| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 0, | |
| "global_step": 154, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006493506493506494, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 9.935064935064936e-06, | |
| "loss": 1.9216, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012987012987012988, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 9.87012987012987e-06, | |
| "loss": 1.9203, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01948051948051948, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 9.805194805194806e-06, | |
| "loss": 1.8268, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.025974025974025976, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 9.740259740259742e-06, | |
| "loss": 1.8553, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.032467532467532464, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 9.675324675324677e-06, | |
| "loss": 1.8987, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03896103896103896, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.610389610389611e-06, | |
| "loss": 1.9061, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.545454545454547e-06, | |
| "loss": 1.8952, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.48051948051948e-06, | |
| "loss": 1.8343, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05844155844155844, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.415584415584416e-06, | |
| "loss": 1.8108, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.350649350649352e-06, | |
| "loss": 1.7731, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 1.8083, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07792207792207792, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.220779220779221e-06, | |
| "loss": 1.8363, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08441558441558442, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.155844155844157e-06, | |
| "loss": 1.7684, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 1.6423, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.09740259740259741, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 9.025974025974027e-06, | |
| "loss": 1.7531, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 8.96103896103896e-06, | |
| "loss": 1.718, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11038961038961038, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 8.896103896103896e-06, | |
| "loss": 1.6784, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.11688311688311688, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 8.831168831168832e-06, | |
| "loss": 1.6809, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12337662337662338, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 8.766233766233767e-06, | |
| "loss": 1.7605, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 8.701298701298701e-06, | |
| "loss": 1.6866, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 8.636363636363637e-06, | |
| "loss": 1.6806, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 1.7233, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.14935064935064934, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 8.506493506493507e-06, | |
| "loss": 1.702, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 8.441558441558442e-06, | |
| "loss": 1.6865, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16233766233766234, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 8.376623376623378e-06, | |
| "loss": 1.7162, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16883116883116883, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 8.311688311688313e-06, | |
| "loss": 1.6615, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.17532467532467533, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 8.246753246753247e-06, | |
| "loss": 1.6482, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 8.181818181818183e-06, | |
| "loss": 1.6257, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.18831168831168832, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 8.116883116883117e-06, | |
| "loss": 1.7048, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.19480519480519481, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 8.051948051948052e-06, | |
| "loss": 1.6033, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2012987012987013, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 7.987012987012988e-06, | |
| "loss": 1.6595, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 7.922077922077924e-06, | |
| "loss": 1.6255, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 1.6825, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.22077922077922077, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.792207792207793e-06, | |
| "loss": 1.6171, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 7.727272727272727e-06, | |
| "loss": 1.6086, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.23376623376623376, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 7.662337662337663e-06, | |
| "loss": 1.5977, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.24025974025974026, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 7.597402597402598e-06, | |
| "loss": 1.6018, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.24675324675324675, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 7.532467532467533e-06, | |
| "loss": 1.6183, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2532467532467532, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.467532467532468e-06, | |
| "loss": 1.5952, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 7.402597402597404e-06, | |
| "loss": 1.564, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2662337662337662, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 7.3376623376623375e-06, | |
| "loss": 1.5618, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.272727272727273e-06, | |
| "loss": 1.5533, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2792207792207792, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 7.207792207792208e-06, | |
| "loss": 1.6762, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 1.5851, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2922077922077922, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.077922077922078e-06, | |
| "loss": 1.5862, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2987012987012987, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 7.012987012987014e-06, | |
| "loss": 1.5626, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3051948051948052, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 6.948051948051948e-06, | |
| "loss": 1.4968, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 6.8831168831168835e-06, | |
| "loss": 1.5374, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 6.818181818181818e-06, | |
| "loss": 1.6145, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3246753246753247, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 6.753246753246754e-06, | |
| "loss": 1.5309, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.33116883116883117, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 6.688311688311689e-06, | |
| "loss": 1.5764, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.33766233766233766, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 6.623376623376624e-06, | |
| "loss": 1.558, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.34415584415584416, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 6.55844155844156e-06, | |
| "loss": 1.51, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.35064935064935066, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 6.493506493506494e-06, | |
| "loss": 1.5634, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 1.5444, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 6.363636363636364e-06, | |
| "loss": 1.5565, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.37012987012987014, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 6.2987012987013e-06, | |
| "loss": 1.4929, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.37662337662337664, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 6.233766233766234e-06, | |
| "loss": 1.5763, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.38311688311688313, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 6.168831168831169e-06, | |
| "loss": 1.5113, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 6.103896103896104e-06, | |
| "loss": 1.5285, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3961038961038961, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 6.03896103896104e-06, | |
| "loss": 1.5784, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4025974025974026, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 5.9740259740259746e-06, | |
| "loss": 1.5411, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 5.90909090909091e-06, | |
| "loss": 1.5138, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 5.844155844155844e-06, | |
| "loss": 1.5787, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.42207792207792205, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 5.77922077922078e-06, | |
| "loss": 1.5045, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 1.528, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.43506493506493504, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 5.64935064935065e-06, | |
| "loss": 1.5024, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.44155844155844154, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 5.584415584415585e-06, | |
| "loss": 1.5593, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.44805194805194803, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 5.5194805194805205e-06, | |
| "loss": 1.4983, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 1.5183, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.461038961038961, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 5.38961038961039e-06, | |
| "loss": 1.5491, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 5.324675324675325e-06, | |
| "loss": 1.43, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.474025974025974, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 5.2597402597402605e-06, | |
| "loss": 1.5414, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4805194805194805, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 5.194805194805194e-06, | |
| "loss": 1.5338, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.487012987012987, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 5.12987012987013e-06, | |
| "loss": 1.5223, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4935064935064935, | |
| "grad_norm": 0.1181640625, | |
| "learning_rate": 5.064935064935065e-06, | |
| "loss": 1.5068, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 5e-06, | |
| "loss": 1.506, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5064935064935064, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 4.935064935064935e-06, | |
| "loss": 1.4808, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.512987012987013, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 4.870129870129871e-06, | |
| "loss": 1.5765, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 4.805194805194806e-06, | |
| "loss": 1.4691, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.525974025974026, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 4.74025974025974e-06, | |
| "loss": 1.4995, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5324675324675324, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 4.675324675324676e-06, | |
| "loss": 1.5206, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.538961038961039, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 4.610389610389611e-06, | |
| "loss": 1.4908, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 1.4625, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.551948051948052, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 4.48051948051948e-06, | |
| "loss": 1.4645, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5584415584415584, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 4.415584415584416e-06, | |
| "loss": 1.5982, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.564935064935065, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 4.350649350649351e-06, | |
| "loss": 1.4427, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 1.4655, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.577922077922078, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 4.220779220779221e-06, | |
| "loss": 1.5021, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5844155844155844, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 4.155844155844157e-06, | |
| "loss": 1.4549, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5909090909090909, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 4.0909090909090915e-06, | |
| "loss": 1.4115, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5974025974025974, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 4.025974025974026e-06, | |
| "loss": 1.4747, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6038961038961039, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 3.961038961038962e-06, | |
| "loss": 1.4906, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6103896103896104, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 3.896103896103897e-06, | |
| "loss": 1.5562, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6168831168831169, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 3.831168831168831e-06, | |
| "loss": 1.5198, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 3.7662337662337666e-06, | |
| "loss": 1.4991, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6298701298701299, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 3.701298701298702e-06, | |
| "loss": 1.4371, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 3.6363636363636366e-06, | |
| "loss": 1.4679, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 1.5392, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 3.506493506493507e-06, | |
| "loss": 1.5854, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6558441558441559, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 3.4415584415584418e-06, | |
| "loss": 1.4859, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6623376623376623, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 3.376623376623377e-06, | |
| "loss": 1.3568, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6688311688311688, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 3.311688311688312e-06, | |
| "loss": 1.5947, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 0.11767578125, | |
| "learning_rate": 3.246753246753247e-06, | |
| "loss": 1.4949, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 3.181818181818182e-06, | |
| "loss": 1.4853, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6883116883116883, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.116883116883117e-06, | |
| "loss": 1.425, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6948051948051948, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 3.051948051948052e-06, | |
| "loss": 1.4564, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7012987012987013, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 2.9870129870129873e-06, | |
| "loss": 1.4962, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7077922077922078, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 2.922077922077922e-06, | |
| "loss": 1.4239, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.11279296875, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 1.4496, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7207792207792207, | |
| "grad_norm": 0.11865234375, | |
| "learning_rate": 2.7922077922077925e-06, | |
| "loss": 1.5018, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 2.7272727272727272e-06, | |
| "loss": 1.4486, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7337662337662337, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 2.6623376623376624e-06, | |
| "loss": 1.4482, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7402597402597403, | |
| "grad_norm": 0.1240234375, | |
| "learning_rate": 2.597402597402597e-06, | |
| "loss": 1.4378, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7467532467532467, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 2.5324675324675324e-06, | |
| "loss": 1.4501, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7532467532467533, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 2.4675324675324676e-06, | |
| "loss": 1.5188, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7597402597402597, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 2.402597402597403e-06, | |
| "loss": 1.4609, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7662337662337663, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 2.337662337662338e-06, | |
| "loss": 1.4207, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7727272727272727, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 2.2727272727272728e-06, | |
| "loss": 1.4122, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 0.1064453125, | |
| "learning_rate": 2.207792207792208e-06, | |
| "loss": 1.4604, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 1.4657, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7922077922077922, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 2.0779220779220784e-06, | |
| "loss": 1.4832, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7987012987012987, | |
| "grad_norm": 0.12451171875, | |
| "learning_rate": 2.012987012987013e-06, | |
| "loss": 1.3874, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8051948051948052, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 1.9480519480519483e-06, | |
| "loss": 1.4578, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8116883116883117, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 1.8831168831168833e-06, | |
| "loss": 1.4518, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 1.8181818181818183e-06, | |
| "loss": 1.446, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8246753246753247, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 1.7532467532467535e-06, | |
| "loss": 1.4921, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 1.6883116883116885e-06, | |
| "loss": 1.4776, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8376623376623377, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.6233766233766235e-06, | |
| "loss": 1.557, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8441558441558441, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 1.5584415584415584e-06, | |
| "loss": 1.4844, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8506493506493507, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 1.4935064935064936e-06, | |
| "loss": 1.5005, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 1.4655, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8636363636363636, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 1.3636363636363636e-06, | |
| "loss": 1.4916, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8701298701298701, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 1.2987012987012986e-06, | |
| "loss": 1.4502, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8766233766233766, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 1.2337662337662338e-06, | |
| "loss": 1.4147, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 1.168831168831169e-06, | |
| "loss": 1.5244, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8896103896103896, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 1.103896103896104e-06, | |
| "loss": 1.4393, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8961038961038961, | |
| "grad_norm": 0.12255859375, | |
| "learning_rate": 1.0389610389610392e-06, | |
| "loss": 1.3782, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9025974025974026, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 9.740259740259742e-07, | |
| "loss": 1.3691, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 1.4324, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9155844155844156, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 8.441558441558442e-07, | |
| "loss": 1.4545, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.922077922077922, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 7.792207792207792e-07, | |
| "loss": 1.4364, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 1.5295, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 6.493506493506493e-07, | |
| "loss": 1.4032, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9415584415584416, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 5.844155844155845e-07, | |
| "loss": 1.483, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.948051948051948, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 5.194805194805196e-07, | |
| "loss": 1.4865, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9545454545454546, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 4.5454545454545457e-07, | |
| "loss": 1.4349, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.961038961038961, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 3.896103896103896e-07, | |
| "loss": 1.4733, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9675324675324676, | |
| "grad_norm": 0.1259765625, | |
| "learning_rate": 3.2467532467532465e-07, | |
| "loss": 1.4665, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.974025974025974, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 2.597402597402598e-07, | |
| "loss": 1.5085, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9805194805194806, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 1.948051948051948e-07, | |
| "loss": 1.4872, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.298701298701299e-07, | |
| "loss": 1.4249, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9935064935064936, | |
| "grad_norm": 0.125, | |
| "learning_rate": 6.493506493506495e-08, | |
| "loss": 1.4824, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.0, | |
| "loss": 1.4958, | |
| "step": 154 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 154, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.350584389810258e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |