| { | |
| "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": 6.127174296898396, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.9401, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.279379818545303, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.9578, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.155471949075422, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.9546, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.509783786725042, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.914, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.158513241148331, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.8715, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.2131358994115065, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.8466, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.226050367652273, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.865, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.577771937096163, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.864, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.623167599675659, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8457, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.527050160700857, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8197, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.102624716762441, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8136, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.956488161214038, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8076, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.0439152252278543, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7803, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.1685940476683236, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7556, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.0240689785171835, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7503, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.2928819373475366, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.729, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.4164456864907744, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.69, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.3725699793149424, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.6951, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.8366753532095268, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.6855, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.1465538133373123, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.6849, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.075307632674618, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.6613, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.6988684299770829, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.6747, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.9714493787718601, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.644, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.891985919303642, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.6657, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8161224477015689, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.6631, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.0138032065364415, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.6376, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.9322543246372766, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6417, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.8792778669615909, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6299, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8516791165400859, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.641, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.910096877131859, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6493, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.6809486267821716, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6171, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.5712595836249316, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6441, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.776294987255976, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6543, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.8338735844134413, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6243, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.043035411580158, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.658, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.2883476307701132, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.5986, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.9775428177217712, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6291, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.9221414688181516, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6161, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.1674233467109725, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6394, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.3151827529333766, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6209, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.6602875158956247, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6092, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.209693597884445, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6044, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.8564978626724826, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6059, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.9667176371713958, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6244, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.3038404950583953, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6197, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.5776512362583406, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6087, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.1299771026828123, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6241, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.8825330302703512, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.575, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.6220734458675133, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.5792, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.8941924474805705, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.5789, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8515287413799947, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.5744, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7295739246999864, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6147, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.9896340612448432, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.5825, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.7968731296372663, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.5982, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.8473201190819475, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.5867, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.8522669928053676, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.5933, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.6642523499928026, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.5895, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.9447909899711103, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.5963, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.9354868335481175, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.605, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7619663125967607, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.5866, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.0314012712506515, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.5885, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.9151236783882004, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.5865, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.8471227870081557, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.5954, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.9364025281474174, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6007, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.6239647771353921, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.5804, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6987882793006224, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.5679, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6536222274331626, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.5793, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.6108070652568574, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.5773, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5376329249639472, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.5708, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.5723695983120836, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.576, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.5612489546722549, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.5863, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5279573298741739, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.5663, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.5456972413498884, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.5743, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.5920011963336029, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.5727, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5924137736201102, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.5906, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.5315159345060548, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.5515, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.5448130385508234, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.553, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5627766543735616, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.5681, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.0902314721704474, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.9397, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.6764273566026807, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.4893, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.6989366464344922, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.4939, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.5984712308926659, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4892, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.6611981019992125, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.4709, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.7119304280321455, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5107, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.8341639299609943, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.4775, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.6013008278056612, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4953, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5325351986332285, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.491, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.591574729073965, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.4899, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.4564319198190788, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.4836, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5921171817193938, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.4838, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5735661374119054, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5133, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5038728311934271, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4914, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6105668396474834, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4959, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.5583635899080047, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.4738, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5675910864314527, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5364, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.5246000930891987, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.4513, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.595775568701185, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5093, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.537625057114193, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4664, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5457120820598083, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.4875, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5788209639971891, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.4852, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5027757411791625, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5046, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6151648484872893, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.4737, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5108975525449087, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.4995, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.48465673501205997, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4923, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6144523725284485, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5119, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5134103664166014, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.4821, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5169708574859718, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4947, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.49399427602513046, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.4575, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.46830723990650547, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4675, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.4892948602155665, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4657, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4659705867934426, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4816, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.5288822331140456, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4904, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.4550231453001598, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.495, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.4709314253174152, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4588, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.4841839351729488, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4908, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.4927616675346394, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4944, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.42222016941763296, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.481, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5304014968907881, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4935, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.4634070395162, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4715, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5005449249150709, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4725, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.44203890512282507, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4933, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.5905462317961483, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4992, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.4700179235687992, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4489, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.6214862613565733, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.4647, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.597441227689741, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4885, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.4421703641594682, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4449, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.548632936519922, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4739, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5182477478345776, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5092, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4693832055404559, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4681, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.471134772511439, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.4657, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.47061727518824764, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.4853, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.4848609542492148, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4806, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.47600708141334247, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4648, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.49874618749657196, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4723, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5010820817729489, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.4864, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5282517919041391, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.4975, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.49497354368495916, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4767, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5349874491169585, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.4799, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.49601914385662216, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4515, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.48539903845418003, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4719, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.5773037415921252, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.4841, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.423423992961836, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4867, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.47315912783747477, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4894, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.5194064604111779, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5029, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.49255881807307733, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.475, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.5054040224736764, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4798, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.43965443597667686, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.4633, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.5403306855152514, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.4727, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.41640296154348905, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4613, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.43306782184014503, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4811, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.40691180534367527, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4757, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.47862874633665475, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.4699, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.36734100256322283, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4548, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.421387281458273, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4555, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.40202318860049147, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.4781, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.43918747499441857, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4829, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.06832201525022, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.821, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6724018505946346, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4022, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.0042520122721768, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.376, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6133578570679781, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3644, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6460706067220617, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3915, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.6480002604849123, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3861, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5165650783628862, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3735, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5363333007643715, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3647, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.49022582267112813, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3727, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.568876932918584, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3788, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.43463309897296404, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3666, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.6232506289851566, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3898, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.459260499681647, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.352, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.6764577581605399, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3827, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.615921639341612, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3837, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5249698208301845, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3444, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.6832518229168759, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3868, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.47719474970385534, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.366, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.6329331040959504, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3503, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.42975162752038704, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3595, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.5457765192704426, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3722, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.3929063944227969, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3652, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5149380999467884, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.396, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4026415666669856, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3369, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.45999864376722943, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3901, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.4227786022400028, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.359, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.39797378073329903, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.363, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.4647950351361235, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.381, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.42239621392148774, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3611, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.407781124975184, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.361, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.44392600912486946, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3762, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.3646776145436372, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3441, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.5255634704631494, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.368, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.42564679382886716, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3849, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.578665353174517, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3813, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.4046458262020165, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3509, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.4726822644393134, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3632, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.3562414066887474, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3573, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.40819579888574514, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3733, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.4105508676422874, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.369, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.3181482279033896, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3404, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3940515584001992, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.353, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.3467294125070774, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3688, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.3668643315480313, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3612, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3384871548085499, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3468, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.38118190078359954, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3687, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.37474949276303504, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3516, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.33667215402526635, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3465, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.37254183613465053, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3798, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.4105441694878146, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3643, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3498792111172931, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3742, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.36295512241551803, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3668, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.3250404395022325, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3693, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3660677675160616, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3813, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3223461134222816, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3504, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.34107872961635605, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3624, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.35004983980353893, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3496, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.39332811631204284, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3665, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.3818260459787085, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3452, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3565781765971555, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3608, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.39311904838580175, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3981, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.3576884830028828, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.338, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3594377131877797, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.379, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3829523748719109, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3748, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.34298823928470473, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3607, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.34470539279946205, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3715, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.3804844743282432, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3833, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.3364382054676476, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3601, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3675352417557565, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3968, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.34263794301820377, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3681, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.31932729135389337, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3507, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3400335543239311, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3776, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.3381815734880397, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.363, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.34354519780372805, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3657, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.31356930583202386, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3442, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.37170955532511163, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3737, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.3302289228219894, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3667, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.33246142971455217, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3572, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.8853626925102991, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.576, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.5901002024383395, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2917, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.40226231376378496, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2593, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.6066754931930282, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2685, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5805612101272846, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2881, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4390136309012013, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2852, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.438962286276724, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2618, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.48424410983628874, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2938, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.4038759158058007, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2621, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4138478148100944, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2622, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.42511018970273834, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2716, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4018571050689614, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2493, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.41843664105107803, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2564, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.40081945454269785, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2507, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.4129960546358239, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2779, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.367376792281747, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2516, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.35990453047844667, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2614, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.4012765867962089, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2711, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3605843047167425, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2558, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3393870255256422, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2519, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3535585436286376, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2586, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.37087144305949266, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2796, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.33385391792456376, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2439, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.32504231557591323, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2601, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.34032580542335433, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2643, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3524199801981708, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2626, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.30132739958164056, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2503, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.332931058952257, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2674, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.32849290437528916, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2486, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3222048116470862, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2764, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.32785670952434887, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2726, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3139594557638986, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2556, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.30921240413514806, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2539, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.31618979646024, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2575, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.31916143258651974, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2711, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.29785051214795993, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2308, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3260773998127286, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2749, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.32499487939835675, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.263, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.31580529073999697, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2548, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.33692159562082324, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2896, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.29675803788412397, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2445, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3258469854979281, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2758, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3145183465597878, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2624, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.31088990954372586, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2499, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.3000757430808575, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2588, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.32017218220503235, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2506, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3549542941067524, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2938, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.29955082134664734, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2594, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3195729470021528, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2705, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.3186042554322007, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2709, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3111145679070632, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2646, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3048849234764353, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.261, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.28143811383155437, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2483, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.3295140493451359, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2887, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3065016740373834, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.257, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.2863651821230842, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2559, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3076244960601254, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2697, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.2974747768765269, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2721, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.31500613410714157, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2679, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.2774038101771903, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2652, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.26690190725725543, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2635, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.28114509105592944, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2692, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.2754785161961236, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2486, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.2828250010986516, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2529, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.27800116848401785, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2497, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.29332530125352657, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2735, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.29780402504597553, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2604, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.2761185699255731, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2318, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.29074758438933623, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.268, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.2882808854217824, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2562, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.2863347306288256, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2599, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.2799071854214936, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2608, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.2885656891912336, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2539, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.28377597656621756, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2761, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.27897989961149744, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2796, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.2834032213775887, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2696, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.2761313523146619, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2442, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.303535749195125, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2823, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.8536769014913224, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3821, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.4722327700895276, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2149, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.38498921836633454, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1959, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.30451487740157634, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2084, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.2924594976858482, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.2093, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.35682821499451467, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.2063, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.4271737714187748, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1911, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.5066048405599893, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.2105, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.3928450654648605, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1986, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.33072122160049067, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2056, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.2855084779262976, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1941, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.32152219694546025, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.2091, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3197418593885838, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.2093, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3156357790948045, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.165, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.35129065972546764, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.2084, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.3046705953972399, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.206, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.28919818997417274, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1958, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2855802471694086, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1888, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.29392542297891233, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2019, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.29978926225855324, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1973, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.298975927568007, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1948, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.3319220811331188, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.2079, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.28476666967796976, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1936, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.27737826122747106, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1965, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.27844826735490935, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.2137, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.2770173756074299, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.2122, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.2740441004409702, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1796, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2771992999909498, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1981, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.259550556848992, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1849, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.25598611684819167, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1859, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.25381206306348647, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1903, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2728923312673308, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.2134, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2636061023029146, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1874, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.25041599207350423, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1896, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.26014568249550973, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.2024, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.25069008167030693, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1976, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.26046478794005623, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1937, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.254565560290067, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1983, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.27005550659324723, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.2038, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.273163366249404, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1921, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.2619817121259986, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1823, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.27121020437769877, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.2006, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.2651145715612084, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1905, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.271061225710045, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.2064, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2529525632809074, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1893, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.26141352795660616, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1888, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.27397774410581366, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.2201, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.26315639969229726, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1993, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.26351018207559485, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1876, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.263204120842595, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1867, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.26777469862607706, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2077, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2420794320461956, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1947, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.2585331730076214, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.2084, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.25358559631623595, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1908, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.2776440643985369, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.2061, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.24888723079412514, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1919, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2591377154910043, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1958, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.24760255928941116, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1811, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2615436625418962, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.2038, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2538044706223925, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1931, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.2522739615085826, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.2001, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.23877598213898313, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1826, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2595228004794417, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1927, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.25294425129891995, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1831, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.2539910597565251, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1994, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.2490380233567818, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1951, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2449309417958549, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1904, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.25406003635581464, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.2069, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2606644805726396, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.2042, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.2558281313979375, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1931, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.25942223661704594, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.2045, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.2515656566757784, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.1994, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.24616389843821132, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1984, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.25839292712691936, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.2154, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2519018943671487, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1807, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2720025954491749, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.2097, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2531329581891392, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1936, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2567372317876449, | |
| "learning_rate": 0.0, | |
| "loss": 0.2016, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 9.802424178211553e+17, | |
| "train_loss": 0.3975551906304482, | |
| "train_runtime": 15848.162, | |
| "train_samples_per_second": 3.155, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "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": 9.802424178211553e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |