| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1235, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004048582995951417, | |
| "grad_norm": 6.101385142561969, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 0.8065, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008097165991902834, | |
| "grad_norm": 6.3682853378091195, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 0.8126, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012145748987854251, | |
| "grad_norm": 6.252651209319461, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 0.8078, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.016194331983805668, | |
| "grad_norm": 5.954238526224445, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 0.7899, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.020242914979757085, | |
| "grad_norm": 5.853176782127661, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 0.8142, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024291497975708502, | |
| "grad_norm": 5.5221985512148075, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 0.7912, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02834008097165992, | |
| "grad_norm": 4.360158856184302, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 0.7549, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.032388663967611336, | |
| "grad_norm": 3.9713399434514405, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 0.7368, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03643724696356275, | |
| "grad_norm": 2.284628794202244, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 0.713, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04048582995951417, | |
| "grad_norm": 2.0708133235012127, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.7295, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044534412955465584, | |
| "grad_norm": 1.9405331880226782, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 0.7104, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.048582995951417005, | |
| "grad_norm": 3.6915977485359184, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 0.7149, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 4.0446734370603075, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 0.7129, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05668016194331984, | |
| "grad_norm": 4.097295482255283, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 0.7043, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06072874493927125, | |
| "grad_norm": 3.976942318244177, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 0.6972, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06477732793522267, | |
| "grad_norm": 3.180043054274617, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 0.6667, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06882591093117409, | |
| "grad_norm": 2.8205154292873944, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 0.6361, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0728744939271255, | |
| "grad_norm": 2.080482117754469, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 0.6243, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 1.3872469301146297, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 0.6222, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08097165991902834, | |
| "grad_norm": 1.1395080609986965, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.5898, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08502024291497975, | |
| "grad_norm": 1.3409115171484052, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 0.5895, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08906882591093117, | |
| "grad_norm": 1.463363523759221, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.5972, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0931174089068826, | |
| "grad_norm": 1.1742251452774415, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 0.5936, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09716599190283401, | |
| "grad_norm": 0.8795747126851504, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.5815, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10121457489878542, | |
| "grad_norm": 1.0546402320434467, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.5959, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 0.787441050106149, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.5773, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10931174089068826, | |
| "grad_norm": 0.7088070067944154, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.5749, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11336032388663968, | |
| "grad_norm": 0.7791487083360543, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.5576, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11740890688259109, | |
| "grad_norm": 0.7824093153642827, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.5547, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1214574898785425, | |
| "grad_norm": 0.5483001744996717, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.5386, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12550607287449392, | |
| "grad_norm": 0.6146620453799485, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5451, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12955465587044535, | |
| "grad_norm": 0.7224742142336095, | |
| "learning_rate": 1.0322580645161291e-05, | |
| "loss": 0.546, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13360323886639677, | |
| "grad_norm": 0.649206934073557, | |
| "learning_rate": 1.0645161290322582e-05, | |
| "loss": 0.5287, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13765182186234817, | |
| "grad_norm": 0.5035326002355579, | |
| "learning_rate": 1.096774193548387e-05, | |
| "loss": 0.5476, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1417004048582996, | |
| "grad_norm": 0.5536890763089765, | |
| "learning_rate": 1.1290322580645164e-05, | |
| "loss": 0.5257, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.145748987854251, | |
| "grad_norm": 0.5763518063018838, | |
| "learning_rate": 1.1612903225806453e-05, | |
| "loss": 0.5396, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14979757085020243, | |
| "grad_norm": 0.4964180998739624, | |
| "learning_rate": 1.1935483870967743e-05, | |
| "loss": 0.5315, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.5010610262073186, | |
| "learning_rate": 1.2258064516129034e-05, | |
| "loss": 0.5068, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 0.5340596786158056, | |
| "learning_rate": 1.2580645161290324e-05, | |
| "loss": 0.5115, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.16194331983805668, | |
| "grad_norm": 0.44312884147518067, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 0.5234, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1659919028340081, | |
| "grad_norm": 0.5740209121862617, | |
| "learning_rate": 1.3225806451612903e-05, | |
| "loss": 0.5209, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1700404858299595, | |
| "grad_norm": 0.5435349384240132, | |
| "learning_rate": 1.3548387096774194e-05, | |
| "loss": 0.5053, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17408906882591094, | |
| "grad_norm": 0.4399305201220378, | |
| "learning_rate": 1.3870967741935486e-05, | |
| "loss": 0.5126, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17813765182186234, | |
| "grad_norm": 0.5467599763460577, | |
| "learning_rate": 1.4193548387096776e-05, | |
| "loss": 0.519, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18218623481781376, | |
| "grad_norm": 0.5023562450449479, | |
| "learning_rate": 1.4516129032258066e-05, | |
| "loss": 0.5066, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1862348178137652, | |
| "grad_norm": 0.3874660646036981, | |
| "learning_rate": 1.4838709677419357e-05, | |
| "loss": 0.4988, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1902834008097166, | |
| "grad_norm": 0.444797572814555, | |
| "learning_rate": 1.5161290322580646e-05, | |
| "loss": 0.5135, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19433198380566802, | |
| "grad_norm": 0.43002801395157386, | |
| "learning_rate": 1.5483870967741936e-05, | |
| "loss": 0.4953, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19838056680161945, | |
| "grad_norm": 0.46934892689543556, | |
| "learning_rate": 1.5806451612903226e-05, | |
| "loss": 0.5022, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.20242914979757085, | |
| "grad_norm": 0.38967090116433073, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.502, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20647773279352227, | |
| "grad_norm": 0.4523460816196144, | |
| "learning_rate": 1.6451612903225807e-05, | |
| "loss": 0.4932, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.3927696616091829, | |
| "learning_rate": 1.6774193548387098e-05, | |
| "loss": 0.4964, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2145748987854251, | |
| "grad_norm": 0.42644363924009554, | |
| "learning_rate": 1.7096774193548388e-05, | |
| "loss": 0.5068, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21862348178137653, | |
| "grad_norm": 0.35084354666395284, | |
| "learning_rate": 1.741935483870968e-05, | |
| "loss": 0.4887, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.22267206477732793, | |
| "grad_norm": 0.418234235425629, | |
| "learning_rate": 1.774193548387097e-05, | |
| "loss": 0.5033, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22672064777327935, | |
| "grad_norm": 0.41869465230109554, | |
| "learning_rate": 1.806451612903226e-05, | |
| "loss": 0.4775, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.3994740676904812, | |
| "learning_rate": 1.838709677419355e-05, | |
| "loss": 0.4808, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23481781376518218, | |
| "grad_norm": 0.42411559390561043, | |
| "learning_rate": 1.870967741935484e-05, | |
| "loss": 0.4777, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2388663967611336, | |
| "grad_norm": 0.42854877578753353, | |
| "learning_rate": 1.903225806451613e-05, | |
| "loss": 0.4855, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.242914979757085, | |
| "grad_norm": 0.48386343327324555, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 0.5037, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24696356275303644, | |
| "grad_norm": 0.6426667044185659, | |
| "learning_rate": 1.967741935483871e-05, | |
| "loss": 0.4905, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.25101214574898784, | |
| "grad_norm": 1.008811243899259, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4819, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2550607287449393, | |
| "grad_norm": 0.8737827700932996, | |
| "learning_rate": 2.0322580645161292e-05, | |
| "loss": 0.4771, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2591093117408907, | |
| "grad_norm": 0.5568170269940472, | |
| "learning_rate": 2.0645161290322582e-05, | |
| "loss": 0.4757, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 0.8171209085495192, | |
| "learning_rate": 2.0967741935483873e-05, | |
| "loss": 0.4909, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.26720647773279355, | |
| "grad_norm": 0.8062055745516644, | |
| "learning_rate": 2.1290322580645163e-05, | |
| "loss": 0.4716, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.27125506072874495, | |
| "grad_norm": 0.78723746237319, | |
| "learning_rate": 2.161290322580645e-05, | |
| "loss": 0.477, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.27530364372469635, | |
| "grad_norm": 0.8942641338925323, | |
| "learning_rate": 2.193548387096774e-05, | |
| "loss": 0.4721, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2793522267206478, | |
| "grad_norm": 0.8097803906022192, | |
| "learning_rate": 2.225806451612903e-05, | |
| "loss": 0.4779, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2834008097165992, | |
| "grad_norm": 1.0220718963039228, | |
| "learning_rate": 2.2580645161290328e-05, | |
| "loss": 0.4698, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2874493927125506, | |
| "grad_norm": 0.761493793204974, | |
| "learning_rate": 2.2903225806451615e-05, | |
| "loss": 0.4676, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.291497975708502, | |
| "grad_norm": 0.7120534542279967, | |
| "learning_rate": 2.3225806451612906e-05, | |
| "loss": 0.4781, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.29554655870445345, | |
| "grad_norm": 0.8687339448487634, | |
| "learning_rate": 2.3548387096774196e-05, | |
| "loss": 0.4685, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29959514170040485, | |
| "grad_norm": 0.7273881519428771, | |
| "learning_rate": 2.3870967741935486e-05, | |
| "loss": 0.4755, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.30364372469635625, | |
| "grad_norm": 0.9144075191468968, | |
| "learning_rate": 2.4193548387096777e-05, | |
| "loss": 0.4723, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.6510085402748221, | |
| "learning_rate": 2.4516129032258067e-05, | |
| "loss": 0.4759, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3117408906882591, | |
| "grad_norm": 0.7992590094526716, | |
| "learning_rate": 2.4838709677419358e-05, | |
| "loss": 0.4768, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 0.6528314332747627, | |
| "learning_rate": 2.5161290322580648e-05, | |
| "loss": 0.4645, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.31983805668016196, | |
| "grad_norm": 0.8565592226281882, | |
| "learning_rate": 2.548387096774194e-05, | |
| "loss": 0.4555, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.32388663967611336, | |
| "grad_norm": 0.6636542555416944, | |
| "learning_rate": 2.5806451612903226e-05, | |
| "loss": 0.4681, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32793522267206476, | |
| "grad_norm": 0.746484679367779, | |
| "learning_rate": 2.6129032258064516e-05, | |
| "loss": 0.4581, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3319838056680162, | |
| "grad_norm": 0.6349200736570594, | |
| "learning_rate": 2.6451612903225806e-05, | |
| "loss": 0.4924, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3360323886639676, | |
| "grad_norm": 1.0013146429921609, | |
| "learning_rate": 2.6774193548387097e-05, | |
| "loss": 0.4734, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.340080971659919, | |
| "grad_norm": 0.7011619015663199, | |
| "learning_rate": 2.7096774193548387e-05, | |
| "loss": 0.4797, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3441295546558704, | |
| "grad_norm": 0.7970412487567091, | |
| "learning_rate": 2.7419354838709678e-05, | |
| "loss": 0.456, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3481781376518219, | |
| "grad_norm": 0.6989041042252208, | |
| "learning_rate": 2.774193548387097e-05, | |
| "loss": 0.4754, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3522267206477733, | |
| "grad_norm": 0.7523317136157248, | |
| "learning_rate": 2.8064516129032262e-05, | |
| "loss": 0.4648, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3562753036437247, | |
| "grad_norm": 0.6939683176596698, | |
| "learning_rate": 2.8387096774193552e-05, | |
| "loss": 0.4654, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3603238866396761, | |
| "grad_norm": 0.7728107275685632, | |
| "learning_rate": 2.8709677419354843e-05, | |
| "loss": 0.4576, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3643724696356275, | |
| "grad_norm": 1.086408119043204, | |
| "learning_rate": 2.9032258064516133e-05, | |
| "loss": 0.4509, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 0.6179157086344635, | |
| "learning_rate": 2.9354838709677423e-05, | |
| "loss": 0.4781, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3724696356275304, | |
| "grad_norm": 0.6387794288607884, | |
| "learning_rate": 2.9677419354838714e-05, | |
| "loss": 0.4766, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3765182186234818, | |
| "grad_norm": 0.5782642062976371, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4483, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3805668016194332, | |
| "grad_norm": 0.6967268180393732, | |
| "learning_rate": 3.032258064516129e-05, | |
| "loss": 0.4704, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.5231195484354747, | |
| "learning_rate": 3.0645161290322585e-05, | |
| "loss": 0.4581, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38866396761133604, | |
| "grad_norm": 0.5457722526905683, | |
| "learning_rate": 3.096774193548387e-05, | |
| "loss": 0.4562, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.39271255060728744, | |
| "grad_norm": 0.5469734170305627, | |
| "learning_rate": 3.1290322580645166e-05, | |
| "loss": 0.4611, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3967611336032389, | |
| "grad_norm": 0.6115844867124165, | |
| "learning_rate": 3.161290322580645e-05, | |
| "loss": 0.467, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4008097165991903, | |
| "grad_norm": 0.7194370938860438, | |
| "learning_rate": 3.193548387096774e-05, | |
| "loss": 0.4719, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4048582995951417, | |
| "grad_norm": 0.7203789433769647, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.4458, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4089068825910931, | |
| "grad_norm": 0.7919826275841038, | |
| "learning_rate": 3.258064516129033e-05, | |
| "loss": 0.4686, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.41295546558704455, | |
| "grad_norm": 0.7107863493372062, | |
| "learning_rate": 3.2903225806451614e-05, | |
| "loss": 0.4631, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.41700404858299595, | |
| "grad_norm": 0.8924113002187805, | |
| "learning_rate": 3.322580645161291e-05, | |
| "loss": 0.4636, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.8717331528375764, | |
| "learning_rate": 3.3548387096774195e-05, | |
| "loss": 0.4682, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4251012145748988, | |
| "grad_norm": 0.994607673691391, | |
| "learning_rate": 3.387096774193549e-05, | |
| "loss": 0.4711, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4291497975708502, | |
| "grad_norm": 1.0017031067646163, | |
| "learning_rate": 3.4193548387096776e-05, | |
| "loss": 0.4666, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4331983805668016, | |
| "grad_norm": 1.2320791323351574, | |
| "learning_rate": 3.451612903225807e-05, | |
| "loss": 0.4595, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.43724696356275305, | |
| "grad_norm": 0.6298519769105961, | |
| "learning_rate": 3.483870967741936e-05, | |
| "loss": 0.4569, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.44129554655870445, | |
| "grad_norm": 0.8306166731535951, | |
| "learning_rate": 3.516129032258065e-05, | |
| "loss": 0.4699, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.44534412955465585, | |
| "grad_norm": 1.0129985427996906, | |
| "learning_rate": 3.548387096774194e-05, | |
| "loss": 0.4626, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4493927125506073, | |
| "grad_norm": 0.9491855610752761, | |
| "learning_rate": 3.5806451612903225e-05, | |
| "loss": 0.4779, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4534412955465587, | |
| "grad_norm": 0.8352838908739086, | |
| "learning_rate": 3.612903225806452e-05, | |
| "loss": 0.4678, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4574898785425101, | |
| "grad_norm": 0.7764695388409772, | |
| "learning_rate": 3.6451612903225805e-05, | |
| "loss": 0.4648, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.6490919579331967, | |
| "learning_rate": 3.67741935483871e-05, | |
| "loss": 0.4616, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.46558704453441296, | |
| "grad_norm": 0.6937558898483563, | |
| "learning_rate": 3.7096774193548386e-05, | |
| "loss": 0.4593, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.46963562753036436, | |
| "grad_norm": 0.8212333533830619, | |
| "learning_rate": 3.741935483870968e-05, | |
| "loss": 0.4465, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 0.5128809116624538, | |
| "learning_rate": 3.7741935483870974e-05, | |
| "loss": 0.448, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4777327935222672, | |
| "grad_norm": 0.6581101093516334, | |
| "learning_rate": 3.806451612903226e-05, | |
| "loss": 0.4445, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4817813765182186, | |
| "grad_norm": 0.7414365250130048, | |
| "learning_rate": 3.8387096774193555e-05, | |
| "loss": 0.4623, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.48582995951417, | |
| "grad_norm": 0.6178561453449056, | |
| "learning_rate": 3.870967741935484e-05, | |
| "loss": 0.4517, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4898785425101215, | |
| "grad_norm": 0.6121650797403921, | |
| "learning_rate": 3.9032258064516135e-05, | |
| "loss": 0.4618, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4939271255060729, | |
| "grad_norm": 0.7590011359354524, | |
| "learning_rate": 3.935483870967742e-05, | |
| "loss": 0.4676, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4979757085020243, | |
| "grad_norm": 0.8897830758475639, | |
| "learning_rate": 3.9677419354838716e-05, | |
| "loss": 0.4775, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5020242914979757, | |
| "grad_norm": 1.0892909307909293, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4623, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5060728744939271, | |
| "grad_norm": 0.9535507473380498, | |
| "learning_rate": 3.9999920040266474e-05, | |
| "loss": 0.4614, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5101214574898786, | |
| "grad_norm": 0.897253954090448, | |
| "learning_rate": 3.999968016170525e-05, | |
| "loss": 0.4553, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5141700404858299, | |
| "grad_norm": 0.6568521014876724, | |
| "learning_rate": 3.9999280366234394e-05, | |
| "loss": 0.4622, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5182186234817814, | |
| "grad_norm": 0.7411642569272227, | |
| "learning_rate": 3.999872065705065e-05, | |
| "loss": 0.4464, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5222672064777328, | |
| "grad_norm": 0.7477138136398503, | |
| "learning_rate": 3.999800103862945e-05, | |
| "loss": 0.4494, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.9066127901534308, | |
| "learning_rate": 3.9997121516724834e-05, | |
| "loss": 0.4488, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5303643724696356, | |
| "grad_norm": 1.0019044648810156, | |
| "learning_rate": 3.999608209836944e-05, | |
| "loss": 0.4523, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5344129554655871, | |
| "grad_norm": 0.5537565105652281, | |
| "learning_rate": 3.999488279187443e-05, | |
| "loss": 0.4606, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.9183468632462392, | |
| "learning_rate": 3.9993523606829425e-05, | |
| "loss": 0.4577, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5425101214574899, | |
| "grad_norm": 1.1508162585519515, | |
| "learning_rate": 3.9992004554102435e-05, | |
| "loss": 0.4521, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5465587044534413, | |
| "grad_norm": 0.4266330283000921, | |
| "learning_rate": 3.999032564583976e-05, | |
| "loss": 0.4512, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5506072874493927, | |
| "grad_norm": 1.0781955883553478, | |
| "learning_rate": 3.998848689546591e-05, | |
| "loss": 0.4606, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5546558704453441, | |
| "grad_norm": 0.6585819857781248, | |
| "learning_rate": 3.9986488317683496e-05, | |
| "loss": 0.4538, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5587044534412956, | |
| "grad_norm": 0.6815419827917019, | |
| "learning_rate": 3.9984329928473067e-05, | |
| "loss": 0.4547, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.562753036437247, | |
| "grad_norm": 0.6586346754937313, | |
| "learning_rate": 3.998201174509306e-05, | |
| "loss": 0.4618, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5668016194331984, | |
| "grad_norm": 0.41166193058364514, | |
| "learning_rate": 3.997953378607961e-05, | |
| "loss": 0.4466, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5708502024291497, | |
| "grad_norm": 0.6050774465408211, | |
| "learning_rate": 3.9976896071246403e-05, | |
| "loss": 0.4455, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5748987854251012, | |
| "grad_norm": 0.6358565209802587, | |
| "learning_rate": 3.997409862168454e-05, | |
| "loss": 0.4588, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.47496703804320733, | |
| "learning_rate": 3.997114145976237e-05, | |
| "loss": 0.4504, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.582995951417004, | |
| "grad_norm": 0.5176882002227109, | |
| "learning_rate": 3.9968024609125245e-05, | |
| "loss": 0.4499, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5870445344129555, | |
| "grad_norm": 0.43448027809410356, | |
| "learning_rate": 3.996474809469545e-05, | |
| "loss": 0.4349, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5910931174089069, | |
| "grad_norm": 0.561159763209722, | |
| "learning_rate": 3.996131194267189e-05, | |
| "loss": 0.4582, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5951417004048583, | |
| "grad_norm": 0.5893782625250995, | |
| "learning_rate": 3.995771618052995e-05, | |
| "loss": 0.4504, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5991902834008097, | |
| "grad_norm": 0.44993713151571274, | |
| "learning_rate": 3.9953960837021256e-05, | |
| "loss": 0.4447, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6032388663967612, | |
| "grad_norm": 0.4975245344069128, | |
| "learning_rate": 3.995004594217343e-05, | |
| "loss": 0.4512, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6072874493927125, | |
| "grad_norm": 0.6154038290598283, | |
| "learning_rate": 3.994597152728987e-05, | |
| "loss": 0.4564, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.611336032388664, | |
| "grad_norm": 0.5038534763863489, | |
| "learning_rate": 3.994173762494947e-05, | |
| "loss": 0.443, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.6749781560159305, | |
| "learning_rate": 3.9937344269006424e-05, | |
| "loss": 0.4469, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6194331983805668, | |
| "grad_norm": 0.6020495480364623, | |
| "learning_rate": 3.993279149458988e-05, | |
| "loss": 0.4406, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6234817813765182, | |
| "grad_norm": 0.5484025659599796, | |
| "learning_rate": 3.9928079338103695e-05, | |
| "loss": 0.4386, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6275303643724697, | |
| "grad_norm": 0.46077771712039817, | |
| "learning_rate": 3.992320783722616e-05, | |
| "loss": 0.4506, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.5613912333326866, | |
| "learning_rate": 3.991817703090965e-05, | |
| "loss": 0.4557, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6356275303643725, | |
| "grad_norm": 0.6008322670977041, | |
| "learning_rate": 3.991298695938038e-05, | |
| "loss": 0.4393, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6396761133603239, | |
| "grad_norm": 0.6217509966833146, | |
| "learning_rate": 3.9907637664138e-05, | |
| "loss": 0.4542, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6437246963562753, | |
| "grad_norm": 0.5619137031216286, | |
| "learning_rate": 3.990212918795534e-05, | |
| "loss": 0.4442, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6477732793522267, | |
| "grad_norm": 0.5531343475777329, | |
| "learning_rate": 3.9896461574878036e-05, | |
| "loss": 0.4509, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6518218623481782, | |
| "grad_norm": 0.5133577961077193, | |
| "learning_rate": 3.989063487022416e-05, | |
| "loss": 0.4428, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6558704453441295, | |
| "grad_norm": 0.6482032868046634, | |
| "learning_rate": 3.9884649120583903e-05, | |
| "loss": 0.4432, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.659919028340081, | |
| "grad_norm": 0.5086534673975972, | |
| "learning_rate": 3.987850437381915e-05, | |
| "loss": 0.4324, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6639676113360324, | |
| "grad_norm": 0.6044028745867187, | |
| "learning_rate": 3.987220067906313e-05, | |
| "loss": 0.4394, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6680161943319838, | |
| "grad_norm": 0.6077367617822153, | |
| "learning_rate": 3.986573808672002e-05, | |
| "loss": 0.4485, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6720647773279352, | |
| "grad_norm": 0.5598169698033242, | |
| "learning_rate": 3.9859116648464534e-05, | |
| "loss": 0.4505, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6761133603238867, | |
| "grad_norm": 0.49776058698695946, | |
| "learning_rate": 3.985233641724152e-05, | |
| "loss": 0.4419, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.680161943319838, | |
| "grad_norm": 0.5329528956995236, | |
| "learning_rate": 3.984539744726554e-05, | |
| "loss": 0.4439, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.5807918507676314, | |
| "learning_rate": 3.983829979402038e-05, | |
| "loss": 0.433, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6882591093117408, | |
| "grad_norm": 0.6122217109374243, | |
| "learning_rate": 3.983104351425871e-05, | |
| "loss": 0.4471, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.5429306778918767, | |
| "learning_rate": 3.982362866600154e-05, | |
| "loss": 0.441, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6963562753036437, | |
| "grad_norm": 0.489513994110238, | |
| "learning_rate": 3.9816055308537806e-05, | |
| "loss": 0.4488, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7004048582995951, | |
| "grad_norm": 0.45384818989648934, | |
| "learning_rate": 3.980832350242388e-05, | |
| "loss": 0.4477, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7044534412955465, | |
| "grad_norm": 0.3992001588085601, | |
| "learning_rate": 3.980043330948306e-05, | |
| "loss": 0.4501, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.708502024291498, | |
| "grad_norm": 0.4610443516775671, | |
| "learning_rate": 3.9792384792805124e-05, | |
| "loss": 0.4452, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7125506072874493, | |
| "grad_norm": 0.41482612865883206, | |
| "learning_rate": 3.978417801674579e-05, | |
| "loss": 0.4436, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7165991902834008, | |
| "grad_norm": 0.4113947746096098, | |
| "learning_rate": 3.9775813046926244e-05, | |
| "loss": 0.4383, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7206477732793523, | |
| "grad_norm": 0.4622833506838898, | |
| "learning_rate": 3.976728995023254e-05, | |
| "loss": 0.4422, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7246963562753036, | |
| "grad_norm": 0.4100084112599449, | |
| "learning_rate": 3.975860879481514e-05, | |
| "loss": 0.446, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.728744939271255, | |
| "grad_norm": 0.539825590128671, | |
| "learning_rate": 3.974976965008833e-05, | |
| "loss": 0.4519, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7327935222672065, | |
| "grad_norm": 0.447354260958095, | |
| "learning_rate": 3.9740772586729674e-05, | |
| "loss": 0.4334, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.4852542766250877, | |
| "learning_rate": 3.973161767667946e-05, | |
| "loss": 0.4429, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7408906882591093, | |
| "grad_norm": 0.5543148762965059, | |
| "learning_rate": 3.972230499314009e-05, | |
| "loss": 0.4229, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7449392712550608, | |
| "grad_norm": 0.608838109360762, | |
| "learning_rate": 3.9712834610575545e-05, | |
| "loss": 0.4382, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7489878542510121, | |
| "grad_norm": 0.49714624600691604, | |
| "learning_rate": 3.9703206604710746e-05, | |
| "loss": 0.4212, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7530364372469636, | |
| "grad_norm": 0.46181183808022375, | |
| "learning_rate": 3.969342105253098e-05, | |
| "loss": 0.4373, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.757085020242915, | |
| "grad_norm": 0.42902823617283126, | |
| "learning_rate": 3.9683478032281246e-05, | |
| "loss": 0.4519, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7611336032388664, | |
| "grad_norm": 0.41210717532441055, | |
| "learning_rate": 3.967337762346568e-05, | |
| "loss": 0.442, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7651821862348178, | |
| "grad_norm": 0.4778847454347151, | |
| "learning_rate": 3.9663119906846885e-05, | |
| "loss": 0.429, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.4297003471933121, | |
| "learning_rate": 3.965270496444528e-05, | |
| "loss": 0.4385, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7732793522267206, | |
| "grad_norm": 0.458740187901282, | |
| "learning_rate": 3.9642132879538465e-05, | |
| "loss": 0.4449, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7773279352226721, | |
| "grad_norm": 0.4621590773935913, | |
| "learning_rate": 3.9631403736660566e-05, | |
| "loss": 0.4651, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7813765182186235, | |
| "grad_norm": 0.4386036216440961, | |
| "learning_rate": 3.962051762160151e-05, | |
| "loss": 0.4376, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7854251012145749, | |
| "grad_norm": 0.36655084147233413, | |
| "learning_rate": 3.960947462140638e-05, | |
| "loss": 0.4245, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.44635970811271947, | |
| "learning_rate": 3.959827482437472e-05, | |
| "loss": 0.4343, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7935222672064778, | |
| "grad_norm": 0.4108358658054732, | |
| "learning_rate": 3.958691832005981e-05, | |
| "loss": 0.4383, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7975708502024291, | |
| "grad_norm": 0.4489609292710576, | |
| "learning_rate": 3.9575405199267944e-05, | |
| "loss": 0.4294, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8016194331983806, | |
| "grad_norm": 0.5163438678081123, | |
| "learning_rate": 3.956373555405774e-05, | |
| "loss": 0.4338, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.805668016194332, | |
| "grad_norm": 0.4222812681143289, | |
| "learning_rate": 3.955190947773937e-05, | |
| "loss": 0.4272, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8097165991902834, | |
| "grad_norm": 0.509819520464986, | |
| "learning_rate": 3.9539927064873824e-05, | |
| "loss": 0.4484, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8137651821862348, | |
| "grad_norm": 0.4330058062742453, | |
| "learning_rate": 3.952778841127214e-05, | |
| "loss": 0.4491, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8178137651821862, | |
| "grad_norm": 0.4666592469719419, | |
| "learning_rate": 3.95154936139947e-05, | |
| "loss": 0.4368, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8218623481781376, | |
| "grad_norm": 0.48667157279909096, | |
| "learning_rate": 3.950304277135034e-05, | |
| "loss": 0.433, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8259109311740891, | |
| "grad_norm": 0.535577330906739, | |
| "learning_rate": 3.94904359828957e-05, | |
| "loss": 0.4501, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8299595141700404, | |
| "grad_norm": 0.5864644910243487, | |
| "learning_rate": 3.94776733494343e-05, | |
| "loss": 0.4559, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8340080971659919, | |
| "grad_norm": 0.5008666962588201, | |
| "learning_rate": 3.9464754973015826e-05, | |
| "loss": 0.4329, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8380566801619433, | |
| "grad_norm": 0.5004309093241015, | |
| "learning_rate": 3.9451680956935275e-05, | |
| "loss": 0.4265, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.4656427425098152, | |
| "learning_rate": 3.943845140573213e-05, | |
| "loss": 0.4532, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.45850069450600806, | |
| "learning_rate": 3.942506642518952e-05, | |
| "loss": 0.4305, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8502024291497976, | |
| "grad_norm": 0.5087949945159534, | |
| "learning_rate": 3.9411526122333396e-05, | |
| "loss": 0.4325, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.854251012145749, | |
| "grad_norm": 0.5052537277267231, | |
| "learning_rate": 3.939783060543167e-05, | |
| "loss": 0.4161, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8582995951417004, | |
| "grad_norm": 0.4607466811254976, | |
| "learning_rate": 3.938397998399333e-05, | |
| "loss": 0.4381, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8623481781376519, | |
| "grad_norm": 0.4451193622397571, | |
| "learning_rate": 3.936997436876756e-05, | |
| "loss": 0.4437, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8663967611336032, | |
| "grad_norm": 0.42944989238556625, | |
| "learning_rate": 3.935581387174289e-05, | |
| "loss": 0.4443, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8704453441295547, | |
| "grad_norm": 0.4766507566679157, | |
| "learning_rate": 3.9341498606146295e-05, | |
| "loss": 0.4323, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8744939271255061, | |
| "grad_norm": 0.4697803706216493, | |
| "learning_rate": 3.932702868644224e-05, | |
| "loss": 0.4303, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8785425101214575, | |
| "grad_norm": 0.545680067159056, | |
| "learning_rate": 3.931240422833181e-05, | |
| "loss": 0.4476, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8825910931174089, | |
| "grad_norm": 0.5263629892735494, | |
| "learning_rate": 3.9297625348751806e-05, | |
| "loss": 0.4321, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8866396761133604, | |
| "grad_norm": 0.4684985976839844, | |
| "learning_rate": 3.9282692165873744e-05, | |
| "loss": 0.4479, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8906882591093117, | |
| "grad_norm": 0.44539822465280443, | |
| "learning_rate": 3.9267604799102945e-05, | |
| "loss": 0.4224, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.4491326816696796, | |
| "learning_rate": 3.9252363369077615e-05, | |
| "loss": 0.4403, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8987854251012146, | |
| "grad_norm": 0.3972207906628645, | |
| "learning_rate": 3.9236967997667796e-05, | |
| "loss": 0.4376, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.902834008097166, | |
| "grad_norm": 0.5182581851342648, | |
| "learning_rate": 3.9221418807974496e-05, | |
| "loss": 0.4367, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9068825910931174, | |
| "grad_norm": 0.6036272964013776, | |
| "learning_rate": 3.92057159243286e-05, | |
| "loss": 0.435, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9109311740890689, | |
| "grad_norm": 0.5001066734851095, | |
| "learning_rate": 3.9189859472289956e-05, | |
| "loss": 0.4328, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9149797570850202, | |
| "grad_norm": 0.5154625660560965, | |
| "learning_rate": 3.9173849578646324e-05, | |
| "loss": 0.435, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9190283400809717, | |
| "grad_norm": 0.5633679897485616, | |
| "learning_rate": 3.9157686371412395e-05, | |
| "loss": 0.4398, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.6949421827936314, | |
| "learning_rate": 3.914136997982874e-05, | |
| "loss": 0.4466, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9271255060728745, | |
| "grad_norm": 0.5877313012378997, | |
| "learning_rate": 3.912490053436079e-05, | |
| "loss": 0.4211, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9311740890688259, | |
| "grad_norm": 0.3967264435386232, | |
| "learning_rate": 3.9108278166697805e-05, | |
| "loss": 0.4465, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9352226720647774, | |
| "grad_norm": 0.6655263616834635, | |
| "learning_rate": 3.909150300975178e-05, | |
| "loss": 0.4427, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9392712550607287, | |
| "grad_norm": 0.7288094578336717, | |
| "learning_rate": 3.907457519765642e-05, | |
| "loss": 0.4441, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9433198380566802, | |
| "grad_norm": 0.6026974537641574, | |
| "learning_rate": 3.905749486576607e-05, | |
| "loss": 0.4359, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.6415040310062975, | |
| "learning_rate": 3.90402621506546e-05, | |
| "loss": 0.43, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.951417004048583, | |
| "grad_norm": 0.5646134478334399, | |
| "learning_rate": 3.902287719011435e-05, | |
| "loss": 0.43, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9554655870445344, | |
| "grad_norm": 0.45333265565302266, | |
| "learning_rate": 3.900534012315499e-05, | |
| "loss": 0.4175, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9595141700404858, | |
| "grad_norm": 0.5154819086776291, | |
| "learning_rate": 3.8987651090002454e-05, | |
| "loss": 0.4244, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9635627530364372, | |
| "grad_norm": 0.5209423949434497, | |
| "learning_rate": 3.896981023209777e-05, | |
| "loss": 0.4335, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9676113360323887, | |
| "grad_norm": 0.5009256669619865, | |
| "learning_rate": 3.8951817692095966e-05, | |
| "loss": 0.4373, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.97165991902834, | |
| "grad_norm": 0.3937611541076344, | |
| "learning_rate": 3.893367361386491e-05, | |
| "loss": 0.4299, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9757085020242915, | |
| "grad_norm": 0.3949175043224698, | |
| "learning_rate": 3.891537814248417e-05, | |
| "loss": 0.4332, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.979757085020243, | |
| "grad_norm": 0.4526934407307224, | |
| "learning_rate": 3.889693142424384e-05, | |
| "loss": 0.4376, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9838056680161943, | |
| "grad_norm": 0.4138857340930773, | |
| "learning_rate": 3.8878333606643405e-05, | |
| "loss": 0.4238, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9878542510121457, | |
| "grad_norm": 0.40755115435487915, | |
| "learning_rate": 3.885958483839049e-05, | |
| "loss": 0.4345, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9919028340080972, | |
| "grad_norm": 0.5211427650321693, | |
| "learning_rate": 3.8840685269399786e-05, | |
| "loss": 0.4386, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9959514170040485, | |
| "grad_norm": 0.47243481420854416, | |
| "learning_rate": 3.882163505079171e-05, | |
| "loss": 0.427, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.35773320816419935, | |
| "learning_rate": 3.880243433489132e-05, | |
| "loss": 0.4421, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.0040485829959513, | |
| "grad_norm": 0.489404032132759, | |
| "learning_rate": 3.878308327522702e-05, | |
| "loss": 0.3931, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.008097165991903, | |
| "grad_norm": 0.5194114427567074, | |
| "learning_rate": 3.876358202652936e-05, | |
| "loss": 0.3959, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0121457489878543, | |
| "grad_norm": 0.5335032850492061, | |
| "learning_rate": 3.8743930744729835e-05, | |
| "loss": 0.3952, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0161943319838056, | |
| "grad_norm": 0.5481296457522375, | |
| "learning_rate": 3.8724129586959535e-05, | |
| "loss": 0.3863, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0202429149797572, | |
| "grad_norm": 0.5056352007507187, | |
| "learning_rate": 3.870417871154801e-05, | |
| "loss": 0.3977, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0242914979757085, | |
| "grad_norm": 0.5019917492463581, | |
| "learning_rate": 3.868407827802192e-05, | |
| "loss": 0.3936, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.0283400809716599, | |
| "grad_norm": 0.48748516119975405, | |
| "learning_rate": 3.8663828447103804e-05, | |
| "loss": 0.3933, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0323886639676114, | |
| "grad_norm": 0.5955515970364735, | |
| "learning_rate": 3.8643429380710765e-05, | |
| "loss": 0.3914, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0364372469635628, | |
| "grad_norm": 0.5214207176430747, | |
| "learning_rate": 3.862288124195319e-05, | |
| "loss": 0.3829, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.040485829959514, | |
| "grad_norm": 0.4148523499741428, | |
| "learning_rate": 3.8602184195133464e-05, | |
| "loss": 0.3863, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0445344129554657, | |
| "grad_norm": 0.5463832054574079, | |
| "learning_rate": 3.8581338405744604e-05, | |
| "loss": 0.3806, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.048582995951417, | |
| "grad_norm": 0.44635202447266076, | |
| "learning_rate": 3.8560344040468996e-05, | |
| "loss": 0.4074, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.5703324051333623, | |
| "learning_rate": 3.853920126717703e-05, | |
| "loss": 0.3771, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.05668016194332, | |
| "grad_norm": 0.47510025528077565, | |
| "learning_rate": 3.851791025492574e-05, | |
| "loss": 0.3963, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0607287449392713, | |
| "grad_norm": 0.4745896328095981, | |
| "learning_rate": 3.8496471173957515e-05, | |
| "loss": 0.396, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0647773279352226, | |
| "grad_norm": 0.5903498819665342, | |
| "learning_rate": 3.847488419569866e-05, | |
| "loss": 0.3947, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0688259109311742, | |
| "grad_norm": 0.33713672077518875, | |
| "learning_rate": 3.8453149492758077e-05, | |
| "loss": 0.3759, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0728744939271255, | |
| "grad_norm": 0.5424267396088538, | |
| "learning_rate": 3.8431267238925886e-05, | |
| "loss": 0.3855, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.50444999504061, | |
| "learning_rate": 3.840923760917199e-05, | |
| "loss": 0.4013, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0809716599190284, | |
| "grad_norm": 0.34702453167902747, | |
| "learning_rate": 3.838706077964473e-05, | |
| "loss": 0.3903, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0850202429149798, | |
| "grad_norm": 0.4791636493982939, | |
| "learning_rate": 3.836473692766944e-05, | |
| "loss": 0.3871, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0890688259109311, | |
| "grad_norm": 0.39062152686321006, | |
| "learning_rate": 3.834226623174705e-05, | |
| "loss": 0.3974, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0931174089068827, | |
| "grad_norm": 0.4160506094890953, | |
| "learning_rate": 3.831964887155264e-05, | |
| "loss": 0.4077, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.097165991902834, | |
| "grad_norm": 0.4199690629243598, | |
| "learning_rate": 3.829688502793403e-05, | |
| "loss": 0.3844, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.1012145748987854, | |
| "grad_norm": 0.3755403615409549, | |
| "learning_rate": 3.827397488291029e-05, | |
| "loss": 0.3945, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.3650392745331637, | |
| "learning_rate": 3.8250918619670343e-05, | |
| "loss": 0.385, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.1093117408906883, | |
| "grad_norm": 0.37992470987208904, | |
| "learning_rate": 3.822771642257145e-05, | |
| "loss": 0.3751, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1133603238866396, | |
| "grad_norm": 0.4393800605164766, | |
| "learning_rate": 3.820436847713776e-05, | |
| "loss": 0.3873, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.117408906882591, | |
| "grad_norm": 0.38000912270284193, | |
| "learning_rate": 3.8180874970058826e-05, | |
| "loss": 0.3983, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.1214574898785425, | |
| "grad_norm": 0.45167750569377296, | |
| "learning_rate": 3.8157236089188103e-05, | |
| "loss": 0.3834, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.125506072874494, | |
| "grad_norm": 0.4064914698707632, | |
| "learning_rate": 3.813345202354145e-05, | |
| "loss": 0.3899, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1295546558704452, | |
| "grad_norm": 0.4072039562549536, | |
| "learning_rate": 3.810952296329563e-05, | |
| "loss": 0.397, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1336032388663968, | |
| "grad_norm": 0.4047869994476749, | |
| "learning_rate": 3.808544909978676e-05, | |
| "loss": 0.3899, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1376518218623481, | |
| "grad_norm": 0.37835024508460385, | |
| "learning_rate": 3.806123062550882e-05, | |
| "loss": 0.389, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1417004048582995, | |
| "grad_norm": 0.35444301642283504, | |
| "learning_rate": 3.803686773411208e-05, | |
| "loss": 0.3999, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.145748987854251, | |
| "grad_norm": 0.37878337553488317, | |
| "learning_rate": 3.801236062040158e-05, | |
| "loss": 0.4035, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1497975708502024, | |
| "grad_norm": 0.42698828865059557, | |
| "learning_rate": 3.798770948033553e-05, | |
| "loss": 0.3909, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.42242009639081335, | |
| "learning_rate": 3.796291451102381e-05, | |
| "loss": 0.3887, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.3593453913882006, | |
| "learning_rate": 3.793797591072631e-05, | |
| "loss": 0.3735, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1619433198380567, | |
| "grad_norm": 0.4108280914030158, | |
| "learning_rate": 3.791289387885144e-05, | |
| "loss": 0.3811, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.165991902834008, | |
| "grad_norm": 0.37551376257989316, | |
| "learning_rate": 3.788766861595443e-05, | |
| "loss": 0.3796, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1700404858299596, | |
| "grad_norm": 0.32613943848608257, | |
| "learning_rate": 3.7862300323735835e-05, | |
| "loss": 0.3928, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.174089068825911, | |
| "grad_norm": 0.4448144414092819, | |
| "learning_rate": 3.783678920503983e-05, | |
| "loss": 0.3849, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1781376518218623, | |
| "grad_norm": 0.47043713909203294, | |
| "learning_rate": 3.781113546385265e-05, | |
| "loss": 0.4081, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1821862348178138, | |
| "grad_norm": 0.4680091682878849, | |
| "learning_rate": 3.7785339305300916e-05, | |
| "loss": 0.3984, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1862348178137652, | |
| "grad_norm": 0.4023765190873225, | |
| "learning_rate": 3.7759400935650025e-05, | |
| "loss": 0.3945, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1902834008097165, | |
| "grad_norm": 0.38144894356699754, | |
| "learning_rate": 3.77333205623025e-05, | |
| "loss": 0.3892, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.194331983805668, | |
| "grad_norm": 0.44058388368238127, | |
| "learning_rate": 3.7707098393796295e-05, | |
| "loss": 0.3884, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1983805668016194, | |
| "grad_norm": 0.3482381346597427, | |
| "learning_rate": 3.7680734639803185e-05, | |
| "loss": 0.411, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.2024291497975708, | |
| "grad_norm": 0.4891004496435386, | |
| "learning_rate": 3.765422951112704e-05, | |
| "loss": 0.3901, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2064777327935223, | |
| "grad_norm": 0.41626057575954595, | |
| "learning_rate": 3.762758321970216e-05, | |
| "loss": 0.3937, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.46072926189600255, | |
| "learning_rate": 3.7600795978591584e-05, | |
| "loss": 0.3882, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.214574898785425, | |
| "grad_norm": 0.42631031327225977, | |
| "learning_rate": 3.757386800198538e-05, | |
| "loss": 0.3833, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2186234817813766, | |
| "grad_norm": 0.49438512966616854, | |
| "learning_rate": 3.7546799505198925e-05, | |
| "loss": 0.4005, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.222672064777328, | |
| "grad_norm": 0.6655475010690797, | |
| "learning_rate": 3.7519590704671197e-05, | |
| "loss": 0.3897, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2267206477732793, | |
| "grad_norm": 0.4671513933579039, | |
| "learning_rate": 3.749224181796305e-05, | |
| "loss": 0.3917, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.47136953509053764, | |
| "learning_rate": 3.746475306375545e-05, | |
| "loss": 0.3834, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2348178137651822, | |
| "grad_norm": 0.5716161153617269, | |
| "learning_rate": 3.743712466184774e-05, | |
| "loss": 0.3858, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2388663967611335, | |
| "grad_norm": 0.4529930708455496, | |
| "learning_rate": 3.7409356833155885e-05, | |
| "loss": 0.3969, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.242914979757085, | |
| "grad_norm": 0.5238643459111187, | |
| "learning_rate": 3.73814497997107e-05, | |
| "loss": 0.3952, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2469635627530364, | |
| "grad_norm": 0.576758963609638, | |
| "learning_rate": 3.7353403784656095e-05, | |
| "loss": 0.4054, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2510121457489878, | |
| "grad_norm": 0.5284448658645027, | |
| "learning_rate": 3.732521901224724e-05, | |
| "loss": 0.3923, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2550607287449393, | |
| "grad_norm": 0.49936732089952646, | |
| "learning_rate": 3.7296895707848844e-05, | |
| "loss": 0.4021, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2591093117408907, | |
| "grad_norm": 0.5425601070835686, | |
| "learning_rate": 3.7268434097933275e-05, | |
| "loss": 0.3939, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.709082480501563, | |
| "learning_rate": 3.723983441007882e-05, | |
| "loss": 0.4061, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.2672064777327936, | |
| "grad_norm": 0.48452390727942074, | |
| "learning_rate": 3.721109687296781e-05, | |
| "loss": 0.3911, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.271255060728745, | |
| "grad_norm": 0.4112368108613164, | |
| "learning_rate": 3.718222171638484e-05, | |
| "loss": 0.397, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2753036437246963, | |
| "grad_norm": 0.477866968628677, | |
| "learning_rate": 3.715320917121488e-05, | |
| "loss": 0.3917, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2793522267206479, | |
| "grad_norm": 0.490753759548234, | |
| "learning_rate": 3.712405946944149e-05, | |
| "loss": 0.4087, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.2834008097165992, | |
| "grad_norm": 0.33252721718395123, | |
| "learning_rate": 3.709477284414488e-05, | |
| "loss": 0.3926, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2874493927125505, | |
| "grad_norm": 0.3766663534123375, | |
| "learning_rate": 3.7065349529500136e-05, | |
| "loss": 0.393, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.291497975708502, | |
| "grad_norm": 0.4035154829076413, | |
| "learning_rate": 3.7035789760775305e-05, | |
| "loss": 0.3919, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2955465587044535, | |
| "grad_norm": 0.4350500611344918, | |
| "learning_rate": 3.70060937743295e-05, | |
| "loss": 0.3882, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2995951417004048, | |
| "grad_norm": 0.40088918708950566, | |
| "learning_rate": 3.6976261807611045e-05, | |
| "loss": 0.3808, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.3036437246963564, | |
| "grad_norm": 0.3779613817323891, | |
| "learning_rate": 3.694629409915555e-05, | |
| "loss": 0.389, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.4630836866809073, | |
| "learning_rate": 3.691619088858401e-05, | |
| "loss": 0.3816, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.311740890688259, | |
| "grad_norm": 0.5054631747778533, | |
| "learning_rate": 3.68859524166009e-05, | |
| "loss": 0.4069, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.3623197818598726, | |
| "learning_rate": 3.6855578924992226e-05, | |
| "loss": 0.3974, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.319838056680162, | |
| "grad_norm": 0.5290275357371962, | |
| "learning_rate": 3.6825070656623626e-05, | |
| "loss": 0.3845, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.3238866396761133, | |
| "grad_norm": 0.5367955758633686, | |
| "learning_rate": 3.67944278554384e-05, | |
| "loss": 0.3971, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3279352226720649, | |
| "grad_norm": 0.39326317761018825, | |
| "learning_rate": 3.676365076645557e-05, | |
| "loss": 0.3781, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3319838056680162, | |
| "grad_norm": 0.5090523704221434, | |
| "learning_rate": 3.673273963576791e-05, | |
| "loss": 0.3873, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3360323886639676, | |
| "grad_norm": 0.35881444521764744, | |
| "learning_rate": 3.670169471054001e-05, | |
| "loss": 0.3828, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3400809716599191, | |
| "grad_norm": 0.3915253461090147, | |
| "learning_rate": 3.6670516239006254e-05, | |
| "loss": 0.3935, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3441295546558705, | |
| "grad_norm": 0.3972511424267767, | |
| "learning_rate": 3.663920447046888e-05, | |
| "loss": 0.3958, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3481781376518218, | |
| "grad_norm": 0.4075856636804219, | |
| "learning_rate": 3.6607759655295954e-05, | |
| "loss": 0.405, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3522267206477734, | |
| "grad_norm": 0.4393096137096971, | |
| "learning_rate": 3.657618204491937e-05, | |
| "loss": 0.3917, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3562753036437247, | |
| "grad_norm": 0.29581095484721986, | |
| "learning_rate": 3.6544471891832875e-05, | |
| "loss": 0.3869, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.360323886639676, | |
| "grad_norm": 0.38823402992139017, | |
| "learning_rate": 3.6512629449589983e-05, | |
| "loss": 0.3929, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3643724696356276, | |
| "grad_norm": 0.322159532099556, | |
| "learning_rate": 3.6480654972802045e-05, | |
| "loss": 0.3877, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.35202599196819706, | |
| "learning_rate": 3.644854871713611e-05, | |
| "loss": 0.3932, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3724696356275303, | |
| "grad_norm": 0.35817304437398095, | |
| "learning_rate": 3.6416310939312924e-05, | |
| "loss": 0.3747, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.376518218623482, | |
| "grad_norm": 0.3115854363797023, | |
| "learning_rate": 3.638394189710493e-05, | |
| "loss": 0.3982, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3805668016194332, | |
| "grad_norm": 0.3219694899573126, | |
| "learning_rate": 3.635144184933412e-05, | |
| "loss": 0.3831, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.2880711495115437, | |
| "learning_rate": 3.6318811055869994e-05, | |
| "loss": 0.3791, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3886639676113361, | |
| "grad_norm": 0.2831792364698526, | |
| "learning_rate": 3.628604977762752e-05, | |
| "loss": 0.388, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3927125506072875, | |
| "grad_norm": 0.3098540887814925, | |
| "learning_rate": 3.6253158276565004e-05, | |
| "loss": 0.3858, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.3967611336032388, | |
| "grad_norm": 0.3502085572117459, | |
| "learning_rate": 3.6220136815682016e-05, | |
| "loss": 0.3869, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.4008097165991904, | |
| "grad_norm": 0.34737898928774086, | |
| "learning_rate": 3.618698565901727e-05, | |
| "loss": 0.393, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.4048582995951417, | |
| "grad_norm": 0.3728932348039422, | |
| "learning_rate": 3.615370507164654e-05, | |
| "loss": 0.385, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.408906882591093, | |
| "grad_norm": 0.4059119491399108, | |
| "learning_rate": 3.61202953196805e-05, | |
| "loss": 0.3785, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.4129554655870447, | |
| "grad_norm": 0.4834308723434812, | |
| "learning_rate": 3.6086756670262654e-05, | |
| "loss": 0.397, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.417004048582996, | |
| "grad_norm": 0.4391647809868069, | |
| "learning_rate": 3.6053089391567135e-05, | |
| "loss": 0.3782, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.3625201194727759, | |
| "learning_rate": 3.601929375279662e-05, | |
| "loss": 0.3791, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.425101214574899, | |
| "grad_norm": 0.4758971904753753, | |
| "learning_rate": 3.598537002418012e-05, | |
| "loss": 0.3919, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4291497975708503, | |
| "grad_norm": 0.3205222000770008, | |
| "learning_rate": 3.5951318476970885e-05, | |
| "loss": 0.3754, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4331983805668016, | |
| "grad_norm": 0.41452065063552807, | |
| "learning_rate": 3.591713938344416e-05, | |
| "loss": 0.3871, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4372469635627532, | |
| "grad_norm": 0.3092826045173501, | |
| "learning_rate": 3.588283301689507e-05, | |
| "loss": 0.3952, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4412955465587045, | |
| "grad_norm": 0.40523495874743687, | |
| "learning_rate": 3.5848399651636424e-05, | |
| "loss": 0.3823, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4453441295546559, | |
| "grad_norm": 0.32873696185908946, | |
| "learning_rate": 3.5813839562996473e-05, | |
| "loss": 0.3851, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4493927125506074, | |
| "grad_norm": 0.3603411737394211, | |
| "learning_rate": 3.5779153027316775e-05, | |
| "loss": 0.396, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4534412955465588, | |
| "grad_norm": 0.38491141759527053, | |
| "learning_rate": 3.574434032194994e-05, | |
| "loss": 0.39, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.45748987854251, | |
| "grad_norm": 0.3524137499084572, | |
| "learning_rate": 3.5709401725257436e-05, | |
| "loss": 0.3966, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.4047475236643499, | |
| "learning_rate": 3.5674337516607346e-05, | |
| "loss": 0.3873, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.465587044534413, | |
| "grad_norm": 0.40935884468006206, | |
| "learning_rate": 3.563914797637216e-05, | |
| "loss": 0.3978, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4696356275303644, | |
| "grad_norm": 0.4300708987029868, | |
| "learning_rate": 3.560383338592649e-05, | |
| "loss": 0.4043, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.39180596466885825, | |
| "learning_rate": 3.556839402764487e-05, | |
| "loss": 0.3932, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4777327935222673, | |
| "grad_norm": 0.40861579066347803, | |
| "learning_rate": 3.553283018489946e-05, | |
| "loss": 0.3786, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4817813765182186, | |
| "grad_norm": 0.308108991878252, | |
| "learning_rate": 3.54971421420578e-05, | |
| "loss": 0.3967, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.48582995951417, | |
| "grad_norm": 0.3901308567310717, | |
| "learning_rate": 3.5461330184480524e-05, | |
| "loss": 0.398, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4898785425101215, | |
| "grad_norm": 0.4158922417839989, | |
| "learning_rate": 3.54253945985191e-05, | |
| "loss": 0.3816, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4939271255060729, | |
| "grad_norm": 0.3786762443502148, | |
| "learning_rate": 3.5389335671513524e-05, | |
| "loss": 0.3838, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4979757085020242, | |
| "grad_norm": 0.3238921667358378, | |
| "learning_rate": 3.5353153691789986e-05, | |
| "loss": 0.3836, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.5020242914979756, | |
| "grad_norm": 0.32901505880202087, | |
| "learning_rate": 3.531684894865866e-05, | |
| "loss": 0.3958, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.5060728744939271, | |
| "grad_norm": 0.33558906847363706, | |
| "learning_rate": 3.528042173241128e-05, | |
| "loss": 0.3847, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.5101214574898787, | |
| "grad_norm": 0.3616385849341088, | |
| "learning_rate": 3.524387233431892e-05, | |
| "loss": 0.3833, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.5141700404858298, | |
| "grad_norm": 0.38016259574073274, | |
| "learning_rate": 3.520720104662958e-05, | |
| "loss": 0.387, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.5182186234817814, | |
| "grad_norm": 0.3202199958830996, | |
| "learning_rate": 3.5170408162565904e-05, | |
| "loss": 0.3916, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.522267206477733, | |
| "grad_norm": 0.3760513966077639, | |
| "learning_rate": 3.513349397632281e-05, | |
| "loss": 0.3954, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.36076446857607203, | |
| "learning_rate": 3.5096458783065145e-05, | |
| "loss": 0.3826, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5303643724696356, | |
| "grad_norm": 0.34279122533580525, | |
| "learning_rate": 3.505930287892533e-05, | |
| "loss": 0.3875, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.5344129554655872, | |
| "grad_norm": 0.3127405914991782, | |
| "learning_rate": 3.502202656100099e-05, | |
| "loss": 0.384, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.3357796305828238, | |
| "learning_rate": 3.498463012735256e-05, | |
| "loss": 0.3876, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.54251012145749, | |
| "grad_norm": 0.5009788990301356, | |
| "learning_rate": 3.494711387700094e-05, | |
| "loss": 0.3929, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5465587044534415, | |
| "grad_norm": 0.4647308297167197, | |
| "learning_rate": 3.490947810992505e-05, | |
| "loss": 0.3908, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5506072874493926, | |
| "grad_norm": 0.3274403049092232, | |
| "learning_rate": 3.48717231270595e-05, | |
| "loss": 0.3985, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5546558704453441, | |
| "grad_norm": 0.48893260301806657, | |
| "learning_rate": 3.483384923029211e-05, | |
| "loss": 0.3882, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5587044534412957, | |
| "grad_norm": 0.4636724937003933, | |
| "learning_rate": 3.479585672246156e-05, | |
| "loss": 0.3881, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5627530364372468, | |
| "grad_norm": 0.3627209272965772, | |
| "learning_rate": 3.4757745907354924e-05, | |
| "loss": 0.3834, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5668016194331984, | |
| "grad_norm": 0.39381953783070467, | |
| "learning_rate": 3.471951708970527e-05, | |
| "loss": 0.3866, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5708502024291497, | |
| "grad_norm": 0.4707124858005391, | |
| "learning_rate": 3.468117057518921e-05, | |
| "loss": 0.3948, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.574898785425101, | |
| "grad_norm": 0.4043040309143367, | |
| "learning_rate": 3.4642706670424426e-05, | |
| "loss": 0.4008, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.36022934759338293, | |
| "learning_rate": 3.460412568296731e-05, | |
| "loss": 0.3986, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.582995951417004, | |
| "grad_norm": 0.46849565929545633, | |
| "learning_rate": 3.456542792131039e-05, | |
| "loss": 0.3865, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5870445344129553, | |
| "grad_norm": 0.40174360839764856, | |
| "learning_rate": 3.4526613694879936e-05, | |
| "loss": 0.3704, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.591093117408907, | |
| "grad_norm": 0.4236523755552526, | |
| "learning_rate": 3.448768331403348e-05, | |
| "loss": 0.4035, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5951417004048583, | |
| "grad_norm": 0.4483074103782301, | |
| "learning_rate": 3.44486370900573e-05, | |
| "loss": 0.3736, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5991902834008096, | |
| "grad_norm": 0.3909089583391225, | |
| "learning_rate": 3.440947533516396e-05, | |
| "loss": 0.3857, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.6032388663967612, | |
| "grad_norm": 0.48753000890800025, | |
| "learning_rate": 3.4370198362489816e-05, | |
| "loss": 0.3925, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.6072874493927125, | |
| "grad_norm": 0.47673257052359674, | |
| "learning_rate": 3.43308064860925e-05, | |
| "loss": 0.4027, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.6113360323886639, | |
| "grad_norm": 0.39296407466767763, | |
| "learning_rate": 3.4291300020948393e-05, | |
| "loss": 0.3873, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.40033117783263766, | |
| "learning_rate": 3.425167928295015e-05, | |
| "loss": 0.3894, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.6194331983805668, | |
| "grad_norm": 0.4098657621878783, | |
| "learning_rate": 3.421194458890411e-05, | |
| "loss": 0.3902, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.623481781376518, | |
| "grad_norm": 0.3578622184662877, | |
| "learning_rate": 3.4172096256527865e-05, | |
| "loss": 0.3783, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.6275303643724697, | |
| "grad_norm": 0.32501667848108334, | |
| "learning_rate": 3.4132134604447595e-05, | |
| "loss": 0.377, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.365206293862368, | |
| "learning_rate": 3.409205995219561e-05, | |
| "loss": 0.4011, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6356275303643724, | |
| "grad_norm": 0.37946827555924656, | |
| "learning_rate": 3.4051872620207765e-05, | |
| "loss": 0.3908, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.639676113360324, | |
| "grad_norm": 0.37592001704054057, | |
| "learning_rate": 3.40115729298209e-05, | |
| "loss": 0.3851, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6437246963562753, | |
| "grad_norm": 0.3606085351968378, | |
| "learning_rate": 3.397116120327025e-05, | |
| "loss": 0.3869, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6477732793522266, | |
| "grad_norm": 0.38568756801366105, | |
| "learning_rate": 3.3930637763686916e-05, | |
| "loss": 0.3829, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6518218623481782, | |
| "grad_norm": 0.39619531342131614, | |
| "learning_rate": 3.389000293509524e-05, | |
| "loss": 0.4142, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6558704453441295, | |
| "grad_norm": 0.32868546615218874, | |
| "learning_rate": 3.3849257042410225e-05, | |
| "loss": 0.3813, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6599190283400809, | |
| "grad_norm": 0.39645455388977996, | |
| "learning_rate": 3.380840041143494e-05, | |
| "loss": 0.3846, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6639676113360324, | |
| "grad_norm": 0.33896040291245627, | |
| "learning_rate": 3.376743336885792e-05, | |
| "loss": 0.4005, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6680161943319838, | |
| "grad_norm": 0.3603353037159437, | |
| "learning_rate": 3.372635624225055e-05, | |
| "loss": 0.3872, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6720647773279351, | |
| "grad_norm": 0.3411784074221887, | |
| "learning_rate": 3.3685169360064445e-05, | |
| "loss": 0.389, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6761133603238867, | |
| "grad_norm": 0.3446668950941737, | |
| "learning_rate": 3.3643873051628804e-05, | |
| "loss": 0.3835, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.680161943319838, | |
| "grad_norm": 0.29623518870348714, | |
| "learning_rate": 3.3602467647147814e-05, | |
| "loss": 0.3973, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.32459733667804375, | |
| "learning_rate": 3.356095347769799e-05, | |
| "loss": 0.3788, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.688259109311741, | |
| "grad_norm": 0.3174788391290475, | |
| "learning_rate": 3.351933087522552e-05, | |
| "loss": 0.3868, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.34380977922028744, | |
| "learning_rate": 3.3477600172543626e-05, | |
| "loss": 0.3865, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6963562753036436, | |
| "grad_norm": 0.27108593153818517, | |
| "learning_rate": 3.3435761703329894e-05, | |
| "loss": 0.3647, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.7004048582995952, | |
| "grad_norm": 0.3334694962519658, | |
| "learning_rate": 3.3393815802123606e-05, | |
| "loss": 0.391, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.7044534412955465, | |
| "grad_norm": 0.3326760921076082, | |
| "learning_rate": 3.335176280432308e-05, | |
| "loss": 0.3765, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.708502024291498, | |
| "grad_norm": 0.29777546559269175, | |
| "learning_rate": 3.330960304618295e-05, | |
| "loss": 0.3766, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.7125506072874495, | |
| "grad_norm": 0.3316533966438082, | |
| "learning_rate": 3.326733686481153e-05, | |
| "loss": 0.3916, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.7165991902834008, | |
| "grad_norm": 0.28754217148702593, | |
| "learning_rate": 3.3224964598168084e-05, | |
| "loss": 0.3887, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.7206477732793521, | |
| "grad_norm": 0.3383702712498491, | |
| "learning_rate": 3.318248658506012e-05, | |
| "loss": 0.4, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7246963562753037, | |
| "grad_norm": 0.3280791077348472, | |
| "learning_rate": 3.313990316514069e-05, | |
| "loss": 0.3872, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.728744939271255, | |
| "grad_norm": 0.35426289205074896, | |
| "learning_rate": 3.309721467890571e-05, | |
| "loss": 0.3854, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.7327935222672064, | |
| "grad_norm": 0.29445532910074973, | |
| "learning_rate": 3.305442146769115e-05, | |
| "loss": 0.402, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.4623248409032944, | |
| "learning_rate": 3.30115238736704e-05, | |
| "loss": 0.3809, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7408906882591093, | |
| "grad_norm": 0.4335493128248124, | |
| "learning_rate": 3.296852223985148e-05, | |
| "loss": 0.3876, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7449392712550607, | |
| "grad_norm": 0.43962523783665125, | |
| "learning_rate": 3.2925416910074315e-05, | |
| "loss": 0.3898, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7489878542510122, | |
| "grad_norm": 0.3801262492101867, | |
| "learning_rate": 3.288220822900796e-05, | |
| "loss": 0.3872, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7530364372469636, | |
| "grad_norm": 0.5190571853749719, | |
| "learning_rate": 3.283889654214788e-05, | |
| "loss": 0.3962, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.757085020242915, | |
| "grad_norm": 0.3779245438146458, | |
| "learning_rate": 3.279548219581317e-05, | |
| "loss": 0.381, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7611336032388665, | |
| "grad_norm": 0.48760617262851896, | |
| "learning_rate": 3.275196553714379e-05, | |
| "loss": 0.3949, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7651821862348178, | |
| "grad_norm": 0.43329073846463334, | |
| "learning_rate": 3.270834691409778e-05, | |
| "loss": 0.3874, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.444796689789912, | |
| "learning_rate": 3.266462667544849e-05, | |
| "loss": 0.3744, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7732793522267207, | |
| "grad_norm": 0.46269679490679894, | |
| "learning_rate": 3.262080517078178e-05, | |
| "loss": 0.39, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.777327935222672, | |
| "grad_norm": 0.4043516213535615, | |
| "learning_rate": 3.257688275049323e-05, | |
| "loss": 0.3853, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.7813765182186234, | |
| "grad_norm": 0.4998129432859283, | |
| "learning_rate": 3.253285976578535e-05, | |
| "loss": 0.3887, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.785425101214575, | |
| "grad_norm": 0.42088421350821303, | |
| "learning_rate": 3.2488736568664756e-05, | |
| "loss": 0.3887, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.5167823124330323, | |
| "learning_rate": 3.244451351193935e-05, | |
| "loss": 0.3868, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.7935222672064777, | |
| "grad_norm": 0.41339960002075354, | |
| "learning_rate": 3.240019094921551e-05, | |
| "loss": 0.3809, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7975708502024292, | |
| "grad_norm": 0.46348146629138504, | |
| "learning_rate": 3.235576923489527e-05, | |
| "loss": 0.3919, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.8016194331983806, | |
| "grad_norm": 0.3579035659823205, | |
| "learning_rate": 3.2311248724173476e-05, | |
| "loss": 0.3952, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.805668016194332, | |
| "grad_norm": 0.4184546956223676, | |
| "learning_rate": 3.226662977303494e-05, | |
| "loss": 0.3913, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.8097165991902835, | |
| "grad_norm": 0.4026493520395328, | |
| "learning_rate": 3.2221912738251626e-05, | |
| "loss": 0.3945, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.8137651821862348, | |
| "grad_norm": 0.3502719158062957, | |
| "learning_rate": 3.2177097977379736e-05, | |
| "loss": 0.3884, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.8178137651821862, | |
| "grad_norm": 0.43550769188927974, | |
| "learning_rate": 3.21321858487569e-05, | |
| "loss": 0.3865, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.8218623481781377, | |
| "grad_norm": 0.33613896301628027, | |
| "learning_rate": 3.208717671149931e-05, | |
| "loss": 0.3895, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.825910931174089, | |
| "grad_norm": 0.37992851722412707, | |
| "learning_rate": 3.204207092549882e-05, | |
| "loss": 0.3856, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.8299595141700404, | |
| "grad_norm": 0.3199635802752801, | |
| "learning_rate": 3.19968688514201e-05, | |
| "loss": 0.399, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.834008097165992, | |
| "grad_norm": 0.3955007962263193, | |
| "learning_rate": 3.195157085069774e-05, | |
| "loss": 0.3943, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8380566801619433, | |
| "grad_norm": 0.3442709230852013, | |
| "learning_rate": 3.190617728553332e-05, | |
| "loss": 0.3839, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.3391676357291445, | |
| "learning_rate": 3.1860688518892606e-05, | |
| "loss": 0.3856, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.41853750809830254, | |
| "learning_rate": 3.181510491450254e-05, | |
| "loss": 0.3902, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8502024291497976, | |
| "grad_norm": 0.3884553314941186, | |
| "learning_rate": 3.176942683684842e-05, | |
| "loss": 0.3795, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.854251012145749, | |
| "grad_norm": 0.2897106381248419, | |
| "learning_rate": 3.1723654651170934e-05, | |
| "loss": 0.3941, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8582995951417005, | |
| "grad_norm": 0.4123167055854753, | |
| "learning_rate": 3.167778872346327e-05, | |
| "loss": 0.3881, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8623481781376519, | |
| "grad_norm": 0.323974598236217, | |
| "learning_rate": 3.163182942046815e-05, | |
| "loss": 0.3813, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8663967611336032, | |
| "grad_norm": 0.37906940752988455, | |
| "learning_rate": 3.158577710967494e-05, | |
| "loss": 0.3852, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8704453441295548, | |
| "grad_norm": 0.3168366369124471, | |
| "learning_rate": 3.153963215931669e-05, | |
| "loss": 0.4034, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.874493927125506, | |
| "grad_norm": 0.4039000320607232, | |
| "learning_rate": 3.14933949383672e-05, | |
| "loss": 0.3851, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8785425101214575, | |
| "grad_norm": 0.3549706850829124, | |
| "learning_rate": 3.144706581653805e-05, | |
| "loss": 0.3825, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.882591093117409, | |
| "grad_norm": 0.30683877741077203, | |
| "learning_rate": 3.140064516427566e-05, | |
| "loss": 0.384, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8866396761133604, | |
| "grad_norm": 0.38176696714984587, | |
| "learning_rate": 3.135413335275833e-05, | |
| "loss": 0.3872, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8906882591093117, | |
| "grad_norm": 0.3583138972832103, | |
| "learning_rate": 3.130753075389327e-05, | |
| "loss": 0.4001, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.29859567950142824, | |
| "learning_rate": 3.1260837740313625e-05, | |
| "loss": 0.3906, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8987854251012146, | |
| "grad_norm": 0.3929110989120314, | |
| "learning_rate": 3.121405468537547e-05, | |
| "loss": 0.3927, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.902834008097166, | |
| "grad_norm": 0.33521357778116223, | |
| "learning_rate": 3.116718196315488e-05, | |
| "loss": 0.389, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.9068825910931175, | |
| "grad_norm": 0.29117052737685384, | |
| "learning_rate": 3.1120219948444894e-05, | |
| "loss": 0.3755, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.9109311740890689, | |
| "grad_norm": 0.3798367844515094, | |
| "learning_rate": 3.107316901675252e-05, | |
| "loss": 0.3928, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.9149797570850202, | |
| "grad_norm": 0.32524256123597844, | |
| "learning_rate": 3.1026029544295745e-05, | |
| "loss": 0.4009, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.9190283400809718, | |
| "grad_norm": 0.3037237035920648, | |
| "learning_rate": 3.097880190800057e-05, | |
| "loss": 0.3822, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.35293975086881035, | |
| "learning_rate": 3.093148648549788e-05, | |
| "loss": 0.386, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.9271255060728745, | |
| "grad_norm": 0.28886531419356076, | |
| "learning_rate": 3.088408365512055e-05, | |
| "loss": 0.3792, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.931174089068826, | |
| "grad_norm": 0.3370498393033456, | |
| "learning_rate": 3.083659379590034e-05, | |
| "loss": 0.3948, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.9352226720647774, | |
| "grad_norm": 0.3900550080332791, | |
| "learning_rate": 3.0789017287564913e-05, | |
| "loss": 0.3728, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9392712550607287, | |
| "grad_norm": 0.2633722785912497, | |
| "learning_rate": 3.074135451053475e-05, | |
| "loss": 0.3776, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9433198380566803, | |
| "grad_norm": 0.4075560602464425, | |
| "learning_rate": 3.069360584592015e-05, | |
| "loss": 0.3843, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.3893270884247527, | |
| "learning_rate": 3.064577167551816e-05, | |
| "loss": 0.3801, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.951417004048583, | |
| "grad_norm": 0.3438120510197475, | |
| "learning_rate": 3.059785238180953e-05, | |
| "loss": 0.3985, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9554655870445345, | |
| "grad_norm": 0.3622404182728857, | |
| "learning_rate": 3.054984834795566e-05, | |
| "loss": 0.3803, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9595141700404857, | |
| "grad_norm": 0.34399615983051063, | |
| "learning_rate": 3.0501759957795537e-05, | |
| "loss": 0.3695, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9635627530364372, | |
| "grad_norm": 0.377585607772437, | |
| "learning_rate": 3.0453587595842627e-05, | |
| "loss": 0.3865, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9676113360323888, | |
| "grad_norm": 0.47638877098953475, | |
| "learning_rate": 3.0405331647281856e-05, | |
| "loss": 0.3812, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.97165991902834, | |
| "grad_norm": 0.31920833731313136, | |
| "learning_rate": 3.0356992497966508e-05, | |
| "loss": 0.3863, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9757085020242915, | |
| "grad_norm": 0.5255050491274422, | |
| "learning_rate": 3.0308570534415137e-05, | |
| "loss": 0.3814, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.979757085020243, | |
| "grad_norm": 0.37155051068084183, | |
| "learning_rate": 3.0260066143808467e-05, | |
| "loss": 0.3746, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9838056680161942, | |
| "grad_norm": 0.4199977766000606, | |
| "learning_rate": 3.021147971398632e-05, | |
| "loss": 0.3863, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9878542510121457, | |
| "grad_norm": 0.4324980748048768, | |
| "learning_rate": 3.0162811633444478e-05, | |
| "loss": 0.3817, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9919028340080973, | |
| "grad_norm": 0.3830698168113854, | |
| "learning_rate": 3.0114062291331635e-05, | |
| "loss": 0.3845, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9959514170040484, | |
| "grad_norm": 0.35303692377771745, | |
| "learning_rate": 3.0065232077446217e-05, | |
| "loss": 0.3756, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.44041625248468275, | |
| "learning_rate": 3.001632138223332e-05, | |
| "loss": 0.3768, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.0040485829959516, | |
| "grad_norm": 0.4905320753419372, | |
| "learning_rate": 2.9967330596781565e-05, | |
| "loss": 0.3346, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0080971659919027, | |
| "grad_norm": 0.6659083611061798, | |
| "learning_rate": 2.9918260112819952e-05, | |
| "loss": 0.339, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.0121457489878543, | |
| "grad_norm": 0.5713750276941627, | |
| "learning_rate": 2.9869110322714778e-05, | |
| "loss": 0.3368, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.016194331983806, | |
| "grad_norm": 0.6666963474925839, | |
| "learning_rate": 2.9819881619466447e-05, | |
| "loss": 0.3212, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.020242914979757, | |
| "grad_norm": 0.42320880236920166, | |
| "learning_rate": 2.9770574396706364e-05, | |
| "loss": 0.3292, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.0242914979757085, | |
| "grad_norm": 0.6344111137047872, | |
| "learning_rate": 2.972118904869376e-05, | |
| "loss": 0.3324, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02834008097166, | |
| "grad_norm": 0.5806931865612909, | |
| "learning_rate": 2.9671725970312563e-05, | |
| "loss": 0.3355, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.032388663967611, | |
| "grad_norm": 0.40052304841703823, | |
| "learning_rate": 2.9622185557068238e-05, | |
| "loss": 0.3217, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.0364372469635628, | |
| "grad_norm": 0.43570999082501316, | |
| "learning_rate": 2.95725682050846e-05, | |
| "loss": 0.3332, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0404858299595143, | |
| "grad_norm": 0.4678776015512857, | |
| "learning_rate": 2.9522874311100684e-05, | |
| "loss": 0.3238, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0445344129554655, | |
| "grad_norm": 0.4898913894040103, | |
| "learning_rate": 2.9473104272467526e-05, | |
| "loss": 0.3329, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.048582995951417, | |
| "grad_norm": 0.4688020933065666, | |
| "learning_rate": 2.942325848714505e-05, | |
| "loss": 0.3327, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.5570062978464078, | |
| "learning_rate": 2.9373337353698804e-05, | |
| "loss": 0.3326, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.0566801619433197, | |
| "grad_norm": 0.4149241253310497, | |
| "learning_rate": 2.932334127129686e-05, | |
| "loss": 0.3179, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.0607287449392713, | |
| "grad_norm": 0.47030486913541675, | |
| "learning_rate": 2.9273270639706547e-05, | |
| "loss": 0.3298, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.064777327935223, | |
| "grad_norm": 0.4315059299349867, | |
| "learning_rate": 2.922312585929131e-05, | |
| "loss": 0.3318, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.068825910931174, | |
| "grad_norm": 0.392789406491033, | |
| "learning_rate": 2.9172907331007478e-05, | |
| "loss": 0.3356, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.0728744939271255, | |
| "grad_norm": 0.4185766688517101, | |
| "learning_rate": 2.9122615456401053e-05, | |
| "loss": 0.33, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.3636167616368414, | |
| "learning_rate": 2.9072250637604535e-05, | |
| "loss": 0.3213, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.080971659919028, | |
| "grad_norm": 0.394771434381802, | |
| "learning_rate": 2.9021813277333667e-05, | |
| "loss": 0.3389, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.08502024291498, | |
| "grad_norm": 0.40348670198932024, | |
| "learning_rate": 2.8971303778884244e-05, | |
| "loss": 0.3313, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.0890688259109313, | |
| "grad_norm": 0.34433075472385816, | |
| "learning_rate": 2.8920722546128863e-05, | |
| "loss": 0.3361, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.0931174089068825, | |
| "grad_norm": 0.3430983434266528, | |
| "learning_rate": 2.887006998351372e-05, | |
| "loss": 0.3338, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.097165991902834, | |
| "grad_norm": 0.3598303196653174, | |
| "learning_rate": 2.881934649605535e-05, | |
| "loss": 0.3272, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.1012145748987856, | |
| "grad_norm": 0.3501592583990438, | |
| "learning_rate": 2.8768552489337408e-05, | |
| "loss": 0.323, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.3391956638261742, | |
| "learning_rate": 2.8717688369507425e-05, | |
| "loss": 0.3277, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.1093117408906883, | |
| "grad_norm": 0.3313290743601798, | |
| "learning_rate": 2.8666754543273535e-05, | |
| "loss": 0.3342, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.11336032388664, | |
| "grad_norm": 0.3551723443001552, | |
| "learning_rate": 2.8615751417901273e-05, | |
| "loss": 0.3319, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.117408906882591, | |
| "grad_norm": 0.3176899000673088, | |
| "learning_rate": 2.856467940121025e-05, | |
| "loss": 0.3252, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.1214574898785425, | |
| "grad_norm": 0.351983621578224, | |
| "learning_rate": 2.8513538901570964e-05, | |
| "loss": 0.3186, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.125506072874494, | |
| "grad_norm": 0.26300566853600116, | |
| "learning_rate": 2.846233032790149e-05, | |
| "loss": 0.3391, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.1295546558704452, | |
| "grad_norm": 0.3460489420127684, | |
| "learning_rate": 2.84110540896642e-05, | |
| "loss": 0.3361, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.133603238866397, | |
| "grad_norm": 0.3412356109033047, | |
| "learning_rate": 2.8359710596862555e-05, | |
| "loss": 0.3341, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.1376518218623484, | |
| "grad_norm": 0.2978698229407348, | |
| "learning_rate": 2.8308300260037734e-05, | |
| "loss": 0.3248, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.1417004048582995, | |
| "grad_norm": 0.3861880332540838, | |
| "learning_rate": 2.8256823490265424e-05, | |
| "loss": 0.3361, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.145748987854251, | |
| "grad_norm": 0.40811812397080865, | |
| "learning_rate": 2.8205280699152515e-05, | |
| "loss": 0.3273, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.1497975708502026, | |
| "grad_norm": 0.3296779875223969, | |
| "learning_rate": 2.8153672298833777e-05, | |
| "loss": 0.3306, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.3916528969456899, | |
| "learning_rate": 2.8101998701968615e-05, | |
| "loss": 0.3405, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.3467827311451583, | |
| "learning_rate": 2.805026032173773e-05, | |
| "loss": 0.3286, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.161943319838057, | |
| "grad_norm": 0.3351502040888187, | |
| "learning_rate": 2.7998457571839832e-05, | |
| "loss": 0.3198, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.165991902834008, | |
| "grad_norm": 0.27845879423161785, | |
| "learning_rate": 2.7946590866488322e-05, | |
| "loss": 0.3191, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.1700404858299596, | |
| "grad_norm": 0.3426579010559102, | |
| "learning_rate": 2.7894660620408e-05, | |
| "loss": 0.3278, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.174089068825911, | |
| "grad_norm": 0.27829210460027803, | |
| "learning_rate": 2.784266724883173e-05, | |
| "loss": 0.333, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.1781376518218623, | |
| "grad_norm": 0.2917400337967327, | |
| "learning_rate": 2.7790611167497128e-05, | |
| "loss": 0.3218, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.182186234817814, | |
| "grad_norm": 0.3530236745914095, | |
| "learning_rate": 2.7738492792643224e-05, | |
| "loss": 0.3218, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.1862348178137654, | |
| "grad_norm": 0.31198894576751757, | |
| "learning_rate": 2.768631254100716e-05, | |
| "loss": 0.3364, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.1902834008097165, | |
| "grad_norm": 0.27044941744776047, | |
| "learning_rate": 2.7634070829820836e-05, | |
| "loss": 0.3257, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.194331983805668, | |
| "grad_norm": 0.3758775066687365, | |
| "learning_rate": 2.7581768076807592e-05, | |
| "loss": 0.3288, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.1983805668016196, | |
| "grad_norm": 0.2667910211314267, | |
| "learning_rate": 2.7529404700178832e-05, | |
| "loss": 0.3183, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.2024291497975708, | |
| "grad_norm": 0.30642786490894797, | |
| "learning_rate": 2.7476981118630735e-05, | |
| "loss": 0.3266, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.2064777327935223, | |
| "grad_norm": 0.3219606823831628, | |
| "learning_rate": 2.742449775134085e-05, | |
| "loss": 0.3185, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.26162978892559713, | |
| "learning_rate": 2.7371955017964788e-05, | |
| "loss": 0.3267, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.214574898785425, | |
| "grad_norm": 0.2923695839526627, | |
| "learning_rate": 2.7319353338632845e-05, | |
| "loss": 0.3413, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.2186234817813766, | |
| "grad_norm": 0.2892698800742035, | |
| "learning_rate": 2.7266693133946645e-05, | |
| "loss": 0.3296, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.2226720647773277, | |
| "grad_norm": 0.31077283972589775, | |
| "learning_rate": 2.7213974824975785e-05, | |
| "loss": 0.3346, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.2267206477732793, | |
| "grad_norm": 0.34238927098742217, | |
| "learning_rate": 2.7161198833254464e-05, | |
| "loss": 0.3222, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.27514531021613514, | |
| "learning_rate": 2.7108365580778095e-05, | |
| "loss": 0.3302, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.234817813765182, | |
| "grad_norm": 0.29698954099951713, | |
| "learning_rate": 2.705547548999996e-05, | |
| "loss": 0.3359, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.2388663967611335, | |
| "grad_norm": 0.287875736157255, | |
| "learning_rate": 2.7002528983827817e-05, | |
| "loss": 0.3339, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.242914979757085, | |
| "grad_norm": 0.278907928405062, | |
| "learning_rate": 2.6949526485620517e-05, | |
| "loss": 0.3323, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.246963562753036, | |
| "grad_norm": 0.2956271640769045, | |
| "learning_rate": 2.6896468419184623e-05, | |
| "loss": 0.3263, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.251012145748988, | |
| "grad_norm": 0.30114449312703795, | |
| "learning_rate": 2.6843355208771024e-05, | |
| "loss": 0.3424, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.2550607287449393, | |
| "grad_norm": 0.2642899430993717, | |
| "learning_rate": 2.6790187279071536e-05, | |
| "loss": 0.3281, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.2591093117408905, | |
| "grad_norm": 0.2921147485963999, | |
| "learning_rate": 2.67369650552155e-05, | |
| "loss": 0.3261, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.2562091766333612, | |
| "learning_rate": 2.6683688962766407e-05, | |
| "loss": 0.3247, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.2672064777327936, | |
| "grad_norm": 0.29593240266081067, | |
| "learning_rate": 2.6630359427718468e-05, | |
| "loss": 0.3269, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2712550607287447, | |
| "grad_norm": 0.2969448875239638, | |
| "learning_rate": 2.6576976876493225e-05, | |
| "loss": 0.3316, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.2753036437246963, | |
| "grad_norm": 0.25581770497370526, | |
| "learning_rate": 2.652354173593614e-05, | |
| "loss": 0.3395, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.279352226720648, | |
| "grad_norm": 0.2976114289232262, | |
| "learning_rate": 2.647005443331317e-05, | |
| "loss": 0.3366, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.283400809716599, | |
| "grad_norm": 0.3414035096041696, | |
| "learning_rate": 2.6416515396307354e-05, | |
| "loss": 0.3416, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.2874493927125505, | |
| "grad_norm": 0.31578143268232334, | |
| "learning_rate": 2.636292505301542e-05, | |
| "loss": 0.3424, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.291497975708502, | |
| "grad_norm": 0.35560157457902386, | |
| "learning_rate": 2.630928383194431e-05, | |
| "loss": 0.3399, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.2955465587044532, | |
| "grad_norm": 0.2934293627407647, | |
| "learning_rate": 2.625559216200781e-05, | |
| "loss": 0.3396, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.299595141700405, | |
| "grad_norm": 0.29207776824775, | |
| "learning_rate": 2.620185047252308e-05, | |
| "loss": 0.322, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.3036437246963564, | |
| "grad_norm": 0.28579242806912275, | |
| "learning_rate": 2.614805919320723e-05, | |
| "loss": 0.3345, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.32581609376107856, | |
| "learning_rate": 2.6094218754173895e-05, | |
| "loss": 0.3345, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.311740890688259, | |
| "grad_norm": 0.3099470504500979, | |
| "learning_rate": 2.60403295859298e-05, | |
| "loss": 0.3357, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.2722645446017704, | |
| "learning_rate": 2.5986392119371297e-05, | |
| "loss": 0.3227, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.3198380566801617, | |
| "grad_norm": 0.29623830639721893, | |
| "learning_rate": 2.5932406785780928e-05, | |
| "loss": 0.3339, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.3238866396761133, | |
| "grad_norm": 0.33514446426993705, | |
| "learning_rate": 2.5878374016823987e-05, | |
| "loss": 0.3276, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.327935222672065, | |
| "grad_norm": 0.29283975885101365, | |
| "learning_rate": 2.5824294244545045e-05, | |
| "loss": 0.3396, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.331983805668016, | |
| "grad_norm": 0.33541721983309547, | |
| "learning_rate": 2.577016790136453e-05, | |
| "loss": 0.3248, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.3360323886639676, | |
| "grad_norm": 0.28603298871653543, | |
| "learning_rate": 2.5715995420075225e-05, | |
| "loss": 0.3452, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.340080971659919, | |
| "grad_norm": 0.3273485568628812, | |
| "learning_rate": 2.5661777233838864e-05, | |
| "loss": 0.326, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.3441295546558703, | |
| "grad_norm": 0.2824340716281727, | |
| "learning_rate": 2.5607513776182616e-05, | |
| "loss": 0.3309, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.348178137651822, | |
| "grad_norm": 0.325134861081836, | |
| "learning_rate": 2.5553205480995633e-05, | |
| "loss": 0.3349, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3522267206477734, | |
| "grad_norm": 0.35114866287428154, | |
| "learning_rate": 2.5498852782525604e-05, | |
| "loss": 0.3218, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.3562753036437245, | |
| "grad_norm": 0.27788294858814966, | |
| "learning_rate": 2.5444456115375252e-05, | |
| "loss": 0.3287, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.360323886639676, | |
| "grad_norm": 0.3555968405754811, | |
| "learning_rate": 2.539001591449889e-05, | |
| "loss": 0.3352, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.3643724696356276, | |
| "grad_norm": 0.3023869585387261, | |
| "learning_rate": 2.53355326151989e-05, | |
| "loss": 0.328, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.3525165649379669, | |
| "learning_rate": 2.5281006653122288e-05, | |
| "loss": 0.3379, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.3724696356275303, | |
| "grad_norm": 0.3824943788343438, | |
| "learning_rate": 2.5226438464257217e-05, | |
| "loss": 0.331, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.376518218623482, | |
| "grad_norm": 0.267180641404731, | |
| "learning_rate": 2.5171828484929447e-05, | |
| "loss": 0.3225, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.380566801619433, | |
| "grad_norm": 0.37604433154174277, | |
| "learning_rate": 2.5117177151798934e-05, | |
| "loss": 0.334, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.34151722523851497, | |
| "learning_rate": 2.506248490185627e-05, | |
| "loss": 0.337, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.388663967611336, | |
| "grad_norm": 0.27970225840310814, | |
| "learning_rate": 2.500775217241923e-05, | |
| "loss": 0.3181, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.3927125506072873, | |
| "grad_norm": 0.3327647004807472, | |
| "learning_rate": 2.4952979401129266e-05, | |
| "loss": 0.3466, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.396761133603239, | |
| "grad_norm": 0.28054336029585736, | |
| "learning_rate": 2.4898167025948e-05, | |
| "loss": 0.3197, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.4008097165991904, | |
| "grad_norm": 0.29903087903912107, | |
| "learning_rate": 2.4843315485153703e-05, | |
| "loss": 0.3345, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.4048582995951415, | |
| "grad_norm": 0.313320844136896, | |
| "learning_rate": 2.4788425217337856e-05, | |
| "loss": 0.3326, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.408906882591093, | |
| "grad_norm": 0.3098691420892224, | |
| "learning_rate": 2.4733496661401572e-05, | |
| "loss": 0.343, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.4129554655870447, | |
| "grad_norm": 0.2945040328297694, | |
| "learning_rate": 2.4678530256552114e-05, | |
| "loss": 0.3373, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.417004048582996, | |
| "grad_norm": 0.2938346539188775, | |
| "learning_rate": 2.46235264422994e-05, | |
| "loss": 0.3364, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.33269531292059923, | |
| "learning_rate": 2.4568485658452452e-05, | |
| "loss": 0.3367, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.425101214574899, | |
| "grad_norm": 0.3191963615398019, | |
| "learning_rate": 2.451340834511592e-05, | |
| "loss": 0.3315, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.42914979757085, | |
| "grad_norm": 0.33918674848434394, | |
| "learning_rate": 2.4458294942686528e-05, | |
| "loss": 0.3143, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4331983805668016, | |
| "grad_norm": 0.3141175952294302, | |
| "learning_rate": 2.440314589184958e-05, | |
| "loss": 0.3166, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.437246963562753, | |
| "grad_norm": 0.3429203571298003, | |
| "learning_rate": 2.434796163357541e-05, | |
| "loss": 0.3317, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.4412955465587043, | |
| "grad_norm": 0.30031977048381203, | |
| "learning_rate": 2.429274260911588e-05, | |
| "loss": 0.3268, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.445344129554656, | |
| "grad_norm": 0.3503053837738292, | |
| "learning_rate": 2.423748926000084e-05, | |
| "loss": 0.3335, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.4493927125506074, | |
| "grad_norm": 0.3344971932996651, | |
| "learning_rate": 2.418220202803459e-05, | |
| "loss": 0.3323, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.4534412955465585, | |
| "grad_norm": 0.3212895593852584, | |
| "learning_rate": 2.412688135529237e-05, | |
| "loss": 0.3232, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.45748987854251, | |
| "grad_norm": 0.3195914998392765, | |
| "learning_rate": 2.4071527684116807e-05, | |
| "loss": 0.3305, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.32024991701089606, | |
| "learning_rate": 2.4016141457114372e-05, | |
| "loss": 0.3337, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.465587044534413, | |
| "grad_norm": 0.2968020221209132, | |
| "learning_rate": 2.396072311715187e-05, | |
| "loss": 0.3189, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.4696356275303644, | |
| "grad_norm": 0.3165931601853622, | |
| "learning_rate": 2.3905273107352863e-05, | |
| "loss": 0.3335, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.2780161687543898, | |
| "learning_rate": 2.3849791871094157e-05, | |
| "loss": 0.3231, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.477732793522267, | |
| "grad_norm": 0.3731872692265382, | |
| "learning_rate": 2.379427985200223e-05, | |
| "loss": 0.3361, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.4817813765182186, | |
| "grad_norm": 0.31883911137641857, | |
| "learning_rate": 2.373873749394972e-05, | |
| "loss": 0.3384, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.48582995951417, | |
| "grad_norm": 0.36991789314513995, | |
| "learning_rate": 2.3683165241051836e-05, | |
| "loss": 0.3325, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.4898785425101213, | |
| "grad_norm": 0.3434165099738912, | |
| "learning_rate": 2.3627563537662823e-05, | |
| "loss": 0.3304, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.493927125506073, | |
| "grad_norm": 0.32608687897134486, | |
| "learning_rate": 2.3571932828372433e-05, | |
| "loss": 0.3267, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.4979757085020244, | |
| "grad_norm": 0.34185351694537747, | |
| "learning_rate": 2.351627355800233e-05, | |
| "loss": 0.3311, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.5020242914979756, | |
| "grad_norm": 0.35818975689655874, | |
| "learning_rate": 2.346058617160256e-05, | |
| "loss": 0.3378, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.506072874493927, | |
| "grad_norm": 0.30756360389053966, | |
| "learning_rate": 2.3404871114447976e-05, | |
| "loss": 0.3377, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.5101214574898787, | |
| "grad_norm": 0.2741826942862568, | |
| "learning_rate": 2.334912883203469e-05, | |
| "loss": 0.3358, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.51417004048583, | |
| "grad_norm": 0.31902825532737517, | |
| "learning_rate": 2.3293359770076505e-05, | |
| "loss": 0.3137, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.5182186234817814, | |
| "grad_norm": 0.270609547435234, | |
| "learning_rate": 2.323756437450136e-05, | |
| "loss": 0.3291, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.522267206477733, | |
| "grad_norm": 0.31277459327551005, | |
| "learning_rate": 2.3181743091447754e-05, | |
| "loss": 0.3173, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.29076811923120743, | |
| "learning_rate": 2.3125896367261164e-05, | |
| "loss": 0.3331, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.5303643724696356, | |
| "grad_norm": 0.29777461675615313, | |
| "learning_rate": 2.307002464849052e-05, | |
| "loss": 0.3291, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.534412955465587, | |
| "grad_norm": 0.29550408301812914, | |
| "learning_rate": 2.30141283818846e-05, | |
| "loss": 0.3326, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.292785530031014, | |
| "learning_rate": 2.2958208014388462e-05, | |
| "loss": 0.3327, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.54251012145749, | |
| "grad_norm": 0.28337654125247896, | |
| "learning_rate": 2.2902263993139857e-05, | |
| "loss": 0.3283, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.5465587044534415, | |
| "grad_norm": 0.2991626536137994, | |
| "learning_rate": 2.2846296765465708e-05, | |
| "loss": 0.3446, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.5506072874493926, | |
| "grad_norm": 0.28132736057839575, | |
| "learning_rate": 2.2790306778878465e-05, | |
| "loss": 0.3232, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.554655870445344, | |
| "grad_norm": 0.2854294603324085, | |
| "learning_rate": 2.273429448107257e-05, | |
| "loss": 0.3372, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.5587044534412957, | |
| "grad_norm": 0.282074624226103, | |
| "learning_rate": 2.2678260319920868e-05, | |
| "loss": 0.3324, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.562753036437247, | |
| "grad_norm": 0.2832763264030283, | |
| "learning_rate": 2.262220474347102e-05, | |
| "loss": 0.3242, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.5668016194331984, | |
| "grad_norm": 0.2765722069056518, | |
| "learning_rate": 2.2566128199941918e-05, | |
| "loss": 0.3303, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.57085020242915, | |
| "grad_norm": 0.3428522644831727, | |
| "learning_rate": 2.2510031137720104e-05, | |
| "loss": 0.3365, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.574898785425101, | |
| "grad_norm": 0.29604598466515913, | |
| "learning_rate": 2.2453914005356206e-05, | |
| "loss": 0.3277, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.31943084966254776, | |
| "learning_rate": 2.2397777251561302e-05, | |
| "loss": 0.3456, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.582995951417004, | |
| "grad_norm": 0.3293158311723465, | |
| "learning_rate": 2.2341621325203393e-05, | |
| "loss": 0.338, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.5870445344129553, | |
| "grad_norm": 0.26339690173098596, | |
| "learning_rate": 2.2285446675303764e-05, | |
| "loss": 0.3245, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.591093117408907, | |
| "grad_norm": 0.2840984230184787, | |
| "learning_rate": 2.2229253751033423e-05, | |
| "loss": 0.3168, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.5951417004048585, | |
| "grad_norm": 0.2987933957887933, | |
| "learning_rate": 2.2173043001709492e-05, | |
| "loss": 0.3287, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.5991902834008096, | |
| "grad_norm": 0.2743948782747588, | |
| "learning_rate": 2.2116814876791627e-05, | |
| "loss": 0.3361, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.603238866396761, | |
| "grad_norm": 0.2867224751254625, | |
| "learning_rate": 2.2060569825878413e-05, | |
| "loss": 0.3312, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.6072874493927127, | |
| "grad_norm": 0.2909967078303511, | |
| "learning_rate": 2.2004308298703777e-05, | |
| "loss": 0.3302, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.611336032388664, | |
| "grad_norm": 0.26860299241822516, | |
| "learning_rate": 2.1948030745133396e-05, | |
| "loss": 0.3307, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.28236767780929545, | |
| "learning_rate": 2.1891737615161083e-05, | |
| "loss": 0.3419, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.619433198380567, | |
| "grad_norm": 0.24579502333087286, | |
| "learning_rate": 2.1835429358905218e-05, | |
| "loss": 0.3309, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.623481781376518, | |
| "grad_norm": 0.28280422844647946, | |
| "learning_rate": 2.17791064266051e-05, | |
| "loss": 0.329, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.6275303643724697, | |
| "grad_norm": 0.2709783972014291, | |
| "learning_rate": 2.172276926861741e-05, | |
| "loss": 0.335, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.27777679251435666, | |
| "learning_rate": 2.166641833541255e-05, | |
| "loss": 0.3429, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.6356275303643724, | |
| "grad_norm": 0.28730662644633803, | |
| "learning_rate": 2.161005407757109e-05, | |
| "loss": 0.3282, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.639676113360324, | |
| "grad_norm": 0.26919120032685423, | |
| "learning_rate": 2.155367694578013e-05, | |
| "loss": 0.3351, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.6437246963562755, | |
| "grad_norm": 0.26572997535353293, | |
| "learning_rate": 2.1497287390829712e-05, | |
| "loss": 0.3382, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.6477732793522266, | |
| "grad_norm": 0.27014282870621026, | |
| "learning_rate": 2.1440885863609208e-05, | |
| "loss": 0.3279, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.651821862348178, | |
| "grad_norm": 0.27138363986717434, | |
| "learning_rate": 2.1384472815103737e-05, | |
| "loss": 0.3205, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.6558704453441297, | |
| "grad_norm": 0.24635082325703325, | |
| "learning_rate": 2.1328048696390534e-05, | |
| "loss": 0.3283, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.659919028340081, | |
| "grad_norm": 0.2347561348659333, | |
| "learning_rate": 2.1271613958635342e-05, | |
| "loss": 0.3306, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.6639676113360324, | |
| "grad_norm": 0.2512017516911489, | |
| "learning_rate": 2.1215169053088825e-05, | |
| "loss": 0.3483, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.668016194331984, | |
| "grad_norm": 0.27106283147143995, | |
| "learning_rate": 2.1158714431082937e-05, | |
| "loss": 0.3392, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.672064777327935, | |
| "grad_norm": 0.25973986549358685, | |
| "learning_rate": 2.110225054402733e-05, | |
| "loss": 0.3279, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.6761133603238867, | |
| "grad_norm": 0.26177156556041936, | |
| "learning_rate": 2.1045777843405747e-05, | |
| "loss": 0.3301, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.6801619433198383, | |
| "grad_norm": 0.26092053579645796, | |
| "learning_rate": 2.09892967807724e-05, | |
| "loss": 0.3239, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.29808578846775596, | |
| "learning_rate": 2.093280780774835e-05, | |
| "loss": 0.3313, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.688259109311741, | |
| "grad_norm": 0.27937495667312917, | |
| "learning_rate": 2.087631137601793e-05, | |
| "loss": 0.3298, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.25480800511907087, | |
| "learning_rate": 2.0819807937325104e-05, | |
| "loss": 0.3243, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.6963562753036436, | |
| "grad_norm": 0.27751979962559625, | |
| "learning_rate": 2.0763297943469847e-05, | |
| "loss": 0.3234, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.700404858299595, | |
| "grad_norm": 0.25967281653887453, | |
| "learning_rate": 2.070678184630458e-05, | |
| "loss": 0.3383, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.7044534412955468, | |
| "grad_norm": 0.22797317868275216, | |
| "learning_rate": 2.06502600977305e-05, | |
| "loss": 0.3381, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.708502024291498, | |
| "grad_norm": 0.26225734063359357, | |
| "learning_rate": 2.0593733149694008e-05, | |
| "loss": 0.3364, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.7125506072874495, | |
| "grad_norm": 0.22279119858130528, | |
| "learning_rate": 2.0537201454183074e-05, | |
| "loss": 0.3344, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.716599190283401, | |
| "grad_norm": 0.24898708610272782, | |
| "learning_rate": 2.0480665463223626e-05, | |
| "loss": 0.3299, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.720647773279352, | |
| "grad_norm": 0.2535151764627004, | |
| "learning_rate": 2.042412562887594e-05, | |
| "loss": 0.3259, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.7246963562753037, | |
| "grad_norm": 0.23970239238786517, | |
| "learning_rate": 2.0367582403231033e-05, | |
| "loss": 0.331, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.7287449392712553, | |
| "grad_norm": 0.26924585705479587, | |
| "learning_rate": 2.0311036238407023e-05, | |
| "loss": 0.3237, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.7327935222672064, | |
| "grad_norm": 0.2708983749062139, | |
| "learning_rate": 2.0254487586545534e-05, | |
| "loss": 0.3301, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.28226545442659107, | |
| "learning_rate": 2.0197936899808088e-05, | |
| "loss": 0.3314, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.7408906882591095, | |
| "grad_norm": 0.2627070591262798, | |
| "learning_rate": 2.0141384630372466e-05, | |
| "loss": 0.3329, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.7449392712550607, | |
| "grad_norm": 0.2938425693605566, | |
| "learning_rate": 2.0084831230429106e-05, | |
| "loss": 0.3303, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.748987854251012, | |
| "grad_norm": 0.2614493049870086, | |
| "learning_rate": 2.0028277152177492e-05, | |
| "loss": 0.3347, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.753036437246964, | |
| "grad_norm": 0.25517201595752537, | |
| "learning_rate": 1.9971722847822518e-05, | |
| "loss": 0.3322, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.757085020242915, | |
| "grad_norm": 0.26405898460030025, | |
| "learning_rate": 1.99151687695709e-05, | |
| "loss": 0.3265, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7611336032388665, | |
| "grad_norm": 0.2219838555471208, | |
| "learning_rate": 1.9858615369627537e-05, | |
| "loss": 0.332, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.765182186234818, | |
| "grad_norm": 0.28969302647913453, | |
| "learning_rate": 1.9802063100191916e-05, | |
| "loss": 0.3386, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.2707685091274618, | |
| "learning_rate": 1.974551241345447e-05, | |
| "loss": 0.3374, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.7732793522267207, | |
| "grad_norm": 0.2740998373804592, | |
| "learning_rate": 1.9688963761592984e-05, | |
| "loss": 0.3147, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.7773279352226723, | |
| "grad_norm": 0.23392484016990725, | |
| "learning_rate": 1.963241759676897e-05, | |
| "loss": 0.3272, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.7813765182186234, | |
| "grad_norm": 0.2749915418624844, | |
| "learning_rate": 1.9575874371124066e-05, | |
| "loss": 0.3369, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.785425101214575, | |
| "grad_norm": 0.25171687857212854, | |
| "learning_rate": 1.951933453677638e-05, | |
| "loss": 0.3372, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.23974473187207254, | |
| "learning_rate": 1.9462798545816932e-05, | |
| "loss": 0.3461, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.7935222672064777, | |
| "grad_norm": 0.27038349599114875, | |
| "learning_rate": 1.9406266850305996e-05, | |
| "loss": 0.3287, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7975708502024292, | |
| "grad_norm": 0.24204999540377042, | |
| "learning_rate": 1.9349739902269503e-05, | |
| "loss": 0.3293, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.801619433198381, | |
| "grad_norm": 0.24032946378418582, | |
| "learning_rate": 1.9293218153695425e-05, | |
| "loss": 0.3356, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.805668016194332, | |
| "grad_norm": 0.27917339719659645, | |
| "learning_rate": 1.9236702056530153e-05, | |
| "loss": 0.341, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.8097165991902835, | |
| "grad_norm": 0.2925872944812625, | |
| "learning_rate": 1.9180192062674907e-05, | |
| "loss": 0.3285, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.813765182186235, | |
| "grad_norm": 0.27727129643644366, | |
| "learning_rate": 1.9123688623982076e-05, | |
| "loss": 0.3211, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.817813765182186, | |
| "grad_norm": 0.331909864570572, | |
| "learning_rate": 1.906719219225165e-05, | |
| "loss": 0.3396, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.8218623481781377, | |
| "grad_norm": 0.26061339901924546, | |
| "learning_rate": 1.9010703219227603e-05, | |
| "loss": 0.3295, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.8259109311740893, | |
| "grad_norm": 0.2752563286741443, | |
| "learning_rate": 1.8954222156594256e-05, | |
| "loss": 0.3337, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.8299595141700404, | |
| "grad_norm": 0.2743669242856445, | |
| "learning_rate": 1.8897749455972673e-05, | |
| "loss": 0.326, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.834008097165992, | |
| "grad_norm": 0.2516021038736473, | |
| "learning_rate": 1.884128556891707e-05, | |
| "loss": 0.3352, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8380566801619436, | |
| "grad_norm": 0.2842065480645892, | |
| "learning_rate": 1.878483094691119e-05, | |
| "loss": 0.3291, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.29535861739432795, | |
| "learning_rate": 1.8728386041364664e-05, | |
| "loss": 0.3284, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.26737825285808126, | |
| "learning_rate": 1.867195130360947e-05, | |
| "loss": 0.3326, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.850202429149798, | |
| "grad_norm": 0.2820126586350078, | |
| "learning_rate": 1.8615527184896263e-05, | |
| "loss": 0.3139, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.854251012145749, | |
| "grad_norm": 0.2854689608392354, | |
| "learning_rate": 1.8559114136390795e-05, | |
| "loss": 0.3428, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.8582995951417005, | |
| "grad_norm": 0.28831184733949367, | |
| "learning_rate": 1.8502712609170298e-05, | |
| "loss": 0.3387, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.862348178137652, | |
| "grad_norm": 0.299770833656906, | |
| "learning_rate": 1.8446323054219876e-05, | |
| "loss": 0.3219, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.866396761133603, | |
| "grad_norm": 0.2882508245978294, | |
| "learning_rate": 1.838994592242891e-05, | |
| "loss": 0.3526, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.8704453441295548, | |
| "grad_norm": 0.3091631974347986, | |
| "learning_rate": 1.8333581664587453e-05, | |
| "loss": 0.3292, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.8744939271255063, | |
| "grad_norm": 0.26764934837841564, | |
| "learning_rate": 1.8277230731382593e-05, | |
| "loss": 0.3212, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.8785425101214575, | |
| "grad_norm": 0.28197410374939086, | |
| "learning_rate": 1.82208935733949e-05, | |
| "loss": 0.3308, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.882591093117409, | |
| "grad_norm": 0.2837854967732637, | |
| "learning_rate": 1.8164570641094793e-05, | |
| "loss": 0.3175, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.8866396761133606, | |
| "grad_norm": 0.28779656067716375, | |
| "learning_rate": 1.810826238483892e-05, | |
| "loss": 0.3341, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.8906882591093117, | |
| "grad_norm": 0.30911233424346685, | |
| "learning_rate": 1.8051969254866608e-05, | |
| "loss": 0.3327, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.29518562720221864, | |
| "learning_rate": 1.7995691701296226e-05, | |
| "loss": 0.3321, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.898785425101215, | |
| "grad_norm": 0.27152383410754743, | |
| "learning_rate": 1.7939430174121593e-05, | |
| "loss": 0.3259, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.902834008097166, | |
| "grad_norm": 0.32105485895020824, | |
| "learning_rate": 1.788318512320838e-05, | |
| "loss": 0.3325, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.9068825910931175, | |
| "grad_norm": 0.2708585668739488, | |
| "learning_rate": 1.782695699829051e-05, | |
| "loss": 0.345, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.910931174089069, | |
| "grad_norm": 0.27011822124966445, | |
| "learning_rate": 1.7770746248966587e-05, | |
| "loss": 0.3282, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.91497975708502, | |
| "grad_norm": 0.23745213815221908, | |
| "learning_rate": 1.7714553324696243e-05, | |
| "loss": 0.3202, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.919028340080972, | |
| "grad_norm": 0.2716762856597137, | |
| "learning_rate": 1.7658378674796614e-05, | |
| "loss": 0.3312, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.25795133864755615, | |
| "learning_rate": 1.76022227484387e-05, | |
| "loss": 0.3209, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.9271255060728745, | |
| "grad_norm": 0.287667880502641, | |
| "learning_rate": 1.7546085994643807e-05, | |
| "loss": 0.3492, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.931174089068826, | |
| "grad_norm": 0.26233359876994067, | |
| "learning_rate": 1.7489968862279902e-05, | |
| "loss": 0.3224, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.9352226720647776, | |
| "grad_norm": 0.3058454942061301, | |
| "learning_rate": 1.743387180005809e-05, | |
| "loss": 0.332, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.9392712550607287, | |
| "grad_norm": 0.23819721563890364, | |
| "learning_rate": 1.737779525652899e-05, | |
| "loss": 0.3293, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.9433198380566803, | |
| "grad_norm": 0.2638265578834233, | |
| "learning_rate": 1.732173968007914e-05, | |
| "loss": 0.3391, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.2609486995708828, | |
| "learning_rate": 1.7265705518927437e-05, | |
| "loss": 0.3285, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.951417004048583, | |
| "grad_norm": 0.2474014062998719, | |
| "learning_rate": 1.7209693221121542e-05, | |
| "loss": 0.3372, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.9554655870445345, | |
| "grad_norm": 0.2347750005874664, | |
| "learning_rate": 1.7153703234534302e-05, | |
| "loss": 0.3276, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.9595141700404857, | |
| "grad_norm": 0.22778826232674326, | |
| "learning_rate": 1.709773600686015e-05, | |
| "loss": 0.3404, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.9635627530364372, | |
| "grad_norm": 0.23065760257268164, | |
| "learning_rate": 1.7041791985611548e-05, | |
| "loss": 0.3302, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.967611336032389, | |
| "grad_norm": 0.23465329348458622, | |
| "learning_rate": 1.69858716181154e-05, | |
| "loss": 0.3241, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.97165991902834, | |
| "grad_norm": 0.23809998816292338, | |
| "learning_rate": 1.692997535150948e-05, | |
| "loss": 0.3336, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.9757085020242915, | |
| "grad_norm": 0.23816301192760372, | |
| "learning_rate": 1.687410363273884e-05, | |
| "loss": 0.3322, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.979757085020243, | |
| "grad_norm": 0.26644394831158347, | |
| "learning_rate": 1.6818256908552257e-05, | |
| "loss": 0.3299, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.983805668016194, | |
| "grad_norm": 0.24181628354822626, | |
| "learning_rate": 1.6762435625498646e-05, | |
| "loss": 0.3286, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.9878542510121457, | |
| "grad_norm": 0.25261458605196024, | |
| "learning_rate": 1.67066402299235e-05, | |
| "loss": 0.3277, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.9919028340080973, | |
| "grad_norm": 0.2404145346706741, | |
| "learning_rate": 1.6650871167965313e-05, | |
| "loss": 0.33, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.9959514170040484, | |
| "grad_norm": 0.24193521202728022, | |
| "learning_rate": 1.6595128885552028e-05, | |
| "loss": 0.3324, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.23228489022389434, | |
| "learning_rate": 1.653941382839745e-05, | |
| "loss": 0.3365, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.0040485829959516, | |
| "grad_norm": 0.39924217092668923, | |
| "learning_rate": 1.6483726441997673e-05, | |
| "loss": 0.278, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.0080971659919027, | |
| "grad_norm": 0.40400626782648835, | |
| "learning_rate": 1.642806717162757e-05, | |
| "loss": 0.2764, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.0121457489878543, | |
| "grad_norm": 0.4470713742847986, | |
| "learning_rate": 1.637243646233718e-05, | |
| "loss": 0.2879, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.016194331983806, | |
| "grad_norm": 0.4774957053442416, | |
| "learning_rate": 1.6316834758948174e-05, | |
| "loss": 0.2736, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.020242914979757, | |
| "grad_norm": 0.4011225248841658, | |
| "learning_rate": 1.6261262506050282e-05, | |
| "loss": 0.2757, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.0242914979757085, | |
| "grad_norm": 0.38244217372736267, | |
| "learning_rate": 1.620572014799777e-05, | |
| "loss": 0.2802, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.02834008097166, | |
| "grad_norm": 0.39048150826873546, | |
| "learning_rate": 1.6150208128905857e-05, | |
| "loss": 0.2908, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.032388663967611, | |
| "grad_norm": 0.39247070539891293, | |
| "learning_rate": 1.6094726892647147e-05, | |
| "loss": 0.2772, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.0364372469635628, | |
| "grad_norm": 0.3041882473791223, | |
| "learning_rate": 1.6039276882848135e-05, | |
| "loss": 0.2881, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0404858299595143, | |
| "grad_norm": 0.36848588943244975, | |
| "learning_rate": 1.5983858542885635e-05, | |
| "loss": 0.2819, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.0445344129554655, | |
| "grad_norm": 0.3244333880886919, | |
| "learning_rate": 1.5928472315883203e-05, | |
| "loss": 0.2766, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.048582995951417, | |
| "grad_norm": 0.32305939216095836, | |
| "learning_rate": 1.5873118644707633e-05, | |
| "loss": 0.2754, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "grad_norm": 0.3532407139642198, | |
| "learning_rate": 1.5817797971965413e-05, | |
| "loss": 0.2699, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.0566801619433197, | |
| "grad_norm": 0.26423615349541224, | |
| "learning_rate": 1.576251073999917e-05, | |
| "loss": 0.2764, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.0607287449392713, | |
| "grad_norm": 0.3315841084603363, | |
| "learning_rate": 1.5707257390884126e-05, | |
| "loss": 0.2834, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.064777327935223, | |
| "grad_norm": 0.2891016351650457, | |
| "learning_rate": 1.5652038366424595e-05, | |
| "loss": 0.2792, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.068825910931174, | |
| "grad_norm": 0.25881601927471987, | |
| "learning_rate": 1.5596854108150424e-05, | |
| "loss": 0.266, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.0728744939271255, | |
| "grad_norm": 0.2932754568591462, | |
| "learning_rate": 1.5541705057313476e-05, | |
| "loss": 0.2802, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.25384294658331785, | |
| "learning_rate": 1.5486591654884086e-05, | |
| "loss": 0.2729, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.080971659919028, | |
| "grad_norm": 0.2773608555118775, | |
| "learning_rate": 1.543151434154755e-05, | |
| "loss": 0.2816, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.08502024291498, | |
| "grad_norm": 0.29250269139918106, | |
| "learning_rate": 1.5376473557700615e-05, | |
| "loss": 0.2795, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.0890688259109313, | |
| "grad_norm": 0.2531474739528881, | |
| "learning_rate": 1.5321469743447893e-05, | |
| "loss": 0.2739, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.0931174089068825, | |
| "grad_norm": 0.29624980817083113, | |
| "learning_rate": 1.5266503338598434e-05, | |
| "loss": 0.2707, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.097165991902834, | |
| "grad_norm": 0.2391376005027395, | |
| "learning_rate": 1.5211574782662149e-05, | |
| "loss": 0.2804, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1012145748987856, | |
| "grad_norm": 0.27115582163367963, | |
| "learning_rate": 1.5156684514846307e-05, | |
| "loss": 0.2818, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.1052631578947367, | |
| "grad_norm": 0.25703871435464876, | |
| "learning_rate": 1.5101832974052012e-05, | |
| "loss": 0.2767, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.1093117408906883, | |
| "grad_norm": 0.2673893168078373, | |
| "learning_rate": 1.5047020598870737e-05, | |
| "loss": 0.2756, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.11336032388664, | |
| "grad_norm": 0.2999934337632074, | |
| "learning_rate": 1.4992247827580778e-05, | |
| "loss": 0.2785, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.117408906882591, | |
| "grad_norm": 0.26229562309035226, | |
| "learning_rate": 1.493751509814374e-05, | |
| "loss": 0.2849, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.1214574898785425, | |
| "grad_norm": 0.289832678285472, | |
| "learning_rate": 1.4882822848201073e-05, | |
| "loss": 0.2698, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.125506072874494, | |
| "grad_norm": 0.2841154611662794, | |
| "learning_rate": 1.4828171515070553e-05, | |
| "loss": 0.2711, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.1295546558704452, | |
| "grad_norm": 0.2504572826963693, | |
| "learning_rate": 1.4773561535742793e-05, | |
| "loss": 0.2709, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.133603238866397, | |
| "grad_norm": 0.2607181425163649, | |
| "learning_rate": 1.4718993346877715e-05, | |
| "loss": 0.2737, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.1376518218623484, | |
| "grad_norm": 0.23851315934997303, | |
| "learning_rate": 1.466446738480111e-05, | |
| "loss": 0.2706, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.1417004048582995, | |
| "grad_norm": 0.2603180501914701, | |
| "learning_rate": 1.460998408550112e-05, | |
| "loss": 0.2771, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.145748987854251, | |
| "grad_norm": 0.26247481255143906, | |
| "learning_rate": 1.4555543884624751e-05, | |
| "loss": 0.2794, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.1497975708502026, | |
| "grad_norm": 0.2544189021287105, | |
| "learning_rate": 1.4501147217474402e-05, | |
| "loss": 0.2686, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 0.2578408695999606, | |
| "learning_rate": 1.444679451900437e-05, | |
| "loss": 0.2689, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.24145312062029148, | |
| "learning_rate": 1.4392486223817397e-05, | |
| "loss": 0.27, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.161943319838057, | |
| "grad_norm": 0.2630532864366628, | |
| "learning_rate": 1.4338222766161143e-05, | |
| "loss": 0.2738, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.165991902834008, | |
| "grad_norm": 0.23377195454250474, | |
| "learning_rate": 1.4284004579924777e-05, | |
| "loss": 0.2841, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.1700404858299596, | |
| "grad_norm": 0.23632055447038902, | |
| "learning_rate": 1.4229832098635479e-05, | |
| "loss": 0.2798, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.174089068825911, | |
| "grad_norm": 0.23637284054425747, | |
| "learning_rate": 1.4175705755454963e-05, | |
| "loss": 0.254, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.1781376518218623, | |
| "grad_norm": 0.25612895412374814, | |
| "learning_rate": 1.412162598317602e-05, | |
| "loss": 0.2863, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.182186234817814, | |
| "grad_norm": 0.25080048003655375, | |
| "learning_rate": 1.406759321421907e-05, | |
| "loss": 0.2682, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.1862348178137654, | |
| "grad_norm": 0.2454013573645321, | |
| "learning_rate": 1.401360788062871e-05, | |
| "loss": 0.2698, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.1902834008097165, | |
| "grad_norm": 0.21293400011933686, | |
| "learning_rate": 1.3959670414070204e-05, | |
| "loss": 0.2803, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.194331983805668, | |
| "grad_norm": 0.240562820839914, | |
| "learning_rate": 1.3905781245826108e-05, | |
| "loss": 0.2692, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.1983805668016196, | |
| "grad_norm": 0.2291946867383788, | |
| "learning_rate": 1.3851940806792778e-05, | |
| "loss": 0.2652, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.2024291497975708, | |
| "grad_norm": 0.22758973259064605, | |
| "learning_rate": 1.379814952747693e-05, | |
| "loss": 0.2724, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.2064777327935223, | |
| "grad_norm": 0.2337506394113515, | |
| "learning_rate": 1.3744407837992193e-05, | |
| "loss": 0.2776, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 0.24896981291148973, | |
| "learning_rate": 1.3690716168055692e-05, | |
| "loss": 0.2692, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.214574898785425, | |
| "grad_norm": 0.2273781684399269, | |
| "learning_rate": 1.363707494698459e-05, | |
| "loss": 0.2749, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.2186234817813766, | |
| "grad_norm": 0.2425236538094465, | |
| "learning_rate": 1.358348460369265e-05, | |
| "loss": 0.2756, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.2226720647773277, | |
| "grad_norm": 0.2308014862589458, | |
| "learning_rate": 1.352994556668684e-05, | |
| "loss": 0.2777, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.2267206477732793, | |
| "grad_norm": 0.23192716296256027, | |
| "learning_rate": 1.3476458264063865e-05, | |
| "loss": 0.28, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.2125079213095375, | |
| "learning_rate": 1.342302312350678e-05, | |
| "loss": 0.2733, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.234817813765182, | |
| "grad_norm": 0.22484645274528445, | |
| "learning_rate": 1.3369640572281537e-05, | |
| "loss": 0.2774, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.2388663967611335, | |
| "grad_norm": 0.22247604450563166, | |
| "learning_rate": 1.3316311037233596e-05, | |
| "loss": 0.2766, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.242914979757085, | |
| "grad_norm": 0.21025118190788517, | |
| "learning_rate": 1.326303494478451e-05, | |
| "loss": 0.2676, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.246963562753036, | |
| "grad_norm": 0.2340051835347287, | |
| "learning_rate": 1.3209812720928472e-05, | |
| "loss": 0.2794, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.251012145748988, | |
| "grad_norm": 0.2254045211991394, | |
| "learning_rate": 1.315664479122898e-05, | |
| "loss": 0.2796, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.2550607287449393, | |
| "grad_norm": 0.23150852849786832, | |
| "learning_rate": 1.3103531580815378e-05, | |
| "loss": 0.2764, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.2591093117408905, | |
| "grad_norm": 0.24517999950678815, | |
| "learning_rate": 1.3050473514379488e-05, | |
| "loss": 0.2623, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.263157894736842, | |
| "grad_norm": 0.2284065158911635, | |
| "learning_rate": 1.2997471016172188e-05, | |
| "loss": 0.2688, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.2672064777327936, | |
| "grad_norm": 0.24551094745502228, | |
| "learning_rate": 1.2944524510000042e-05, | |
| "loss": 0.2908, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.2712550607287447, | |
| "grad_norm": 0.2203242067400843, | |
| "learning_rate": 1.2891634419221908e-05, | |
| "loss": 0.2711, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.2753036437246963, | |
| "grad_norm": 0.22609280441057272, | |
| "learning_rate": 1.2838801166745545e-05, | |
| "loss": 0.2671, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.279352226720648, | |
| "grad_norm": 0.22619004839505413, | |
| "learning_rate": 1.2786025175024216e-05, | |
| "loss": 0.2847, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.283400809716599, | |
| "grad_norm": 0.22288116095292684, | |
| "learning_rate": 1.2733306866053357e-05, | |
| "loss": 0.274, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.2874493927125505, | |
| "grad_norm": 0.23960228115663865, | |
| "learning_rate": 1.2680646661367163e-05, | |
| "loss": 0.2616, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.291497975708502, | |
| "grad_norm": 0.22907465142551603, | |
| "learning_rate": 1.2628044982035219e-05, | |
| "loss": 0.2643, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.2955465587044532, | |
| "grad_norm": 0.24631669796755626, | |
| "learning_rate": 1.2575502248659155e-05, | |
| "loss": 0.2868, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.299595141700405, | |
| "grad_norm": 0.22391230277988403, | |
| "learning_rate": 1.2523018881369269e-05, | |
| "loss": 0.2726, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.3036437246963564, | |
| "grad_norm": 0.2446732284707899, | |
| "learning_rate": 1.2470595299821176e-05, | |
| "loss": 0.2854, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.23364981531023354, | |
| "learning_rate": 1.2418231923192416e-05, | |
| "loss": 0.2892, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.311740890688259, | |
| "grad_norm": 0.2424961510140671, | |
| "learning_rate": 1.2365929170179165e-05, | |
| "loss": 0.2672, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.3157894736842106, | |
| "grad_norm": 0.22529081958658861, | |
| "learning_rate": 1.2313687458992848e-05, | |
| "loss": 0.2783, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.3198380566801617, | |
| "grad_norm": 0.23647736617154086, | |
| "learning_rate": 1.2261507207356785e-05, | |
| "loss": 0.2742, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.3238866396761133, | |
| "grad_norm": 0.22389449289349847, | |
| "learning_rate": 1.220938883250288e-05, | |
| "loss": 0.2775, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.327935222672065, | |
| "grad_norm": 0.20533052701848864, | |
| "learning_rate": 1.215733275116827e-05, | |
| "loss": 0.2799, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.331983805668016, | |
| "grad_norm": 0.22744567472097166, | |
| "learning_rate": 1.2105339379592005e-05, | |
| "loss": 0.2793, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.3360323886639676, | |
| "grad_norm": 0.22344308552333722, | |
| "learning_rate": 1.2053409133511681e-05, | |
| "loss": 0.2805, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.340080971659919, | |
| "grad_norm": 0.2190081330743894, | |
| "learning_rate": 1.2001542428160173e-05, | |
| "loss": 0.2693, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.3441295546558703, | |
| "grad_norm": 0.2423037418204456, | |
| "learning_rate": 1.1949739678262278e-05, | |
| "loss": 0.2763, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.348178137651822, | |
| "grad_norm": 0.24104433103647746, | |
| "learning_rate": 1.189800129803139e-05, | |
| "loss": 0.2629, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.3522267206477734, | |
| "grad_norm": 0.23405981978204357, | |
| "learning_rate": 1.1846327701166228e-05, | |
| "loss": 0.2749, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.3562753036437245, | |
| "grad_norm": 0.2304175611849135, | |
| "learning_rate": 1.1794719300847493e-05, | |
| "loss": 0.2707, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.360323886639676, | |
| "grad_norm": 0.24370184596640457, | |
| "learning_rate": 1.1743176509734582e-05, | |
| "loss": 0.2906, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.3643724696356276, | |
| "grad_norm": 0.23883636006734774, | |
| "learning_rate": 1.1691699739962275e-05, | |
| "loss": 0.2744, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.3684210526315788, | |
| "grad_norm": 0.23055074021339023, | |
| "learning_rate": 1.1640289403137456e-05, | |
| "loss": 0.2775, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.3724696356275303, | |
| "grad_norm": 0.2276190504071774, | |
| "learning_rate": 1.1588945910335802e-05, | |
| "loss": 0.2695, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.376518218623482, | |
| "grad_norm": 0.22487309753583623, | |
| "learning_rate": 1.1537669672098517e-05, | |
| "loss": 0.2664, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.380566801619433, | |
| "grad_norm": 0.2452658051808123, | |
| "learning_rate": 1.1486461098429036e-05, | |
| "loss": 0.2715, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.2399581894249173, | |
| "learning_rate": 1.1435320598789747e-05, | |
| "loss": 0.2808, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.388663967611336, | |
| "grad_norm": 0.22738993204221872, | |
| "learning_rate": 1.1384248582098736e-05, | |
| "loss": 0.2756, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.3927125506072873, | |
| "grad_norm": 0.23866936808073125, | |
| "learning_rate": 1.1333245456726467e-05, | |
| "loss": 0.2751, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.396761133603239, | |
| "grad_norm": 0.23107554910851374, | |
| "learning_rate": 1.128231163049258e-05, | |
| "loss": 0.2883, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.4008097165991904, | |
| "grad_norm": 0.24047373929132068, | |
| "learning_rate": 1.123144751066259e-05, | |
| "loss": 0.2704, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.4048582995951415, | |
| "grad_norm": 0.2120472570710982, | |
| "learning_rate": 1.1180653503944657e-05, | |
| "loss": 0.2701, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.408906882591093, | |
| "grad_norm": 0.20545371612277163, | |
| "learning_rate": 1.1129930016486287e-05, | |
| "loss": 0.2682, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.4129554655870447, | |
| "grad_norm": 0.23995594010514407, | |
| "learning_rate": 1.107927745387114e-05, | |
| "loss": 0.2807, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.417004048582996, | |
| "grad_norm": 0.22098774101980007, | |
| "learning_rate": 1.1028696221115766e-05, | |
| "loss": 0.2693, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.2171324136751636, | |
| "learning_rate": 1.0978186722666341e-05, | |
| "loss": 0.2718, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.425101214574899, | |
| "grad_norm": 0.23196262645073182, | |
| "learning_rate": 1.0927749362395475e-05, | |
| "loss": 0.2678, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.42914979757085, | |
| "grad_norm": 0.22931709227450953, | |
| "learning_rate": 1.0877384543598954e-05, | |
| "loss": 0.2778, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.4331983805668016, | |
| "grad_norm": 0.21204621784410627, | |
| "learning_rate": 1.0827092668992529e-05, | |
| "loss": 0.2883, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.437246963562753, | |
| "grad_norm": 0.23240873407707752, | |
| "learning_rate": 1.077687414070869e-05, | |
| "loss": 0.2775, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.4412955465587043, | |
| "grad_norm": 0.21566561803790474, | |
| "learning_rate": 1.0726729360293451e-05, | |
| "loss": 0.2631, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.445344129554656, | |
| "grad_norm": 0.2333333419037628, | |
| "learning_rate": 1.0676658728703147e-05, | |
| "loss": 0.2841, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.4493927125506074, | |
| "grad_norm": 0.22355132202453654, | |
| "learning_rate": 1.0626662646301198e-05, | |
| "loss": 0.2734, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.4534412955465585, | |
| "grad_norm": 0.22759660309676935, | |
| "learning_rate": 1.0576741512854958e-05, | |
| "loss": 0.276, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.45748987854251, | |
| "grad_norm": 0.20910642038052116, | |
| "learning_rate": 1.052689572753247e-05, | |
| "loss": 0.2651, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.22769659560374206, | |
| "learning_rate": 1.0477125688899328e-05, | |
| "loss": 0.2732, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.465587044534413, | |
| "grad_norm": 0.216183671279783, | |
| "learning_rate": 1.0427431794915406e-05, | |
| "loss": 0.2711, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.4696356275303644, | |
| "grad_norm": 0.21816301405634342, | |
| "learning_rate": 1.0377814442931769e-05, | |
| "loss": 0.2672, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.473684210526316, | |
| "grad_norm": 0.24550934828693932, | |
| "learning_rate": 1.0328274029687439e-05, | |
| "loss": 0.2835, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.477732793522267, | |
| "grad_norm": 0.22790414798699687, | |
| "learning_rate": 1.0278810951306253e-05, | |
| "loss": 0.2808, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.4817813765182186, | |
| "grad_norm": 0.2412276620393712, | |
| "learning_rate": 1.0229425603293647e-05, | |
| "loss": 0.2733, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.48582995951417, | |
| "grad_norm": 0.24401712268006648, | |
| "learning_rate": 1.0180118380533558e-05, | |
| "loss": 0.2735, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.4898785425101213, | |
| "grad_norm": 0.22498367700082653, | |
| "learning_rate": 1.0130889677285225e-05, | |
| "loss": 0.2761, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.493927125506073, | |
| "grad_norm": 0.23509040400350317, | |
| "learning_rate": 1.008173988718005e-05, | |
| "loss": 0.2695, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.4979757085020244, | |
| "grad_norm": 0.22121177406946624, | |
| "learning_rate": 1.0032669403218439e-05, | |
| "loss": 0.2761, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.5020242914979756, | |
| "grad_norm": 0.21270017932458063, | |
| "learning_rate": 9.983678617766677e-06, | |
| "loss": 0.2732, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.506072874493927, | |
| "grad_norm": 0.22807917282472465, | |
| "learning_rate": 9.934767922553788e-06, | |
| "loss": 0.2711, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.5101214574898787, | |
| "grad_norm": 0.2192306063510848, | |
| "learning_rate": 9.885937708668371e-06, | |
| "loss": 0.2839, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.51417004048583, | |
| "grad_norm": 0.26621889453968445, | |
| "learning_rate": 9.837188366555524e-06, | |
| "loss": 0.2713, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.5182186234817814, | |
| "grad_norm": 0.2160119217935681, | |
| "learning_rate": 9.788520286013691e-06, | |
| "loss": 0.2663, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.522267206477733, | |
| "grad_norm": 0.2256337929624642, | |
| "learning_rate": 9.739933856191537e-06, | |
| "loss": 0.2739, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.526315789473684, | |
| "grad_norm": 0.22980682675238387, | |
| "learning_rate": 9.691429465584866e-06, | |
| "loss": 0.2962, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.5303643724696356, | |
| "grad_norm": 0.23032343063047414, | |
| "learning_rate": 9.643007502033492e-06, | |
| "loss": 0.2828, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.534412955465587, | |
| "grad_norm": 0.22088264379096936, | |
| "learning_rate": 9.594668352718152e-06, | |
| "loss": 0.2875, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.23111588240282321, | |
| "learning_rate": 9.546412404157383e-06, | |
| "loss": 0.2723, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.54251012145749, | |
| "grad_norm": 0.22255573105516932, | |
| "learning_rate": 9.498240042204471e-06, | |
| "loss": 0.2668, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.5465587044534415, | |
| "grad_norm": 0.24789852528370937, | |
| "learning_rate": 9.450151652044339e-06, | |
| "loss": 0.28, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.5506072874493926, | |
| "grad_norm": 0.2174962403250783, | |
| "learning_rate": 9.402147618190472e-06, | |
| "loss": 0.2745, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.554655870445344, | |
| "grad_norm": 0.21096639453938534, | |
| "learning_rate": 9.354228324481841e-06, | |
| "loss": 0.2833, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.5587044534412957, | |
| "grad_norm": 0.24262124620791237, | |
| "learning_rate": 9.306394154079852e-06, | |
| "loss": 0.2765, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.562753036437247, | |
| "grad_norm": 0.22988509259285667, | |
| "learning_rate": 9.258645489465254e-06, | |
| "loss": 0.286, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.5668016194331984, | |
| "grad_norm": 0.22661313233850686, | |
| "learning_rate": 9.21098271243509e-06, | |
| "loss": 0.281, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.57085020242915, | |
| "grad_norm": 0.22815324480634108, | |
| "learning_rate": 9.163406204099659e-06, | |
| "loss": 0.2761, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.574898785425101, | |
| "grad_norm": 0.24370392530531737, | |
| "learning_rate": 9.115916344879452e-06, | |
| "loss": 0.2747, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.5789473684210527, | |
| "grad_norm": 0.22138451189195568, | |
| "learning_rate": 9.068513514502126e-06, | |
| "loss": 0.2889, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.582995951417004, | |
| "grad_norm": 0.24362885954081398, | |
| "learning_rate": 9.021198091999439e-06, | |
| "loss": 0.2801, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.5870445344129553, | |
| "grad_norm": 0.23495858517592913, | |
| "learning_rate": 8.973970455704251e-06, | |
| "loss": 0.2689, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.591093117408907, | |
| "grad_norm": 0.23108291210969312, | |
| "learning_rate": 8.926830983247495e-06, | |
| "loss": 0.2793, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.5951417004048585, | |
| "grad_norm": 0.22134344587393162, | |
| "learning_rate": 8.879780051555116e-06, | |
| "loss": 0.2824, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.5991902834008096, | |
| "grad_norm": 0.23330899904625127, | |
| "learning_rate": 8.83281803684512e-06, | |
| "loss": 0.2719, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.603238866396761, | |
| "grad_norm": 0.22334546800876995, | |
| "learning_rate": 8.785945314624532e-06, | |
| "loss": 0.2526, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.6072874493927127, | |
| "grad_norm": 0.20732606645524118, | |
| "learning_rate": 8.739162259686385e-06, | |
| "loss": 0.2769, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.611336032388664, | |
| "grad_norm": 0.21958369933588684, | |
| "learning_rate": 8.692469246106734e-06, | |
| "loss": 0.2769, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 0.22120704271413164, | |
| "learning_rate": 8.645866647241677e-06, | |
| "loss": 0.266, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.619433198380567, | |
| "grad_norm": 0.20989721694212635, | |
| "learning_rate": 8.599354835724348e-06, | |
| "loss": 0.2773, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.623481781376518, | |
| "grad_norm": 0.2273951270621597, | |
| "learning_rate": 8.552934183461958e-06, | |
| "loss": 0.268, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.6275303643724697, | |
| "grad_norm": 0.19423155281509807, | |
| "learning_rate": 8.506605061632805e-06, | |
| "loss": 0.2715, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 0.2135132476375263, | |
| "learning_rate": 8.460367840683308e-06, | |
| "loss": 0.2778, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 3.6356275303643724, | |
| "grad_norm": 0.2213423949151078, | |
| "learning_rate": 8.414222890325065e-06, | |
| "loss": 0.2745, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.639676113360324, | |
| "grad_norm": 0.19310076412233565, | |
| "learning_rate": 8.368170579531855e-06, | |
| "loss": 0.2704, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 3.6437246963562755, | |
| "grad_norm": 0.21103371027078982, | |
| "learning_rate": 8.322211276536733e-06, | |
| "loss": 0.2783, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.6477732793522266, | |
| "grad_norm": 0.24115030001662396, | |
| "learning_rate": 8.27634534882907e-06, | |
| "loss": 0.271, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 3.651821862348178, | |
| "grad_norm": 0.2025342289300406, | |
| "learning_rate": 8.230573163151585e-06, | |
| "loss": 0.2727, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.6558704453441297, | |
| "grad_norm": 0.22212999208507023, | |
| "learning_rate": 8.184895085497466e-06, | |
| "loss": 0.2763, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 3.659919028340081, | |
| "grad_norm": 0.2268337873433904, | |
| "learning_rate": 8.139311481107401e-06, | |
| "loss": 0.2834, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.6639676113360324, | |
| "grad_norm": 0.2121832507583236, | |
| "learning_rate": 8.093822714466686e-06, | |
| "loss": 0.2817, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.668016194331984, | |
| "grad_norm": 0.21378488808559495, | |
| "learning_rate": 8.048429149302271e-06, | |
| "loss": 0.2702, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.672064777327935, | |
| "grad_norm": 0.20852072002688057, | |
| "learning_rate": 8.003131148579902e-06, | |
| "loss": 0.2777, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 3.6761133603238867, | |
| "grad_norm": 0.2157655073352012, | |
| "learning_rate": 7.957929074501182e-06, | |
| "loss": 0.2749, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.6801619433198383, | |
| "grad_norm": 0.2140234756837412, | |
| "learning_rate": 7.912823288500695e-06, | |
| "loss": 0.2866, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.2228329071787837, | |
| "learning_rate": 7.867814151243102e-06, | |
| "loss": 0.2759, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.688259109311741, | |
| "grad_norm": 0.2092824939842251, | |
| "learning_rate": 7.822902022620267e-06, | |
| "loss": 0.2745, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.21982182099716843, | |
| "learning_rate": 7.778087261748377e-06, | |
| "loss": 0.2673, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.6963562753036436, | |
| "grad_norm": 0.23840869154631206, | |
| "learning_rate": 7.733370226965056e-06, | |
| "loss": 0.2833, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 3.700404858299595, | |
| "grad_norm": 0.21367833527000166, | |
| "learning_rate": 7.688751275826527e-06, | |
| "loss": 0.2665, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.7044534412955468, | |
| "grad_norm": 0.20892862840622667, | |
| "learning_rate": 7.644230765104735e-06, | |
| "loss": 0.28, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.708502024291498, | |
| "grad_norm": 0.21840475901040854, | |
| "learning_rate": 7.599809050784502e-06, | |
| "loss": 0.2704, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.7125506072874495, | |
| "grad_norm": 0.20189744865832676, | |
| "learning_rate": 7.555486488060661e-06, | |
| "loss": 0.2777, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 3.716599190283401, | |
| "grad_norm": 0.209374690768919, | |
| "learning_rate": 7.511263431335249e-06, | |
| "loss": 0.2738, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.720647773279352, | |
| "grad_norm": 0.20944660973773835, | |
| "learning_rate": 7.467140234214656e-06, | |
| "loss": 0.2781, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 3.7246963562753037, | |
| "grad_norm": 0.21567700916456187, | |
| "learning_rate": 7.423117249506777e-06, | |
| "loss": 0.2686, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.7287449392712553, | |
| "grad_norm": 0.1997748571983223, | |
| "learning_rate": 7.379194829218231e-06, | |
| "loss": 0.2694, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 3.7327935222672064, | |
| "grad_norm": 0.2206341413304694, | |
| "learning_rate": 7.335373324551518e-06, | |
| "loss": 0.2764, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 3.736842105263158, | |
| "grad_norm": 0.2218863406166018, | |
| "learning_rate": 7.291653085902224e-06, | |
| "loss": 0.2746, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 3.7408906882591095, | |
| "grad_norm": 0.20211362905564034, | |
| "learning_rate": 7.248034462856213e-06, | |
| "loss": 0.2801, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 3.7449392712550607, | |
| "grad_norm": 0.20987211932043912, | |
| "learning_rate": 7.20451780418683e-06, | |
| "loss": 0.2704, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.748987854251012, | |
| "grad_norm": 0.20265192919453726, | |
| "learning_rate": 7.161103457852128e-06, | |
| "loss": 0.2851, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 3.753036437246964, | |
| "grad_norm": 0.21088645314203078, | |
| "learning_rate": 7.117791770992046e-06, | |
| "loss": 0.2822, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 3.757085020242915, | |
| "grad_norm": 0.1940611806662089, | |
| "learning_rate": 7.074583089925688e-06, | |
| "loss": 0.2796, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 3.7611336032388665, | |
| "grad_norm": 0.21668429020684668, | |
| "learning_rate": 7.0314777601485155e-06, | |
| "loss": 0.2826, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 3.765182186234818, | |
| "grad_norm": 0.201594705846371, | |
| "learning_rate": 6.988476126329604e-06, | |
| "loss": 0.2825, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.18915395714699687, | |
| "learning_rate": 6.945578532308857e-06, | |
| "loss": 0.2703, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 3.7732793522267207, | |
| "grad_norm": 0.19567255160874517, | |
| "learning_rate": 6.902785321094301e-06, | |
| "loss": 0.2737, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 3.7773279352226723, | |
| "grad_norm": 0.19763801531864406, | |
| "learning_rate": 6.8600968348593115e-06, | |
| "loss": 0.2712, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 3.7813765182186234, | |
| "grad_norm": 0.20766573476369898, | |
| "learning_rate": 6.817513414939892e-06, | |
| "loss": 0.2787, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 3.785425101214575, | |
| "grad_norm": 0.18775341428898487, | |
| "learning_rate": 6.775035401831924e-06, | |
| "loss": 0.2635, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.7894736842105265, | |
| "grad_norm": 0.19808177771735358, | |
| "learning_rate": 6.732663135188473e-06, | |
| "loss": 0.2777, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 3.7935222672064777, | |
| "grad_norm": 0.19770812640816338, | |
| "learning_rate": 6.690396953817055e-06, | |
| "loss": 0.2776, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 3.7975708502024292, | |
| "grad_norm": 0.20864644162460622, | |
| "learning_rate": 6.648237195676927e-06, | |
| "loss": 0.2734, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 3.801619433198381, | |
| "grad_norm": 0.21389128538277963, | |
| "learning_rate": 6.606184197876395e-06, | |
| "loss": 0.2743, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.805668016194332, | |
| "grad_norm": 0.23421952006711108, | |
| "learning_rate": 6.564238296670109e-06, | |
| "loss": 0.2764, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.8097165991902835, | |
| "grad_norm": 0.21660789076307418, | |
| "learning_rate": 6.52239982745638e-06, | |
| "loss": 0.273, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 3.813765182186235, | |
| "grad_norm": 0.2063598034913621, | |
| "learning_rate": 6.480669124774484e-06, | |
| "loss": 0.2754, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 3.817813765182186, | |
| "grad_norm": 0.2186263984524299, | |
| "learning_rate": 6.439046522302013e-06, | |
| "loss": 0.2844, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 3.8218623481781377, | |
| "grad_norm": 0.2081002490238555, | |
| "learning_rate": 6.397532352852191e-06, | |
| "loss": 0.2679, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 3.8259109311740893, | |
| "grad_norm": 0.203782544875315, | |
| "learning_rate": 6.356126948371202e-06, | |
| "loss": 0.2616, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.8299595141700404, | |
| "grad_norm": 0.20593075661834678, | |
| "learning_rate": 6.314830639935559e-06, | |
| "loss": 0.2712, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.834008097165992, | |
| "grad_norm": 0.21781329034997357, | |
| "learning_rate": 6.273643757749448e-06, | |
| "loss": 0.2771, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 3.8380566801619436, | |
| "grad_norm": 0.20779406550290436, | |
| "learning_rate": 6.2325666311420856e-06, | |
| "loss": 0.2694, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 3.8421052631578947, | |
| "grad_norm": 0.19383841924557269, | |
| "learning_rate": 6.1915995885650675e-06, | |
| "loss": 0.2819, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.20880381684797428, | |
| "learning_rate": 6.150742957589784e-06, | |
| "loss": 0.2669, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.850202429149798, | |
| "grad_norm": 0.19889643655056133, | |
| "learning_rate": 6.109997064904774e-06, | |
| "loss": 0.2778, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.854251012145749, | |
| "grad_norm": 0.19172122571504077, | |
| "learning_rate": 6.069362236313092e-06, | |
| "loss": 0.2736, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 3.8582995951417005, | |
| "grad_norm": 0.19254612672380758, | |
| "learning_rate": 6.0288387967297565e-06, | |
| "loss": 0.2773, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 3.862348178137652, | |
| "grad_norm": 0.20444502303307238, | |
| "learning_rate": 5.988427070179108e-06, | |
| "loss": 0.2698, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.866396761133603, | |
| "grad_norm": 0.20236913170767598, | |
| "learning_rate": 5.948127379792234e-06, | |
| "loss": 0.2716, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.8704453441295548, | |
| "grad_norm": 0.19552813311486603, | |
| "learning_rate": 5.90794004780439e-06, | |
| "loss": 0.2894, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 3.8744939271255063, | |
| "grad_norm": 0.20466257946665156, | |
| "learning_rate": 5.867865395552405e-06, | |
| "loss": 0.2757, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 3.8785425101214575, | |
| "grad_norm": 0.2025478376911895, | |
| "learning_rate": 5.8279037434721364e-06, | |
| "loss": 0.2736, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 3.882591093117409, | |
| "grad_norm": 0.19680613129078978, | |
| "learning_rate": 5.788055411095892e-06, | |
| "loss": 0.28, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 3.8866396761133606, | |
| "grad_norm": 0.19560062334078354, | |
| "learning_rate": 5.7483207170498625e-06, | |
| "loss": 0.2781, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.8906882591093117, | |
| "grad_norm": 0.2164265129472508, | |
| "learning_rate": 5.708699979051613e-06, | |
| "loss": 0.2832, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 3.8947368421052633, | |
| "grad_norm": 0.19337548102252972, | |
| "learning_rate": 5.66919351390751e-06, | |
| "loss": 0.2791, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 3.898785425101215, | |
| "grad_norm": 0.18733683728332556, | |
| "learning_rate": 5.629801637510188e-06, | |
| "loss": 0.2687, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.902834008097166, | |
| "grad_norm": 0.20548080056581378, | |
| "learning_rate": 5.590524664836043e-06, | |
| "loss": 0.2875, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 3.9068825910931175, | |
| "grad_norm": 0.20420767608904994, | |
| "learning_rate": 5.551362909942708e-06, | |
| "loss": 0.2789, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.910931174089069, | |
| "grad_norm": 0.1998133032699797, | |
| "learning_rate": 5.51231668596653e-06, | |
| "loss": 0.275, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 3.91497975708502, | |
| "grad_norm": 0.1887436120855873, | |
| "learning_rate": 5.47338630512007e-06, | |
| "loss": 0.2823, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 3.919028340080972, | |
| "grad_norm": 0.1934820158072468, | |
| "learning_rate": 5.434572078689617e-06, | |
| "loss": 0.2816, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.18086812977124364, | |
| "learning_rate": 5.395874317032694e-06, | |
| "loss": 0.2765, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 3.9271255060728745, | |
| "grad_norm": 0.2085256572222119, | |
| "learning_rate": 5.357293329575573e-06, | |
| "loss": 0.2818, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.931174089068826, | |
| "grad_norm": 0.1952218109868174, | |
| "learning_rate": 5.318829424810797e-06, | |
| "loss": 0.2763, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 3.9352226720647776, | |
| "grad_norm": 0.18760283935912553, | |
| "learning_rate": 5.280482910294728e-06, | |
| "loss": 0.2981, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 3.9392712550607287, | |
| "grad_norm": 0.18824303338452889, | |
| "learning_rate": 5.242254092645078e-06, | |
| "loss": 0.2789, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 3.9433198380566803, | |
| "grad_norm": 0.19299507854868517, | |
| "learning_rate": 5.2041432775384445e-06, | |
| "loss": 0.2712, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 0.18812156995670393, | |
| "learning_rate": 5.166150769707892e-06, | |
| "loss": 0.2746, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.951417004048583, | |
| "grad_norm": 0.19876812684493866, | |
| "learning_rate": 5.128276872940503e-06, | |
| "loss": 0.2736, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 3.9554655870445345, | |
| "grad_norm": 0.1984249913462269, | |
| "learning_rate": 5.090521890074953e-06, | |
| "loss": 0.2585, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 3.9595141700404857, | |
| "grad_norm": 0.19549708383489883, | |
| "learning_rate": 5.0528861229990675e-06, | |
| "loss": 0.2775, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 3.9635627530364372, | |
| "grad_norm": 0.2002879527252906, | |
| "learning_rate": 5.015369872647442e-06, | |
| "loss": 0.2769, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 3.967611336032389, | |
| "grad_norm": 0.19830447793458372, | |
| "learning_rate": 4.9779734389990195e-06, | |
| "loss": 0.2724, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.97165991902834, | |
| "grad_norm": 0.1934556203927066, | |
| "learning_rate": 4.940697121074675e-06, | |
| "loss": 0.2623, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 3.9757085020242915, | |
| "grad_norm": 0.201797934170134, | |
| "learning_rate": 4.903541216934864e-06, | |
| "loss": 0.2771, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 3.979757085020243, | |
| "grad_norm": 0.19730390580307233, | |
| "learning_rate": 4.8665060236772e-06, | |
| "loss": 0.263, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 3.983805668016194, | |
| "grad_norm": 0.21152470033142184, | |
| "learning_rate": 4.829591837434102e-06, | |
| "loss": 0.2743, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 3.9878542510121457, | |
| "grad_norm": 0.19345473816011505, | |
| "learning_rate": 4.792798953370424e-06, | |
| "loss": 0.2604, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.9919028340080973, | |
| "grad_norm": 0.18352532537019378, | |
| "learning_rate": 4.756127665681083e-06, | |
| "loss": 0.2883, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 3.9959514170040484, | |
| "grad_norm": 0.20188842862821677, | |
| "learning_rate": 4.719578267588725e-06, | |
| "loss": 0.2818, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.2005208370116757, | |
| "learning_rate": 4.6831510513413525e-06, | |
| "loss": 0.2647, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.004048582995951, | |
| "grad_norm": 0.34901361176910795, | |
| "learning_rate": 4.6468463082100176e-06, | |
| "loss": 0.2352, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.008097165991903, | |
| "grad_norm": 0.28026085249919386, | |
| "learning_rate": 4.610664328486485e-06, | |
| "loss": 0.2399, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.012145748987854, | |
| "grad_norm": 0.24971356135155917, | |
| "learning_rate": 4.5746054014809e-06, | |
| "loss": 0.2311, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.016194331983805, | |
| "grad_norm": 0.4465353515122549, | |
| "learning_rate": 4.5386698155194766e-06, | |
| "loss": 0.2362, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.020242914979757, | |
| "grad_norm": 0.31758343768789943, | |
| "learning_rate": 4.502857857942204e-06, | |
| "loss": 0.2273, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.0242914979757085, | |
| "grad_norm": 0.2362498938907069, | |
| "learning_rate": 4.467169815100549e-06, | |
| "loss": 0.2202, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.02834008097166, | |
| "grad_norm": 0.36777959223969975, | |
| "learning_rate": 4.431605972355137e-06, | |
| "loss": 0.2501, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.032388663967612, | |
| "grad_norm": 0.34339793551998204, | |
| "learning_rate": 4.396166614073516e-06, | |
| "loss": 0.2487, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.036437246963563, | |
| "grad_norm": 0.2256810088427771, | |
| "learning_rate": 4.360852023627846e-06, | |
| "loss": 0.2382, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.040485829959514, | |
| "grad_norm": 0.2567921070304409, | |
| "learning_rate": 4.325662483392656e-06, | |
| "loss": 0.2505, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.044534412955466, | |
| "grad_norm": 0.32743512328836627, | |
| "learning_rate": 4.29059827474257e-06, | |
| "loss": 0.2464, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.048582995951417, | |
| "grad_norm": 0.25595562058871, | |
| "learning_rate": 4.255659678050063e-06, | |
| "loss": 0.2401, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.052631578947368, | |
| "grad_norm": 0.21685571611557763, | |
| "learning_rate": 4.22084697268323e-06, | |
| "loss": 0.2384, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.05668016194332, | |
| "grad_norm": 0.21800976976026637, | |
| "learning_rate": 4.186160437003535e-06, | |
| "loss": 0.2251, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.060728744939271, | |
| "grad_norm": 0.2462370004707933, | |
| "learning_rate": 4.151600348363582e-06, | |
| "loss": 0.2413, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.064777327935222, | |
| "grad_norm": 0.23692170289670073, | |
| "learning_rate": 4.11716698310493e-06, | |
| "loss": 0.2346, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.068825910931174, | |
| "grad_norm": 0.22771946366312953, | |
| "learning_rate": 4.082860616555848e-06, | |
| "loss": 0.2407, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.0728744939271255, | |
| "grad_norm": 0.20097238594565542, | |
| "learning_rate": 4.048681523029125e-06, | |
| "loss": 0.2399, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 0.23661690382205344, | |
| "learning_rate": 4.014629975819882e-06, | |
| "loss": 0.2405, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.080971659919029, | |
| "grad_norm": 0.22501314380470708, | |
| "learning_rate": 3.9807062472033855e-06, | |
| "loss": 0.2347, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.08502024291498, | |
| "grad_norm": 0.19736340770490957, | |
| "learning_rate": 3.94691060843287e-06, | |
| "loss": 0.2203, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.089068825910931, | |
| "grad_norm": 0.1939977602090196, | |
| "learning_rate": 3.913243329737353e-06, | |
| "loss": 0.2258, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.093117408906883, | |
| "grad_norm": 0.20891483858135515, | |
| "learning_rate": 3.879704680319503e-06, | |
| "loss": 0.2391, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.097165991902834, | |
| "grad_norm": 0.21624606014501344, | |
| "learning_rate": 3.846294928353471e-06, | |
| "loss": 0.2343, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.101214574898785, | |
| "grad_norm": 0.18985148101160904, | |
| "learning_rate": 3.8130143409827345e-06, | |
| "loss": 0.2451, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.105263157894737, | |
| "grad_norm": 0.20200750142001705, | |
| "learning_rate": 3.7798631843179887e-06, | |
| "loss": 0.2337, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.109311740890688, | |
| "grad_norm": 0.1952003585038046, | |
| "learning_rate": 3.7468417234349973e-06, | |
| "loss": 0.2262, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.113360323886639, | |
| "grad_norm": 0.21104994911457123, | |
| "learning_rate": 3.7139502223724823e-06, | |
| "loss": 0.2505, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.117408906882591, | |
| "grad_norm": 0.18851960803719955, | |
| "learning_rate": 3.6811889441300095e-06, | |
| "loss": 0.2438, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.1214574898785425, | |
| "grad_norm": 0.19541768464571233, | |
| "learning_rate": 3.648558150665886e-06, | |
| "loss": 0.2249, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.125506072874494, | |
| "grad_norm": 0.20003318999108644, | |
| "learning_rate": 3.616058102895075e-06, | |
| "loss": 0.2409, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.129554655870446, | |
| "grad_norm": 0.20548394482610832, | |
| "learning_rate": 3.583689060687079e-06, | |
| "loss": 0.2285, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.133603238866397, | |
| "grad_norm": 0.195907497654985, | |
| "learning_rate": 3.5514512828639023e-06, | |
| "loss": 0.2402, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.137651821862348, | |
| "grad_norm": 0.19946844938954647, | |
| "learning_rate": 3.51934502719796e-06, | |
| "loss": 0.2301, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.1417004048583, | |
| "grad_norm": 0.1915020746780249, | |
| "learning_rate": 3.4873705504100185e-06, | |
| "loss": 0.2472, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.145748987854251, | |
| "grad_norm": 0.1880958502933102, | |
| "learning_rate": 3.4555281081671345e-06, | |
| "loss": 0.2438, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.149797570850202, | |
| "grad_norm": 0.19392911139468222, | |
| "learning_rate": 3.4238179550806326e-06, | |
| "loss": 0.2276, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 0.17617470932096232, | |
| "learning_rate": 3.3922403447040497e-06, | |
| "loss": 0.2393, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.157894736842105, | |
| "grad_norm": 0.19849139036565494, | |
| "learning_rate": 3.3607955295311247e-06, | |
| "loss": 0.2436, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.161943319838056, | |
| "grad_norm": 0.19509757750816298, | |
| "learning_rate": 3.32948376099375e-06, | |
| "loss": 0.2367, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.165991902834008, | |
| "grad_norm": 1.9603219010308814, | |
| "learning_rate": 3.2983052894599976e-06, | |
| "loss": 0.2267, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.17004048582996, | |
| "grad_norm": 0.2020585007229797, | |
| "learning_rate": 3.2672603642320945e-06, | |
| "loss": 0.2438, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.174089068825911, | |
| "grad_norm": 0.19701666653504407, | |
| "learning_rate": 3.236349233544438e-06, | |
| "loss": 0.2424, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.178137651821863, | |
| "grad_norm": 0.19143696784653322, | |
| "learning_rate": 3.205572144561604e-06, | |
| "loss": 0.2298, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.182186234817814, | |
| "grad_norm": 0.19862036317003293, | |
| "learning_rate": 3.174929343376374e-06, | |
| "loss": 0.2344, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.186234817813765, | |
| "grad_norm": 0.19626853144870698, | |
| "learning_rate": 3.1444210750077777e-06, | |
| "loss": 0.2346, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.190283400809717, | |
| "grad_norm": 0.19804487815338548, | |
| "learning_rate": 3.1140475833991046e-06, | |
| "loss": 0.2334, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.194331983805668, | |
| "grad_norm": 0.18566654140179967, | |
| "learning_rate": 3.0838091114159895e-06, | |
| "loss": 0.2364, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.198380566801619, | |
| "grad_norm": 0.18753530171457222, | |
| "learning_rate": 3.0537059008444524e-06, | |
| "loss": 0.231, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.202429149797571, | |
| "grad_norm": 0.19275391340300002, | |
| "learning_rate": 3.0237381923889563e-06, | |
| "loss": 0.2404, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.206477732793522, | |
| "grad_norm": 0.19723367294399533, | |
| "learning_rate": 2.9939062256704997e-06, | |
| "loss": 0.2306, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.2388019235530486, | |
| "learning_rate": 2.964210239224696e-06, | |
| "loss": 0.2398, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.2145748987854255, | |
| "grad_norm": 0.1907092343760149, | |
| "learning_rate": 2.934650470499869e-06, | |
| "loss": 0.2327, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 4.218623481781377, | |
| "grad_norm": 0.18095080134005673, | |
| "learning_rate": 2.905227155855128e-06, | |
| "loss": 0.2358, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 4.222672064777328, | |
| "grad_norm": 0.19567390577062463, | |
| "learning_rate": 2.875940530558519e-06, | |
| "loss": 0.2198, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 4.22672064777328, | |
| "grad_norm": 0.1928437303598785, | |
| "learning_rate": 2.8467908287851154e-06, | |
| "loss": 0.2322, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.18660728932848583, | |
| "learning_rate": 2.8177782836151603e-06, | |
| "loss": 0.2332, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.234817813765182, | |
| "grad_norm": 0.20177066285569314, | |
| "learning_rate": 2.7889031270321874e-06, | |
| "loss": 0.2598, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 4.238866396761134, | |
| "grad_norm": 0.19010552023939392, | |
| "learning_rate": 2.760165589921184e-06, | |
| "loss": 0.2437, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 4.242914979757085, | |
| "grad_norm": 0.34686627730737707, | |
| "learning_rate": 2.7315659020667305e-06, | |
| "loss": 0.2336, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 4.246963562753036, | |
| "grad_norm": 0.35876802287701526, | |
| "learning_rate": 2.703104292151164e-06, | |
| "loss": 0.2289, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 4.251012145748988, | |
| "grad_norm": 0.18512792965274252, | |
| "learning_rate": 2.6747809877527607e-06, | |
| "loss": 0.2474, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.255060728744939, | |
| "grad_norm": 0.1822108350511778, | |
| "learning_rate": 2.6465962153439082e-06, | |
| "loss": 0.2319, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.2591093117408905, | |
| "grad_norm": 0.23671894042336536, | |
| "learning_rate": 2.618550200289303e-06, | |
| "loss": 0.2457, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.2631578947368425, | |
| "grad_norm": 0.17853352338888404, | |
| "learning_rate": 2.59064316684412e-06, | |
| "loss": 0.2347, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.267206477732794, | |
| "grad_norm": 0.18754812225701892, | |
| "learning_rate": 2.562875338152264e-06, | |
| "loss": 0.2427, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.271255060728745, | |
| "grad_norm": 0.18229744178420046, | |
| "learning_rate": 2.5352469362445552e-06, | |
| "loss": 0.2516, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.275303643724697, | |
| "grad_norm": 0.1873373385256184, | |
| "learning_rate": 2.5077581820369523e-06, | |
| "loss": 0.2336, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.279352226720648, | |
| "grad_norm": 0.18790315977376124, | |
| "learning_rate": 2.480409295328803e-06, | |
| "loss": 0.2354, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.283400809716599, | |
| "grad_norm": 0.17553840227904918, | |
| "learning_rate": 2.45320049480108e-06, | |
| "loss": 0.2329, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.287449392712551, | |
| "grad_norm": 0.18914943578272142, | |
| "learning_rate": 2.4261319980146293e-06, | |
| "loss": 0.2342, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.291497975708502, | |
| "grad_norm": 0.1927186477453819, | |
| "learning_rate": 2.399204021408421e-06, | |
| "loss": 0.2291, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.295546558704453, | |
| "grad_norm": 0.18835358871894303, | |
| "learning_rate": 2.3724167802978395e-06, | |
| "loss": 0.2389, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.299595141700405, | |
| "grad_norm": 0.1801282260936399, | |
| "learning_rate": 2.3457704888729625e-06, | |
| "loss": 0.2353, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.303643724696356, | |
| "grad_norm": 0.18123694043556582, | |
| "learning_rate": 2.3192653601968162e-06, | |
| "loss": 0.23, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 0.18287544440886083, | |
| "learning_rate": 2.2929016062037057e-06, | |
| "loss": 0.2435, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.3117408906882595, | |
| "grad_norm": 0.19674297179866498, | |
| "learning_rate": 2.2666794376975055e-06, | |
| "loss": 0.2526, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.315789473684211, | |
| "grad_norm": 0.18550005942629805, | |
| "learning_rate": 2.24059906434998e-06, | |
| "loss": 0.2364, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.319838056680162, | |
| "grad_norm": 0.18270331606856008, | |
| "learning_rate": 2.214660694699091e-06, | |
| "loss": 0.2373, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.323886639676114, | |
| "grad_norm": 0.181844989181954, | |
| "learning_rate": 2.188864536147355e-06, | |
| "loss": 0.2366, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.327935222672065, | |
| "grad_norm": 0.18621918175338212, | |
| "learning_rate": 2.1632107949601757e-06, | |
| "loss": 0.236, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.331983805668016, | |
| "grad_norm": 0.18610220660601892, | |
| "learning_rate": 2.137699676264171e-06, | |
| "loss": 0.2402, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.336032388663968, | |
| "grad_norm": 0.1844360052658574, | |
| "learning_rate": 2.1123313840455738e-06, | |
| "loss": 0.2431, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.340080971659919, | |
| "grad_norm": 0.17369427616769065, | |
| "learning_rate": 2.08710612114857e-06, | |
| "loss": 0.2359, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.34412955465587, | |
| "grad_norm": 0.1812085021458126, | |
| "learning_rate": 2.0620240892736953e-06, | |
| "loss": 0.2376, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.348178137651822, | |
| "grad_norm": 0.18441778093942737, | |
| "learning_rate": 2.0370854889762e-06, | |
| "loss": 0.2234, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.352226720647773, | |
| "grad_norm": 0.18630008118907107, | |
| "learning_rate": 2.0122905196644726e-06, | |
| "loss": 0.2309, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.3562753036437245, | |
| "grad_norm": 0.17860225325814108, | |
| "learning_rate": 1.987639379598427e-06, | |
| "loss": 0.2472, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.3603238866396765, | |
| "grad_norm": 0.18129318678940054, | |
| "learning_rate": 1.963132265887919e-06, | |
| "loss": 0.2494, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.364372469635628, | |
| "grad_norm": 0.18256816981846238, | |
| "learning_rate": 1.9387693744911784e-06, | |
| "loss": 0.2466, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.368421052631579, | |
| "grad_norm": 0.18179787923456514, | |
| "learning_rate": 1.9145509002132368e-06, | |
| "loss": 0.2419, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.372469635627531, | |
| "grad_norm": 0.18722227467899588, | |
| "learning_rate": 1.8904770367043723e-06, | |
| "loss": 0.2266, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.376518218623482, | |
| "grad_norm": 0.18757486909018992, | |
| "learning_rate": 1.8665479764585504e-06, | |
| "loss": 0.2338, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.380566801619433, | |
| "grad_norm": 0.17953777950629707, | |
| "learning_rate": 1.8427639108118977e-06, | |
| "loss": 0.2412, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.18051859013006652, | |
| "learning_rate": 1.8191250299411755e-06, | |
| "loss": 0.2434, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.388663967611336, | |
| "grad_norm": 0.17241810219037845, | |
| "learning_rate": 1.7956315228622446e-06, | |
| "loss": 0.2341, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.392712550607287, | |
| "grad_norm": 0.1726861489652147, | |
| "learning_rate": 1.7722835774285553e-06, | |
| "loss": 0.2302, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.396761133603239, | |
| "grad_norm": 0.18634612245788362, | |
| "learning_rate": 1.7490813803296624e-06, | |
| "loss": 0.2435, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.40080971659919, | |
| "grad_norm": 0.17529836929697934, | |
| "learning_rate": 1.726025117089718e-06, | |
| "loss": 0.2292, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.4048582995951415, | |
| "grad_norm": 0.17881885603095843, | |
| "learning_rate": 1.70311497206598e-06, | |
| "loss": 0.2392, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.4089068825910935, | |
| "grad_norm": 0.1822943477307572, | |
| "learning_rate": 1.6803511284473638e-06, | |
| "loss": 0.2372, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.412955465587045, | |
| "grad_norm": 0.18035598212283552, | |
| "learning_rate": 1.6577337682529538e-06, | |
| "loss": 0.2425, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.417004048582996, | |
| "grad_norm": 0.174887977911496, | |
| "learning_rate": 1.6352630723305617e-06, | |
| "loss": 0.2351, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.421052631578947, | |
| "grad_norm": 0.17922397220298025, | |
| "learning_rate": 1.612939220355272e-06, | |
| "loss": 0.2288, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.425101214574899, | |
| "grad_norm": 0.17382103722261033, | |
| "learning_rate": 1.5907623908280112e-06, | |
| "loss": 0.2468, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.42914979757085, | |
| "grad_norm": 0.17989767378126673, | |
| "learning_rate": 1.568732761074121e-06, | |
| "loss": 0.2324, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.433198380566802, | |
| "grad_norm": 0.17826122430542657, | |
| "learning_rate": 1.5468505072419237e-06, | |
| "loss": 0.2447, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.437246963562753, | |
| "grad_norm": 0.17137704963708736, | |
| "learning_rate": 1.5251158043013448e-06, | |
| "loss": 0.2409, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.441295546558704, | |
| "grad_norm": 0.18471581973402268, | |
| "learning_rate": 1.5035288260424885e-06, | |
| "loss": 0.2363, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.445344129554655, | |
| "grad_norm": 0.1758135802578888, | |
| "learning_rate": 1.4820897450742623e-06, | |
| "loss": 0.2311, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.449392712550607, | |
| "grad_norm": 0.17431564594391172, | |
| "learning_rate": 1.4607987328229767e-06, | |
| "loss": 0.2434, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.4534412955465585, | |
| "grad_norm": 0.1721067322896035, | |
| "learning_rate": 1.4396559595310056e-06, | |
| "loss": 0.2357, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.4574898785425106, | |
| "grad_norm": 0.18034134487081338, | |
| "learning_rate": 1.4186615942553973e-06, | |
| "loss": 0.237, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.17957614133497454, | |
| "learning_rate": 1.3978158048665424e-06, | |
| "loss": 0.2413, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.465587044534413, | |
| "grad_norm": 0.17695796217724707, | |
| "learning_rate": 1.3771187580468115e-06, | |
| "loss": 0.2388, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.469635627530364, | |
| "grad_norm": 0.17700606908255098, | |
| "learning_rate": 1.3565706192892392e-06, | |
| "loss": 0.2398, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.17586789410679177, | |
| "learning_rate": 1.3361715528961993e-06, | |
| "loss": 0.2342, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.477732793522267, | |
| "grad_norm": 0.17058485228103823, | |
| "learning_rate": 1.3159217219780817e-06, | |
| "loss": 0.23, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.481781376518219, | |
| "grad_norm": 0.17858334090985933, | |
| "learning_rate": 1.2958212884519949e-06, | |
| "loss": 0.2317, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.48582995951417, | |
| "grad_norm": 0.1765085838342427, | |
| "learning_rate": 1.2758704130404675e-06, | |
| "loss": 0.2259, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.489878542510121, | |
| "grad_norm": 0.17699723906808632, | |
| "learning_rate": 1.2560692552701738e-06, | |
| "loss": 0.2387, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.493927125506072, | |
| "grad_norm": 0.17712626601931053, | |
| "learning_rate": 1.2364179734706382e-06, | |
| "loss": 0.2499, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.497975708502024, | |
| "grad_norm": 0.17969917679745095, | |
| "learning_rate": 1.2169167247729852e-06, | |
| "loss": 0.2462, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.502024291497976, | |
| "grad_norm": 0.18365872209391254, | |
| "learning_rate": 1.1975656651086864e-06, | |
| "loss": 0.2337, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.506072874493928, | |
| "grad_norm": 0.18028493127037828, | |
| "learning_rate": 1.1783649492082926e-06, | |
| "loss": 0.2383, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.510121457489879, | |
| "grad_norm": 0.17608465667557727, | |
| "learning_rate": 1.1593147306002183e-06, | |
| "loss": 0.2463, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.51417004048583, | |
| "grad_norm": 0.17185864726136982, | |
| "learning_rate": 1.140415161609505e-06, | |
| "loss": 0.2356, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.518218623481781, | |
| "grad_norm": 0.17449697641513148, | |
| "learning_rate": 1.121666393356604e-06, | |
| "loss": 0.2396, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.522267206477733, | |
| "grad_norm": 0.16947045962158258, | |
| "learning_rate": 1.1030685757561632e-06, | |
| "loss": 0.235, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.526315789473684, | |
| "grad_norm": 0.18450531453804087, | |
| "learning_rate": 1.0846218575158373e-06, | |
| "loss": 0.2312, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.530364372469636, | |
| "grad_norm": 0.17973009873227763, | |
| "learning_rate": 1.066326386135097e-06, | |
| "loss": 0.2376, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 4.534412955465587, | |
| "grad_norm": 0.1797331838826402, | |
| "learning_rate": 1.0481823079040398e-06, | |
| "loss": 0.2461, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 0.1697177464879968, | |
| "learning_rate": 1.0301897679022344e-06, | |
| "loss": 0.2445, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 4.5425101214574894, | |
| "grad_norm": 0.17700962092559594, | |
| "learning_rate": 1.0123489099975491e-06, | |
| "loss": 0.242, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 4.5465587044534415, | |
| "grad_norm": 0.17334177848267815, | |
| "learning_rate": 9.9465987684501e-07, | |
| "loss": 0.2578, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 4.550607287449393, | |
| "grad_norm": 0.17609802586171333, | |
| "learning_rate": 9.771228098856534e-07, | |
| "loss": 0.2407, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 4.554655870445345, | |
| "grad_norm": 0.17178327435464702, | |
| "learning_rate": 9.597378493454013e-07, | |
| "loss": 0.24, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.558704453441296, | |
| "grad_norm": 0.1804529203128621, | |
| "learning_rate": 9.425051342339353e-07, | |
| "loss": 0.2281, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 4.562753036437247, | |
| "grad_norm": 0.17300338233862067, | |
| "learning_rate": 9.254248023435864e-07, | |
| "loss": 0.2372, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 4.566801619433198, | |
| "grad_norm": 0.1771435744528256, | |
| "learning_rate": 9.084969902482288e-07, | |
| "loss": 0.2321, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 4.57085020242915, | |
| "grad_norm": 0.17914927800813327, | |
| "learning_rate": 8.917218333021993e-07, | |
| "loss": 0.2383, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 4.574898785425101, | |
| "grad_norm": 0.17386056872768427, | |
| "learning_rate": 8.750994656392109e-07, | |
| "loss": 0.2296, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "grad_norm": 0.1703276864212741, | |
| "learning_rate": 8.58630020171265e-07, | |
| "loss": 0.239, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 4.582995951417004, | |
| "grad_norm": 0.17729196810763223, | |
| "learning_rate": 8.423136285876099e-07, | |
| "loss": 0.2416, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 4.587044534412955, | |
| "grad_norm": 0.1709830344422447, | |
| "learning_rate": 8.26150421353682e-07, | |
| "loss": 0.2368, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 4.5910931174089065, | |
| "grad_norm": 0.17636238784997302, | |
| "learning_rate": 8.101405277100549e-07, | |
| "loss": 0.2423, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 4.5951417004048585, | |
| "grad_norm": 0.1703771134162661, | |
| "learning_rate": 7.942840756714077e-07, | |
| "loss": 0.2521, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.59919028340081, | |
| "grad_norm": 0.17828973489236247, | |
| "learning_rate": 7.785811920255093e-07, | |
| "loss": 0.2315, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 4.603238866396762, | |
| "grad_norm": 0.17160909304927813, | |
| "learning_rate": 7.630320023322024e-07, | |
| "loss": 0.2248, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 4.607287449392713, | |
| "grad_norm": 0.1701188322939676, | |
| "learning_rate": 7.476366309223903e-07, | |
| "loss": 0.2244, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 4.611336032388664, | |
| "grad_norm": 0.1695070676556434, | |
| "learning_rate": 7.323952008970537e-07, | |
| "loss": 0.2314, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.16998370758077708, | |
| "learning_rate": 7.173078341262618e-07, | |
| "loss": 0.2352, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.619433198380567, | |
| "grad_norm": 0.17082208554168157, | |
| "learning_rate": 7.023746512481989e-07, | |
| "loss": 0.2225, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 4.623481781376518, | |
| "grad_norm": 0.18354930900085697, | |
| "learning_rate": 6.875957716681902e-07, | |
| "loss": 0.2373, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 4.62753036437247, | |
| "grad_norm": 0.17296118209462852, | |
| "learning_rate": 6.729713135577687e-07, | |
| "loss": 0.2358, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 4.631578947368421, | |
| "grad_norm": 0.18056376812491076, | |
| "learning_rate": 6.585013938537143e-07, | |
| "loss": 0.2372, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 4.635627530364372, | |
| "grad_norm": 0.17369800901570573, | |
| "learning_rate": 6.441861282571115e-07, | |
| "loss": 0.2433, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.6396761133603235, | |
| "grad_norm": 0.17003001472787777, | |
| "learning_rate": 6.300256312324471e-07, | |
| "loss": 0.2381, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 4.6437246963562755, | |
| "grad_norm": 0.16924419271030886, | |
| "learning_rate": 6.160200160066776e-07, | |
| "loss": 0.2357, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 4.647773279352227, | |
| "grad_norm": 0.17449275073757212, | |
| "learning_rate": 6.021693945683327e-07, | |
| "loss": 0.2316, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 4.651821862348179, | |
| "grad_norm": 0.18046239845599288, | |
| "learning_rate": 5.884738776666044e-07, | |
| "loss": 0.2385, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 4.65587044534413, | |
| "grad_norm": 0.18015096278751758, | |
| "learning_rate": 5.749335748104878e-07, | |
| "loss": 0.2334, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.659919028340081, | |
| "grad_norm": 0.17556752991640817, | |
| "learning_rate": 5.615485942678778e-07, | |
| "loss": 0.2529, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 4.663967611336032, | |
| "grad_norm": 0.17569119663538257, | |
| "learning_rate": 5.483190430647289e-07, | |
| "loss": 0.2292, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 4.668016194331984, | |
| "grad_norm": 0.17568198229876364, | |
| "learning_rate": 5.352450269841747e-07, | |
| "loss": 0.2299, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 4.672064777327935, | |
| "grad_norm": 0.182003405495038, | |
| "learning_rate": 5.22326650565701e-07, | |
| "loss": 0.2364, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 4.676113360323887, | |
| "grad_norm": 0.17479256688637493, | |
| "learning_rate": 5.095640171043048e-07, | |
| "loss": 0.2312, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.680161943319838, | |
| "grad_norm": 0.17833271328514191, | |
| "learning_rate": 4.969572286496594e-07, | |
| "loss": 0.2286, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.684210526315789, | |
| "grad_norm": 0.17351623459162255, | |
| "learning_rate": 4.845063860053079e-07, | |
| "loss": 0.2472, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 4.6882591093117405, | |
| "grad_norm": 0.1809762252337556, | |
| "learning_rate": 4.7221158872785997e-07, | |
| "loss": 0.2328, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 0.18058908161078924, | |
| "learning_rate": 4.600729351261857e-07, | |
| "loss": 0.2376, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 4.696356275303644, | |
| "grad_norm": 0.16550641078021414, | |
| "learning_rate": 4.4809052226063355e-07, | |
| "loss": 0.2398, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.700404858299595, | |
| "grad_norm": 0.17778038385934725, | |
| "learning_rate": 4.3626444594226046e-07, | |
| "loss": 0.2241, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 4.704453441295547, | |
| "grad_norm": 0.1729159012155342, | |
| "learning_rate": 4.245948007320588e-07, | |
| "loss": 0.2379, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 4.708502024291498, | |
| "grad_norm": 0.1813133114343326, | |
| "learning_rate": 4.130816799401971e-07, | |
| "loss": 0.2438, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 4.712550607287449, | |
| "grad_norm": 0.16612997181133593, | |
| "learning_rate": 4.017251756252827e-07, | |
| "loss": 0.2361, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 4.716599190283401, | |
| "grad_norm": 0.17415164419468962, | |
| "learning_rate": 3.9052537859362473e-07, | |
| "loss": 0.2326, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.720647773279352, | |
| "grad_norm": 0.17203932028692365, | |
| "learning_rate": 3.7948237839849687e-07, | |
| "loss": 0.252, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 4.724696356275303, | |
| "grad_norm": 0.17637262052308045, | |
| "learning_rate": 3.6859626333943797e-07, | |
| "loss": 0.2418, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 4.728744939271255, | |
| "grad_norm": 0.18113641671612418, | |
| "learning_rate": 3.578671204615347e-07, | |
| "loss": 0.2292, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 4.732793522267206, | |
| "grad_norm": 0.1713122660003078, | |
| "learning_rate": 3.472950355547267e-07, | |
| "loss": 0.242, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.17561999279835686, | |
| "learning_rate": 3.3688009315312244e-07, | |
| "loss": 0.2369, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.7408906882591095, | |
| "grad_norm": 0.17344529001872222, | |
| "learning_rate": 3.266223765343224e-07, | |
| "loss": 0.2295, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.744939271255061, | |
| "grad_norm": 0.1742807192637025, | |
| "learning_rate": 3.16521967718757e-07, | |
| "loss": 0.2409, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 4.748987854251012, | |
| "grad_norm": 0.1668646149845011, | |
| "learning_rate": 3.0657894746902726e-07, | |
| "loss": 0.2313, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 4.753036437246964, | |
| "grad_norm": 0.18186369095104868, | |
| "learning_rate": 2.9679339528925654e-07, | |
| "loss": 0.2444, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 4.757085020242915, | |
| "grad_norm": 0.17399068835716935, | |
| "learning_rate": 2.871653894244597e-07, | |
| "loss": 0.2342, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.761133603238866, | |
| "grad_norm": 0.17886774319171816, | |
| "learning_rate": 2.7769500685991266e-07, | |
| "loss": 0.2296, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 4.765182186234818, | |
| "grad_norm": 0.1719541067419508, | |
| "learning_rate": 2.683823233205485e-07, | |
| "loss": 0.2537, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.18304695736829807, | |
| "learning_rate": 2.592274132703287e-07, | |
| "loss": 0.226, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 4.77327935222672, | |
| "grad_norm": 0.18594538149272874, | |
| "learning_rate": 2.502303499116754e-07, | |
| "loss": 0.2203, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 4.777327935222672, | |
| "grad_norm": 0.1707654265297572, | |
| "learning_rate": 2.413912051848666e-07, | |
| "loss": 0.2369, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.781376518218623, | |
| "grad_norm": 0.1762106596342796, | |
| "learning_rate": 2.3271004976746613e-07, | |
| "loss": 0.2269, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 4.7854251012145745, | |
| "grad_norm": 0.18646862992407273, | |
| "learning_rate": 2.2418695307376393e-07, | |
| "loss": 0.2132, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 4.7894736842105265, | |
| "grad_norm": 0.17289663383070386, | |
| "learning_rate": 2.1582198325420968e-07, | |
| "loss": 0.2304, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 4.793522267206478, | |
| "grad_norm": 0.17403795320980903, | |
| "learning_rate": 2.0761520719488447e-07, | |
| "loss": 0.2189, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 4.797570850202429, | |
| "grad_norm": 0.1671515386911494, | |
| "learning_rate": 1.9956669051694798e-07, | |
| "loss": 0.2473, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.801619433198381, | |
| "grad_norm": 0.1724034479062511, | |
| "learning_rate": 1.916764975761254e-07, | |
| "loss": 0.2254, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 4.805668016194332, | |
| "grad_norm": 0.17702289126408005, | |
| "learning_rate": 1.839446914621923e-07, | |
| "loss": 0.2436, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 4.809716599190283, | |
| "grad_norm": 0.18093848207404278, | |
| "learning_rate": 1.7637133399845962e-07, | |
| "loss": 0.2454, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.813765182186235, | |
| "grad_norm": 0.17137498851495445, | |
| "learning_rate": 1.6895648574129397e-07, | |
| "loss": 0.2347, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 4.817813765182186, | |
| "grad_norm": 0.16767834534738935, | |
| "learning_rate": 1.6170020597962465e-07, | |
| "loss": 0.2382, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.821862348178137, | |
| "grad_norm": 0.18761021744955653, | |
| "learning_rate": 1.5460255273447078e-07, | |
| "loss": 0.2352, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 4.825910931174089, | |
| "grad_norm": 0.17374524945188438, | |
| "learning_rate": 1.4766358275847715e-07, | |
| "loss": 0.2254, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 4.82995951417004, | |
| "grad_norm": 0.17703403132524048, | |
| "learning_rate": 1.4088335153546573e-07, | |
| "loss": 0.2307, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 4.834008097165992, | |
| "grad_norm": 0.18179355460883742, | |
| "learning_rate": 1.3426191327998496e-07, | |
| "loss": 0.2445, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 4.838056680161944, | |
| "grad_norm": 0.167996797129519, | |
| "learning_rate": 1.277993209368744e-07, | |
| "loss": 0.2245, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.842105263157895, | |
| "grad_norm": 0.1748779867432469, | |
| "learning_rate": 1.214956261808542e-07, | |
| "loss": 0.2428, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.17358979202821462, | |
| "learning_rate": 1.1535087941609846e-07, | |
| "loss": 0.2323, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 4.850202429149798, | |
| "grad_norm": 0.17811854691996942, | |
| "learning_rate": 1.0936512977583802e-07, | |
| "loss": 0.2334, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 4.854251012145749, | |
| "grad_norm": 0.17454277778994018, | |
| "learning_rate": 1.0353842512196732e-07, | |
| "loss": 0.2341, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 4.8582995951417, | |
| "grad_norm": 0.17038222697897304, | |
| "learning_rate": 9.787081204466253e-08, | |
| "loss": 0.2354, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.862348178137652, | |
| "grad_norm": 0.17180473577432628, | |
| "learning_rate": 9.236233586200627e-08, | |
| "loss": 0.2335, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 4.866396761133603, | |
| "grad_norm": 0.16974288263430048, | |
| "learning_rate": 8.70130406196279e-08, | |
| "loss": 0.2291, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 4.870445344129554, | |
| "grad_norm": 0.16717454589957534, | |
| "learning_rate": 8.182296909035047e-08, | |
| "loss": 0.2477, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 4.874493927125506, | |
| "grad_norm": 0.16803846441727288, | |
| "learning_rate": 7.679216277384439e-08, | |
| "loss": 0.2446, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 4.8785425101214575, | |
| "grad_norm": 0.17267285804447116, | |
| "learning_rate": 7.19206618963053e-08, | |
| "loss": 0.2304, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.882591093117409, | |
| "grad_norm": 0.1698222234357237, | |
| "learning_rate": 6.720850541012347e-08, | |
| "loss": 0.2359, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 4.886639676113361, | |
| "grad_norm": 0.17075152975807809, | |
| "learning_rate": 6.265573099357714e-08, | |
| "loss": 0.2366, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 4.890688259109312, | |
| "grad_norm": 0.1741988305299127, | |
| "learning_rate": 5.826237505053067e-08, | |
| "loss": 0.2348, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 4.894736842105263, | |
| "grad_norm": 0.2713797825394198, | |
| "learning_rate": 5.4028472710137e-08, | |
| "loss": 0.245, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 4.898785425101215, | |
| "grad_norm": 0.1743184330600498, | |
| "learning_rate": 4.9954057826571146e-08, | |
| "loss": 0.2314, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.902834008097166, | |
| "grad_norm": 0.16673112476055824, | |
| "learning_rate": 4.6039162978743776e-08, | |
| "loss": 0.2375, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 4.906882591093117, | |
| "grad_norm": 0.17132973284020717, | |
| "learning_rate": 4.228381947004812e-08, | |
| "loss": 0.2534, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 4.910931174089069, | |
| "grad_norm": 0.17002882947009398, | |
| "learning_rate": 3.868805732811343e-08, | |
| "loss": 0.2256, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 4.91497975708502, | |
| "grad_norm": 0.1728581005003658, | |
| "learning_rate": 3.525190530455636e-08, | |
| "loss": 0.2241, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 4.919028340080971, | |
| "grad_norm": 0.1722444609426268, | |
| "learning_rate": 3.197539087475887e-08, | |
| "loss": 0.2305, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.17201892433455365, | |
| "learning_rate": 2.8858540237641786e-08, | |
| "loss": 0.2411, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 4.9271255060728745, | |
| "grad_norm": 0.17399059383464058, | |
| "learning_rate": 2.5901378315458247e-08, | |
| "loss": 0.2317, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 4.931174089068826, | |
| "grad_norm": 0.16639310988917208, | |
| "learning_rate": 2.3103928753598347e-08, | |
| "loss": 0.2433, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 4.935222672064778, | |
| "grad_norm": 0.17567046937674669, | |
| "learning_rate": 2.0466213920393718e-08, | |
| "loss": 0.2263, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 4.939271255060729, | |
| "grad_norm": 0.17781435956700592, | |
| "learning_rate": 1.7988254906942115e-08, | |
| "loss": 0.228, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.94331983805668, | |
| "grad_norm": 0.17450734596540923, | |
| "learning_rate": 1.5670071526936447e-08, | |
| "loss": 0.235, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "grad_norm": 0.1717517713574551, | |
| "learning_rate": 1.3511682316509344e-08, | |
| "loss": 0.2376, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 4.951417004048583, | |
| "grad_norm": 0.17450416942706073, | |
| "learning_rate": 1.1513104534086605e-08, | |
| "loss": 0.2362, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 4.955465587044534, | |
| "grad_norm": 0.18125874658141625, | |
| "learning_rate": 9.67435416023843e-09, | |
| "loss": 0.2358, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 4.959514170040486, | |
| "grad_norm": 0.17081481611090057, | |
| "learning_rate": 7.995445897566178e-09, | |
| "loss": 0.2266, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.963562753036437, | |
| "grad_norm": 0.17431918127673854, | |
| "learning_rate": 6.476393170578021e-09, | |
| "loss": 0.2302, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 4.967611336032388, | |
| "grad_norm": 0.17227793773209477, | |
| "learning_rate": 5.11720812557348e-09, | |
| "loss": 0.2358, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 4.97165991902834, | |
| "grad_norm": 0.16855752718610692, | |
| "learning_rate": 3.917901630563492e-09, | |
| "loss": 0.2398, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 4.9757085020242915, | |
| "grad_norm": 0.17803642135180503, | |
| "learning_rate": 2.878483275170485e-09, | |
| "loss": 0.2452, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 4.979757085020243, | |
| "grad_norm": 0.17724843919219005, | |
| "learning_rate": 1.998961370555108e-09, | |
| "loss": 0.2306, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.983805668016195, | |
| "grad_norm": 0.16463323120865625, | |
| "learning_rate": 1.2793429493518361e-09, | |
| "loss": 0.2484, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 4.987854251012146, | |
| "grad_norm": 0.17011563272459723, | |
| "learning_rate": 7.196337656112384e-10, | |
| "loss": 0.2362, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 4.991902834008097, | |
| "grad_norm": 0.16960012058485313, | |
| "learning_rate": 3.198382947511292e-10, | |
| "loss": 0.231, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 4.995951417004049, | |
| "grad_norm": 0.17132324228005585, | |
| "learning_rate": 7.995973352770137e-11, | |
| "loss": 0.2446, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.16812119084894125, | |
| "learning_rate": 0.0, | |
| "loss": 0.2394, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1235, | |
| "total_flos": 4.764433372703359e+18, | |
| "train_loss": 0.3434084688603637, | |
| "train_runtime": 29328.3301, | |
| "train_samples_per_second": 5.386, | |
| "train_steps_per_second": 0.042 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1235, | |
| "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": 4.764433372703359e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |