| { | |
| "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": 6.26086426888324, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.8129, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.306126196946495, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.8096, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.049902944307282, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.8005, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.581033978240449, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.7907, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 3.942080964560779, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.7535, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.2720108435500013, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.7213, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 1.9957489865732976, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.7091, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 3.9880184178313756, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.7235, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.134129876470508, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.722, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.040723427362194, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.6764, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.511278950630208, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.6641, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.04652322868078, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.659, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.700051238383732, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6418, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.083434689509928, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.6139, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.5750836960749037, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.5965, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.0469056582986285, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.576, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.1108288971121183, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.578, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.9744239138790556, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5675, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.9220896822862445, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.5505, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.8128162224874632, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.5658, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.793777177297142, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.5386, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.8819542871590188, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.5535, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7112160131811152, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.5345, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.6848683941727483, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.5245, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8308277413611194, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.5432, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.7166377631078487, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.5338, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7246735587689199, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.5168, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.6577215931633504, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.5133, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.6714903521175072, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.5197, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.9606729922520589, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.511, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.1045233118646731, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.524, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.5562721421946714, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.5087, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.1570687322980913, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.5183, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.9992075232151465, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.5063, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7821336226559921, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.4987, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.0958653690079694, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.4899, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.050726072651429, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.5098, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.8030844999253013, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.4903, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.8388731052947013, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4972, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.8604621256633986, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.4914, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.8801076063495301, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.4835, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.1413703134026665, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.4959, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.6061203310398915, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.5122, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.9893197541434355, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.4954, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.924107374836722, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.4792, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7958395428756192, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.4721, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.5976463856922977, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.4991, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.8891228069914984, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.484, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.7486579775074256, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.4806, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.689950390874166, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.4761, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.6635298145139862, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.4813, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.5756346897629063, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.467, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.5887287924193134, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.4662, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6513157378202401, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.475, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.5410036701444721, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.4777, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.603332014727978, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.4681, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.4735325755723473, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.4672, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.6427613435338064, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.4755, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.5537027209490739, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.4901, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.6630289443889019, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.482, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.6544428610710412, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.4707, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6409838460131501, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.4889, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.488579279900872, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.4589, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.4974794335870708, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.4566, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.44850050607551334, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.4695, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.4318026593887461, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.456, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.5280679906275694, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.47, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.4740963366928391, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.4646, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.7768935267731776, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.4745, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.5887497688015558, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.4642, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.43778520837922263, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.4848, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5363121760846471, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.4616, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.4230201233797813, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.4613, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.44963541176853844, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.4805, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.4121600740410282, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.4647, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.40497074659042975, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.4671, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.5380169210181003, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.4505, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5437647888231878, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.4671, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.9976388816566887, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.8213, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.337624475987189, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.4323, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.5579400847247054, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.4214, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.1458289822955747, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.431, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7446181410102658, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.4159, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8309838581252051, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.4366, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6963341993849291, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.4214, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.695419039133152, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4271, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5481189704661072, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4251, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.5768669301290407, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.4337, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.468541931616848, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.4082, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.608239023782051, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.4314, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5322706943724788, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.3962, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5340730142381851, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4208, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6875198256217474, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4374, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.42963225384243486, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.4023, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5721832511828568, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.4258, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.545262021457018, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.4112, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.469053130863097, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.4216, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.48626971536934105, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4179, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.41745533751099595, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.4092, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.4823367072462575, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.4119, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.34984497917575696, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.3963, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.45972763708591186, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.4267, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.35980745963881966, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.4059, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.3720153536752175, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4266, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.37143311579857646, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.4031, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.3601009232400128, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.3988, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.4223113010417236, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4209, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.5040612427890517, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.4262, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.4908426087725137, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4092, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5194314432651223, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4211, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4463945773201381, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4101, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.5058488897063651, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4145, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.3697744664886333, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4237, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.394330733943469, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4037, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.32882807790446394, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4186, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.36758690369630964, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.3971, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.3169830964789753, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4028, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.41625257503517865, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4027, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.32806172248556953, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.406, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.3990352735831593, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4184, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.3885296426251057, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4084, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.41972015625197684, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4028, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.45111807851415764, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4146, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.3446066909528985, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.4177, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.3886771179276535, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4169, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.3909950788862754, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4122, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.4291465946905848, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4339, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.310377815108431, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.417, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.3959938592783724, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4026, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.36306977924576295, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.4228, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.329711144386943, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.3952, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.3311456262958043, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4188, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.2993116141924671, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4106, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.3524552286579906, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.392, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.3348047451882289, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.4113, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.3393922814496209, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.419, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.34158527336617184, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4077, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.33903515559921366, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.3928, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.33763227417315095, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4125, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.4123593762432786, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4179, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.32949267792081377, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.3994, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.3541120831264513, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4136, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.33322735007871573, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.3936, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.30983332956750553, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4169, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.49879144333474923, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.4121, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4646713382478522, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4118, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.3104966361932239, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.4105, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.3583676948729214, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.3934, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.33092514151217584, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4141, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.3947621380345138, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4155, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.33405675959805153, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.404, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.31026534662599714, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.4238, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4540129758975968, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4171, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.3611934451004751, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4135, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.3191263302244131, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.4207, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.371887726964349, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4149, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.7661076195189961, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.7209, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6756175882724555, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3625, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.6013233843098078, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3635, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.49070675239840633, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3532, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.47594570931276653, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3561, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.518454166439218, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.36, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5047683183963112, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3542, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.45110241464652145, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3584, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.455010567366992, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.346, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.43419897690106, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3487, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.4086284973931011, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3519, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.4545475240907437, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3523, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.3932042397039675, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3689, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.49011922723355117, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3401, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.42452784372905494, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.361, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.4875456809152096, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3481, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.4600757798725378, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3475, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.43748332283638214, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3604, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.38758759987248037, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3489, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.34231617556230487, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3543, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.3716159784904776, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3458, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.342666147689472, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3564, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.3846729531167565, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3568, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.3205900109984895, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3412, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.36184027443570566, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3509, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.3093073268958482, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3432, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.31667158807142654, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3494, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.3392100712415189, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3471, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.28699959427384886, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3544, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3336275236602359, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3452, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.2804343842386502, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3645, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.31518104010616965, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.343, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.27996376207708157, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3569, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.2965769893905253, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3329, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.26317626608265243, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3365, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.29154412541317937, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3669, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.2798990365906198, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3499, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.295446446891741, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3518, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.296953241435807, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.357, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.27160667329712757, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3389, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.29400036593157836, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3581, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3096705333266336, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3453, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.2841646159899552, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3444, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.29925039811502735, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3562, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.27848935671309805, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3441, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.4558022364402173, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3388, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.27513723883526936, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3474, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.2807643317471773, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3597, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.3015119534908792, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3534, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.2583371811112862, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3429, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.280270636158833, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3441, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.2658476030961622, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3521, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.3080659016298943, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3526, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.2645838054710768, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3574, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.32375249436467296, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3693, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.2509414398523086, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3409, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.29343084500718414, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3471, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.2537999172260796, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3488, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.2722879111842384, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3604, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.2533663517141614, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.358, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.2385568309371461, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3475, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.2784557914807277, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3546, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.2626932688074936, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3477, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.2694769209282753, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3561, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.2543362748656524, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3556, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.24736750387840983, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3546, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.23127142543154042, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3609, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.22674444955989653, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3498, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.23290060279978336, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3457, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.2549748694355307, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3426, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.2262553721650218, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3599, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.2346560857354473, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.35, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.2408055930303229, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3384, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.26371867879882593, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3452, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.2835669238821544, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3456, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.242703575638523, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3528, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.32772191884210533, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3578, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.22839292378297754, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.349, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.6845755903569339, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.6036, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.4274811270142624, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2857, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.40058256768267225, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2999, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5490842811691296, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2971, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.38933804326080973, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2906, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.39727802749489594, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.292, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.4062541863689099, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2989, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.3949101758171211, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.3135, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.3037084889514547, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2915, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3710360244499918, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.291, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.33312464166417016, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2953, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.29620888257421235, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2983, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3503208824510386, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2945, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.3049700287136054, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2877, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.275436406656346, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.3044, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3458038950891842, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2902, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.24871517556891068, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2841, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.28555004331414563, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2993, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.27585435222398164, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.3023, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.2813540945801244, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.3035, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.2507080815755503, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2928, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.26689371993105343, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2968, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.24545635079682496, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2869, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.2451712206741302, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2933, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.2635687563890725, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2993, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.2377811570030084, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2901, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.23514220407784397, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2998, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.2622686706972494, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2934, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.2131634198520768, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2884, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.2135494506893715, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2995, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.25177194921546125, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.3009, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.214612186988692, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2867, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.22063711931034696, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2832, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.2436271406012483, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2922, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.22062891628128503, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.3015, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.21344146156440286, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2867, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.23100049601921757, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.3004, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.214101439545734, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.295, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.20794476614303853, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2918, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.23018049557158482, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.301, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.21095144272789018, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2918, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.22104754466348284, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2983, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.22182522555571718, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2902, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.20304197572956137, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2872, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.22926371651418037, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.286, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.20594738414878283, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.3014, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.2247319859188061, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2939, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.21839262953103478, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2898, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.21010571904360598, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2941, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.20499647273214588, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.3016, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.20133132335608683, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2867, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.2054724813277891, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2985, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.19620421805072635, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2899, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.20071876753426632, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2995, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.19485234179402647, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2883, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.20808018827311356, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2977, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.19849511942767112, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2967, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.19655958764733017, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2874, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.19920714072720627, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2992, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.20218706805831987, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.289, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.195678207037672, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2965, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.19674347574842688, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2936, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.21097636511291962, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2991, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.19480805276821342, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2929, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.1998821977604018, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2967, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.19307124742011486, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.295, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.19837949300665553, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.3029, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.20064408017383228, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2967, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.19757642343325554, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2939, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.1904462205942121, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2903, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.1951297142788012, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.3045, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.19659784207414205, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2874, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.18270222203561715, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2899, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.1872369922460286, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2953, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.18916624601200407, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2926, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.18663346424419813, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2914, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.1832307325493174, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2873, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.1995855841053093, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.3036, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.5517743863537016, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.5558, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.2923387972764061, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2564, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.25012589459497103, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.2547, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.22164350858619136, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2611, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.2382447257522135, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.2625, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.2913127722951209, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.2646, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.29877340888722387, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.2612, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.2751712256733569, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.269, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.2095888709785107, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.2573, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.22422092339217775, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2617, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.24820103669858176, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.2651, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.2345049271144045, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.2615, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.2177580501613954, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.259, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.1990219075957119, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.2523, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.2030911641348855, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.2601, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.205832985975619, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.2509, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2008565099091807, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.251, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2422687558295086, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.2572, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.1998308955194785, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2575, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.20415272918531643, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.2645, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.18343984703646293, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.2427, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.18763363252615764, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.2687, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.19523006276372218, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.2611, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.1857641913701162, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.2485, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.18496102935605904, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.2623, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.18534231107502755, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.2592, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.1866155428365046, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.2517, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.17737753594146918, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.2548, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.18397130545364818, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.2536, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.17944405075028114, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.266, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.1779091949018619, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.2558, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.16867706493615456, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.2491, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.1734551720399595, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.2569, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.1764926364905321, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.2539, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.18984736887209516, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.2613, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.17626489856720834, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.2584, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.17076698753063432, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.2581, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.16960163130210817, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.2508, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.17490550833750718, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.2618, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.17704349735274724, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.2512, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.18212194484772012, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.2574, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.17477010769672657, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.2531, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.17209174695788668, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.2565, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.17690735792287773, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.2606, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.17177068214386296, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.2504, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.17112327427567636, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.2556, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.18234932025829617, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.2675, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.18478620491241385, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.2628, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.18593627419226674, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.2657, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.1710709562845122, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.2532, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.17942533213998382, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2632, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.17280219162209093, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.259, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.16969765450973692, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.2579, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.16776314571978587, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.2405, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.16931866093976752, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.2606, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.17320370817469904, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.2606, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.1710630432846978, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.2585, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.1700250057574657, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.2528, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.16546593226024825, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.2493, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.17615131431186276, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.2701, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.17109240058864536, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.2596, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.17205662981023936, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.2601, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.17385098373464983, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.2587, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.168280815451522, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.2578, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.16552107694374615, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.255, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.16739008008249578, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.2534, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.16803972466872988, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.2592, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.17196163587901453, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.2521, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.16777309745457525, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.2628, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.16928680274151056, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.2524, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.16791362511268418, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.247, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.16991628005399909, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.2688, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.17487631947964238, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.2552, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.165727453833623, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.2548, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.16523828658109455, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.2432, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.16779266015284725, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.252, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.16292862292350446, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.2493, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.174807807958931, | |
| "learning_rate": 0.0, | |
| "loss": 0.2659, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 2.0551971820588237e+18, | |
| "train_loss": 0.37441407300722906, | |
| "train_runtime": 60245.2785, | |
| "train_samples_per_second": 0.83, | |
| "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": 2.0551971820588237e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |