| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.992, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 7.164950271369886, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.3363, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 7.188107120097212, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.3668, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 7.014852719900384, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.3602, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 6.71890357153603, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.397, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.800329079971533, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 1.3217, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.4546024067291605, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.2951, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 3.2046867552172387, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 1.2951, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 6.259700539472787, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 1.2834, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 5.79283760537252, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.2979, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 3.323447917543639, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 1.2498, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 2.624690816053142, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 1.2504, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.3480180564150426, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 1.2002, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.7451214613149895, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.1911, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.7716947114358235, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 1.1649, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.8112661073479859, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 1.1177, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.6764859206158345, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 1.139, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.675576451603029, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 1.1366, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.3712292219002176, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 1.125, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.4343496147047017, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 1.1298, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.3369281283006227, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 1.0938, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.3466095451129891, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 1.0978, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.2034784374931198, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 1.1105, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.0576073442646439, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 1.1096, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.2764711529321302, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 1.0818, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.277188018725038, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 1.0748, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.9052874765803237, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 1.0936, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.9285230519282399, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 1.0798, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 1.0913136287963536, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 1.0361, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.987490071274066, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 1.0932, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.5345332620628627, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 1.0336, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.4537328686413111, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 1.0465, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 1.1992537731436224, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 1.0715, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.8195530418744121, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 1.0396, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.43084635089086, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 1.042, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.713246366337894, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 1.0475, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.5820093171614773, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 1.0605, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.9872032855155564, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 1.0559, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.2187695868975368, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 1.0412, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 2.8104170842167084, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0263, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.974795391335889, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 1.0578, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 3.0061420983785814, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 1.0367, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 2.234576933635993, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 1.0522, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 2.930414055237407, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 1.0347, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 2.1383162093699886, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 1.0315, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 2.7023871568398934, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 1.0417, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 2.302068857065522, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 1.0378, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 2.2518183298195633, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 1.0191, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.7058378144032798, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 1.0221, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 2.4001286805205897, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 1.0168, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.745040133077506, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 1.0323, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 2.394929338775231, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 1.0347, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.9658035952344797, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 1.0125, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 2.172581028650137, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 1.0094, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.7176404079809477, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 1.0313, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 2.5841075304197836, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 1.0282, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 2.1292911215091608, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.9907, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.9208118574219684, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 1.0071, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 1.7464011697929016, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 1.0177, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 2.0529524231611225, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 1.0304, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.7021682966396998, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 1.038, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.9747322953497688, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 1.0284, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 1.6689660541089448, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 1.0574, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.9085286458081447, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 1.0087, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 1.9709070582656916, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.9993, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.591558411797372, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 1.0088, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 1.4618449025325562, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.9835, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 1.7429881827620637, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 1.0082, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 1.3506682914473163, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.9945, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.656027049558351, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 1.0353, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.7143921151842663, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.9827, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.5137349609023163, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 1.0058, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 1.2238054151082596, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.9927, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 1.597047832030457, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 1.0173, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 1.1367117675736156, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.9917, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.7295177017370893, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.9916, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 1.5271702941759078, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 1.0035, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 1.5877604836924284, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.9891, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 1.4352731097580478, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.989, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.3888660377630004, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.9824, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.4253317466807294, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.9637, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 1.1372357091540986, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.9629, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 2.177557113796439, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.9612, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 1.4698348788124762, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.9808, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 1.9893633729836133, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.9676, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.9079842946365686, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.9899, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 1.4698172966343532, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.9841, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 2.181229449884897, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 1.0057, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 1.6664501354091148, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 1.0046, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 2.1565677883215777, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.96, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 1.6006610148803861, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.9694, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 2.110368158751775, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.9462, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 1.5375648712898362, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.995, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 2.286421323310093, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.9653, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 1.8818328490981036, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.9723, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 1.8736741214509196, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.9558, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 1.8558538404600857, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.9625, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 1.6576430950495782, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.9642, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 1.2710501333025392, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.9553, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 1.78947951017059, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.955, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.0046908558018692, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.9468, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 2.299298500737952, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.968, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 1.8842926680362198, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.9662, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 1.4904718458319735, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.97, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 1.381065853784866, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.9536, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 1.5496353050270812, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.9558, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 1.1603265668878235, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.9609, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 1.2181512609568441, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.9596, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 1.3585472014887268, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.988, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.9687424270712927, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.9635, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 1.8016675392121388, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 1.0002, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 1.0325596481559047, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.9543, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 2.11023778130976, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.976, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 1.8317117372988703, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.9568, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 1.6389149116237727, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.9429, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 1.6411010273756654, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.9811, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 1.5624507733685191, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.9488, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 1.4028502840139594, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.9774, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 1.45349200251987, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.9713, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 1.1690141658307711, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.9523, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 1.4077553945852521, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.9787, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 1.2481301943687173, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.9652, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 1.0932124477672736, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.985, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.9729353275653629, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.9495, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 1.1936981245321536, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.9416, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.1404518208695538, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.9732, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.6993203038771096, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.9546, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 1.0608627344320947, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.9506, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 1.7559845007888086, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.9704, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.8570482025795857, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.9588, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 1.3809795673071772, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.9664, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 1.192097090708205, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.9706, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 1.1728749824185387, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.9635, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 1.0254236166458253, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.9582, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 1.2944670798178768, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.9466, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 1.0852845114386136, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.9404, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.9893542167214449, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.9293, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.7962698125683273, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.9788, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 1.129314297991463, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.9905, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 1.3418335710887281, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.9825, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.8451710152152404, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.9652, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 1.4833314561327016, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 1.0017, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 1.0188391468540687, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.9528, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 1.443999127402159, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.9648, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 1.165886746905586, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.9602, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 1.2079424801859484, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.9825, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 1.2611231901729372, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.9643, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 1.0288902206989285, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.9328, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 1.0461884712233733, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.9528, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.9764378382966373, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.9669, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.9833853602027257, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.963, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.8554676094946236, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.9441, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.7965412812926458, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.9649, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.9371774811571757, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.9622, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 1.037438714597397, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.9518, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.878475264261868, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.9643, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 1.0254072538419412, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.9792, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.9151915442462214, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.9135, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.9062819935142862, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.9157, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.8551937877393435, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.9183, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.7217580214942478, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.9225, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6970821745484023, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.9046, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.8150297794550617, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.9099, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 1.0789140351767412, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.9061, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.6695054091011142, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.8724, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.8912183959068726, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.8924, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.8952149424633479, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.949, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.8568616957578176, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.898, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.885359304863275, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.913, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.6221998110176642, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.8939, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.8040519441441258, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.9381, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.48174211651046894, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.8788, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.8955190741167064, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.9172, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.6441267133742979, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.9211, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.687305439646437, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.8969, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.8125615608399682, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.9284, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.5040263914384315, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.8979, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.8319313400045124, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.942, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.6132934273719209, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.8899, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5865334955701278, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.926, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.7732127603042257, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.886, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.5574267505313263, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.9249, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.6333327806281892, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.9061, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.5020924086621534, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.9265, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.5241553499368118, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.884, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.5258104025467343, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.9194, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.5519940504823936, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.9251, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.5670815561170646, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.9099, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.5376431573871367, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.9009, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.4851049322508697, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.9129, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.5819962293805395, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.9187, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.5883683504527734, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.8951, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.548510698596303, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.9298, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.5228904411928151, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.898, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.575577920181799, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.8989, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.5208065019798606, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.926, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.5107247184771793, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.9153, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.5335951342144748, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.8897, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.5860719237335125, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.9154, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.5177265087659593, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.9325, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.6831142541202275, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.9142, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.4211155013693619, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.8686, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.5365682906160393, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.9129, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.431427523186253, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.8861, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.4404972668402265, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.9203, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.47743559291027765, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.9343, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.4558343377906512, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.9106, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.39241064439181733, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.9009, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.47270467980687986, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.8912, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.46530768142793816, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.9117, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.48579142956579957, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.9373, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.4162821538935548, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.9212, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.3821235129761479, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.9141, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.43441851563720163, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.8972, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.43903447514949184, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.9417, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4253635143936466, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.9116, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.47906081248536403, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.9312, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.3541426747199045, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.909, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.47818528671289545, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.9112, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.4144535504609497, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.8721, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.43492017910491865, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.9252, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.4301181164359425, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.9063, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.4931312326878231, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.9167, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.4015444541710621, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.9101, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.44697717687099564, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.9046, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.4796171218523403, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.916, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.4174639548693177, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.9176, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.4828788324452908, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.9161, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.5296042245535623, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.9267, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.39341157273773414, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.8952, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.4867706387396608, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.8949, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.41150861042633097, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.9065, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.4804245933056599, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.908, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.47155109150840635, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.9149, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.41988354058518246, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.9124, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.5206364046784432, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.8634, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.47266998679777616, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.8606, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.4252137751225095, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.8525, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5915679374031652, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.8498, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.47699219597688836, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.8618, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.6592655555285256, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.8982, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.437527168280848, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.861, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5671023725917032, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.8756, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.5152227819137052, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.8565, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.5099608377571634, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.859, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.535828604166928, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.8328, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.6271639739397001, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.8649, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.5170187988729998, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.8534, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.5604711208756066, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.839, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.46003342269477776, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.8627, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.44069293934722, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.8343, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.47139653442059765, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.8396, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.44274943545342926, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.838, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.4446513964424396, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.8496, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.42024050689042824, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.8459, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.44970253836320634, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.8423, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3908082563889526, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.8603, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.44204728995375253, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.8624, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.39179597824316936, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.8564, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3873248319127629, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.8661, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3755311209951693, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.8525, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3815379605586814, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.894, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.3827290487935651, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.8639, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3326056667036131, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.906, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.38488995845668356, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.8711, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3632298422675201, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.845, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.35453520894961466, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.8624, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.32953957759546015, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.8519, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3276699060632302, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.873, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3233589411073006, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.8526, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3276242141439983, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.8782, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.2908392270971308, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.8677, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.32101686331927526, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.8619, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.29856386828391446, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.8436, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.29320793611313534, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.8547, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.30872912114867584, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.8574, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.2921591079218018, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.8443, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3218532315765348, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.8526, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3171241132362057, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.8579, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.2872908250335455, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.8562, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.350077788650057, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.8479, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3116741039475533, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.8732, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.28668625704559364, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.8516, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3501064179219289, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.8655, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.2970474993126229, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.837, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3142514067554776, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.8523, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3300454131595622, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.8377, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.29744641348974693, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.857, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.26017399042245387, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.8352, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.29774399017842684, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.8605, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.2851307609607096, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.864, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3122120869813163, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.8516, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.2725614367896525, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.8539, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.3041428970002878, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.8433, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.2962055684130299, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.8839, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.26556281288752653, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.8824, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.3087677932385054, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.86, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.2976254233929599, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.8697, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3368870805303976, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.8446, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.3080605977489481, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.8719, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2583076669175364, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.8335, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.2688227436632698, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.8399, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.30820761804417557, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.8588, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.25802816355778335, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.8446, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.2648279837111969, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.8721, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.2954452997850354, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.8466, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.3027270659239321, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.8763, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.2578743886460077, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.8577, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.30473416398596653, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.8437, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.27298782554650586, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.8637, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.24731095049296872, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.8539, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.2586851765365903, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.8646, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.28192115508091664, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.8698, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.27155186929289393, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.8655, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.30480349046261057, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.8359, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.3065397546371416, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.8298, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.2904874325690977, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.8333, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.31198706491295003, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.835, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.2910259951258636, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.8202, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.3288089012425444, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.8103, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.32811717287677095, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.8257, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.28764869118645825, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.8043, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.31959650351172064, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.8006, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.33362308577416316, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.8404, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.2834617108041555, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.8388, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.2533960898658581, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.8303, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.29493402774300675, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.8299, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.29655168229324996, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.8143, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.2801717023680744, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.8071, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.26013042587351004, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.8273, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.25289256849529507, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.8356, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.26652875378307894, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.8183, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.254895411910539, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.8328, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.24616930188144143, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.8431, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.24696132991823286, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.8261, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.2532209451928927, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.829, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.24801648513526872, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.7862, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.24573150494837506, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.8073, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.23995671560779114, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.8266, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.22540022534837623, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.8182, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.25432847437819195, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.7879, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2400887912803633, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.8394, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.22354837344195608, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.8054, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.23224982137410188, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.8376, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.23804222174647305, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.8241, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2421572512693195, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.8178, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.220427549174197, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.8341, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.23422350399546596, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.8432, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.20313110973495466, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.8114, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.23307256758969278, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.8429, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.21715061006497335, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.8355, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.21335257281523295, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.8429, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.22437415987408385, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.8248, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.21994430486110494, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.8149, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.21229529645116088, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.8274, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.2192597408311037, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.828, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.2067951876371041, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.8124, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.20838661681232568, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.813, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.20981136104892134, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.8389, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.24112083727959965, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.8391, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.21012942288526443, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.841, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.22972959579752636, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.822, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.23044166772319077, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.8477, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.20672004802105884, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.8185, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.19619532089387265, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.804, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.20143198283831878, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.8113, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.20364983676918752, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.7937, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.1988768827715147, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.8025, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.22254590075781017, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.8296, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.19598061434947578, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.8346, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.20365695768440017, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.8117, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.19474775299560979, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.8226, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.19148044990981916, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.8105, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.1957001259407145, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.8343, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.2177790330920898, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.8425, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.20739383482377585, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.8516, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2085874255217451, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.835, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.19899767805414778, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.8206, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.1930846911869035, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.8076, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2110519348031712, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.8048, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.1990929906361697, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.8476, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.19778093878994266, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.8043, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.18696362303365713, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.824, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.19002846780082872, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.8195, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.19480939612643827, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.8199, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.21157164649110938, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.8315, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.1969527400953149, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.8267, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.1981097493846016, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.8399, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.19966199390953818, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.8407, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.20181333023611767, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.8258, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.201562055113913, | |
| "learning_rate": 0.0, | |
| "loss": 0.8238, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 2.5463658746048676e+18, | |
| "train_loss": 0.9288770392919198, | |
| "train_runtime": 36740.926, | |
| "train_samples_per_second": 1.361, | |
| "train_steps_per_second": 0.011 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.5463658746048676e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |