| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 504, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003968253968253968, | |
| "grad_norm": 305.6814880371094, | |
| "learning_rate": 3.846153846153847e-07, | |
| "loss": 3.0274, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007936507936507936, | |
| "grad_norm": 258.49066162109375, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 3.0629, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.011904761904761904, | |
| "grad_norm": 213.76901245117188, | |
| "learning_rate": 1.153846153846154e-06, | |
| "loss": 3.1315, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.015873015873015872, | |
| "grad_norm": 267.2046813964844, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 3.1417, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01984126984126984, | |
| "grad_norm": 265.2721862792969, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 3.2616, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.023809523809523808, | |
| "grad_norm": 289.385986328125, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 2.9708, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.027777777777777776, | |
| "grad_norm": 279.978515625, | |
| "learning_rate": 2.6923076923076923e-06, | |
| "loss": 2.7235, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.031746031746031744, | |
| "grad_norm": 391.742431640625, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 2.2668, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 305.847412109375, | |
| "learning_rate": 3.4615384615384617e-06, | |
| "loss": 2.1289, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03968253968253968, | |
| "grad_norm": 279.4045104980469, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 2.0271, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04365079365079365, | |
| "grad_norm": 59.683937072753906, | |
| "learning_rate": 4.230769230769231e-06, | |
| "loss": 1.5848, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 61.3370361328125, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.5105, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.051587301587301584, | |
| "grad_norm": 56.375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.4317, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 107.44894409179688, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 1.4334, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05952380952380952, | |
| "grad_norm": 120.30421447753906, | |
| "learning_rate": 5.769230769230769e-06, | |
| "loss": 1.3853, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 46.73382568359375, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.3136, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06746031746031746, | |
| "grad_norm": 20.867332458496094, | |
| "learning_rate": 6.538461538461539e-06, | |
| "loss": 1.2486, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 15.966314315795898, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 1.2147, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07539682539682539, | |
| "grad_norm": 38.716041564941406, | |
| "learning_rate": 7.307692307692308e-06, | |
| "loss": 1.1702, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07936507936507936, | |
| "grad_norm": 10.495340347290039, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.1272, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 1649.7857666015625, | |
| "learning_rate": 8.076923076923077e-06, | |
| "loss": 1.0522, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0873015873015873, | |
| "grad_norm": 11.990230560302734, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 1.2053, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.09126984126984126, | |
| "grad_norm": 9.10935115814209, | |
| "learning_rate": 8.846153846153847e-06, | |
| "loss": 1.2156, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 7.118689060211182, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.1402, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0992063492063492, | |
| "grad_norm": 10.714500427246094, | |
| "learning_rate": 9.615384615384616e-06, | |
| "loss": 1.0423, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10317460317460317, | |
| "grad_norm": 13.593528747558594, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6624, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 16.534046173095703, | |
| "learning_rate": 9.999892010284378e-06, | |
| "loss": 0.6708, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 6.320470333099365, | |
| "learning_rate": 9.999568045802216e-06, | |
| "loss": 0.7056, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11507936507936507, | |
| "grad_norm": 6.473380088806152, | |
| "learning_rate": 9.999028120547456e-06, | |
| "loss": 0.5976, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11904761904761904, | |
| "grad_norm": 10.304583549499512, | |
| "learning_rate": 9.99827225784264e-06, | |
| "loss": 0.6378, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12301587301587301, | |
| "grad_norm": 4.470282077789307, | |
| "learning_rate": 9.99730049033793e-06, | |
| "loss": 0.5887, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 5.942732334136963, | |
| "learning_rate": 9.996112860009689e-06, | |
| "loss": 0.6002, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13095238095238096, | |
| "grad_norm": 22.721927642822266, | |
| "learning_rate": 9.994709418158652e-06, | |
| "loss": 0.5524, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1349206349206349, | |
| "grad_norm": 9.322417259216309, | |
| "learning_rate": 9.993090225407743e-06, | |
| "loss": 0.654, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 2.5783913135528564, | |
| "learning_rate": 9.991255351699422e-06, | |
| "loss": 0.5709, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 25.98549461364746, | |
| "learning_rate": 9.98920487629269e-06, | |
| "loss": 0.6553, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14682539682539683, | |
| "grad_norm": 36.62657165527344, | |
| "learning_rate": 9.986938887759643e-06, | |
| "loss": 0.6571, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15079365079365079, | |
| "grad_norm": 16.053661346435547, | |
| "learning_rate": 9.98445748398167e-06, | |
| "loss": 0.6355, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15476190476190477, | |
| "grad_norm": 2.3475632667541504, | |
| "learning_rate": 9.981760772145201e-06, | |
| "loss": 0.683, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 11.760564804077148, | |
| "learning_rate": 9.978848868737099e-06, | |
| "loss": 0.6856, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1626984126984127, | |
| "grad_norm": 15.502592086791992, | |
| "learning_rate": 9.975721899539607e-06, | |
| "loss": 0.6661, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 4.046666622161865, | |
| "learning_rate": 9.972379999624935e-06, | |
| "loss": 0.7245, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17063492063492064, | |
| "grad_norm": 2.2085020542144775, | |
| "learning_rate": 9.968823313349412e-06, | |
| "loss": 0.6414, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1746031746031746, | |
| "grad_norm": 35.867584228515625, | |
| "learning_rate": 9.96505199434725e-06, | |
| "loss": 0.651, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 6.324635028839111, | |
| "learning_rate": 9.961066205523917e-06, | |
| "loss": 0.6871, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.18253968253968253, | |
| "grad_norm": 2.5837783813476562, | |
| "learning_rate": 9.956866119049095e-06, | |
| "loss": 0.7055, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1865079365079365, | |
| "grad_norm": 2.535799503326416, | |
| "learning_rate": 9.952451916349242e-06, | |
| "loss": 0.6302, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 3.493901491165161, | |
| "learning_rate": 9.947823788099754e-06, | |
| "loss": 0.5705, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19444444444444445, | |
| "grad_norm": 2.7961313724517822, | |
| "learning_rate": 9.942981934216731e-06, | |
| "loss": 0.5889, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1984126984126984, | |
| "grad_norm": 6.1772661209106445, | |
| "learning_rate": 9.937926563848345e-06, | |
| "loss": 0.6446, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20238095238095238, | |
| "grad_norm": 2.4115350246429443, | |
| "learning_rate": 9.9326578953658e-06, | |
| "loss": 0.5302, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.20634920634920634, | |
| "grad_norm": 2.7692086696624756, | |
| "learning_rate": 9.9271761563539e-06, | |
| "loss": 0.5113, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.21031746031746032, | |
| "grad_norm": 2.8092377185821533, | |
| "learning_rate": 9.921481583601218e-06, | |
| "loss": 0.5752, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 2.240276575088501, | |
| "learning_rate": 9.915574423089872e-06, | |
| "loss": 0.5055, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.21825396825396826, | |
| "grad_norm": 2.1406800746917725, | |
| "learning_rate": 9.909454929984894e-06, | |
| "loss": 0.5151, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 2.25881028175354, | |
| "learning_rate": 9.903123368623216e-06, | |
| "loss": 0.4876, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2261904761904762, | |
| "grad_norm": 2.4127275943756104, | |
| "learning_rate": 9.896580012502238e-06, | |
| "loss": 0.5779, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23015873015873015, | |
| "grad_norm": 2.5743143558502197, | |
| "learning_rate": 9.889825144268029e-06, | |
| "loss": 0.5217, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.23412698412698413, | |
| "grad_norm": 2.4898812770843506, | |
| "learning_rate": 9.882859055703109e-06, | |
| "loss": 0.4978, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 2.4116387367248535, | |
| "learning_rate": 9.875682047713847e-06, | |
| "loss": 0.5185, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24206349206349206, | |
| "grad_norm": 2.632223129272461, | |
| "learning_rate": 9.868294430317464e-06, | |
| "loss": 0.5603, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.24603174603174602, | |
| "grad_norm": 2.6133947372436523, | |
| "learning_rate": 9.860696522628638e-06, | |
| "loss": 0.5581, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.720884084701538, | |
| "learning_rate": 9.852888652845729e-06, | |
| "loss": 0.5631, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 2.845902442932129, | |
| "learning_rate": 9.84487115823659e-06, | |
| "loss": 0.5493, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.25793650793650796, | |
| "grad_norm": 2.9875378608703613, | |
| "learning_rate": 9.836644385124006e-06, | |
| "loss": 0.5451, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2619047619047619, | |
| "grad_norm": 2.9919440746307373, | |
| "learning_rate": 9.828208688870736e-06, | |
| "loss": 0.566, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.26587301587301587, | |
| "grad_norm": 3.1801774501800537, | |
| "learning_rate": 9.81956443386415e-06, | |
| "loss": 0.6042, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2698412698412698, | |
| "grad_norm": 3.2077016830444336, | |
| "learning_rate": 9.810711993500506e-06, | |
| "loss": 0.5147, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.27380952380952384, | |
| "grad_norm": 3.3942503929138184, | |
| "learning_rate": 9.801651750168815e-06, | |
| "loss": 0.5269, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 3.4853150844573975, | |
| "learning_rate": 9.792384095234312e-06, | |
| "loss": 0.5143, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.28174603174603174, | |
| "grad_norm": 3.6130869388580322, | |
| "learning_rate": 9.782909429021568e-06, | |
| "loss": 0.5095, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 3.2448408603668213, | |
| "learning_rate": 9.773228160797187e-06, | |
| "loss": 0.5097, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2896825396825397, | |
| "grad_norm": 3.151676654815674, | |
| "learning_rate": 9.76334070875213e-06, | |
| "loss": 0.5123, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29365079365079366, | |
| "grad_norm": 3.153676986694336, | |
| "learning_rate": 9.753247499983649e-06, | |
| "loss": 0.4614, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2976190476190476, | |
| "grad_norm": 3.421731472015381, | |
| "learning_rate": 9.742948970476845e-06, | |
| "loss": 0.4604, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.30158730158730157, | |
| "grad_norm": 2.7017500400543213, | |
| "learning_rate": 9.732445565085823e-06, | |
| "loss": 0.3889, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3055555555555556, | |
| "grad_norm": 3.0560805797576904, | |
| "learning_rate": 9.721737737514492e-06, | |
| "loss": 0.38, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.30952380952380953, | |
| "grad_norm": 3.138594388961792, | |
| "learning_rate": 9.71082595029695e-06, | |
| "loss": 0.4207, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.3134920634920635, | |
| "grad_norm": 3.3207204341888428, | |
| "learning_rate": 9.699710674777519e-06, | |
| "loss": 0.3807, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 3.461073398590088, | |
| "learning_rate": 9.688392391090374e-06, | |
| "loss": 0.4303, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 3.502265691757202, | |
| "learning_rate": 9.676871588138812e-06, | |
| "loss": 0.3855, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3253968253968254, | |
| "grad_norm": 3.644130229949951, | |
| "learning_rate": 9.665148763574123e-06, | |
| "loss": 0.3942, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.32936507936507936, | |
| "grad_norm": 3.7463479042053223, | |
| "learning_rate": 9.653224423774107e-06, | |
| "loss": 0.3772, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 3.7819101810455322, | |
| "learning_rate": 9.64109908382119e-06, | |
| "loss": 0.3621, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3373015873015873, | |
| "grad_norm": 3.8595614433288574, | |
| "learning_rate": 9.628773267480177e-06, | |
| "loss": 0.4279, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3412698412698413, | |
| "grad_norm": 3.950291156768799, | |
| "learning_rate": 9.616247507175624e-06, | |
| "loss": 0.4054, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.34523809523809523, | |
| "grad_norm": 4.081280708312988, | |
| "learning_rate": 9.603522343968852e-06, | |
| "loss": 0.4287, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3492063492063492, | |
| "grad_norm": 4.042370319366455, | |
| "learning_rate": 9.590598327534563e-06, | |
| "loss": 0.3955, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3531746031746032, | |
| "grad_norm": 4.129544734954834, | |
| "learning_rate": 9.577476016137105e-06, | |
| "loss": 0.3687, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 4.139583110809326, | |
| "learning_rate": 9.56415597660634e-06, | |
| "loss": 0.4016, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3611111111111111, | |
| "grad_norm": 4.293641090393066, | |
| "learning_rate": 9.550638784313187e-06, | |
| "loss": 0.3928, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.36507936507936506, | |
| "grad_norm": 4.329298496246338, | |
| "learning_rate": 9.536925023144742e-06, | |
| "loss": 0.361, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.36904761904761907, | |
| "grad_norm": 4.435636043548584, | |
| "learning_rate": 9.523015285479076e-06, | |
| "loss": 0.4067, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.373015873015873, | |
| "grad_norm": 4.418957710266113, | |
| "learning_rate": 9.508910172159635e-06, | |
| "loss": 0.3716, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.376984126984127, | |
| "grad_norm": 4.311522483825684, | |
| "learning_rate": 9.494610292469287e-06, | |
| "loss": 0.352, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 4.45173454284668, | |
| "learning_rate": 9.48011626410401e-06, | |
| "loss": 0.3867, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.38492063492063494, | |
| "grad_norm": 3.7801730632781982, | |
| "learning_rate": 9.465428713146206e-06, | |
| "loss": 0.3834, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 3.4346587657928467, | |
| "learning_rate": 9.450548274037652e-06, | |
| "loss": 0.3654, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 3.401343584060669, | |
| "learning_rate": 9.435475589552107e-06, | |
| "loss": 0.3409, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3968253968253968, | |
| "grad_norm": 3.916692018508911, | |
| "learning_rate": 9.420211310767534e-06, | |
| "loss": 0.3656, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4007936507936508, | |
| "grad_norm": 2.716371536254883, | |
| "learning_rate": 9.40475609703798e-06, | |
| "loss": 0.2815, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.40476190476190477, | |
| "grad_norm": 3.4734158515930176, | |
| "learning_rate": 9.389110615965102e-06, | |
| "loss": 0.2986, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.4087301587301587, | |
| "grad_norm": 3.0159623622894287, | |
| "learning_rate": 9.37327554336932e-06, | |
| "loss": 0.2361, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.4126984126984127, | |
| "grad_norm": 3.1320643424987793, | |
| "learning_rate": 9.35725156326063e-06, | |
| "loss": 0.2744, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 3.046394109725952, | |
| "learning_rate": 9.341039367809056e-06, | |
| "loss": 0.2967, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.42063492063492064, | |
| "grad_norm": 3.066723108291626, | |
| "learning_rate": 9.324639657314742e-06, | |
| "loss": 0.3036, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4246031746031746, | |
| "grad_norm": 3.0018229484558105, | |
| "learning_rate": 9.308053140177722e-06, | |
| "loss": 0.2581, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 3.130518913269043, | |
| "learning_rate": 9.291280532867301e-06, | |
| "loss": 0.2584, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.43253968253968256, | |
| "grad_norm": 3.050171136856079, | |
| "learning_rate": 9.27432255989112e-06, | |
| "loss": 0.2878, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4365079365079365, | |
| "grad_norm": 3.036541700363159, | |
| "learning_rate": 9.257179953763846e-06, | |
| "loss": 0.2726, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.44047619047619047, | |
| "grad_norm": 2.9555277824401855, | |
| "learning_rate": 9.239853454975548e-06, | |
| "loss": 0.2653, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 2.7957818508148193, | |
| "learning_rate": 9.222343811959694e-06, | |
| "loss": 0.2756, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.44841269841269843, | |
| "grad_norm": 2.9211173057556152, | |
| "learning_rate": 9.204651781060832e-06, | |
| "loss": 0.2328, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4523809523809524, | |
| "grad_norm": 2.892176628112793, | |
| "learning_rate": 9.186778126501916e-06, | |
| "loss": 0.2504, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.45634920634920634, | |
| "grad_norm": 2.7730042934417725, | |
| "learning_rate": 9.168723620351298e-06, | |
| "loss": 0.2552, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4603174603174603, | |
| "grad_norm": 2.6634700298309326, | |
| "learning_rate": 9.150489042489368e-06, | |
| "loss": 0.2219, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 2.5813324451446533, | |
| "learning_rate": 9.13207518057488e-06, | |
| "loss": 0.2533, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.46825396825396826, | |
| "grad_norm": 2.44559383392334, | |
| "learning_rate": 9.113482830010918e-06, | |
| "loss": 0.2116, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4722222222222222, | |
| "grad_norm": 2.425828456878662, | |
| "learning_rate": 9.094712793910541e-06, | |
| "loss": 0.2384, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 2.380803108215332, | |
| "learning_rate": 9.075765883062093e-06, | |
| "loss": 0.2306, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4801587301587302, | |
| "grad_norm": 2.1514198780059814, | |
| "learning_rate": 9.056642915894182e-06, | |
| "loss": 0.2213, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.48412698412698413, | |
| "grad_norm": 1.8206262588500977, | |
| "learning_rate": 9.037344718440321e-06, | |
| "loss": 0.2463, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4880952380952381, | |
| "grad_norm": 1.6986949443817139, | |
| "learning_rate": 9.017872124303255e-06, | |
| "loss": 0.2395, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.49206349206349204, | |
| "grad_norm": 1.7301899194717407, | |
| "learning_rate": 8.99822597461894e-06, | |
| "loss": 0.1952, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.49603174603174605, | |
| "grad_norm": 1.7308430671691895, | |
| "learning_rate": 8.978407118020226e-06, | |
| "loss": 0.204, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.4037597179412842, | |
| "learning_rate": 8.958416410600188e-06, | |
| "loss": 0.2135, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.503968253968254, | |
| "grad_norm": 1.2156530618667603, | |
| "learning_rate": 8.938254715875152e-06, | |
| "loss": 0.1781, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 1.2453689575195312, | |
| "learning_rate": 8.917922904747385e-06, | |
| "loss": 0.1722, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5119047619047619, | |
| "grad_norm": 1.406607985496521, | |
| "learning_rate": 8.897421855467491e-06, | |
| "loss": 0.2184, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5158730158730159, | |
| "grad_norm": 1.4896661043167114, | |
| "learning_rate": 8.876752453596462e-06, | |
| "loss": 0.2587, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5198412698412699, | |
| "grad_norm": 1.3468254804611206, | |
| "learning_rate": 8.85591559196743e-06, | |
| "loss": 0.237, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 1.1635011434555054, | |
| "learning_rate": 8.834912170647102e-06, | |
| "loss": 0.2071, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5277777777777778, | |
| "grad_norm": 1.126786470413208, | |
| "learning_rate": 8.813743096896872e-06, | |
| "loss": 0.1919, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5317460317460317, | |
| "grad_norm": 1.1282813549041748, | |
| "learning_rate": 8.792409285133644e-06, | |
| "loss": 0.1919, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 1.122883915901184, | |
| "learning_rate": 8.770911656890325e-06, | |
| "loss": 0.2193, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5396825396825397, | |
| "grad_norm": 1.201058030128479, | |
| "learning_rate": 8.749251140776016e-06, | |
| "loss": 0.2037, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5436507936507936, | |
| "grad_norm": 1.0873547792434692, | |
| "learning_rate": 8.727428672435911e-06, | |
| "loss": 0.2093, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5476190476190477, | |
| "grad_norm": 1.2122111320495605, | |
| "learning_rate": 8.705445194510868e-06, | |
| "loss": 0.1961, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5515873015873016, | |
| "grad_norm": 1.259795069694519, | |
| "learning_rate": 8.6833016565967e-06, | |
| "loss": 0.2266, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 1.1738680601119995, | |
| "learning_rate": 8.660999015203152e-06, | |
| "loss": 0.1874, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5595238095238095, | |
| "grad_norm": 1.2351460456848145, | |
| "learning_rate": 8.638538233712581e-06, | |
| "loss": 0.2007, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5634920634920635, | |
| "grad_norm": 1.1309882402420044, | |
| "learning_rate": 8.615920282338355e-06, | |
| "loss": 0.2051, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5674603174603174, | |
| "grad_norm": 1.409136414527893, | |
| "learning_rate": 8.593146138082925e-06, | |
| "loss": 0.2601, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.2396869659423828, | |
| "learning_rate": 8.570216784695637e-06, | |
| "loss": 0.1965, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5753968253968254, | |
| "grad_norm": 1.27550208568573, | |
| "learning_rate": 8.54713321263023e-06, | |
| "loss": 0.1877, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5793650793650794, | |
| "grad_norm": 1.3119581937789917, | |
| "learning_rate": 8.52389641900206e-06, | |
| "loss": 0.1929, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 1.3157020807266235, | |
| "learning_rate": 8.50050740754502e-06, | |
| "loss": 0.2354, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5873015873015873, | |
| "grad_norm": 1.3332029581069946, | |
| "learning_rate": 8.476967188568187e-06, | |
| "loss": 0.1879, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5912698412698413, | |
| "grad_norm": 1.2403169870376587, | |
| "learning_rate": 8.453276778912186e-06, | |
| "loss": 0.2004, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5952380952380952, | |
| "grad_norm": 1.1858034133911133, | |
| "learning_rate": 8.429437201905254e-06, | |
| "loss": 0.1704, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5992063492063492, | |
| "grad_norm": 1.1488471031188965, | |
| "learning_rate": 8.405449487319049e-06, | |
| "loss": 0.1762, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6031746031746031, | |
| "grad_norm": 1.2257866859436035, | |
| "learning_rate": 8.38131467132416e-06, | |
| "loss": 0.1596, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 1.250268816947937, | |
| "learning_rate": 8.357033796445356e-06, | |
| "loss": 0.182, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 1.146741509437561, | |
| "learning_rate": 8.332607911516545e-06, | |
| "loss": 0.1512, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6150793650793651, | |
| "grad_norm": 1.4444941282272339, | |
| "learning_rate": 8.308038071635475e-06, | |
| "loss": 0.2736, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 1.263741374015808, | |
| "learning_rate": 8.283325338118154e-06, | |
| "loss": 0.1721, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.623015873015873, | |
| "grad_norm": 1.392531394958496, | |
| "learning_rate": 8.258470778453005e-06, | |
| "loss": 0.211, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.626984126984127, | |
| "grad_norm": 1.3893020153045654, | |
| "learning_rate": 8.233475466254766e-06, | |
| "loss": 0.1752, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6309523809523809, | |
| "grad_norm": 1.3871208429336548, | |
| "learning_rate": 8.208340481218094e-06, | |
| "loss": 0.1839, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 1.8047232627868652, | |
| "learning_rate": 8.183066909070946e-06, | |
| "loss": 0.1654, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6388888888888888, | |
| "grad_norm": 1.3427791595458984, | |
| "learning_rate": 8.15765584152767e-06, | |
| "loss": 0.1762, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 1.3540451526641846, | |
| "learning_rate": 8.132108376241849e-06, | |
| "loss": 0.1527, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6468253968253969, | |
| "grad_norm": 1.4374990463256836, | |
| "learning_rate": 8.106425616758886e-06, | |
| "loss": 0.2038, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6507936507936508, | |
| "grad_norm": 1.5352795124053955, | |
| "learning_rate": 8.08060867246834e-06, | |
| "loss": 0.2226, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6547619047619048, | |
| "grad_norm": 1.600878357887268, | |
| "learning_rate": 8.054658658555998e-06, | |
| "loss": 0.2067, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6587301587301587, | |
| "grad_norm": 1.453999638557434, | |
| "learning_rate": 8.028576695955711e-06, | |
| "loss": 0.2039, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6626984126984127, | |
| "grad_norm": 1.4652482271194458, | |
| "learning_rate": 8.002363911300966e-06, | |
| "loss": 0.1811, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.4747813940048218, | |
| "learning_rate": 7.976021436876232e-06, | |
| "loss": 0.1635, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6706349206349206, | |
| "grad_norm": 1.5624871253967285, | |
| "learning_rate": 7.949550410568033e-06, | |
| "loss": 0.1756, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6746031746031746, | |
| "grad_norm": 1.5503687858581543, | |
| "learning_rate": 7.92295197581581e-06, | |
| "loss": 0.1552, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 1.6321192979812622, | |
| "learning_rate": 7.89622728156253e-06, | |
| "loss": 0.1696, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6825396825396826, | |
| "grad_norm": 1.6671695709228516, | |
| "learning_rate": 7.869377482205042e-06, | |
| "loss": 0.2371, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6865079365079365, | |
| "grad_norm": 1.3252042531967163, | |
| "learning_rate": 7.842403737544226e-06, | |
| "loss": 0.1724, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6904761904761905, | |
| "grad_norm": 1.343565821647644, | |
| "learning_rate": 7.815307212734888e-06, | |
| "loss": 0.1434, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 1.3573709726333618, | |
| "learning_rate": 7.788089078235432e-06, | |
| "loss": 0.177, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 1.2602770328521729, | |
| "learning_rate": 7.7607505097573e-06, | |
| "loss": 0.179, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7023809523809523, | |
| "grad_norm": 1.1597180366516113, | |
| "learning_rate": 7.733292688214182e-06, | |
| "loss": 0.1414, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7063492063492064, | |
| "grad_norm": 1.2210685014724731, | |
| "learning_rate": 7.705716799671019e-06, | |
| "loss": 0.1728, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7103174603174603, | |
| "grad_norm": 1.3051807880401611, | |
| "learning_rate": 7.678024035292757e-06, | |
| "loss": 0.1506, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.152103304862976, | |
| "learning_rate": 7.650215591292888e-06, | |
| "loss": 0.1362, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7182539682539683, | |
| "grad_norm": 1.2233182191848755, | |
| "learning_rate": 7.622292668881805e-06, | |
| "loss": 0.1705, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 1.1889208555221558, | |
| "learning_rate": 7.594256474214883e-06, | |
| "loss": 0.1446, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7261904761904762, | |
| "grad_norm": 1.2984225749969482, | |
| "learning_rate": 7.566108218340399e-06, | |
| "loss": 0.1267, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7301587301587301, | |
| "grad_norm": 1.2752000093460083, | |
| "learning_rate": 7.537849117147212e-06, | |
| "loss": 0.1707, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7341269841269841, | |
| "grad_norm": 1.2290748357772827, | |
| "learning_rate": 7.509480391312243e-06, | |
| "loss": 0.1177, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7380952380952381, | |
| "grad_norm": 1.3750877380371094, | |
| "learning_rate": 7.481003266247745e-06, | |
| "loss": 0.1851, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.7420634920634921, | |
| "grad_norm": 1.4225335121154785, | |
| "learning_rate": 7.452418972048372e-06, | |
| "loss": 0.1759, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.746031746031746, | |
| "grad_norm": 1.2389686107635498, | |
| "learning_rate": 7.4237287434380485e-06, | |
| "loss": 0.1747, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.340498685836792, | |
| "learning_rate": 7.394933819716625e-06, | |
| "loss": 0.1812, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.753968253968254, | |
| "grad_norm": 1.1293567419052124, | |
| "learning_rate": 7.366035444706346e-06, | |
| "loss": 0.134, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7579365079365079, | |
| "grad_norm": 1.3577476739883423, | |
| "learning_rate": 7.337034866698138e-06, | |
| "loss": 0.1548, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 1.2981313467025757, | |
| "learning_rate": 7.307933338397667e-06, | |
| "loss": 0.2098, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7658730158730159, | |
| "grad_norm": 1.856937289237976, | |
| "learning_rate": 7.278732116871239e-06, | |
| "loss": 0.1628, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7698412698412699, | |
| "grad_norm": 1.2699828147888184, | |
| "learning_rate": 7.249432463491498e-06, | |
| "loss": 0.1349, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7738095238095238, | |
| "grad_norm": 1.333633542060852, | |
| "learning_rate": 7.220035643882938e-06, | |
| "loss": 0.1463, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 1.2837098836898804, | |
| "learning_rate": 7.190542927867234e-06, | |
| "loss": 0.1325, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7817460317460317, | |
| "grad_norm": 1.3769762516021729, | |
| "learning_rate": 7.160955589408395e-06, | |
| "loss": 0.2226, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 1.2040393352508545, | |
| "learning_rate": 7.131274906557725e-06, | |
| "loss": 0.1816, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7896825396825397, | |
| "grad_norm": 1.1478785276412964, | |
| "learning_rate": 7.101502161398626e-06, | |
| "loss": 0.1429, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 1.1133449077606201, | |
| "learning_rate": 7.0716386399912075e-06, | |
| "loss": 0.1401, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7976190476190477, | |
| "grad_norm": 1.5019123554229736, | |
| "learning_rate": 7.041685632316748e-06, | |
| "loss": 0.1438, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8015873015873016, | |
| "grad_norm": 0.9408583641052246, | |
| "learning_rate": 7.0116444322219575e-06, | |
| "loss": 0.113, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8055555555555556, | |
| "grad_norm": 0.9584061503410339, | |
| "learning_rate": 6.981516337363099e-06, | |
| "loss": 0.107, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 0.997330904006958, | |
| "learning_rate": 6.95130264914993e-06, | |
| "loss": 0.1247, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8134920634920635, | |
| "grad_norm": 1.0142338275909424, | |
| "learning_rate": 6.9210046726894885e-06, | |
| "loss": 0.1126, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8174603174603174, | |
| "grad_norm": 1.03889799118042, | |
| "learning_rate": 6.890623716729724e-06, | |
| "loss": 0.1493, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 1.0274310111999512, | |
| "learning_rate": 6.860161093602949e-06, | |
| "loss": 0.1438, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 0.9676130414009094, | |
| "learning_rate": 6.829618119169169e-06, | |
| "loss": 0.1448, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8293650793650794, | |
| "grad_norm": 1.1232084035873413, | |
| "learning_rate": 6.798996112759233e-06, | |
| "loss": 0.151, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 1.0271536111831665, | |
| "learning_rate": 6.768296397117848e-06, | |
| "loss": 0.1719, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8373015873015873, | |
| "grad_norm": 1.134443998336792, | |
| "learning_rate": 6.737520298346438e-06, | |
| "loss": 0.1271, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8412698412698413, | |
| "grad_norm": 0.9489789009094238, | |
| "learning_rate": 6.706669145845863e-06, | |
| "loss": 0.1111, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8452380952380952, | |
| "grad_norm": 1.1553804874420166, | |
| "learning_rate": 6.6757442722590015e-06, | |
| "loss": 0.1782, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8492063492063492, | |
| "grad_norm": 1.0384173393249512, | |
| "learning_rate": 6.6447470134131685e-06, | |
| "loss": 0.1247, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8531746031746031, | |
| "grad_norm": 0.9292367696762085, | |
| "learning_rate": 6.613678708262439e-06, | |
| "loss": 0.1046, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 1.168887972831726, | |
| "learning_rate": 6.5825406988297815e-06, | |
| "loss": 0.147, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8611111111111112, | |
| "grad_norm": 1.090557336807251, | |
| "learning_rate": 6.551334330149114e-06, | |
| "loss": 0.149, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8650793650793651, | |
| "grad_norm": 0.9753419756889343, | |
| "learning_rate": 6.520060950207186e-06, | |
| "loss": 0.1442, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8690476190476191, | |
| "grad_norm": 1.0562981367111206, | |
| "learning_rate": 6.488721909885359e-06, | |
| "loss": 0.1778, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.873015873015873, | |
| "grad_norm": 0.8327993750572205, | |
| "learning_rate": 6.457318562901257e-06, | |
| "loss": 0.0964, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.876984126984127, | |
| "grad_norm": 1.0785108804702759, | |
| "learning_rate": 6.425852265750282e-06, | |
| "loss": 0.1567, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8809523809523809, | |
| "grad_norm": 1.1328438520431519, | |
| "learning_rate": 6.394324377647028e-06, | |
| "loss": 0.1609, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.8849206349206349, | |
| "grad_norm": 1.1382204294204712, | |
| "learning_rate": 6.362736260466561e-06, | |
| "loss": 0.1328, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.111432433128357, | |
| "learning_rate": 6.331089278685599e-06, | |
| "loss": 0.1856, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 1.0427393913269043, | |
| "learning_rate": 6.299384799323568e-06, | |
| "loss": 0.1165, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8968253968253969, | |
| "grad_norm": 1.0258524417877197, | |
| "learning_rate": 6.267624191883551e-06, | |
| "loss": 0.1369, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9007936507936508, | |
| "grad_norm": 0.8920129537582397, | |
| "learning_rate": 6.235808828293135e-06, | |
| "loss": 0.1233, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 0.95817631483078, | |
| "learning_rate": 6.203940082845144e-06, | |
| "loss": 0.162, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9087301587301587, | |
| "grad_norm": 0.7560166120529175, | |
| "learning_rate": 6.172019332138285e-06, | |
| "loss": 0.1046, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9126984126984127, | |
| "grad_norm": 0.9044355154037476, | |
| "learning_rate": 6.140047955017672e-06, | |
| "loss": 0.1508, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 0.7884669303894043, | |
| "learning_rate": 6.108027332515276e-06, | |
| "loss": 0.0952, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9206349206349206, | |
| "grad_norm": 0.9973843097686768, | |
| "learning_rate": 6.075958847790262e-06, | |
| "loss": 0.1333, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9246031746031746, | |
| "grad_norm": 0.8128191828727722, | |
| "learning_rate": 6.043843886069251e-06, | |
| "loss": 0.1174, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.9058467149734497, | |
| "learning_rate": 6.011683834586474e-06, | |
| "loss": 0.1275, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.9325396825396826, | |
| "grad_norm": 0.9947108626365662, | |
| "learning_rate": 5.979480082523858e-06, | |
| "loss": 0.1598, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9365079365079365, | |
| "grad_norm": 0.9292263388633728, | |
| "learning_rate": 5.947234020951015e-06, | |
| "loss": 0.1333, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9404761904761905, | |
| "grad_norm": 0.953533411026001, | |
| "learning_rate": 5.914947042765149e-06, | |
| "loss": 0.1541, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 0.8998634219169617, | |
| "learning_rate": 5.882620542630901e-06, | |
| "loss": 0.1422, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9484126984126984, | |
| "grad_norm": 0.9365549683570862, | |
| "learning_rate": 5.850255916920093e-06, | |
| "loss": 0.1368, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.8849459290504456, | |
| "learning_rate": 5.817854563651415e-06, | |
| "loss": 0.1432, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9563492063492064, | |
| "grad_norm": 0.8897712230682373, | |
| "learning_rate": 5.785417882430035e-06, | |
| "loss": 0.1152, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.9603174603174603, | |
| "grad_norm": 1.0064246654510498, | |
| "learning_rate": 5.752947274387147e-06, | |
| "loss": 0.1333, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 0.8676203489303589, | |
| "learning_rate": 5.720444142119445e-06, | |
| "loss": 0.152, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9682539682539683, | |
| "grad_norm": 0.8367963433265686, | |
| "learning_rate": 5.687909889628529e-06, | |
| "loss": 0.1167, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9722222222222222, | |
| "grad_norm": 0.9383995532989502, | |
| "learning_rate": 5.6553459222602714e-06, | |
| "loss": 0.1528, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9761904761904762, | |
| "grad_norm": 0.7976069450378418, | |
| "learning_rate": 5.622753646644102e-06, | |
| "loss": 0.0996, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.9801587301587301, | |
| "grad_norm": 0.9432264566421509, | |
| "learning_rate": 5.59013447063225e-06, | |
| "loss": 0.1604, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.9841269841269841, | |
| "grad_norm": 0.9246891736984253, | |
| "learning_rate": 5.557489803238934e-06, | |
| "loss": 0.1195, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.9880952380952381, | |
| "grad_norm": 0.8265438675880432, | |
| "learning_rate": 5.524821054579491e-06, | |
| "loss": 0.1336, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.9920634920634921, | |
| "grad_norm": 0.8477950096130371, | |
| "learning_rate": 5.492129635809473e-06, | |
| "loss": 0.1135, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.996031746031746, | |
| "grad_norm": 0.9364736080169678, | |
| "learning_rate": 5.459416959063688e-06, | |
| "loss": 0.1371, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8975462317466736, | |
| "learning_rate": 5.426684437395196e-06, | |
| "loss": 0.1724, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.003968253968254, | |
| "grad_norm": 0.820711612701416, | |
| "learning_rate": 5.393933484714284e-06, | |
| "loss": 0.1094, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.007936507936508, | |
| "grad_norm": 0.8223680853843689, | |
| "learning_rate": 5.361165515727374e-06, | |
| "loss": 0.1155, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0119047619047619, | |
| "grad_norm": 0.8105658292770386, | |
| "learning_rate": 5.328381945875933e-06, | |
| "loss": 0.1194, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0158730158730158, | |
| "grad_norm": 0.8975909948348999, | |
| "learning_rate": 5.295584191275308e-06, | |
| "loss": 0.1238, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.0198412698412698, | |
| "grad_norm": 0.9527161121368408, | |
| "learning_rate": 5.26277366865358e-06, | |
| "loss": 0.1529, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0238095238095237, | |
| "grad_norm": 0.8438770174980164, | |
| "learning_rate": 5.229951795290353e-06, | |
| "loss": 0.1192, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.0277777777777777, | |
| "grad_norm": 0.7387480139732361, | |
| "learning_rate": 5.197119988955534e-06, | |
| "loss": 0.0773, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0317460317460316, | |
| "grad_norm": 0.7979804873466492, | |
| "learning_rate": 5.164279667848094e-06, | |
| "loss": 0.0989, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.0357142857142858, | |
| "grad_norm": 0.7561360001564026, | |
| "learning_rate": 5.131432250534809e-06, | |
| "loss": 0.0717, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0396825396825398, | |
| "grad_norm": 1.0588268041610718, | |
| "learning_rate": 5.0985791558889785e-06, | |
| "loss": 0.125, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0436507936507937, | |
| "grad_norm": 0.9475772380828857, | |
| "learning_rate": 5.065721803029146e-06, | |
| "loss": 0.1267, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": 0.8831772804260254, | |
| "learning_rate": 5.032861611257783e-06, | |
| "loss": 0.0973, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0515873015873016, | |
| "grad_norm": 0.8899000287055969, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1004, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0555555555555556, | |
| "grad_norm": 0.9299007654190063, | |
| "learning_rate": 4.967138388742218e-06, | |
| "loss": 0.1132, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0595238095238095, | |
| "grad_norm": 0.8108347654342651, | |
| "learning_rate": 4.934278196970857e-06, | |
| "loss": 0.1039, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0634920634920635, | |
| "grad_norm": 0.9785858392715454, | |
| "learning_rate": 4.9014208441110215e-06, | |
| "loss": 0.133, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0674603174603174, | |
| "grad_norm": 0.7768881320953369, | |
| "learning_rate": 4.868567749465192e-06, | |
| "loss": 0.0861, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 0.7783759832382202, | |
| "learning_rate": 4.835720332151907e-06, | |
| "loss": 0.0903, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0753968253968254, | |
| "grad_norm": 0.796172022819519, | |
| "learning_rate": 4.802880011044467e-06, | |
| "loss": 0.0909, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.0793650793650793, | |
| "grad_norm": 0.804815948009491, | |
| "learning_rate": 4.770048204709648e-06, | |
| "loss": 0.0861, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.0833333333333333, | |
| "grad_norm": 0.8001012206077576, | |
| "learning_rate": 4.73722633134642e-06, | |
| "loss": 0.085, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.0873015873015872, | |
| "grad_norm": 0.8862425088882446, | |
| "learning_rate": 4.7044158087246926e-06, | |
| "loss": 0.1066, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.0912698412698412, | |
| "grad_norm": 0.8834551572799683, | |
| "learning_rate": 4.67161805412407e-06, | |
| "loss": 0.1017, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.0952380952380953, | |
| "grad_norm": 0.8457803726196289, | |
| "learning_rate": 4.6388344842726266e-06, | |
| "loss": 0.091, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.0992063492063493, | |
| "grad_norm": 0.7590792775154114, | |
| "learning_rate": 4.606066515285719e-06, | |
| "loss": 0.1015, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.1031746031746033, | |
| "grad_norm": 0.6717273592948914, | |
| "learning_rate": 4.573315562604804e-06, | |
| "loss": 0.0824, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1071428571428572, | |
| "grad_norm": 0.7835940718650818, | |
| "learning_rate": 4.540583040936313e-06, | |
| "loss": 0.0928, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.7849730253219604, | |
| "learning_rate": 4.5078703641905275e-06, | |
| "loss": 0.0807, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1150793650793651, | |
| "grad_norm": 0.7564438581466675, | |
| "learning_rate": 4.4751789454205105e-06, | |
| "loss": 0.0856, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.119047619047619, | |
| "grad_norm": 0.8760502934455872, | |
| "learning_rate": 4.442510196761068e-06, | |
| "loss": 0.0931, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.123015873015873, | |
| "grad_norm": 0.8113685846328735, | |
| "learning_rate": 4.409865529367751e-06, | |
| "loss": 0.0844, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.126984126984127, | |
| "grad_norm": 0.8088273406028748, | |
| "learning_rate": 4.377246353355899e-06, | |
| "loss": 0.0856, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.130952380952381, | |
| "grad_norm": 0.8084850907325745, | |
| "learning_rate": 4.34465407773973e-06, | |
| "loss": 0.0851, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1349206349206349, | |
| "grad_norm": 0.8789322376251221, | |
| "learning_rate": 4.312090110371473e-06, | |
| "loss": 0.1142, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1388888888888888, | |
| "grad_norm": 0.8025081753730774, | |
| "learning_rate": 4.279555857880558e-06, | |
| "loss": 0.0774, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.8897087574005127, | |
| "learning_rate": 4.247052725612853e-06, | |
| "loss": 0.091, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1468253968253967, | |
| "grad_norm": 0.8720001578330994, | |
| "learning_rate": 4.214582117569966e-06, | |
| "loss": 0.0987, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.1507936507936507, | |
| "grad_norm": 0.8932520151138306, | |
| "learning_rate": 4.182145436348587e-06, | |
| "loss": 0.1009, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1547619047619047, | |
| "grad_norm": 0.8310272693634033, | |
| "learning_rate": 4.1497440830799084e-06, | |
| "loss": 0.0893, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1587301587301586, | |
| "grad_norm": 0.8611903190612793, | |
| "learning_rate": 4.1173794573691e-06, | |
| "loss": 0.1108, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1626984126984128, | |
| "grad_norm": 0.8152250647544861, | |
| "learning_rate": 4.0850529572348505e-06, | |
| "loss": 0.0861, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 0.7950680255889893, | |
| "learning_rate": 4.052765979048986e-06, | |
| "loss": 0.0903, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.1706349206349207, | |
| "grad_norm": 0.7983028292655945, | |
| "learning_rate": 4.0205199174761435e-06, | |
| "loss": 0.0864, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1746031746031746, | |
| "grad_norm": 0.8648856282234192, | |
| "learning_rate": 3.988316165413528e-06, | |
| "loss": 0.0897, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.1785714285714286, | |
| "grad_norm": 0.866005539894104, | |
| "learning_rate": 3.956156113930752e-06, | |
| "loss": 0.0821, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.1825396825396826, | |
| "grad_norm": 0.7894255518913269, | |
| "learning_rate": 3.924041152209739e-06, | |
| "loss": 0.0951, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.1865079365079365, | |
| "grad_norm": 0.830491304397583, | |
| "learning_rate": 3.891972667484726e-06, | |
| "loss": 0.0921, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.7861407399177551, | |
| "learning_rate": 3.859952044982329e-06, | |
| "loss": 0.0976, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1944444444444444, | |
| "grad_norm": 0.7552138566970825, | |
| "learning_rate": 3.827980667861716e-06, | |
| "loss": 0.1001, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.1984126984126984, | |
| "grad_norm": 0.7325167059898376, | |
| "learning_rate": 3.7960599171548572e-06, | |
| "loss": 0.0825, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2023809523809523, | |
| "grad_norm": 0.7873362898826599, | |
| "learning_rate": 3.764191171706867e-06, | |
| "loss": 0.1011, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2063492063492063, | |
| "grad_norm": 0.7138109803199768, | |
| "learning_rate": 3.732375808116451e-06, | |
| "loss": 0.0782, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2103174603174602, | |
| "grad_norm": 0.8072618842124939, | |
| "learning_rate": 3.7006152006764336e-06, | |
| "loss": 0.0998, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 0.8527789115905762, | |
| "learning_rate": 3.6689107213144025e-06, | |
| "loss": 0.0843, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.2182539682539684, | |
| "grad_norm": 0.8525322079658508, | |
| "learning_rate": 3.6372637395334416e-06, | |
| "loss": 0.1162, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.8509613871574402, | |
| "learning_rate": 3.6056756223529734e-06, | |
| "loss": 0.0914, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2261904761904763, | |
| "grad_norm": 0.6690645813941956, | |
| "learning_rate": 3.574147734249719e-06, | |
| "loss": 0.0754, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2301587301587302, | |
| "grad_norm": 0.8752093315124512, | |
| "learning_rate": 3.542681437098745e-06, | |
| "loss": 0.1008, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2341269841269842, | |
| "grad_norm": 0.943699061870575, | |
| "learning_rate": 3.5112780901146426e-06, | |
| "loss": 0.103, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 0.8705005049705505, | |
| "learning_rate": 3.479939049792817e-06, | |
| "loss": 0.0988, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.242063492063492, | |
| "grad_norm": 0.7615220546722412, | |
| "learning_rate": 3.448665669850888e-06, | |
| "loss": 0.0854, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.246031746031746, | |
| "grad_norm": 0.9106603860855103, | |
| "learning_rate": 3.4174593011702197e-06, | |
| "loss": 0.1208, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.7418195009231567, | |
| "learning_rate": 3.386321291737563e-06, | |
| "loss": 0.0778, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.253968253968254, | |
| "grad_norm": 0.9659190773963928, | |
| "learning_rate": 3.3552529865868323e-06, | |
| "loss": 0.1019, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.257936507936508, | |
| "grad_norm": 0.7192327976226807, | |
| "learning_rate": 3.3242557277410015e-06, | |
| "loss": 0.0753, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2619047619047619, | |
| "grad_norm": 0.8058218359947205, | |
| "learning_rate": 3.2933308541541365e-06, | |
| "loss": 0.099, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.2658730158730158, | |
| "grad_norm": 0.7373734712600708, | |
| "learning_rate": 3.2624797016535626e-06, | |
| "loss": 0.0808, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2698412698412698, | |
| "grad_norm": 0.8924863934516907, | |
| "learning_rate": 3.2317036028821523e-06, | |
| "loss": 0.1209, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2738095238095237, | |
| "grad_norm": 0.7899354696273804, | |
| "learning_rate": 3.201003887240768e-06, | |
| "loss": 0.0876, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.2777777777777777, | |
| "grad_norm": 0.843604326248169, | |
| "learning_rate": 3.1703818808308327e-06, | |
| "loss": 0.0733, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.2817460317460316, | |
| "grad_norm": 0.9317384958267212, | |
| "learning_rate": 3.1398389063970512e-06, | |
| "loss": 0.0935, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.8000175356864929, | |
| "learning_rate": 3.1093762832702775e-06, | |
| "loss": 0.1022, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.2896825396825398, | |
| "grad_norm": 0.8414785265922546, | |
| "learning_rate": 3.0789953273105123e-06, | |
| "loss": 0.0983, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.2936507936507937, | |
| "grad_norm": 0.9642720222473145, | |
| "learning_rate": 3.048697350850073e-06, | |
| "loss": 0.1247, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.2976190476190477, | |
| "grad_norm": 0.8774341940879822, | |
| "learning_rate": 3.0184836626369034e-06, | |
| "loss": 0.1028, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3015873015873016, | |
| "grad_norm": 0.7191782593727112, | |
| "learning_rate": 2.988355567778043e-06, | |
| "loss": 0.085, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3055555555555556, | |
| "grad_norm": 0.7950049638748169, | |
| "learning_rate": 2.9583143676832526e-06, | |
| "loss": 0.0973, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3095238095238095, | |
| "grad_norm": 0.7327785491943359, | |
| "learning_rate": 2.9283613600087933e-06, | |
| "loss": 0.0662, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3134920634920635, | |
| "grad_norm": 0.7112066745758057, | |
| "learning_rate": 2.8984978386013767e-06, | |
| "loss": 0.0883, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3174603174603174, | |
| "grad_norm": 0.7717375755310059, | |
| "learning_rate": 2.8687250934422774e-06, | |
| "loss": 0.0837, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3214285714285714, | |
| "grad_norm": 0.9309632778167725, | |
| "learning_rate": 2.839044410591606e-06, | |
| "loss": 0.1291, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3253968253968254, | |
| "grad_norm": 0.7958415746688843, | |
| "learning_rate": 2.809457072132766e-06, | |
| "loss": 0.1007, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3293650793650793, | |
| "grad_norm": 0.718254566192627, | |
| "learning_rate": 2.779964356117063e-06, | |
| "loss": 0.0811, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.8122881054878235, | |
| "learning_rate": 2.750567536508504e-06, | |
| "loss": 0.0934, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3373015873015874, | |
| "grad_norm": 0.7767068147659302, | |
| "learning_rate": 2.7212678831287627e-06, | |
| "loss": 0.0938, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.3412698412698414, | |
| "grad_norm": 0.781796395778656, | |
| "learning_rate": 2.692066661602333e-06, | |
| "loss": 0.0844, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3452380952380953, | |
| "grad_norm": 0.865658164024353, | |
| "learning_rate": 2.662965133301862e-06, | |
| "loss": 0.0981, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.3492063492063493, | |
| "grad_norm": 0.8029084801673889, | |
| "learning_rate": 2.633964555293654e-06, | |
| "loss": 0.0904, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3531746031746033, | |
| "grad_norm": 0.7445881366729736, | |
| "learning_rate": 2.605066180283378e-06, | |
| "loss": 0.0756, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 0.8395903706550598, | |
| "learning_rate": 2.576271256561953e-06, | |
| "loss": 0.0873, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3611111111111112, | |
| "grad_norm": 0.8283248543739319, | |
| "learning_rate": 2.5475810279516287e-06, | |
| "loss": 0.0985, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3650793650793651, | |
| "grad_norm": 0.8155176043510437, | |
| "learning_rate": 2.5189967337522574e-06, | |
| "loss": 0.0862, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.369047619047619, | |
| "grad_norm": 0.9677881598472595, | |
| "learning_rate": 2.49051960868776e-06, | |
| "loss": 0.1088, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.373015873015873, | |
| "grad_norm": 0.8105742931365967, | |
| "learning_rate": 2.46215088285279e-06, | |
| "loss": 0.0877, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.376984126984127, | |
| "grad_norm": 0.7642203569412231, | |
| "learning_rate": 2.433891781659603e-06, | |
| "loss": 0.0717, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 0.8333291411399841, | |
| "learning_rate": 2.4057435257851173e-06, | |
| "loss": 0.1073, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.3849206349206349, | |
| "grad_norm": 0.7504364848136902, | |
| "learning_rate": 2.377707331118196e-06, | |
| "loss": 0.0951, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 0.6735277771949768, | |
| "learning_rate": 2.349784408707112e-06, | |
| "loss": 0.0821, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.3928571428571428, | |
| "grad_norm": 0.6802200675010681, | |
| "learning_rate": 2.3219759647072467e-06, | |
| "loss": 0.0737, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.3968253968253967, | |
| "grad_norm": 0.8249236941337585, | |
| "learning_rate": 2.2942832003289823e-06, | |
| "loss": 0.0873, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4007936507936507, | |
| "grad_norm": 0.7745435237884521, | |
| "learning_rate": 2.2667073117858185e-06, | |
| "loss": 0.1097, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4047619047619047, | |
| "grad_norm": 0.6764716506004333, | |
| "learning_rate": 2.2392494902427027e-06, | |
| "loss": 0.0728, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4087301587301586, | |
| "grad_norm": 0.6814324259757996, | |
| "learning_rate": 2.2119109217645697e-06, | |
| "loss": 0.0839, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4126984126984126, | |
| "grad_norm": 0.7345415949821472, | |
| "learning_rate": 2.1846927872651135e-06, | |
| "loss": 0.0808, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4166666666666667, | |
| "grad_norm": 0.8106203675270081, | |
| "learning_rate": 2.1575962624557754e-06, | |
| "loss": 0.0974, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4206349206349207, | |
| "grad_norm": 0.8489047884941101, | |
| "learning_rate": 2.1306225177949584e-06, | |
| "loss": 0.1057, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4246031746031746, | |
| "grad_norm": 0.6826899647712708, | |
| "learning_rate": 2.1037727184374705e-06, | |
| "loss": 0.0713, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.8070489168167114, | |
| "learning_rate": 2.07704802418419e-06, | |
| "loss": 0.0952, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4325396825396826, | |
| "grad_norm": 0.9150393009185791, | |
| "learning_rate": 2.050449589431969e-06, | |
| "loss": 0.1096, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.4365079365079365, | |
| "grad_norm": 0.7656403183937073, | |
| "learning_rate": 2.023978563123771e-06, | |
| "loss": 0.0806, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4404761904761905, | |
| "grad_norm": 0.8073910474777222, | |
| "learning_rate": 1.997636088699035e-06, | |
| "loss": 0.088, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.8604737520217896, | |
| "learning_rate": 1.9714233040442915e-06, | |
| "loss": 0.0957, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4484126984126984, | |
| "grad_norm": 0.73807293176651, | |
| "learning_rate": 1.9453413414440043e-06, | |
| "loss": 0.0894, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4523809523809523, | |
| "grad_norm": 0.9138429164886475, | |
| "learning_rate": 1.919391327531663e-06, | |
| "loss": 0.1196, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.4563492063492063, | |
| "grad_norm": 0.7921180725097656, | |
| "learning_rate": 1.8935743832411163e-06, | |
| "loss": 0.077, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4603174603174602, | |
| "grad_norm": 0.8694080710411072, | |
| "learning_rate": 1.8678916237581524e-06, | |
| "loss": 0.0918, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4642857142857144, | |
| "grad_norm": 0.9354726672172546, | |
| "learning_rate": 1.8423441584723312e-06, | |
| "loss": 0.0981, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4682539682539684, | |
| "grad_norm": 0.8500098586082458, | |
| "learning_rate": 1.816933090929055e-06, | |
| "loss": 0.0937, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.4722222222222223, | |
| "grad_norm": 0.8204169273376465, | |
| "learning_rate": 1.791659518781908e-06, | |
| "loss": 0.0824, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.4761904761904763, | |
| "grad_norm": 0.9199097156524658, | |
| "learning_rate": 1.7665245337452368e-06, | |
| "loss": 0.0999, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.4801587301587302, | |
| "grad_norm": 0.8253099918365479, | |
| "learning_rate": 1.7415292215469948e-06, | |
| "loss": 0.082, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.4841269841269842, | |
| "grad_norm": 0.8107463717460632, | |
| "learning_rate": 1.716674661881848e-06, | |
| "loss": 0.0986, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.4880952380952381, | |
| "grad_norm": 0.9667322039604187, | |
| "learning_rate": 1.6919619283645262e-06, | |
| "loss": 0.1033, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.492063492063492, | |
| "grad_norm": 0.7419166564941406, | |
| "learning_rate": 1.667392088483456e-06, | |
| "loss": 0.0841, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.496031746031746, | |
| "grad_norm": 0.8369618058204651, | |
| "learning_rate": 1.6429662035546451e-06, | |
| "loss": 0.0829, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.6413199305534363, | |
| "learning_rate": 1.6186853286758397e-06, | |
| "loss": 0.0731, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.503968253968254, | |
| "grad_norm": 0.5977425575256348, | |
| "learning_rate": 1.5945505126809524e-06, | |
| "loss": 0.0578, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.507936507936508, | |
| "grad_norm": 0.6303495168685913, | |
| "learning_rate": 1.570562798094747e-06, | |
| "loss": 0.0639, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.5119047619047619, | |
| "grad_norm": 0.7885125279426575, | |
| "learning_rate": 1.5467232210878153e-06, | |
| "loss": 0.1052, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5158730158730158, | |
| "grad_norm": 0.827327311038971, | |
| "learning_rate": 1.5230328114318127e-06, | |
| "loss": 0.0915, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5198412698412698, | |
| "grad_norm": 0.6350776553153992, | |
| "learning_rate": 1.4994925924549797e-06, | |
| "loss": 0.0652, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 0.7374575734138489, | |
| "learning_rate": 1.4761035809979395e-06, | |
| "loss": 0.0689, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5277777777777777, | |
| "grad_norm": 0.7057892680168152, | |
| "learning_rate": 1.452866787369771e-06, | |
| "loss": 0.0894, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5317460317460316, | |
| "grad_norm": 0.7495548725128174, | |
| "learning_rate": 1.4297832153043657e-06, | |
| "loss": 0.0833, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5357142857142856, | |
| "grad_norm": 0.762456476688385, | |
| "learning_rate": 1.4068538619170763e-06, | |
| "loss": 0.0724, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5396825396825395, | |
| "grad_norm": 0.7710168361663818, | |
| "learning_rate": 1.3840797176616467e-06, | |
| "loss": 0.0768, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.5436507936507935, | |
| "grad_norm": 0.8323285579681396, | |
| "learning_rate": 1.3614617662874197e-06, | |
| "loss": 0.0988, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5476190476190477, | |
| "grad_norm": 0.8215609192848206, | |
| "learning_rate": 1.3390009847968505e-06, | |
| "loss": 0.0873, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.5515873015873016, | |
| "grad_norm": 0.7752273678779602, | |
| "learning_rate": 1.316698343403302e-06, | |
| "loss": 0.1002, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.9786981344223022, | |
| "learning_rate": 1.2945548054891322e-06, | |
| "loss": 0.1182, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.5595238095238095, | |
| "grad_norm": 0.7958163619041443, | |
| "learning_rate": 1.27257132756409e-06, | |
| "loss": 0.0807, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5634920634920635, | |
| "grad_norm": 0.8421275019645691, | |
| "learning_rate": 1.2507488592239848e-06, | |
| "loss": 0.0903, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5674603174603174, | |
| "grad_norm": 0.8981534242630005, | |
| "learning_rate": 1.2290883431096778e-06, | |
| "loss": 0.0847, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.7970631718635559, | |
| "learning_rate": 1.2075907148663579e-06, | |
| "loss": 0.0818, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.5753968253968254, | |
| "grad_norm": 0.7949286103248596, | |
| "learning_rate": 1.186256903103129e-06, | |
| "loss": 0.0729, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.5793650793650795, | |
| "grad_norm": 0.8888193964958191, | |
| "learning_rate": 1.1650878293528994e-06, | |
| "loss": 0.0978, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.5833333333333335, | |
| "grad_norm": 0.8030498623847961, | |
| "learning_rate": 1.1440844080325703e-06, | |
| "loss": 0.09, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.5873015873015874, | |
| "grad_norm": 0.7452394366264343, | |
| "learning_rate": 1.1232475464035386e-06, | |
| "loss": 0.087, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5912698412698414, | |
| "grad_norm": 0.6083802580833435, | |
| "learning_rate": 1.10257814453251e-06, | |
| "loss": 0.0609, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.5952380952380953, | |
| "grad_norm": 0.8018980026245117, | |
| "learning_rate": 1.0820770952526155e-06, | |
| "loss": 0.0889, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.5992063492063493, | |
| "grad_norm": 0.6214452981948853, | |
| "learning_rate": 1.0617452841248494e-06, | |
| "loss": 0.0827, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6031746031746033, | |
| "grad_norm": 0.7810742855072021, | |
| "learning_rate": 1.0415835893998116e-06, | |
| "loss": 0.0927, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.6071428571428572, | |
| "grad_norm": 0.7865670323371887, | |
| "learning_rate": 1.0215928819797744e-06, | |
| "loss": 0.096, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6111111111111112, | |
| "grad_norm": 0.657080352306366, | |
| "learning_rate": 1.0017740253810608e-06, | |
| "loss": 0.0582, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6150793650793651, | |
| "grad_norm": 0.6832369565963745, | |
| "learning_rate": 9.821278756967467e-07, | |
| "loss": 0.076, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 0.94068843126297, | |
| "learning_rate": 9.62655281559679e-07, | |
| "loss": 0.1121, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.623015873015873, | |
| "grad_norm": 0.7479247450828552, | |
| "learning_rate": 9.433570841058187e-07, | |
| "loss": 0.0771, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.626984126984127, | |
| "grad_norm": 0.7667514681816101, | |
| "learning_rate": 9.242341169379077e-07, | |
| "loss": 0.0884, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.630952380952381, | |
| "grad_norm": 0.7788687944412231, | |
| "learning_rate": 9.052872060894613e-07, | |
| "loss": 0.0844, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6349206349206349, | |
| "grad_norm": 0.8176941871643066, | |
| "learning_rate": 8.865171699890835e-07, | |
| "loss": 0.0785, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6388888888888888, | |
| "grad_norm": 0.8092814087867737, | |
| "learning_rate": 8.679248194251211e-07, | |
| "loss": 0.0961, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 0.7338768243789673, | |
| "learning_rate": 8.495109575106331e-07, | |
| "loss": 0.0825, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.6468253968253967, | |
| "grad_norm": 0.7372951507568359, | |
| "learning_rate": 8.312763796487038e-07, | |
| "loss": 0.0738, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6507936507936507, | |
| "grad_norm": 0.7479267716407776, | |
| "learning_rate": 8.132218734980852e-07, | |
| "loss": 0.0823, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.6547619047619047, | |
| "grad_norm": 0.7982333898544312, | |
| "learning_rate": 7.953482189391687e-07, | |
| "loss": 0.088, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6587301587301586, | |
| "grad_norm": 0.8214784860610962, | |
| "learning_rate": 7.776561880403072e-07, | |
| "loss": 0.0857, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6626984126984126, | |
| "grad_norm": 1.0360147953033447, | |
| "learning_rate": 7.601465450244528e-07, | |
| "loss": 0.1566, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.7793018221855164, | |
| "learning_rate": 7.42820046236154e-07, | |
| "loss": 0.0806, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.6706349206349205, | |
| "grad_norm": 0.8804833889007568, | |
| "learning_rate": 7.256774401088817e-07, | |
| "loss": 0.0945, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.6746031746031746, | |
| "grad_norm": 0.8523043394088745, | |
| "learning_rate": 7.087194671326986e-07, | |
| "loss": 0.0796, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.6785714285714286, | |
| "grad_norm": 0.9247816801071167, | |
| "learning_rate": 6.91946859822279e-07, | |
| "loss": 0.1049, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.6825396825396826, | |
| "grad_norm": 0.8288869857788086, | |
| "learning_rate": 6.753603426852589e-07, | |
| "loss": 0.1076, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.6865079365079365, | |
| "grad_norm": 0.6493294835090637, | |
| "learning_rate": 6.589606321909464e-07, | |
| "loss": 0.0786, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.6904761904761905, | |
| "grad_norm": 0.7904691100120544, | |
| "learning_rate": 6.427484367393699e-07, | |
| "loss": 0.0992, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.6944444444444444, | |
| "grad_norm": 0.7053451538085938, | |
| "learning_rate": 6.267244566306801e-07, | |
| "loss": 0.0788, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.6984126984126984, | |
| "grad_norm": 0.7589640021324158, | |
| "learning_rate": 6.108893840348995e-07, | |
| "loss": 0.1059, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.7023809523809523, | |
| "grad_norm": 0.7016319632530212, | |
| "learning_rate": 5.952439029620222e-07, | |
| "loss": 0.0731, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7063492063492065, | |
| "grad_norm": 0.6430158615112305, | |
| "learning_rate": 5.797886892324695e-07, | |
| "loss": 0.0657, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7103174603174605, | |
| "grad_norm": 0.668388843536377, | |
| "learning_rate": 5.645244104478947e-07, | |
| "loss": 0.0747, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.8092103004455566, | |
| "learning_rate": 5.494517259623478e-07, | |
| "loss": 0.1113, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7182539682539684, | |
| "grad_norm": 0.7932750582695007, | |
| "learning_rate": 5.34571286853795e-07, | |
| "loss": 0.1037, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.7222222222222223, | |
| "grad_norm": 0.8262640833854675, | |
| "learning_rate": 5.198837358959901e-07, | |
| "loss": 0.1064, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7261904761904763, | |
| "grad_norm": 0.6715550422668457, | |
| "learning_rate": 5.05389707530714e-07, | |
| "loss": 0.0635, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7301587301587302, | |
| "grad_norm": 0.6762986183166504, | |
| "learning_rate": 4.91089827840367e-07, | |
| "loss": 0.0774, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7341269841269842, | |
| "grad_norm": 0.851775586605072, | |
| "learning_rate": 4.769847145209244e-07, | |
| "loss": 0.1099, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7380952380952381, | |
| "grad_norm": 0.774852454662323, | |
| "learning_rate": 4.6307497685525894e-07, | |
| "loss": 0.0877, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.742063492063492, | |
| "grad_norm": 0.7787303328514099, | |
| "learning_rate": 4.4936121568681546e-07, | |
| "loss": 0.0844, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.746031746031746, | |
| "grad_norm": 0.737352192401886, | |
| "learning_rate": 4.3584402339366174e-07, | |
| "loss": 0.0744, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.7330744862556458, | |
| "learning_rate": 4.225239838628981e-07, | |
| "loss": 0.0744, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.753968253968254, | |
| "grad_norm": 0.8031402230262756, | |
| "learning_rate": 4.0940167246543595e-07, | |
| "loss": 0.0877, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.757936507936508, | |
| "grad_norm": 0.8259134888648987, | |
| "learning_rate": 3.964776560311484e-07, | |
| "loss": 0.0943, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 0.816821277141571, | |
| "learning_rate": 3.8375249282437743e-07, | |
| "loss": 0.0934, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.7658730158730158, | |
| "grad_norm": 0.8658815026283264, | |
| "learning_rate": 3.71226732519826e-07, | |
| "loss": 0.1, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.7698412698412698, | |
| "grad_norm": 0.8272645473480225, | |
| "learning_rate": 3.589009161788104e-07, | |
| "loss": 0.0839, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.7738095238095237, | |
| "grad_norm": 0.7262760400772095, | |
| "learning_rate": 3.4677557622589175e-07, | |
| "loss": 0.0757, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.7805841565132141, | |
| "learning_rate": 3.3485123642587657e-07, | |
| "loss": 0.0805, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.7817460317460316, | |
| "grad_norm": 0.7673043608665466, | |
| "learning_rate": 3.2312841186118937e-07, | |
| "loss": 0.0804, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.8263194561004639, | |
| "learning_rate": 3.116076089096265e-07, | |
| "loss": 0.0858, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.7896825396825395, | |
| "grad_norm": 0.7008450627326965, | |
| "learning_rate": 3.0028932522248256e-07, | |
| "loss": 0.0781, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.7936507936507935, | |
| "grad_norm": 0.6593312621116638, | |
| "learning_rate": 2.8917404970305096e-07, | |
| "loss": 0.0697, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.7976190476190477, | |
| "grad_norm": 0.7077324986457825, | |
| "learning_rate": 2.782622624855097e-07, | |
| "loss": 0.0848, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8015873015873016, | |
| "grad_norm": 0.6854461431503296, | |
| "learning_rate": 2.6755443491417786e-07, | |
| "loss": 0.0808, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8055555555555556, | |
| "grad_norm": 0.6377874612808228, | |
| "learning_rate": 2.570510295231571e-07, | |
| "loss": 0.0648, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 0.7600693106651306, | |
| "learning_rate": 2.467525000163523e-07, | |
| "loss": 0.0826, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8134920634920635, | |
| "grad_norm": 0.7715564370155334, | |
| "learning_rate": 2.36659291247871e-07, | |
| "loss": 0.0894, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.8174603174603174, | |
| "grad_norm": 0.6611568331718445, | |
| "learning_rate": 2.2677183920281342e-07, | |
| "loss": 0.0555, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8214285714285714, | |
| "grad_norm": 0.7414703965187073, | |
| "learning_rate": 2.1709057097843266e-07, | |
| "loss": 0.0709, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8253968253968254, | |
| "grad_norm": 0.7198565006256104, | |
| "learning_rate": 2.0761590476568893e-07, | |
| "loss": 0.068, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8293650793650795, | |
| "grad_norm": 0.8624417185783386, | |
| "learning_rate": 1.9834824983118673e-07, | |
| "loss": 0.1105, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 0.6496530175209045, | |
| "learning_rate": 1.892880064994934e-07, | |
| "loss": 0.0619, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.8373015873015874, | |
| "grad_norm": 0.8599086999893188, | |
| "learning_rate": 1.8043556613585143e-07, | |
| "loss": 0.1006, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8412698412698414, | |
| "grad_norm": 0.8874303102493286, | |
| "learning_rate": 1.7179131112926628e-07, | |
| "loss": 0.0978, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.8452380952380953, | |
| "grad_norm": 0.7611068487167358, | |
| "learning_rate": 1.6335561487599406e-07, | |
| "loss": 0.0767, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8492063492063493, | |
| "grad_norm": 1.0293042659759521, | |
| "learning_rate": 1.551288417634106e-07, | |
| "loss": 0.1407, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8531746031746033, | |
| "grad_norm": 0.8029921650886536, | |
| "learning_rate": 1.471113471542712e-07, | |
| "loss": 0.0847, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.8207499980926514, | |
| "learning_rate": 1.3930347737136195e-07, | |
| "loss": 0.0991, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8611111111111112, | |
| "grad_norm": 0.9645194411277771, | |
| "learning_rate": 1.3170556968253756e-07, | |
| "loss": 0.1227, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.8650793650793651, | |
| "grad_norm": 0.8458879590034485, | |
| "learning_rate": 1.2431795228615372e-07, | |
| "loss": 0.0852, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.869047619047619, | |
| "grad_norm": 0.8573436737060547, | |
| "learning_rate": 1.1714094429689127e-07, | |
| "loss": 0.0837, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.873015873015873, | |
| "grad_norm": 0.8495934009552002, | |
| "learning_rate": 1.1017485573197151e-07, | |
| "loss": 0.0934, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.876984126984127, | |
| "grad_norm": 0.7013949155807495, | |
| "learning_rate": 1.0341998749776316e-07, | |
| "loss": 0.0655, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.880952380952381, | |
| "grad_norm": 0.8161298036575317, | |
| "learning_rate": 9.687663137678605e-08, | |
| "loss": 0.0976, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.8849206349206349, | |
| "grad_norm": 0.8887423872947693, | |
| "learning_rate": 9.054507001510727e-08, | |
| "loss": 0.1053, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.6531678438186646, | |
| "learning_rate": 8.442557691013042e-08, | |
| "loss": 0.0793, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.8928571428571428, | |
| "grad_norm": 0.8312963247299194, | |
| "learning_rate": 7.851841639878399e-08, | |
| "loss": 0.0814, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.8968253968253967, | |
| "grad_norm": 0.7785887718200684, | |
| "learning_rate": 7.282384364610207e-08, | |
| "loss": 0.1005, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9007936507936507, | |
| "grad_norm": 0.617367684841156, | |
| "learning_rate": 6.734210463420099e-08, | |
| "loss": 0.0615, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 0.6540130376815796, | |
| "learning_rate": 6.207343615165562e-08, | |
| "loss": 0.0687, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9087301587301586, | |
| "grad_norm": 0.68092942237854, | |
| "learning_rate": 5.701806578327029e-08, | |
| "loss": 0.0654, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.9126984126984126, | |
| "grad_norm": 0.7034346461296082, | |
| "learning_rate": 5.21762119002478e-08, | |
| "loss": 0.073, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9166666666666665, | |
| "grad_norm": 0.668262779712677, | |
| "learning_rate": 4.7548083650759134e-08, | |
| "loss": 0.0647, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9206349206349205, | |
| "grad_norm": 0.6120609641075134, | |
| "learning_rate": 4.31338809509052e-08, | |
| "loss": 0.0563, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9246031746031746, | |
| "grad_norm": 0.7227481007575989, | |
| "learning_rate": 3.8933794476083143e-08, | |
| "loss": 0.0718, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.7431139945983887, | |
| "learning_rate": 3.494800565275125e-08, | |
| "loss": 0.0701, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.9325396825396826, | |
| "grad_norm": 0.698998212814331, | |
| "learning_rate": 3.1176686650589147e-08, | |
| "loss": 0.0754, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9365079365079365, | |
| "grad_norm": 0.8017215132713318, | |
| "learning_rate": 2.7620000375064848e-08, | |
| "loss": 0.0947, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.9404761904761905, | |
| "grad_norm": 0.7906326651573181, | |
| "learning_rate": 2.4278100460393138e-08, | |
| "loss": 0.0708, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9444444444444444, | |
| "grad_norm": 0.881683886051178, | |
| "learning_rate": 2.115113126290258e-08, | |
| "loss": 0.1025, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9484126984126984, | |
| "grad_norm": 0.7980960607528687, | |
| "learning_rate": 1.8239227854799368e-08, | |
| "loss": 0.0723, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9523809523809523, | |
| "grad_norm": 0.8099157214164734, | |
| "learning_rate": 1.554251601833201e-08, | |
| "loss": 0.0885, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9563492063492065, | |
| "grad_norm": 0.7876162528991699, | |
| "learning_rate": 1.3061112240357887e-08, | |
| "loss": 0.0717, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.9603174603174605, | |
| "grad_norm": 0.7510733008384705, | |
| "learning_rate": 1.0795123707312283e-08, | |
| "loss": 0.0739, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.9642857142857144, | |
| "grad_norm": 0.8528300523757935, | |
| "learning_rate": 8.744648300578196e-09, | |
| "loss": 0.0963, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.9682539682539684, | |
| "grad_norm": 0.8461951613426208, | |
| "learning_rate": 6.9097745922580564e-09, | |
| "loss": 0.082, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.9722222222222223, | |
| "grad_norm": 0.7651507258415222, | |
| "learning_rate": 5.2905818413478975e-09, | |
| "loss": 0.0667, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.9761904761904763, | |
| "grad_norm": 0.7861319184303284, | |
| "learning_rate": 3.887139990313427e-09, | |
| "loss": 0.084, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.9801587301587302, | |
| "grad_norm": 0.7902892231941223, | |
| "learning_rate": 2.699509662069666e-09, | |
| "loss": 0.0977, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.9841269841269842, | |
| "grad_norm": 0.6281541585922241, | |
| "learning_rate": 1.7277421573608234e-09, | |
| "loss": 0.0703, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9880952380952381, | |
| "grad_norm": 0.6765737533569336, | |
| "learning_rate": 9.71879452545399e-10, | |
| "loss": 0.0772, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.992063492063492, | |
| "grad_norm": 0.927233099937439, | |
| "learning_rate": 4.3195419778319095e-10, | |
| "loss": 0.1074, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.996031746031746, | |
| "grad_norm": 0.7625772356987, | |
| "learning_rate": 1.0798971562364647e-10, | |
| "loss": 0.0793, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.616499662399292, | |
| "learning_rate": 0.0, | |
| "loss": 0.0544, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 504, | |
| "total_flos": 54880760659968.0, | |
| "train_loss": 0.26880344517144655, | |
| "train_runtime": 2787.2623, | |
| "train_samples_per_second": 2.887, | |
| "train_steps_per_second": 0.181 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 504, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 54880760659968.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |