| { | |
| "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": 6.917312863904752, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.8778, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 7.115675932218564, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.8978, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 7.143090959616597, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.8836, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 6.547261440554129, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.8536, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 5.080727646487042, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.8178, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.9238509462148383, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.7305, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.1194750359479344, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.7348, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 3.762748319278985, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.7398, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.939983995799894, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.6939, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.4957667148675045, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.6766, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.466339084180279, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.6567, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.1065964129262578, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.6419, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.129844525633678, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6066, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.6341650255331666, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.5788, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.4851280434276903, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.5903, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.7306485929486228, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.5908, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.1989593380058583, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.5816, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.5709794458589184, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5451, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.2712863753195993, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.5311, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.7593234560614043, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.5728, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.0229247297773885, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.5259, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.0575624137381636, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.536, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7079152316748596, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.5312, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.9714120507188312, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.5134, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7949976651512598, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.5191, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.9651032085990672, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.521, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.8248971552417219, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.5346, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.6876941086023203, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.5197, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.7726526234427976, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.5334, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.8622968341305126, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.5174, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.7178794371288015, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.4888, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.6291964033066837, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.5014, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.6304774189343574, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.4776, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.6509761608633534, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.4812, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7024262808757761, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.4929, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.7469272638066381, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.515, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.7063033600086017, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.5154, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.6891954146829927, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.5247, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.729694958902066, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4879, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.6293498574203988, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.5072, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.7067898835951391, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.4859, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.6792215516455709, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.4762, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.7112534998939477, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.4966, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.8917671882383743, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.4879, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7164453252644027, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.4682, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.9202895733189819, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.4737, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.6666361003438368, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.4692, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.8993138978977713, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.5134, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.7819898848123767, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.4615, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.7700009217327806, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.4936, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.7749980808284274, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.478, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7193450266009187, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.4747, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.6600121021852756, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.4628, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.5917480390981291, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.5044, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.6527439856884594, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.4969, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.6245050670150339, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.4634, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.5683956077196995, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.4987, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.6824717383950719, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.465, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.8133420587406068, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.4777, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5905406037427718, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.4701, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.5285981231850914, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.4747, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6448941053978279, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.4866, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6070514524417691, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.4769, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.636962417422208, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.4619, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.6305102055156264, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.4775, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.5996436733494288, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.4796, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.836070659374884, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.4589, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.7862525734167326, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.4888, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.66679339503122, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.4614, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.7166546763180385, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.4771, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.9045232117210558, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.4738, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5864164936190818, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.4758, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.8376736221861006, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.4869, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.5930708646985897, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.4792, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6959142572580448, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.467, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.7695011224043343, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.4637, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.7781393617335021, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.4686, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.6016159368333164, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.477, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.7325602634576774, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.3618, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.579918445686115, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.3828, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.6300933561194503, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.3577, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.7166897812510176, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4064, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.6134723000268245, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.3388, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6485054609510611, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.3566, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6086765543022894, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.3685, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.6436282196870952, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.343, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.6136703693399181, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.3627, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6597854461545477, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.4019, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.6976091101182352, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.35, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6012919125740401, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.353, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.7099386640811836, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.3568, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5779312698308307, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.3473, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6701143291441497, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.3661, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.5438415019228182, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.3618, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5882049949982814, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.3942, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.4701506172491297, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.3467, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.6001712253990339, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.3432, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5513168302131914, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.3436, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5922895209718582, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.3437, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6108707802859178, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.3461, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5231668478617609, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.3348, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.7410731078341939, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.3477, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.48288301746598766, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.353, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.7512257208564571, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.3887, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5287591533630462, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.3551, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.6514949600652048, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.3529, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5624863544283835, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.3419, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.566762007230502, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.3581, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.5277592599377072, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.343, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5116061122773884, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.3614, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4594112497751582, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.3548, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.4894830836405895, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.3599, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.5627245095789299, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.3465, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5580236130428585, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.365, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5464546445092676, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.3733, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.48289230111879294, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.348, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.48490496766157387, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.389, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5152479812331888, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.3745, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.52389539970339, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.3546, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5087714025796426, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.3496, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.596455971246047, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.3369, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.546204923669321, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.3439, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.6089651252091146, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.3858, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5123329944742488, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.3399, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5460750230352305, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.3433, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.4842152363327416, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.3504, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.4831525746417104, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.3324, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.47203152665523074, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.3363, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4667572772193624, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.3557, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5313968132068019, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.3377, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.4603288687508432, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.3596, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5704329499820529, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.3576, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.49024553172019536, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.3578, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.4441736094109462, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.3883, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5054012189818847, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.3577, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.45990759797573905, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.3384, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.5042972012753414, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.3522, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.47173263555770334, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.3588, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.536218080491284, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.3353, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.5291824051748952, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.3352, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.5388290949320104, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.3624, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.47325434101138114, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.3497, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.49540812516736116, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.3308, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.5034044255655133, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.3219, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.4712459982237337, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.3676, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.43988904870746354, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.3727, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4688533638844716, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.3596, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.4395990789411024, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.3511, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.4629286104729869, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.371, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4299981843603125, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.3613, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.4788981552771838, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.336, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.43615735210231366, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.3673, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.472358999942315, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.3467, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.4528381291272913, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.3604, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.44002931894325426, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.3474, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.4710557835531748, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.3419, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.5984188783781017, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.2575, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.43808210985272245, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.2603, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.021104609817689, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.2435, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6229112357572324, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.2395, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6400787003433385, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.2427, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5219847532700883, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.2442, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5514955488257288, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.2294, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5429649242477618, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.2328, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5566520824690419, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.2235, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5337400616346479, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.2305, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.48911001853812824, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.2234, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.6209324495822574, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.2322, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.45607628895768343, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.2351, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5553683422626101, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.2289, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.4504220323360159, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.2336, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.508634916908271, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.2238, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.44816802882428297, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.2222, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.42908742815813744, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.2303, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4758264147487167, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.2426, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.41247654890650964, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.2296, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.41642953802943083, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.2441, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.4437874842536521, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.2186, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.41784001269201265, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.2318, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.46721551739616907, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.2093, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.41112578910321584, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.2441, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.43979540887477736, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.2214, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.40544235255732436, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.2108, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.40184483388257736, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.2221, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.4308628597440365, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.2341, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.39298579507534864, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.2219, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.43188493844288345, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.2424, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.41130337697754366, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.2222, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.40246809462390837, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.2199, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.4088112010685077, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.233, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.3660759480240648, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.2209, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.3919129516116809, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.228, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.4077936034059262, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.2304, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.417985730697302, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.2573, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4227092312003486, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.2283, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.40115319247096887, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.2333, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.43510045143506826, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.2244, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3880884229030637, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.2461, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.40308061075230467, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.2213, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.42756912689781795, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.2206, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.39943092230727767, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.247, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.37344591996160337, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.2358, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.40046097239336376, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.2403, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.38476391536822985, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.2212, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.43288279571587374, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.2368, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.3700461308028724, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.2618, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.4616637161436342, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.2174, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.39035133902253477, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.2225, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.4615541847206986, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.2451, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.40812141024101817, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.2448, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3790021279522406, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.2268, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.41068999411679585, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.2355, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.36597102281558597, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.2389, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.41906748164633945, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.2338, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.3847727926837576, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.2418, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3602038255117451, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.2533, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.40242562598449944, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.2464, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.36537418924420284, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.2073, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3741508848048974, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.2416, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3625808111319984, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.2271, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.3692557062739783, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.227, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3828871524290137, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.2197, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.38151721911154457, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.233, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.3786750713015652, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.2316, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.36767050954329955, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.2174, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3838174394324037, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.2556, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.4122104081652899, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.2607, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.36751878868243526, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.2224, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.34514347989382804, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.2329, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3823271811821992, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.233, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.37643499161120975, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.2395, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.3987921136655855, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.2301, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.36187712283271656, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.2378, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.36431473793509095, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.2162, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.4342500045335589, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.1861, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.4129767033707882, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.1464, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.3312785368446801, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.177, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.4335543103057223, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.1635, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.4606606835139822, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.1459, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.40242659216003274, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.138, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.3335201884903501, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.139, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.3645495863663169, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.1445, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.3824301187556719, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.146, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3424946141123901, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.1397, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.33139230307805356, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.1349, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.3406508841684583, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.1474, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.36887333789876015, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.1641, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.34568515710581665, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.1456, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.32937953601397224, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.1493, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3358387850386913, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.1372, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.32369053977299683, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.1362, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.34777208110183666, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.1459, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3387384441271983, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.1297, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3305197022844186, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.1467, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.32825629222170216, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.1422, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.36275451831392513, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.1511, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.34376622589938244, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.1662, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3185289435275682, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.1492, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3122326664709875, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.1469, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3299569491313471, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.1467, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3080431216734959, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.1548, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.3078596243161126, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.136, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3192246593041759, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.1465, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3368616103541651, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.1358, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3026632461475973, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.1592, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3106375759183477, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.1301, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.31276027570498793, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.1226, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.2962955540289862, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.1577, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3057229986287228, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.1465, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.31143010677072214, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.1415, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3007146315051674, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.1556, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3028862641231386, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.1379, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.3012569121094226, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.1526, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.29537699389840416, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.1452, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.29288395833603187, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.144, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.2980050111798448, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.1493, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3072752930176877, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.1796, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.29972555622874003, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.1596, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.3002806996999677, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.1453, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.31498127367500756, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.1546, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.29673264593160975, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.1437, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.298778931606412, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.1213, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.30678453495493, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.1519, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.2959759682993311, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.1505, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.29411661056927013, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.1362, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.2892155768652224, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.1539, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.2795468650885671, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.1253, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.3785102022769812, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.1387, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3002694668553819, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.1397, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.29436747743610725, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.1423, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2952396977031661, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.1318, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.30243034458148, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.1579, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.2959833802980389, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.1395, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.29261738770695445, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.1459, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.27437485695959046, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.156, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.2852412068436951, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.1222, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.27664571131499804, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.142, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.29312076194262626, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.1388, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.29420685159823096, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.1356, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.283690838251792, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.1529, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.28242081034865174, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.1399, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.287923553335738, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.1455, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.2860752321041213, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.1288, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.28249363240206893, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.1313, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.27630088292909294, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.1427, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.29211186438565667, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.139, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.29642167879499903, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.134, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.28616488618453645, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.1512, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.286134692196085, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.1485, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.28160417320585635, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.1252, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.2732320650248573, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.1379, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.2890348891935754, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.1382, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.2658873774571514, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.12, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.28829467706893275, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.1113, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.2926462624152583, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.109, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.23217579454357423, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.0929, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.22632856588643657, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.0886, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.2697818741789059, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.12, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.2958760189814652, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1027, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.31239726698428777, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1095, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.30945783794605863, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.112, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.27042311557861265, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1211, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.250810164241747, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.094, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.22766150547994146, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.0977, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.25442781418258187, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.0886, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.24128634253047432, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.111, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.22809175946788002, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1091, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.22838394593196218, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.131, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.24123953281829802, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1065, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2268091319969786, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1019, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.22335799894308225, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.0993, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.21715114999813875, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1006, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.2104223361612852, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1009, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.2106637576403025, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1022, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.2263490487535301, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.0964, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.21501359524388158, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1009, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.22569777910551836, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.0807, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.2077323346879666, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.0933, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.21045077059465855, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1056, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.23590019433974949, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1014, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.22529326295060567, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1053, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.21547965472381023, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1063, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.22313890152734314, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.0981, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2272155930696546, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1042, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.22055667722703767, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.0959, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.22833080435593262, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1113, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.2125777220603205, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.0899, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.22182880401841537, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.0991, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.21638022460250375, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.0874, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.2210865190100104, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.102, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.21375159066828703, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1007, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.21618809403988168, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.0894, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.21470287491838058, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.0838, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.21983713612706263, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.105, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.2215884766371513, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1135, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.21053572943937773, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1011, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2338281134822983, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.0995, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.20823111592940963, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1244, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2061050909395005, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.0858, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.21148070237488217, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1056, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.21288461368812162, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1011, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2057873955256772, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.106, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.21635344730518735, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.0949, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2143613382155315, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.0927, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.20228998947621846, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1128, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.20482983920733092, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.0936, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.21953218401893168, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.0891, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.20244241996307016, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.0932, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.21391720405666895, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1021, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.2182333422260704, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1177, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2084304991983031, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.0955, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.20795283461277356, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.0936, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.21861030107083146, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1057, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.21161038554212513, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1217, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.20071896496622085, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.105, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.22336813996118293, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.0869, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.20626212361882176, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1049, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.20765922740991494, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1028, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.22275529422271872, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1009, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.22135908885017214, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.098, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.20830230664411545, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1017, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.20226748336287717, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.0889, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.19992620933341473, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1047, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.20269374334136603, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.1073, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.2081832474041765, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.0903, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.19899878263209597, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.0797, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.21031388981769522, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.0867, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.20703163671053085, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1016, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2062387617505902, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.0865, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2122944755587596, | |
| "learning_rate": 0.0, | |
| "loss": 0.1011, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 5.358630760129495e+17, | |
| "train_loss": 0.2751285242537657, | |
| "train_runtime": 10768.1535, | |
| "train_samples_per_second": 4.643, | |
| "train_steps_per_second": 0.036 | |
| } | |
| ], | |
| "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": 5.358630760129495e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |