| { | |
| "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": 6.38768377747976, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.004, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.258582313030617, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0257, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.985665520366591, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.9489, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.636517326297908, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9904, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.264245706101825, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9352, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 1.9991275954117318, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.8521, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.057063330139481, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8603, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.532948301348326, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.844, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.571648374378108, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8515, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.6324718694241405, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8523, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.130856478596325, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8306, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.9058328201571118, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8042, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.234559856660955, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.816, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.3838956526730426, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.81, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.15359315430978, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7706, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.7168249519226755, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7386, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.5429335779959272, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7777, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.6836537794679654, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7417, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.1986217048225807, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7038, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.8328351314633629, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.6885, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.129318957914725, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7142, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.8726514906512445, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7203, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7950663565026697, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.681, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8479346405772735, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.6923, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8236586555415657, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.6554, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.74305816516341, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.701, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.9208609397548848, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7066, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.7185465463351426, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6628, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.7320722154302732, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6494, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6246124448211506, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.651, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.0119758481712162, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6652, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.9370832491719673, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6839, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.8874652928362486, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6507, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.8218859501574677, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6648, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.896187514045983, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6458, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.9140506258058744, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6604, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.0638047121900094, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6635, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.8976757040110662, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6469, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.8057543517015057, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6191, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.8599088801840852, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.658, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.8708690765175714, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6547, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.9896097866870105, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6275, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.4058881842156088, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6586, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.7383665740455885, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6536, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.794782140021188, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6087, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 1.065876812121002, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6517, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.7566982782534306, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6114, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.4267330300306866, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.5918, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.90298044852148, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.611, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.9849133136885599, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.66, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8564703727939461, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.5984, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7422660355165284, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.5912, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.7945735365284259, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6023, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.8427619808084357, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.639, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.7471610677291403, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.5753, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.7820386023783767, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.5996, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.7240533372642097, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.5905, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.9415977690764906, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6139, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.7366415923153254, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.5765, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7770006741134806, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.5859, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.6291728961342121, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6061, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6536187481241915, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.5908, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.7747275898784713, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.5811, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.7006886533723703, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.5788, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.6678569510865513, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6027, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.9846446011192492, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6006, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.7452904561114078, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6114, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.5639336370080911, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6084, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5809946766285476, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.569, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.7107793036378137, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.5628, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.7725284572327248, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.5763, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5134240280695541, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.5748, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.850068048768358, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.5786, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.6381550007515633, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.5893, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.631194997702762, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6189, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.5333608293105327, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.5563, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.46618416608244967, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.5667, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5399433985418359, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.5785, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.422931954666897, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.0627, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.8620793988029459, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5006, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 1.0035480157754244, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.4876, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.74402384642657, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4616, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7504227542152153, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5148, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6481248626292647, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.4821, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.7933414266757354, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5377, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.6168555615356381, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4965, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5647770121915517, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4907, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6373697282115588, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5192, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.5838810555646947, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5397, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6618716616735669, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.4778, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5745891026658267, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.4827, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5742857725061112, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4619, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6282355292322966, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4463, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.5586229268083559, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.509, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.6182409635180593, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.4892, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.5842379557897653, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.4891, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.6407435080385205, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5176, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5674851168549228, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4388, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.4370373798690236, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.4693, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5555940899415323, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.4869, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.48386259353958805, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.4818, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.46923320840121546, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.4506, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.49154953920710065, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.4825, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.5085110324574746, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5241, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5640622239731405, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.4872, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.41969289226372886, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.429, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5790215744520885, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5159, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.5535635643061567, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.4507, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.4914302748908893, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4962, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5160847268601544, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4585, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5510544141321211, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4791, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.529156989716788, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4818, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.6821361694520031, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.462, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5364828341298498, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4875, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.6059726318911631, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4916, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.4639785382956982, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4715, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5617164707304216, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4854, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.4509341497151774, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4491, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.5581971345476343, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4761, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.4396554872502241, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4636, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.5081551745282894, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4818, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.44030942475593393, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4695, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5225320694345398, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4987, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5382884415972378, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.4872, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.425755549680754, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4704, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5869377331053153, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4844, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.4054849019509311, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4732, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5864955718853877, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.4979, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.5007082914816513, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5059, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.49989617883471166, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.4766, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.5553023837913968, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.4744, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.4751304849684561, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4787, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.4860406824652901, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4831, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5529779099992289, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4506, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.4741922916577771, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.4895, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5955401552539042, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.4951, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.42728461130766043, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4468, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5586066599747623, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5018, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.5881918814636677, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4617, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.4606032806807354, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4606, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.6327915546377536, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.4668, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.4231999730033915, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4859, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.6503532899965093, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4792, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.41401273422599477, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4384, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5932441783689704, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.4572, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.510162526703894, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5161, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.5015515229501777, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.4333, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.493261895501617, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.4606, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.4975054320068755, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4841, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4705100405972777, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4927, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.44279479903389063, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4781, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5175415204497199, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.4747, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.41094207009488304, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4465, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.5717374553846247, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4543, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.4454540657851889, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.4652, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5789274333136252, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4956, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.360932732047883, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.8455, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6471678885369845, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3698, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.1010801876503231, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3426, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.5478671012752752, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3689, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.5647377087172448, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3663, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.4932489836945577, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3416, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.48540205841811135, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3315, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5174807795622552, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3567, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5333256296520831, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3776, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5253754891439574, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.339, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5078262413294057, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3328, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.47832057264897254, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3531, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.49992512080829216, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3617, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.47199382813481866, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3189, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.47350431538808124, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3563, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.4460489663068142, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3469, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.43926934226876424, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3395, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.48827449877985246, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.363, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.42938341330249835, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3422, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4903673722866839, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3454, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.467933661099253, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3492, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.43543008536315403, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3614, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.4787936572127691, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3429, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.420329638247727, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3504, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.44208451416066646, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3506, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.4090870133819708, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3512, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.43228975998840213, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3424, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.40959568274452934, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.352, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.3928804162298462, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3651, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.41627156367200685, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3457, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.3732016518116201, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3635, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.3786193356769031, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3526, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.3605379826388927, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3435, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.369271052869712, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.341, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.3608832245217376, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3361, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.37645448391339553, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3481, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.3849448155598022, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3405, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.3740822625676976, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3427, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.42673177393558626, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.351, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.3879009366793217, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3593, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.3736431468051782, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3185, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3743315922529099, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3661, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.38029424649360377, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3318, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.42963098647316594, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3533, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.32383838068901993, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3096, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.4294006522277665, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3481, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.4153108193767605, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3377, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.4108555799338753, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3543, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.41215911294084356, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3701, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.41123561764851574, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3189, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.36969385722617637, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3414, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.44684352433144386, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3476, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.39888336849866735, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.336, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3819610470739442, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3683, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.4102809879095424, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3548, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.40309756034069766, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3378, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3895288547606462, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3621, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.4138524427556682, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3509, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.3579982829228515, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3329, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.40158731068860537, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3537, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.3736077145069081, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3474, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.35300898773848505, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3404, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3664482604122401, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3411, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3590675967431455, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3593, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.4353815456208337, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3513, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3527667307642299, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3367, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.40451821487702677, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3689, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.3754056230841885, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.345, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3845662165065078, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3401, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.4189870034072437, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3692, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.35362464659061366, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3444, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.4310174137299933, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3559, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.3894304862386695, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3454, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3731857763462173, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.327, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.4168830200691077, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3704, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.39468531831818743, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3514, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.422985351532148, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3726, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3546397421378005, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3324, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.1211536886700337, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.5717, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.6637458119142641, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2746, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.44859599854710724, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.247, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.661806803341597, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2284, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.709056754631819, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2511, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.479108277772952, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2323, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5116102551817895, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2494, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5083714200894768, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2473, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.49421493660307414, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2488, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.44588300102481987, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2418, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4713645567999502, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2551, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.460787088650932, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.231, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.4329860085506179, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2392, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4709317297123745, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.264, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.34935519044393326, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.212, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.44598855165303875, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.258, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3880171073524762, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2288, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.36853923434398084, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2218, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.41258110155075095, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.24, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3900315714021956, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2223, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.43483326419032314, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.261, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3610601655069623, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2229, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.3729337203613411, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2525, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.38337277032500255, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2254, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3831118500178122, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2353, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.34612571137446324, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2351, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3172633882755389, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2084, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.370021154979822, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.235, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3651904526968984, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2404, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.35069860922593044, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2404, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3410123354817081, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2275, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3473695966967474, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2434, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.30701856753062895, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2048, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.31597605483277935, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.216, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3500971321787639, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2503, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.33178772742723, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2345, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3301926698347677, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2212, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.31542914606528993, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2097, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.33330881894800485, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2506, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.3269559095476107, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2389, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.3331865625735446, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2356, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3172602993531627, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2235, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3346327892198248, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2324, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3334148992149954, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2316, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.3130925147363956, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2278, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.29688822252292935, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2234, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.31274157469020114, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2364, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.31728363817045446, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2353, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.29493005871684974, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2277, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.32353310506678185, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.244, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3137215209736845, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.228, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3386752499736865, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2521, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.3018971527797994, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2332, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.2942932481229599, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2387, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.30570306552033083, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2313, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3150399423985604, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2224, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3302221935857157, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2342, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.31559945847084847, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2255, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.3044828982592592, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2175, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.3145425735544188, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2418, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.29834615104068185, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2198, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.3216821139788341, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2301, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.33195570998632884, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2382, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3316870845486162, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2477, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.3354017771659439, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2146, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.3486277525123898, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2503, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.31306817043389246, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2384, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.29664784914593983, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.209, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.3072113602175979, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2367, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.29414132060912374, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2228, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.29699231892419053, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2186, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.32340472986258056, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2355, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.3152501702677041, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2567, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.28718573938391323, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2299, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.31532727100852215, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2235, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.31063009806949665, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2373, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.3207662021391083, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2626, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.31871439054772693, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2344, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.9242806837654529, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3773, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5686899090416707, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.1913, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.3890083028289445, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.149, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3476476198201339, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1697, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.3252567027969146, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1775, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3717707797334998, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1667, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.37559132763056546, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1564, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.42192082669881686, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1661, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.44180694742724125, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1635, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.39292923497954124, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1707, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.3050392843828178, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.149, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.2983434254450389, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1846, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3277691113749579, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1787, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3023723459089548, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1523, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.32515601362966307, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1626, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.31907547477819487, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1745, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.32582728737410815, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1598, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.31951853667746816, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.174, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.2808304063449876, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1646, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.29307269023615706, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1788, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.29717587639416376, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1741, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.2865082227983273, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1575, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.29553273299453664, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1662, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2825658792371876, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1595, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.2960530087605503, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1632, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.28205608003425314, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1594, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.30770895506647983, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1723, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2798940588335153, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1517, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.28890780352817813, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1586, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.3005271114241149, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1706, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.27077877401227346, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1572, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2799499103736573, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1575, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2793246253072625, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1686, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.2625774919158723, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1524, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.28844056061286116, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1794, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.25248987761145386, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1409, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.2804170341420727, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1699, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.2928328928542396, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.166, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.28082568388356005, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1668, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.279628601967676, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1684, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.27366214292709373, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1572, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.27029648109480015, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1576, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.2787711390722436, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1739, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.2625638005310226, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1538, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.271939304357851, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1595, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2900279321380369, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1727, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.27072106936935053, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1561, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2696615937729465, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.172, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.26997226538815267, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1578, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.28301614351997717, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.171, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.27677289683858325, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1677, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2970339062323758, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1814, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.27154404470071625, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1531, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.27639502951849626, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1663, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.2779753042921285, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1637, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2636730450543714, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1507, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2714984339138379, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1665, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.2889073676404897, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1758, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2587215548660513, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.15, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.29335272620537145, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1742, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.28706317546883137, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1763, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.26370212776520546, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1552, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2763149719603404, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1643, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2597121252966565, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1643, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.2750346651571479, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1804, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.25716822704273445, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1546, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2768725219310522, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1631, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.2693713151128825, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1586, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2787290188273803, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1714, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.27005182332616406, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.155, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.27027417423657907, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1518, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.28504678673425315, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.1737, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.28379595945792574, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1632, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.28401859737499763, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.165, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.29913916097540943, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1734, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2616552661503292, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1389, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2639369469741811, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1451, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.26497802910691404, | |
| "learning_rate": 0.0, | |
| "loss": 0.1618, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 8.107518503315046e+17, | |
| "train_loss": 0.3858068761153099, | |
| "train_runtime": 14601.7781, | |
| "train_samples_per_second": 3.424, | |
| "train_steps_per_second": 0.027 | |
| } | |
| ], | |
| "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": 8.107518503315046e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |