| { | |
| "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": 6.3857031845988805, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.0146, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.513457738231102, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0407, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.389369904722248, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0401, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.657954631643649, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.0027, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.2492750171036775, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9808, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.787538797244276, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9533, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.545431313650447, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.9248, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.368850015054219, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.9499, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.286250499212212, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9464, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.095203710101558, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.9051, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.718972852702716, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8952, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.5516000233573277, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8712, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.643547893866034, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8577, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.6053997700362166, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8136, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.3344448240126532, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7975, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.0761544947032076, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7872, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.1636146263275826, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.8055, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.0969147745284513, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7756, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.9698538367910754, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7499, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.0666601130971711, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7432, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.9401468890973036, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7327, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.8000355819072862, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7404, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.9438590359860037, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7164, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8533247867027401, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7383, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.999532561766686, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7248, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.8006836702962915, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7351, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6805979308241464, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7205, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.998808499998666, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.7096, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.9809835269133407, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7113, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.345020185848282, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6985, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.8199700627632992, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.7008, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.9280100911930844, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6964, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.1968103155254415, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.7062, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.2904515116682622, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6876, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7405587717186609, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6942, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.2333725656317318, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6816, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.9199921652580408, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6851, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.1266206767787048, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6834, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.0983329089323715, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6849, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.2933039671370061, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6792, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.1044414827820608, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6815, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.6229197518771812, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6812, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.0537460648246924, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6669, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.6669868894341158, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6824, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.062285226779462, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6691, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 2.315993547681366, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6761, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 2.0499112390059926, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6644, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.2800447567817648, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6809, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.4208100924571356, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6789, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.2828897919868043, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6727, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.1378068747326247, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6666, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.1703720670666418, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6535, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.8378124740862146, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6743, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.1803403654880005, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6607, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.7014239373149845, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6495, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.9599898972216756, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6731, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.015973863547262, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6732, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 1.3645890344778389, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6505, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.7396969296398802, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6515, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7457766555713855, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6551, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.0213173629581123, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6578, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 1.420903141785154, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6535, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6787691231974609, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.642, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 1.3907521323681888, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6522, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.1489305276103492, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6471, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.82984156577414, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6577, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 1.558153671441526, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6491, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.701248806012898, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6478, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.541542547111201, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6482, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.790406383186093, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6322, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.0598252879306564, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6645, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 1.1772692501202318, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6407, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.5997636093563055, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6454, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 1.1473486204015397, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6412, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.8986351955161239, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6457, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.7846398455496357, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6372, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 1.1094087249281304, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6432, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.6848404586025102, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6306, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.9964224498739199, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.5884, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.9102689953399805, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5942, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 1.1178515136602816, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5925, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.1353848859438942, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5844, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7692431316197359, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.6025, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 1.0299639532071048, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5934, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.096779927244621, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5838, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.5928090493327227, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5901, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.782379693650114, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5894, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.8060013915829145, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5808, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.7145181062083139, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5775, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6962399353140069, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5834, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.7395804826488809, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5827, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5621723281500479, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5791, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.5318029984699583, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5943, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.5733431964660091, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5885, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.6960371666052328, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5955, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.9359896638510405, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5842, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.8480765879835703, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5771, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5400407046866157, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5747, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.8161506433842548, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5766, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5058811630616397, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5837, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.7431695528629912, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5723, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6643391693290811, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5751, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.6011164147503848, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5876, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.4730209286365697, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.595, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5841846391799305, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.6005, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.6632343206084076, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5804, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.6630054629726436, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5769, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.598020909384883, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.588, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.48985433484279217, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.57, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.3863983254888785, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5729, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4740187512818743, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5793, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.42622525168734493, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5841, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.39643940618622864, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5861, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.4242528887242483, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5834, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5421500517599109, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.582, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.5311832845228792, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.572, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4662212069404133, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5783, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.45833419968026556, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5642, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.4497591263859104, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5859, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.6258514869389551, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5711, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.5686703454809979, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5841, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.3681018500089249, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.584, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.42669869438547914, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5868, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5441247186904558, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5764, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.501140527539522, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5794, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.509604255782758, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5887, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.468819660502078, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.5708, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5681474181696031, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5764, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.48439243320979486, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5872, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5120908720823386, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.574, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.6349706050854664, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5654, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5797502972193269, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5671, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.47115877117844046, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5808, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5315156219737786, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5736, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.606597702185915, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5737, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.6363278961587933, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5778, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.37272326213256113, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5686, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5391746666140868, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5716, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.4933426735591773, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5786, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.47449680125291605, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5663, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.46684001938007624, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5657, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.47569872068326147, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5747, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.5626877646957587, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5703, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.5005964045377986, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5702, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.4728803164820471, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5701, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.6279759957965352, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5687, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4414026788106849, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5709, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.39742569125698024, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5677, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.46569036957339405, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5757, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4334250181502594, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5675, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.4204930094090842, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5687, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.4987314412738242, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.571, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.46940615574832384, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5691, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.5577877039238854, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5581, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.42448538136278574, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5789, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.6763612785070511, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5741, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.5722503174192143, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.5143, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.5026703958168597, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.5223, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.8492116119835177, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.5036, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.913996520180068, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.5197, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.578935444154028, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.4952, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5860904629062806, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.5, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.8044025573168775, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.4917, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5906320253924908, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.5018, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6498325035791859, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.4965, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.628304699648312, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.4865, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5998061185741181, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.491, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5981618933467712, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.5045, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.4744854867041487, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.4894, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.49326360205323266, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.4894, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.44948652290190905, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.4981, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5056694712408589, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.4947, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5313951384739188, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4947, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.40427169688811354, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.5054, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5812741282451808, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.4981, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4295515657065, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.5049, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.42951431128473816, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.4896, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.5201501793555885, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.494, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.4290076152050141, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.4951, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.41103474663991013, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.4951, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.3732307708169955, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.4957, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.43299916544822603, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.4962, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.36682921850650274, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.4999, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.3507507395182079, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.4897, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.4403078896816384, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.4942, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3760556219110924, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.5047, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.32749741599683097, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.4835, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.3903898162096965, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.5067, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.38438738834323594, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.4849, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.402863434106592, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.5002, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.39000613279216245, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.5136, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.4419462350894651, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.4922, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.3606703689703085, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.5045, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.39658921683014065, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.4992, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.37008936834657263, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.5001, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.38277633581754456, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.4881, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.34780459495046656, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.489, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.42528032608133876, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.4972, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.3298875279115166, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.4925, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.40522458502738923, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.4987, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.28890557367196196, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.4828, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.35752143768891786, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.495, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.2847955714786017, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.4923, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.34462734916524024, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.4919, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.28997561583741843, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.4876, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.32736744376129595, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.4889, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.29552255599003757, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.495, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.2986535725397418, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.4953, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.32346026994057464, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.4862, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3248091215402218, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.4822, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3069758746580514, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.5123, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.3763288654288441, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.4957, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3201346373064764, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.4953, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.39729074863968356, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.499, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.3304189063909658, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.4959, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3430447918196304, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.4934, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.34316923398695875, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.5064, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.3481950457193192, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.4957, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.34710786210143224, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.4884, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3347710607736284, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.4891, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.3343701183653994, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.4981, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3521372001758378, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.5067, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.3041068550935644, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.4868, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.3210055574329485, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.4953, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.27212342350119617, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.4914, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.2988079253064433, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.5055, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3249357836053954, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.4963, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.30877687777261964, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.5087, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.34825679402688037, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.4852, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3122362652270399, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.4927, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3555980989681846, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.4993, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.2890714695184332, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.4791, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.3340506927436075, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.5169, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.31408243123886154, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.4956, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.40654585096835055, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.4588, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.419884900413749, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.4254, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.6008062571027746, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.433, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.4465527398249857, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.444, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.4193803025980898, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.4313, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4125516108235289, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.4103, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.41021467807578205, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.4212, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.3790875492885813, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.4334, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.42829444758072965, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.4269, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3544697233015643, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.4277, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4040035567489867, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.4312, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.3438830200025999, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.4408, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3457383948694485, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.4318, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.3209468399576312, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.4282, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.31904439275611846, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.435, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3247930622800883, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.4155, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.2873541554352374, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.4364, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.34081150707234453, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.421, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.2948836577244133, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.4368, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3162585544662662, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.4278, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3314459330554049, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.427, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.2776568172512342, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.4361, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.32208249694143454, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.4203, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3027462367980654, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.4248, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.28819685100487374, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.4288, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.2980531747486014, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.4232, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.2890118308205293, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.4139, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.26366169093621045, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.4319, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.2728501624545788, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.4193, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.29681862061736447, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.426, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.2544508537998433, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.4187, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3207370911151253, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.4184, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.2649553530642913, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.4232, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.2575834273446465, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.4211, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.30780701681300127, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.4244, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.28170446467321153, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.4199, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.2637541363350921, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.4214, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.27966307752352326, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.4319, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.2689145271383354, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.4294, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.26132213013856004, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.4279, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.4649247492039518, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.4163, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.25611463232583487, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.4316, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.2566981779171312, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.4159, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.24958050574233248, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.4287, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.24836265500751314, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.4275, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.24692726709965532, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.4267, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.24609086568215463, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.4092, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.23954856666870555, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.4195, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.23364220624396864, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.4329, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.22609657976453879, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.4365, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.2519088471123882, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.406, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.24226192603330354, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.4294, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.24703777332038798, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.4235, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.2416819894481494, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.4236, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.2437920000095238, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.419, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.23327557264379006, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.4334, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2293395410726176, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.4235, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.22677999885899983, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.4176, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.23165615642225584, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.4185, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.25076188779206476, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.413, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.24230610135722463, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.4195, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.22966849116061802, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.4383, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.237876486704717, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.4212, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.23420288846813658, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.424, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.22855408286153545, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.4216, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.22496842605104692, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.4268, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.260580979935843, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.4362, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.23515581942316058, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.4227, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.2301821744711806, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.4115, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.24041427254233036, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.4163, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.22962672133758927, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.412, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.2546876405009508, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.4453, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.251874165605757, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.4226, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.24762103003628663, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.4213, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.23547656543525808, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.4213, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.22124698994060055, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.4322, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.23151356501707346, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.4299, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.2222851779764549, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.4267, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.27313353289247716, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.4036, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.37647735983296937, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.3924, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.28037822372132515, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.3638, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3125956471742338, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.3823, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.48895993038764357, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.3775, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3445166504068549, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.3813, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.30571755702603826, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.3979, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.33530399695171575, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.3839, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.30971433570989015, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.3807, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.31386159719572704, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.3578, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.26836719144538546, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.3817, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.28412212263969033, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.3889, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.29762400347113643, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.3846, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.2860278447290372, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.3701, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.2706472326402335, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.3775, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.26185114842096163, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.3694, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2612562139510835, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.3931, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.25162090008945814, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.3722, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.23993277918222877, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.3979, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.23113009585064762, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.3781, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.2318792430421106, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.3693, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.23481091374966023, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.4017, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.2271355649495333, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.3788, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2242750207050269, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.3809, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.22750939685861177, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.3826, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.22032122669334825, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.3767, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.21396305180442826, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.3805, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.21808076268228624, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.3723, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.22918310419902882, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.3721, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.21912269552519437, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.3674, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.21061760499423257, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.3668, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2135093244604889, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.3667, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.20676744028641914, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.3856, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.21342342355138336, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.4001, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.21295302256781662, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.3966, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.20877331416746198, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.3894, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.21727350357530545, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.3691, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.21847302181712927, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.3887, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.212011319409588, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.3905, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.20316624291873847, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.384, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.20640523424617857, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.3797, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.2054997141217928, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.3737, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.20057769739419745, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.3815, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.20074679817279875, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.3779, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.20327244059916494, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.3756, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2046938820000016, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.385, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.20842525006937218, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.3787, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.19726586055435785, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.4023, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.19706019306135975, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.3768, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.1930540936571029, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.3682, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.20579857751364183, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.3651, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.20820600565350564, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.3761, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.20574772820004417, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.3818, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.202604848883481, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.3799, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.1944897386871682, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.3902, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.1994640180714002, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.3741, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2040530220053012, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.3865, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.20017182295693167, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.3904, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.20525505206304456, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.3801, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.20293579102879616, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.3825, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.19636267825368714, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.378, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.19655957456657486, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.3907, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2063623796074329, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.3663, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.19683990366572185, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.3912, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.19932367917369181, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.3879, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.19792117174200943, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.3678, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.20001620409625334, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.3746, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.1990001961157355, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.3888, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.20240824252767245, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.3595, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.19804532092587782, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.3877, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.20626940327483265, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.3837, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.19646889767438333, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.3817, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.19958394868575255, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.3791, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.20335317670077688, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.3868, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2031273073431524, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.3642, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.19642636644665254, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.372, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2028644971237751, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.3661, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2008515231696296, | |
| "learning_rate": 0.0, | |
| "loss": 0.3784, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 2.1484573977508577e+18, | |
| "train_loss": 0.5224053398156777, | |
| "train_runtime": 63510.1122, | |
| "train_samples_per_second": 0.787, | |
| "train_steps_per_second": 0.006 | |
| } | |
| ], | |
| "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": 2.1484573977508577e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |