| { | |
| "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": 6.359983322132192, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.8126, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.236883186981875, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.814, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.081132028140397, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.8191, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.668604958400867, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.8133, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 3.984095421902308, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.7513, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.2938386048562007, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.7162, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.136228020272299, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.7188, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 3.9342859415032247, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.7306, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.303341642032038, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.7281, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.189626788036456, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.6931, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.6977424607439024, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.6946, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 1.9528035975118787, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.6234, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.6934195819153213, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6436, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.0812677353280016, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.6134, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.5713199933597666, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.6172, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.1010445043760573, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.6118, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.223047115623996, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.5834, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.9228360503417343, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5716, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.9320342580445303, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.5679, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.8240356527249942, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.5736, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.8316434580672932, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.5572, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.8815562730273679, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.5562, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7362694665622409, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.5462, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.7541180145346861, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.5305, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7040375509404285, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.5247, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.6340756538529179, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.5252, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6958981314950035, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.5192, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.8077781822648501, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.5171, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.723654585169182, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.5203, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.8489885382682859, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.5235, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.6940698620517392, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.5044, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.6217730710530112, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.5305, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.5562603359454786, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.51, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.7147585229292516, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.5047, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7224660759293063, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.5134, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.8488495137504695, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.5127, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.8565285462311633, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.5074, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.2109822957438114, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.5013, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.9602640015218507, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4915, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.9104208488686752, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.5079, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.8356274032130077, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.497, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.6654253591360805, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.4907, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.8287601661675987, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.5005, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.5493733282993682, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.4749, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.8695964660521356, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.4731, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7021476240633592, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.4822, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.6104841678452269, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.4875, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.0159722317284792, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.5043, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.6982530575159299, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.4667, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.4779985850760904, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.4602, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.6865127716670445, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.4763, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7584253768331256, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.4899, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.5454083161686872, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.4877, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6102896274065267, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.4685, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.9269879912871652, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.4617, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 1.1257340249970111, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.4727, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.6695889837787722, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.4643, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.6143939704506852, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.4601, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.8776637875070274, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.4625, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.643157657726606, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.4651, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.5918379454902895, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.4653, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6519665071377362, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.4655, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6115185164151091, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.4659, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.504336201948985, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.4545, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.6306961868794007, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.4707, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.5844682485168733, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.4704, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.49427447150230486, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.4664, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.5959803095382642, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.4573, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5400648528517348, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.454, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.4478087177477396, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.4584, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.577232539021707, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.4719, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5662873187815036, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.4506, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.47734231050947984, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.4662, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.5551281281506267, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.4587, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5018738734937632, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.4513, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.6766738014066402, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.4629, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.5814139315964194, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.467, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.4562025035490514, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.4492, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.8965149204428019, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.8549, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.8877021667423628, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.4067, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.7469345525371692, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.422, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.6909334630574211, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4085, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.715311771514669, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.4227, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6339805558009638, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.4022, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6519034041228249, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.4302, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.5839383532820845, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4167, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.6388425182256798, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4375, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6471232210378166, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.4372, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.6591030801594738, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.4327, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.7568379345452619, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.4043, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5468036939234712, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.4131, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.6836001325347488, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4224, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.49310642800841326, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4206, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.5297833153642639, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.3977, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.4937191797345033, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.4286, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.41726983473631724, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.413, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.43987327234478374, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.4153, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.44081383345726843, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4165, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.3998069990570622, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.4182, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.409519412177145, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.3908, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.3520243892017844, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.4313, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.5197610917988033, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.4054, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.42206533070402813, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.4215, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.3974019931313252, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4098, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.356458047651259, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.4197, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.3711276460087547, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.4206, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.35153470431888184, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4119, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.364808408085956, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.4121, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.38303320458111056, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4254, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.41068702666096546, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4074, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.3901562426454199, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4277, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.42884243484932283, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4085, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.386894231646286, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4145, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.31668561970259196, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.3968, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.32902075812484305, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4068, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.34125741115010366, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4003, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.40782564164291635, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4172, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.3970639002784617, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4051, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.44710168287056895, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4025, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.37305231735273864, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4253, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.4575191870828874, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4143, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.3974950499828991, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4038, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.44509982605924653, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.3931, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.42846540192106636, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.3989, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.35659498193954614, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4081, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.4439984896820595, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4067, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.3714135922040878, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4121, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.37640866082634983, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.4045, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.35630214309982017, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4039, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.4694983805754382, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.4047, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.4538247090221982, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.421, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.3457990010829799, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4067, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.40969028078140407, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.417, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.39080928621675726, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4103, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.37356378345607244, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.4088, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.32760832310053006, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.4006, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.33132540113607556, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4056, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.3673725730071437, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.3957, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.41272543630558045, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4175, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.31533090857880175, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4011, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.33562774447617344, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.4002, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.4029034072868709, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4193, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.43645259169895906, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.3925, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.38350974043813657, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.399, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.3469191772004374, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.4057, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.413037919852792, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4155, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.3906487699162059, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.4148, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.31253273494163675, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.4088, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.3704962284582328, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4077, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.3963981569763363, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.411, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.3769340662523909, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4062, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.28576918810777885, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.3902, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4155086452708957, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.3982, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.3656807099796843, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4002, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.3283319942044213, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.3988, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.2979898214690633, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.389, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.6578640332533632, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.7163, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6276138290166481, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3558, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.5806427519128321, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3423, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.5360651621159771, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3552, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6241899824061685, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3695, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.6240271749354501, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3484, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.4698651203309415, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3493, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.45244442969047777, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3418, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5081138455107693, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3523, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.43239977890120557, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3465, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.3743942630723038, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3573, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5374834779582888, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3493, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.4496010722757497, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3593, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5406902045449381, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3527, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.3749430910170632, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3528, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.4878918096006516, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3475, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.41488526233624573, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3399, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.38563994102127985, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3462, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4043557950850195, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3339, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.34682150489546826, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3592, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.3901821995578983, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3528, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.357695692240288, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3488, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.34679658250639256, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3315, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.3938851552079498, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3475, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.35635246490276035, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3603, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.3330770753181355, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3577, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.3427377165721337, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3416, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.31323295185304245, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3507, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.3299051311046998, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3497, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3595518298241771, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3412, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.338583918972112, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3489, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.29123304333734634, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3459, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.3453265760417586, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3603, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.28928604315802964, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.346, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.29686753967627255, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3406, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.2797581880390852, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3395, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.2709550343797354, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3326, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.28505786649266573, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3424, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.26792060176540355, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3487, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.29750424205813947, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3458, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.24622841648839439, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3455, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.2579181710874334, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3447, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.26389436997101645, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3464, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.2745637355827058, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3562, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.2586019880180475, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3447, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.2771691958709938, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3503, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.2683514284460922, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3461, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.2623224453762875, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3563, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.26798711834888805, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3528, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.24803374569122752, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3344, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.2743135581701106, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3579, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.27441924435632714, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3403, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.26277260885154025, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3587, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3111792870132343, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3509, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.2678749446392414, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3374, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.29258888793960974, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3522, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.29727252908118373, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3571, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.2540582538924719, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3414, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.2973041092314046, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3691, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.2729183820100434, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3444, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.24685612293956968, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3407, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.2713841558319105, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3474, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.29346317832271684, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3451, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.2751500335048413, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3534, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.3131315962213258, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3427, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.2760393704826248, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3575, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.2984057898233131, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3558, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.25873254677377144, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3366, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3090566289692919, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3521, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.25712495236896443, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3307, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.2754201109628778, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3551, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.24716470036213756, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3324, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.24775971858961432, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3524, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.2826122711077555, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3462, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.220700744271221, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3435, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.30702401802838897, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.343, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.23866180245254806, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3528, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.2747279023555536, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3404, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.6275703438326738, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.6183, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.4102481508561138, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.3055, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.43168247024643636, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2982, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5145362200673524, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2962, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.4069405027892344, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2927, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4217195357807677, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.3007, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.3915067450901604, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2896, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.35623823056636894, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2997, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.3348264866011541, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2988, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3460399962836094, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2962, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.3390142403340877, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2966, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.2775419916066782, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2884, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3020455411480731, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2958, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.29143406280838136, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2979, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.2773605721610392, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2868, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.2771201557883132, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.3042, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3058812989672803, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.295, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.26644948620615844, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.281, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.32244290216042354, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2949, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.33721010380860006, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.3021, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.26624498864452656, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2888, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.33437354059497354, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2919, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.2549032059134277, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2962, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.2953145638676912, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2905, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.26771077491314305, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2875, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.24705671852350958, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2836, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.2519996762568636, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2752, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.25222327747761675, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.299, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.2395877377506612, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2845, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.21706386426977103, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.284, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.24398855953212573, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.304, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.21784014008953007, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2801, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.22424156851478347, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2952, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.24736090698234994, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2964, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.22156652985603287, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2895, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.23650522578207536, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2817, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.2197214304208983, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.296, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.2291637125070749, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2994, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.23445675359315335, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2906, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.2178666730909089, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.29, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.22961853140309582, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.3044, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.22842100639708415, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.291, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.2087575754950737, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2992, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.21170474889922047, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2993, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.20728406949790718, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2897, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.20410780019350855, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2994, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.20706435213379912, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.3006, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.1960223441576943, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2876, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.18779145248364681, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2878, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.2070342964058405, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2969, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.19542042577017307, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2867, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.19260860342245195, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.297, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.19458929330483496, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2943, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.20833903102346224, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2858, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.19208304964110295, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2896, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.19556996047107, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2984, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.21209551122777193, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2984, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.2071471605961549, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2875, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.19208936593121742, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2858, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.20016935162557226, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2862, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.2118592398449892, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2964, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.20863811485723005, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.3022, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.19602893610076277, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2809, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.20658415901610194, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.3005, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.23161623517518, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2937, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.19817130626613383, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.289, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.20252242230908585, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.3016, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.1978616471182992, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2916, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.21169766557483544, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2926, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.1890897650810334, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2901, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.19811055759863005, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2964, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.19745387303884826, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2959, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.19522488537662167, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2938, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.20366322996147027, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.3082, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.18091441538061614, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2886, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.20436069752457997, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2871, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.20097152129151696, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.3077, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.1843551191345779, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2836, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.5460824481862353, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.5495, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.3017753824451453, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2572, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.2375292518430657, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.2483, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.23664774683808626, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2623, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.25368072041162026, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.2583, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.31280577838980933, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.2644, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.2624155419206717, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.2466, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.2663592020804932, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.2712, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.2353738145029052, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.2652, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.21912441863120785, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2509, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.24403643083674145, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.2571, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.2344818374419897, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.2593, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.23431087561524763, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.267, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.21635236586586454, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.2614, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.20175768393889576, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.2606, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.28374317324396275, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.2588, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.22307690373338168, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.2537, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.21493197188049648, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.2584, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.1996258189868415, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.249, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.20254302356336967, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.265, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.1984426523527175, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.2562, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.20075172254188053, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.2635, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.2083554083066276, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.2602, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.19761602969133163, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.2592, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.20038578394375134, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.2529, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.18796378903155805, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.2588, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.18643578215153453, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.2547, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.18280131625280735, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.2586, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.18099044332943004, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.2553, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.18446010629189014, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.2647, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.1735773109019441, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.2551, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.18369700347736956, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.2575, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.1830132760669594, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.2546, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.18396947401185804, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.2555, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.18026768891458453, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.2614, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.17830880833890075, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.2617, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.1706412694750075, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.2481, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.17512531646418064, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.2627, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.16926167947095738, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.2479, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.17367147044924297, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.2499, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.16975028979589998, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.2556, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.16832507186213594, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.2558, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.17854799680119768, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.2629, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.17257681095565866, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.2479, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.17519770623715808, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.2561, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.17193388209061078, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.2642, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.17081838293319956, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.2493, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.1733013778729429, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.2576, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.17400805239071468, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.2531, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.16805171965408439, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.2577, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.17224652436122115, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2545, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.17081160560488798, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.2545, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.16641458645584337, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.2597, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.16927076526021254, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.2511, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.18009261695853312, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.2642, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.17120033050271458, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.2699, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.17294621953469197, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.2516, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.18449473339291073, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.2515, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.17120844051799233, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.265, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.17354992424054086, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.2505, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.18119084305046249, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.2586, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.17277596365178818, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.2566, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.16399254474246303, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.2539, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.16575064417474455, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.2564, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.16671185550021772, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.26, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.16193950175084013, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.2516, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.1692266245724526, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.2525, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.1688945404481823, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.2597, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.17142911007065878, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.2726, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.17290338820706183, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.2571, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.17609578889352118, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.2665, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.16457390937649688, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.2486, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.17235230139017066, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.2594, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.1705929659828749, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.262, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.17075607232577814, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.2581, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.16793854775671166, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.2508, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.1651736304787536, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.2572, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.16690862558119737, | |
| "learning_rate": 0.0, | |
| "loss": 0.2585, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.8658329858534277e+18, | |
| "train_loss": 0.3731630460956158, | |
| "train_runtime": 19785.2158, | |
| "train_samples_per_second": 2.527, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "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": 1.8658329858534277e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |