339124391938846, | |
| "grad_norm": 0.24037517607212067, | |
| "learning_rate": 8.591836734693878e-05, | |
| "loss": 0.259, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.836692147324531, | |
| "grad_norm": 0.25154754519462585, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.2983, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.8394718554551772, | |
| "grad_norm": 0.2672985792160034, | |
| "learning_rate": 8.551020408163266e-05, | |
| "loss": 0.3657, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.8422515635858234, | |
| "grad_norm": 0.31283143162727356, | |
| "learning_rate": 8.53061224489796e-05, | |
| "loss": 0.3211, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.8450312717164699, | |
| "grad_norm": 0.26797670125961304, | |
| "learning_rate": 8.510204081632653e-05, | |
| "loss": 0.3665, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.847810979847116, | |
| "grad_norm": 0.2569994032382965, | |
| "learning_rate": 8.489795918367347e-05, | |
| "loss": 0.4167, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.8505906879777623, | |
| "grad_norm": 0.261764258146286, | |
| "learning_rate": 8.469387755102041e-05, | |
| "loss": 0.3976, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.8533703961084087, | |
| "grad_norm": 0.28137752413749695, | |
| "learning_rate": 8.448979591836736e-05, | |
| "loss": 0.2909, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.856150104239055, | |
| "grad_norm": 0.2962735891342163, | |
| "learning_rate": 8.428571428571429e-05, | |
| "loss": 0.3961, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.8589298123697011, | |
| "grad_norm": 0.2491084784269333, | |
| "learning_rate": 8.408163265306123e-05, | |
| "loss": 0.334, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.8617095205003475, | |
| "grad_norm": 0.30861160159111023, | |
| "learning_rate": 8.387755102040816e-05, | |
| "loss": 0.3896, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.8644892286309938, | |
| "grad_norm": 0.2767215371131897, | |
| "learning_rate": 8.367346938775511e-05, | |
| "loss": 0.457, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.86726893676164, | |
| "grad_norm": 0.2909168303012848, | |
| "learning_rate": 8.346938775510204e-05, | |
| "loss": 0.4469, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.8700486448922864, | |
| "grad_norm": 0.30907315015792847, | |
| "learning_rate": 8.326530612244899e-05, | |
| "loss": 0.4008, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.8728283530229326, | |
| "grad_norm": 0.26147618889808655, | |
| "learning_rate": 8.306122448979592e-05, | |
| "loss": 0.3456, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.8756080611535788, | |
| "grad_norm": 0.280230313539505, | |
| "learning_rate": 8.285714285714287e-05, | |
| "loss": 0.454, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.8783877692842252, | |
| "grad_norm": 0.2674858570098877, | |
| "learning_rate": 8.26530612244898e-05, | |
| "loss": 0.3247, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.8811674774148714, | |
| "grad_norm": 0.26258382201194763, | |
| "learning_rate": 8.244897959183675e-05, | |
| "loss": 0.3402, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.8839471855455177, | |
| "grad_norm": 0.2922073006629944, | |
| "learning_rate": 8.224489795918367e-05, | |
| "loss": 0.4958, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.886726893676164, | |
| "grad_norm": 0.2971295714378357, | |
| "learning_rate": 8.204081632653062e-05, | |
| "loss": 0.3341, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.8895066018068103, | |
| "grad_norm": 0.27569159865379333, | |
| "learning_rate": 8.183673469387756e-05, | |
| "loss": 0.3873, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.8922863099374565, | |
| "grad_norm": 0.30602556467056274, | |
| "learning_rate": 8.163265306122449e-05, | |
| "loss": 0.2462, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.895066018068103, | |
| "grad_norm": 0.298258900642395, | |
| "learning_rate": 8.142857142857143e-05, | |
| "loss": 0.3863, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.8978457261987491, | |
| "grad_norm": 0.2906138598918915, | |
| "learning_rate": 8.122448979591836e-05, | |
| "loss": 0.4014, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.9006254343293953, | |
| "grad_norm": 0.2641445994377136, | |
| "learning_rate": 8.10204081632653e-05, | |
| "loss": 0.367, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.9034051424600418, | |
| "grad_norm": 0.7982620596885681, | |
| "learning_rate": 8.081632653061225e-05, | |
| "loss": 0.6423, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.906184850590688, | |
| "grad_norm": 0.26612165570259094, | |
| "learning_rate": 8.061224489795919e-05, | |
| "loss": 0.3586, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.9089645587213342, | |
| "grad_norm": 0.27895480394363403, | |
| "learning_rate": 8.040816326530612e-05, | |
| "loss": 0.3828, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.9117442668519806, | |
| "grad_norm": 0.2650473713874817, | |
| "learning_rate": 8.020408163265306e-05, | |
| "loss": 0.403, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.9145239749826268, | |
| "grad_norm": 0.2670430541038513, | |
| "learning_rate": 8e-05, | |
| "loss": 0.2993, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.917303683113273, | |
| "grad_norm": 0.27033767104148865, | |
| "learning_rate": 7.979591836734695e-05, | |
| "loss": 0.3602, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.9200833912439195, | |
| "grad_norm": 0.279803603887558, | |
| "learning_rate": 7.959183673469388e-05, | |
| "loss": 0.4034, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.9228630993745657, | |
| "grad_norm": 0.2839685082435608, | |
| "learning_rate": 7.938775510204082e-05, | |
| "loss": 0.3458, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.9256428075052119, | |
| "grad_norm": 0.2718431353569031, | |
| "learning_rate": 7.918367346938775e-05, | |
| "loss": 0.3323, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.9284225156358583, | |
| "grad_norm": 0.2734295725822449, | |
| "learning_rate": 7.897959183673471e-05, | |
| "loss": 0.2751, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.9312022237665045, | |
| "grad_norm": 0.3010067641735077, | |
| "learning_rate": 7.877551020408164e-05, | |
| "loss": 0.3508, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.9339819318971507, | |
| "grad_norm": 0.3028511703014374, | |
| "learning_rate": 7.857142857142858e-05, | |
| "loss": 0.3794, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.9367616400277972, | |
| "grad_norm": 0.26215803623199463, | |
| "learning_rate": 7.836734693877551e-05, | |
| "loss": 0.3737, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.9395413481584434, | |
| "grad_norm": 0.3025459051132202, | |
| "learning_rate": 7.816326530612245e-05, | |
| "loss": 0.44, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.9423210562890896, | |
| "grad_norm": 0.29042086005210876, | |
| "learning_rate": 7.79591836734694e-05, | |
| "loss": 0.3401, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.945100764419736, | |
| "grad_norm": 0.31652623414993286, | |
| "learning_rate": 7.775510204081632e-05, | |
| "loss": 0.4041, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.9478804725503822, | |
| "grad_norm": 0.2801991403102875, | |
| "learning_rate": 7.755102040816327e-05, | |
| "loss": 0.3173, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.9506601806810284, | |
| "grad_norm": 0.30574724078178406, | |
| "learning_rate": 7.73469387755102e-05, | |
| "loss": 0.3477, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.9534398888116749, | |
| "grad_norm": 0.3101007640361786, | |
| "learning_rate": 7.714285714285715e-05, | |
| "loss": 0.3974, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.956219596942321, | |
| "grad_norm": 0.27257412672042847, | |
| "learning_rate": 7.693877551020408e-05, | |
| "loss": 0.3449, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.9589993050729673, | |
| "grad_norm": 0.2764023542404175, | |
| "learning_rate": 7.673469387755103e-05, | |
| "loss": 0.3517, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.9617790132036137, | |
| "grad_norm": 0.29003384709358215, | |
| "learning_rate": 7.653061224489796e-05, | |
| "loss": 0.3886, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.96455872133426, | |
| "grad_norm": 0.4509872794151306, | |
| "learning_rate": 7.632653061224491e-05, | |
| "loss": 0.4117, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.9673384294649061, | |
| "grad_norm": 0.28539761900901794, | |
| "learning_rate": 7.612244897959184e-05, | |
| "loss": 0.3671, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.9701181375955525, | |
| "grad_norm": 0.27598991990089417, | |
| "learning_rate": 7.591836734693878e-05, | |
| "loss": 0.3327, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.9728978457261988, | |
| "grad_norm": 0.36246633529663086, | |
| "learning_rate": 7.571428571428571e-05, | |
| "loss": 0.4338, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.975677553856845, | |
| "grad_norm": 0.29050305485725403, | |
| "learning_rate": 7.551020408163266e-05, | |
| "loss": 0.5175, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.9784572619874914, | |
| "grad_norm": 0.4746512174606323, | |
| "learning_rate": 7.53061224489796e-05, | |
| "loss": 0.4983, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.9812369701181376, | |
| "grad_norm": 0.27185899019241333, | |
| "learning_rate": 7.510204081632653e-05, | |
| "loss": 0.3781, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.9840166782487838, | |
| "grad_norm": 0.28276991844177246, | |
| "learning_rate": 7.489795918367347e-05, | |
| "loss": 0.3137, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.9867963863794302, | |
| "grad_norm": 0.27965956926345825, | |
| "learning_rate": 7.469387755102041e-05, | |
| "loss": 0.3681, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.9895760945100764, | |
| "grad_norm": 0.245852991938591, | |
| "learning_rate": 7.448979591836736e-05, | |
| "loss": 0.3879, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.9923558026407227, | |
| "grad_norm": 0.28121981024742126, | |
| "learning_rate": 7.428571428571429e-05, | |
| "loss": 0.4085, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.995135510771369, | |
| "grad_norm": 0.2986759543418884, | |
| "learning_rate": 7.408163265306123e-05, | |
| "loss": 0.2768, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.9979152189020153, | |
| "grad_norm": 0.26322054862976074, | |
| "learning_rate": 7.387755102040816e-05, | |
| "loss": 0.3901, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.3256114721298218, | |
| "learning_rate": 7.36734693877551e-05, | |
| "loss": 0.3634, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.0027797081306464, | |
| "grad_norm": 0.27563294768333435, | |
| "learning_rate": 7.346938775510205e-05, | |
| "loss": 0.3962, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.0055594162612924, | |
| "grad_norm": 0.36298710107803345, | |
| "learning_rate": 7.326530612244899e-05, | |
| "loss": 0.3589, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.008339124391939, | |
| "grad_norm": 0.28008803725242615, | |
| "learning_rate": 7.306122448979592e-05, | |
| "loss": 0.3883, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.0111188325225853, | |
| "grad_norm": 0.34973540902137756, | |
| "learning_rate": 7.285714285714286e-05, | |
| "loss": 0.4629, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.0138985406532313, | |
| "grad_norm": 0.38526445627212524, | |
| "learning_rate": 7.26530612244898e-05, | |
| "loss": 0.4442, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.0166782487838777, | |
| "grad_norm": 0.2718683183193207, | |
| "learning_rate": 7.244897959183675e-05, | |
| "loss": 0.3513, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.019457956914524, | |
| "grad_norm": 0.26923590898513794, | |
| "learning_rate": 7.224489795918368e-05, | |
| "loss": 0.326, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.02223766504517, | |
| "grad_norm": 0.2851525545120239, | |
| "learning_rate": 7.204081632653062e-05, | |
| "loss": 0.2804, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.0250173731758165, | |
| "grad_norm": 0.3298304080963135, | |
| "learning_rate": 7.183673469387755e-05, | |
| "loss": 0.4877, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.027797081306463, | |
| "grad_norm": 0.304770290851593, | |
| "learning_rate": 7.163265306122449e-05, | |
| "loss": 0.3777, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.030576789437109, | |
| "grad_norm": 0.2693670094013214, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.3315, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.0333564975677554, | |
| "grad_norm": 0.2542417049407959, | |
| "learning_rate": 7.122448979591836e-05, | |
| "loss": 0.3493, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.036136205698402, | |
| "grad_norm": 0.2779878079891205, | |
| "learning_rate": 7.10204081632653e-05, | |
| "loss": 0.3686, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.038915913829048, | |
| "grad_norm": 0.3229525685310364, | |
| "learning_rate": 7.081632653061225e-05, | |
| "loss": 0.4557, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.0416956219596942, | |
| "grad_norm": 0.30455026030540466, | |
| "learning_rate": 7.061224489795919e-05, | |
| "loss": 0.3291, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.0444753300903407, | |
| "grad_norm": 0.26810574531555176, | |
| "learning_rate": 7.040816326530612e-05, | |
| "loss": 0.3269, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.0472550382209866, | |
| "grad_norm": 0.26769372820854187, | |
| "learning_rate": 7.020408163265306e-05, | |
| "loss": 0.3385, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.050034746351633, | |
| "grad_norm": 0.3141888380050659, | |
| "learning_rate": 7e-05, | |
| "loss": 0.3584, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.0528144544822795, | |
| "grad_norm": 0.31162917613983154, | |
| "learning_rate": 6.979591836734695e-05, | |
| "loss": 0.3418, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.0555941626129255, | |
| "grad_norm": 0.28655368089675903, | |
| "learning_rate": 6.959183673469388e-05, | |
| "loss": 0.3356, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.058373870743572, | |
| "grad_norm": 0.3098087012767792, | |
| "learning_rate": 6.938775510204082e-05, | |
| "loss": 0.4558, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 2.0611535788742184, | |
| "grad_norm": 0.28569385409355164, | |
| "learning_rate": 6.918367346938775e-05, | |
| "loss": 0.2827, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 2.0639332870048643, | |
| "grad_norm": 0.2704540491104126, | |
| "learning_rate": 6.897959183673471e-05, | |
| "loss": 0.3118, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 2.0667129951355108, | |
| "grad_norm": 0.2655661702156067, | |
| "learning_rate": 6.877551020408164e-05, | |
| "loss": 0.3569, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 2.069492703266157, | |
| "grad_norm": 0.292889267206192, | |
| "learning_rate": 6.857142857142858e-05, | |
| "loss": 0.4, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.072272411396803, | |
| "grad_norm": 0.30672675371170044, | |
| "learning_rate": 6.836734693877551e-05, | |
| "loss": 0.4197, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 2.0750521195274496, | |
| "grad_norm": 0.2781777083873749, | |
| "learning_rate": 6.816326530612245e-05, | |
| "loss": 0.3949, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 2.077831827658096, | |
| "grad_norm": 0.2681523561477661, | |
| "learning_rate": 6.79591836734694e-05, | |
| "loss": 0.3073, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 2.080611535788742, | |
| "grad_norm": 0.2789252996444702, | |
| "learning_rate": 6.775510204081633e-05, | |
| "loss": 0.308, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 2.0833912439193885, | |
| "grad_norm": 0.3158692717552185, | |
| "learning_rate": 6.755102040816327e-05, | |
| "loss": 0.4182, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.086170952050035, | |
| "grad_norm": 0.2763406038284302, | |
| "learning_rate": 6.73469387755102e-05, | |
| "loss": 0.3707, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 2.088950660180681, | |
| "grad_norm": 0.2762833833694458, | |
| "learning_rate": 6.714285714285714e-05, | |
| "loss": 0.3287, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 2.0917303683113273, | |
| "grad_norm": 0.26660582423210144, | |
| "learning_rate": 6.693877551020408e-05, | |
| "loss": 0.2388, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 2.0945100764419737, | |
| "grad_norm": 0.28191235661506653, | |
| "learning_rate": 6.673469387755103e-05, | |
| "loss": 0.2989, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 2.0972897845726197, | |
| "grad_norm": 0.29017800092697144, | |
| "learning_rate": 6.653061224489796e-05, | |
| "loss": 0.3203, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.100069492703266, | |
| "grad_norm": 0.2984425127506256, | |
| "learning_rate": 6.63265306122449e-05, | |
| "loss": 0.3803, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 2.1028492008339126, | |
| "grad_norm": 0.30566540360450745, | |
| "learning_rate": 6.612244897959184e-05, | |
| "loss": 0.3248, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 2.1056289089645586, | |
| "grad_norm": 0.2737206220626831, | |
| "learning_rate": 6.591836734693878e-05, | |
| "loss": 0.3448, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 2.108408617095205, | |
| "grad_norm": 0.3234175145626068, | |
| "learning_rate": 6.571428571428571e-05, | |
| "loss": 0.3773, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 2.1111883252258514, | |
| "grad_norm": 0.3321143090724945, | |
| "learning_rate": 6.551020408163266e-05, | |
| "loss": 0.4053, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.1139680333564974, | |
| "grad_norm": 0.3120156228542328, | |
| "learning_rate": 6.530612244897959e-05, | |
| "loss": 0.3614, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 2.116747741487144, | |
| "grad_norm": 0.3069126307964325, | |
| "learning_rate": 6.510204081632654e-05, | |
| "loss": 0.3392, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 2.1195274496177903, | |
| "grad_norm": 0.2996152937412262, | |
| "learning_rate": 6.489795918367347e-05, | |
| "loss": 0.4007, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 2.1223071577484363, | |
| "grad_norm": 0.28609102964401245, | |
| "learning_rate": 6.469387755102042e-05, | |
| "loss": 0.3619, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 2.1250868658790827, | |
| "grad_norm": 0.28463420271873474, | |
| "learning_rate": 6.448979591836734e-05, | |
| "loss": 0.3684, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.127866574009729, | |
| "grad_norm": 0.2721307873725891, | |
| "learning_rate": 6.428571428571429e-05, | |
| "loss": 0.3153, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 2.130646282140375, | |
| "grad_norm": 0.29114431142807007, | |
| "learning_rate": 6.408163265306123e-05, | |
| "loss": 0.3558, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 2.1334259902710215, | |
| "grad_norm": 0.2995449900627136, | |
| "learning_rate": 6.387755102040816e-05, | |
| "loss": 0.3632, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 2.136205698401668, | |
| "grad_norm": 0.28987494111061096, | |
| "learning_rate": 6.36734693877551e-05, | |
| "loss": 0.3928, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 2.138985406532314, | |
| "grad_norm": 0.2857901155948639, | |
| "learning_rate": 6.346938775510203e-05, | |
| "loss": 0.4101, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.1417651146629604, | |
| "grad_norm": 0.2712436020374298, | |
| "learning_rate": 6.326530612244899e-05, | |
| "loss": 0.379, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 2.144544822793607, | |
| "grad_norm": 0.2805217504501343, | |
| "learning_rate": 6.306122448979592e-05, | |
| "loss": 0.3018, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 2.147324530924253, | |
| "grad_norm": 0.29896092414855957, | |
| "learning_rate": 6.285714285714286e-05, | |
| "loss": 0.3065, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 2.1501042390548992, | |
| "grad_norm": 0.27301499247550964, | |
| "learning_rate": 6.265306122448979e-05, | |
| "loss": 0.3213, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 2.1528839471855457, | |
| "grad_norm": 0.2966340482234955, | |
| "learning_rate": 6.244897959183675e-05, | |
| "loss": 0.5333, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.1556636553161916, | |
| "grad_norm": 0.31995972990989685, | |
| "learning_rate": 6.224489795918368e-05, | |
| "loss": 0.3398, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 2.158443363446838, | |
| "grad_norm": 0.2765233516693115, | |
| "learning_rate": 6.204081632653062e-05, | |
| "loss": 0.3369, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 2.1612230715774845, | |
| "grad_norm": 0.2692531645298004, | |
| "learning_rate": 6.183673469387755e-05, | |
| "loss": 0.3347, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 2.1640027797081305, | |
| "grad_norm": 0.2936429977416992, | |
| "learning_rate": 6.163265306122449e-05, | |
| "loss": 0.3794, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 2.166782487838777, | |
| "grad_norm": 0.28063851594924927, | |
| "learning_rate": 6.142857142857143e-05, | |
| "loss": 0.295, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.1695621959694233, | |
| "grad_norm": 0.29556742310523987, | |
| "learning_rate": 6.122448979591838e-05, | |
| "loss": 0.3784, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 2.1723419041000693, | |
| "grad_norm": 0.2919243276119232, | |
| "learning_rate": 6.102040816326531e-05, | |
| "loss": 0.2643, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 2.1751216122307158, | |
| "grad_norm": 0.28946730494499207, | |
| "learning_rate": 6.081632653061224e-05, | |
| "loss": 0.3054, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 2.177901320361362, | |
| "grad_norm": 0.3075507581233978, | |
| "learning_rate": 6.061224489795919e-05, | |
| "loss": 0.3919, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 2.180681028492008, | |
| "grad_norm": 0.27501824498176575, | |
| "learning_rate": 6.040816326530613e-05, | |
| "loss": 0.3273, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.1834607366226546, | |
| "grad_norm": 0.30170246958732605, | |
| "learning_rate": 6.0204081632653065e-05, | |
| "loss": 0.3575, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 2.186240444753301, | |
| "grad_norm": 0.3145295977592468, | |
| "learning_rate": 6e-05, | |
| "loss": 0.3849, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 2.189020152883947, | |
| "grad_norm": 0.2984488308429718, | |
| "learning_rate": 5.979591836734694e-05, | |
| "loss": 0.3593, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 2.1917998610145935, | |
| "grad_norm": 0.27264416217803955, | |
| "learning_rate": 5.959183673469389e-05, | |
| "loss": 0.2832, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 2.19457956914524, | |
| "grad_norm": 0.27420729398727417, | |
| "learning_rate": 5.9387755102040824e-05, | |
| "loss": 0.362, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.197359277275886, | |
| "grad_norm": 0.31026434898376465, | |
| "learning_rate": 5.918367346938776e-05, | |
| "loss": 0.3357, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 2.2001389854065323, | |
| "grad_norm": 0.31472471356391907, | |
| "learning_rate": 5.8979591836734696e-05, | |
| "loss": 0.3327, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 2.2029186935371787, | |
| "grad_norm": 0.2863774597644806, | |
| "learning_rate": 5.877551020408164e-05, | |
| "loss": 0.3819, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 2.2056984016678247, | |
| "grad_norm": 0.35021165013313293, | |
| "learning_rate": 5.8571428571428575e-05, | |
| "loss": 0.3515, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 2.208478109798471, | |
| "grad_norm": 0.2918795347213745, | |
| "learning_rate": 5.836734693877551e-05, | |
| "loss": 0.4171, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.2112578179291176, | |
| "grad_norm": 0.2973978817462921, | |
| "learning_rate": 5.816326530612245e-05, | |
| "loss": 0.3353, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 2.2140375260597636, | |
| "grad_norm": 0.3020112216472626, | |
| "learning_rate": 5.7959183673469384e-05, | |
| "loss": 0.3738, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 2.21681723419041, | |
| "grad_norm": 0.2840433716773987, | |
| "learning_rate": 5.775510204081633e-05, | |
| "loss": 0.3097, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 2.2195969423210564, | |
| "grad_norm": 0.29329997301101685, | |
| "learning_rate": 5.755102040816327e-05, | |
| "loss": 0.3839, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 2.2223766504517024, | |
| "grad_norm": 0.29126960039138794, | |
| "learning_rate": 5.7346938775510206e-05, | |
| "loss": 0.3542, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.225156358582349, | |
| "grad_norm": 0.30318763852119446, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 0.3761, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 2.2279360667129953, | |
| "grad_norm": 0.4189499616622925, | |
| "learning_rate": 5.693877551020409e-05, | |
| "loss": 0.405, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 2.2307157748436413, | |
| "grad_norm": 0.2965310513973236, | |
| "learning_rate": 5.673469387755103e-05, | |
| "loss": 0.3449, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 2.2334954829742877, | |
| "grad_norm": 0.2907022535800934, | |
| "learning_rate": 5.6530612244897964e-05, | |
| "loss": 0.3174, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 2.236275191104934, | |
| "grad_norm": 0.32027149200439453, | |
| "learning_rate": 5.63265306122449e-05, | |
| "loss": 0.3595, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.23905489923558, | |
| "grad_norm": 0.33651795983314514, | |
| "learning_rate": 5.6122448979591836e-05, | |
| "loss": 0.4608, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 2.2418346073662265, | |
| "grad_norm": 0.3164324462413788, | |
| "learning_rate": 5.5918367346938786e-05, | |
| "loss": 0.3974, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 2.244614315496873, | |
| "grad_norm": 0.28678157925605774, | |
| "learning_rate": 5.571428571428572e-05, | |
| "loss": 0.3312, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 2.247394023627519, | |
| "grad_norm": 0.27763816714286804, | |
| "learning_rate": 5.551020408163266e-05, | |
| "loss": 0.2562, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 2.2501737317581654, | |
| "grad_norm": 0.28213679790496826, | |
| "learning_rate": 5.5306122448979594e-05, | |
| "loss": 0.3686, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.252953439888812, | |
| "grad_norm": 0.28878074884414673, | |
| "learning_rate": 5.510204081632653e-05, | |
| "loss": 0.3424, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 2.255733148019458, | |
| "grad_norm": 0.3575231432914734, | |
| "learning_rate": 5.4897959183673473e-05, | |
| "loss": 0.3903, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 2.258512856150104, | |
| "grad_norm": 0.3188578486442566, | |
| "learning_rate": 5.469387755102041e-05, | |
| "loss": 0.2794, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 2.2612925642807506, | |
| "grad_norm": 0.2995283603668213, | |
| "learning_rate": 5.4489795918367346e-05, | |
| "loss": 0.3653, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 2.2640722724113966, | |
| "grad_norm": 0.288529634475708, | |
| "learning_rate": 5.428571428571428e-05, | |
| "loss": 0.2245, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.266851980542043, | |
| "grad_norm": 0.29823359847068787, | |
| "learning_rate": 5.408163265306123e-05, | |
| "loss": 0.4187, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 2.2696316886726895, | |
| "grad_norm": 0.2879655063152313, | |
| "learning_rate": 5.387755102040817e-05, | |
| "loss": 0.2789, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 2.2724113968033355, | |
| "grad_norm": 0.3046426773071289, | |
| "learning_rate": 5.3673469387755104e-05, | |
| "loss": 0.3757, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 2.275191104933982, | |
| "grad_norm": 0.28883394598960876, | |
| "learning_rate": 5.346938775510204e-05, | |
| "loss": 0.3508, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 2.2779708130646283, | |
| "grad_norm": 0.2815608084201813, | |
| "learning_rate": 5.3265306122448976e-05, | |
| "loss": 0.2869, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.2807505211952743, | |
| "grad_norm": 0.27973008155822754, | |
| "learning_rate": 5.3061224489795926e-05, | |
| "loss": 0.3415, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 2.2835302293259208, | |
| "grad_norm": 0.3324487805366516, | |
| "learning_rate": 5.285714285714286e-05, | |
| "loss": 0.3302, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 2.286309937456567, | |
| "grad_norm": 0.30007997155189514, | |
| "learning_rate": 5.26530612244898e-05, | |
| "loss": 0.3831, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 2.289089645587213, | |
| "grad_norm": 0.3399452269077301, | |
| "learning_rate": 5.2448979591836735e-05, | |
| "loss": 0.3788, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 2.2918693537178596, | |
| "grad_norm": 0.28804537653923035, | |
| "learning_rate": 5.224489795918368e-05, | |
| "loss": 0.2934, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.294649061848506, | |
| "grad_norm": 0.34307217597961426, | |
| "learning_rate": 5.2040816326530614e-05, | |
| "loss": 0.3915, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 2.297428769979152, | |
| "grad_norm": 0.26888802647590637, | |
| "learning_rate": 5.1836734693877557e-05, | |
| "loss": 0.2455, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 2.3002084781097984, | |
| "grad_norm": 0.2897135615348816, | |
| "learning_rate": 5.163265306122449e-05, | |
| "loss": 0.2991, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 2.302988186240445, | |
| "grad_norm": 0.2886922359466553, | |
| "learning_rate": 5.142857142857143e-05, | |
| "loss": 0.3311, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 2.305767894371091, | |
| "grad_norm": 0.28240638971328735, | |
| "learning_rate": 5.122448979591837e-05, | |
| "loss": 0.2945, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.3085476025017373, | |
| "grad_norm": 0.29429686069488525, | |
| "learning_rate": 5.102040816326531e-05, | |
| "loss": 0.3996, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 2.3113273106323837, | |
| "grad_norm": 0.30782943964004517, | |
| "learning_rate": 5.0816326530612244e-05, | |
| "loss": 0.3983, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 2.3141070187630297, | |
| "grad_norm": 0.2880532741546631, | |
| "learning_rate": 5.061224489795918e-05, | |
| "loss": 0.4122, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 2.316886726893676, | |
| "grad_norm": 0.29007789492607117, | |
| "learning_rate": 5.040816326530613e-05, | |
| "loss": 0.2972, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 2.3196664350243226, | |
| "grad_norm": 0.31411439180374146, | |
| "learning_rate": 5.0204081632653066e-05, | |
| "loss": 0.3198, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.3224461431549686, | |
| "grad_norm": 0.2791215479373932, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2701, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 2.325225851285615, | |
| "grad_norm": 0.3492055833339691, | |
| "learning_rate": 4.979591836734694e-05, | |
| "loss": 0.2776, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 2.3280055594162614, | |
| "grad_norm": 0.2989236116409302, | |
| "learning_rate": 4.959183673469388e-05, | |
| "loss": 0.3567, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 2.3307852675469074, | |
| "grad_norm": 0.2757430970668793, | |
| "learning_rate": 4.938775510204082e-05, | |
| "loss": 0.311, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 2.333564975677554, | |
| "grad_norm": 0.27763327956199646, | |
| "learning_rate": 4.918367346938776e-05, | |
| "loss": 0.3006, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.3363446838082003, | |
| "grad_norm": 0.2845268249511719, | |
| "learning_rate": 4.89795918367347e-05, | |
| "loss": 0.3055, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 2.3391243919388462, | |
| "grad_norm": 0.29126036167144775, | |
| "learning_rate": 4.877551020408164e-05, | |
| "loss": 0.3033, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 2.3419041000694927, | |
| "grad_norm": 0.29777050018310547, | |
| "learning_rate": 4.8571428571428576e-05, | |
| "loss": 0.2959, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 2.344683808200139, | |
| "grad_norm": 0.27072009444236755, | |
| "learning_rate": 4.836734693877551e-05, | |
| "loss": 0.3122, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 2.347463516330785, | |
| "grad_norm": 0.31056344509124756, | |
| "learning_rate": 4.816326530612245e-05, | |
| "loss": 0.3475, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.3502432244614315, | |
| "grad_norm": 0.2831839323043823, | |
| "learning_rate": 4.795918367346939e-05, | |
| "loss": 0.3048, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 2.353022932592078, | |
| "grad_norm": 0.2710871398448944, | |
| "learning_rate": 4.775510204081633e-05, | |
| "loss": 0.3236, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 2.355802640722724, | |
| "grad_norm": 0.2876143753528595, | |
| "learning_rate": 4.7551020408163263e-05, | |
| "loss": 0.3534, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 2.3585823488533704, | |
| "grad_norm": 0.2987620532512665, | |
| "learning_rate": 4.7346938775510206e-05, | |
| "loss": 0.3324, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 2.361362056984017, | |
| "grad_norm": 0.2859969735145569, | |
| "learning_rate": 4.714285714285714e-05, | |
| "loss": 0.3564, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.364141765114663, | |
| "grad_norm": 0.3041338324546814, | |
| "learning_rate": 4.6938775510204086e-05, | |
| "loss": 0.4652, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 2.366921473245309, | |
| "grad_norm": 0.28285476565361023, | |
| "learning_rate": 4.673469387755102e-05, | |
| "loss": 0.3897, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 2.3697011813759556, | |
| "grad_norm": 0.307158499956131, | |
| "learning_rate": 4.653061224489796e-05, | |
| "loss": 0.3194, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 2.3724808895066016, | |
| "grad_norm": 0.31216737627983093, | |
| "learning_rate": 4.63265306122449e-05, | |
| "loss": 0.3554, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 2.375260597637248, | |
| "grad_norm": 0.3206304907798767, | |
| "learning_rate": 4.612244897959184e-05, | |
| "loss": 0.4214, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.3780403057678945, | |
| "grad_norm": 0.29540154337882996, | |
| "learning_rate": 4.591836734693878e-05, | |
| "loss": 0.3425, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 2.3808200138985405, | |
| "grad_norm": 0.2654211223125458, | |
| "learning_rate": 4.5714285714285716e-05, | |
| "loss": 0.2876, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 2.383599722029187, | |
| "grad_norm": 0.2998722195625305, | |
| "learning_rate": 4.551020408163266e-05, | |
| "loss": 0.3234, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 2.3863794301598333, | |
| "grad_norm": 0.305711030960083, | |
| "learning_rate": 4.5306122448979595e-05, | |
| "loss": 0.3229, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 2.3891591382904793, | |
| "grad_norm": 0.29556146264076233, | |
| "learning_rate": 4.510204081632654e-05, | |
| "loss": 0.3013, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.3919388464211258, | |
| "grad_norm": 0.2814411222934723, | |
| "learning_rate": 4.4897959183673474e-05, | |
| "loss": 0.2751, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 2.394718554551772, | |
| "grad_norm": 0.29568272829055786, | |
| "learning_rate": 4.469387755102041e-05, | |
| "loss": 0.3107, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 2.397498262682418, | |
| "grad_norm": 0.28841209411621094, | |
| "learning_rate": 4.448979591836735e-05, | |
| "loss": 0.3088, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 2.4002779708130646, | |
| "grad_norm": 0.3105250298976898, | |
| "learning_rate": 4.428571428571428e-05, | |
| "loss": 0.3483, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 2.403057678943711, | |
| "grad_norm": 0.287725567817688, | |
| "learning_rate": 4.4081632653061226e-05, | |
| "loss": 0.3556, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.405837387074357, | |
| "grad_norm": 0.33084672689437866, | |
| "learning_rate": 4.387755102040816e-05, | |
| "loss": 0.4709, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 2.4086170952050034, | |
| "grad_norm": 0.3187430799007416, | |
| "learning_rate": 4.3673469387755105e-05, | |
| "loss": 0.3438, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 2.41139680333565, | |
| "grad_norm": 0.29251527786254883, | |
| "learning_rate": 4.346938775510204e-05, | |
| "loss": 0.4519, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 2.414176511466296, | |
| "grad_norm": 0.27169179916381836, | |
| "learning_rate": 4.3265306122448984e-05, | |
| "loss": 0.3039, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 2.4169562195969423, | |
| "grad_norm": 0.2867499589920044, | |
| "learning_rate": 4.306122448979592e-05, | |
| "loss": 0.3444, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.4197359277275887, | |
| "grad_norm": 0.33954930305480957, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.4085, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 2.4225156358582347, | |
| "grad_norm": 0.31500715017318726, | |
| "learning_rate": 4.26530612244898e-05, | |
| "loss": 0.3574, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 2.425295343988881, | |
| "grad_norm": 0.31862321496009827, | |
| "learning_rate": 4.2448979591836735e-05, | |
| "loss": 0.361, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 2.4280750521195276, | |
| "grad_norm": 0.28617167472839355, | |
| "learning_rate": 4.224489795918368e-05, | |
| "loss": 0.362, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 2.4308547602501736, | |
| "grad_norm": 0.3150579333305359, | |
| "learning_rate": 4.2040816326530615e-05, | |
| "loss": 0.281, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.43363446838082, | |
| "grad_norm": 0.28917694091796875, | |
| "learning_rate": 4.183673469387756e-05, | |
| "loss": 0.3215, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 2.4364141765114664, | |
| "grad_norm": 0.29830271005630493, | |
| "learning_rate": 4.1632653061224494e-05, | |
| "loss": 0.2891, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 2.4391938846421124, | |
| "grad_norm": 0.30335041880607605, | |
| "learning_rate": 4.1428571428571437e-05, | |
| "loss": 0.3507, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 2.441973592772759, | |
| "grad_norm": 0.2745303213596344, | |
| "learning_rate": 4.122448979591837e-05, | |
| "loss": 0.366, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 2.4447533009034053, | |
| "grad_norm": 0.26936864852905273, | |
| "learning_rate": 4.102040816326531e-05, | |
| "loss": 0.3247, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.4475330090340512, | |
| "grad_norm": 0.3225851058959961, | |
| "learning_rate": 4.0816326530612245e-05, | |
| "loss": 0.3757, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 2.4503127171646977, | |
| "grad_norm": 0.28156349062919617, | |
| "learning_rate": 4.061224489795918e-05, | |
| "loss": 0.219, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 2.453092425295344, | |
| "grad_norm": 0.46358248591423035, | |
| "learning_rate": 4.0408163265306124e-05, | |
| "loss": 0.3567, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 2.45587213342599, | |
| "grad_norm": 0.2787911295890808, | |
| "learning_rate": 4.020408163265306e-05, | |
| "loss": 0.275, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 2.4586518415566365, | |
| "grad_norm": 0.3119618892669678, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3376, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.461431549687283, | |
| "grad_norm": 0.2867750823497772, | |
| "learning_rate": 3.979591836734694e-05, | |
| "loss": 0.2792, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 2.464211257817929, | |
| "grad_norm": 0.29150545597076416, | |
| "learning_rate": 3.9591836734693876e-05, | |
| "loss": 0.3543, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 2.4669909659485754, | |
| "grad_norm": 0.28618574142456055, | |
| "learning_rate": 3.938775510204082e-05, | |
| "loss": 0.3556, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 2.469770674079222, | |
| "grad_norm": 0.2973514199256897, | |
| "learning_rate": 3.9183673469387755e-05, | |
| "loss": 0.3071, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 2.472550382209868, | |
| "grad_norm": 0.36355265974998474, | |
| "learning_rate": 3.89795918367347e-05, | |
| "loss": 0.37, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.475330090340514, | |
| "grad_norm": 0.28134381771087646, | |
| "learning_rate": 3.8775510204081634e-05, | |
| "loss": 0.3189, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 2.4781097984711606, | |
| "grad_norm": 0.29104146361351013, | |
| "learning_rate": 3.857142857142858e-05, | |
| "loss": 0.3467, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.4808895066018066, | |
| "grad_norm": 0.3080955743789673, | |
| "learning_rate": 3.836734693877551e-05, | |
| "loss": 0.439, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 2.483669214732453, | |
| "grad_norm": 0.3061097264289856, | |
| "learning_rate": 3.8163265306122456e-05, | |
| "loss": 0.2978, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 2.4864489228630995, | |
| "grad_norm": 0.30751508474349976, | |
| "learning_rate": 3.795918367346939e-05, | |
| "loss": 0.3819, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.4892286309937455, | |
| "grad_norm": 0.2971366047859192, | |
| "learning_rate": 3.775510204081633e-05, | |
| "loss": 0.2825, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 2.492008339124392, | |
| "grad_norm": 0.29852065443992615, | |
| "learning_rate": 3.7551020408163264e-05, | |
| "loss": 0.341, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 2.4947880472550383, | |
| "grad_norm": 0.28704988956451416, | |
| "learning_rate": 3.734693877551021e-05, | |
| "loss": 0.3001, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 2.4975677553856843, | |
| "grad_norm": 0.3021228015422821, | |
| "learning_rate": 3.7142857142857143e-05, | |
| "loss": 0.3204, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 2.5003474635163307, | |
| "grad_norm": 0.30266740918159485, | |
| "learning_rate": 3.693877551020408e-05, | |
| "loss": 0.299, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.5003474635163307, | |
| "eval_loss": 0.4358259439468384, | |
| "eval_runtime": 212.5847, | |
| "eval_samples_per_second": 1.693, | |
| "eval_steps_per_second": 1.693, | |
| "step": 900 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1080, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.8347978086418432e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |