| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.8358962588937342, | |
| "eval_steps": 500, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004590314436538903, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 1.9960000000000002e-05, | |
| "loss": 1.2999, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.009180628873077806, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9915555555555557e-05, | |
| "loss": 0.6696, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013770943309616708, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.9871111111111112e-05, | |
| "loss": 0.6397, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.018361257746155612, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.9826666666666668e-05, | |
| "loss": 0.6335, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.022951572182694516, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.9782222222222226e-05, | |
| "loss": 0.6376, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.027541886619233416, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.973777777777778e-05, | |
| "loss": 0.5998, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03213220105577232, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9693333333333337e-05, | |
| "loss": 0.5576, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.036722515492311224, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.9648888888888892e-05, | |
| "loss": 0.5889, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.041312829928850124, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.9604444444444447e-05, | |
| "loss": 0.5732, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.04590314436538903, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9560000000000002e-05, | |
| "loss": 0.5744, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05049345880192793, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.9515555555555558e-05, | |
| "loss": 0.6096, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05508377323846683, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.9471111111111113e-05, | |
| "loss": 0.5568, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.05967408767500574, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.9426666666666668e-05, | |
| "loss": 0.5632, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06426440211154465, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.9382222222222223e-05, | |
| "loss": 0.5978, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06885471654808355, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.933777777777778e-05, | |
| "loss": 0.5439, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07344503098462245, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.9293333333333334e-05, | |
| "loss": 0.5606, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.07803534542116135, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.924888888888889e-05, | |
| "loss": 0.5474, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08262565985770025, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9204444444444444e-05, | |
| "loss": 0.5369, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.08721597429423915, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.916e-05, | |
| "loss": 0.5518, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.09180628873077806, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9115555555555555e-05, | |
| "loss": 0.5175, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09639660316731696, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.9071111111111113e-05, | |
| "loss": 0.4991, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.10098691760385586, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.902666666666667e-05, | |
| "loss": 0.5193, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.10557723204039476, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.8982222222222224e-05, | |
| "loss": 0.5548, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.11016754647693366, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.893777777777778e-05, | |
| "loss": 0.5331, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.11475786091347258, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.8893333333333334e-05, | |
| "loss": 0.5053, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.11934817535001148, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.884888888888889e-05, | |
| "loss": 0.563, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.12393848978655038, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8804444444444445e-05, | |
| "loss": 0.5089, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.1285288042230893, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.876e-05, | |
| "loss": 0.5113, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1331191186596282, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 1.871555555555556e-05, | |
| "loss": 0.5101, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1377094330961671, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.8671111111111114e-05, | |
| "loss": 0.5471, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.142299747532706, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.862666666666667e-05, | |
| "loss": 0.5001, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.1468900619692449, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.8582222222222224e-05, | |
| "loss": 0.5236, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1514803764057838, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.853777777777778e-05, | |
| "loss": 0.5357, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.1560706908423227, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.8493333333333335e-05, | |
| "loss": 0.5317, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.1606610052788616, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.844888888888889e-05, | |
| "loss": 0.5335, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.1652513197154005, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.840444444444445e-05, | |
| "loss": 0.4946, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.1698416341519394, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.8360000000000004e-05, | |
| "loss": 0.5056, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1744319485884783, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.831555555555556e-05, | |
| "loss": 0.4724, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.17902226302501723, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.8271111111111114e-05, | |
| "loss": 0.5489, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.18361257746155613, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.822666666666667e-05, | |
| "loss": 0.495, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.18820289189809503, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.8182222222222225e-05, | |
| "loss": 0.5209, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.19279320633463393, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.813777777777778e-05, | |
| "loss": 0.4774, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.19738352077117283, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.8093333333333335e-05, | |
| "loss": 0.4903, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.20197383520771173, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.804888888888889e-05, | |
| "loss": 0.4869, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.20656414964425063, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.8004444444444446e-05, | |
| "loss": 0.5008, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.21115446408078953, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.796e-05, | |
| "loss": 0.4668, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.21574477851732843, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.7915555555555556e-05, | |
| "loss": 0.4881, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.22033509295386733, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.787111111111111e-05, | |
| "loss": 0.4718, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.22492540739040623, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.7826666666666667e-05, | |
| "loss": 0.4783, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.22951572182694516, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.7782222222222222e-05, | |
| "loss": 0.5052, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23410603626348406, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.7737777777777777e-05, | |
| "loss": 0.4772, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.23869635070002296, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.7693333333333336e-05, | |
| "loss": 0.4855, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.24328666513656186, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.764888888888889e-05, | |
| "loss": 0.4438, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.24787697957310076, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 1.7604444444444446e-05, | |
| "loss": 0.4687, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2524672940096397, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.756e-05, | |
| "loss": 0.4514, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.2570576084461786, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.7515555555555557e-05, | |
| "loss": 0.4628, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2616479228827175, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.7471111111111112e-05, | |
| "loss": 0.4395, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2662382373192564, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.7426666666666667e-05, | |
| "loss": 0.4306, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2708285517557953, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.7382222222222222e-05, | |
| "loss": 0.4691, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2754188661923342, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.733777777777778e-05, | |
| "loss": 0.4755, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2800091806288731, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.7293333333333336e-05, | |
| "loss": 0.424, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.284599495065412, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.724888888888889e-05, | |
| "loss": 0.4662, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.2891898095019509, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.7204444444444446e-05, | |
| "loss": 0.4714, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2937801239384898, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7160000000000002e-05, | |
| "loss": 0.4383, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.2983704383750287, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.7115555555555557e-05, | |
| "loss": 0.4303, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3029607528115676, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.7071111111111112e-05, | |
| "loss": 0.4407, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3075510672481065, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.702666666666667e-05, | |
| "loss": 0.4429, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.3121413816846454, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.6982222222222226e-05, | |
| "loss": 0.4458, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3167316961211843, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.693777777777778e-05, | |
| "loss": 0.4515, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.3213220105577232, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.6893333333333336e-05, | |
| "loss": 0.4844, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3259123249942621, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.6848888888888892e-05, | |
| "loss": 0.4478, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.330502639430801, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.6804444444444447e-05, | |
| "loss": 0.4765, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3350929538673399, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.6760000000000002e-05, | |
| "loss": 0.423, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.3396832683038788, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 1.6715555555555557e-05, | |
| "loss": 0.4159, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.3442735827404177, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.6671111111111113e-05, | |
| "loss": 0.4501, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3488638971769566, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.6626666666666668e-05, | |
| "loss": 0.4407, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.35345421161349555, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.6582222222222223e-05, | |
| "loss": 0.4757, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.35804452605003445, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 1.6537777777777778e-05, | |
| "loss": 0.4475, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.36263484048657335, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.6493333333333334e-05, | |
| "loss": 0.434, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.36722515492311225, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.644888888888889e-05, | |
| "loss": 0.4449, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.37181546935965115, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.6404444444444444e-05, | |
| "loss": 0.3868, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.37640578379619005, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.636e-05, | |
| "loss": 0.4219, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.38099609823272895, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.6315555555555558e-05, | |
| "loss": 0.4183, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.38558641266926785, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.6271111111111113e-05, | |
| "loss": 0.4006, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.39017672710580675, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.6226666666666668e-05, | |
| "loss": 0.4333, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.39476704154234565, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.6182222222222224e-05, | |
| "loss": 0.4475, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.39935735597888455, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.613777777777778e-05, | |
| "loss": 0.4234, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.40394767041542345, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.6093333333333334e-05, | |
| "loss": 0.4239, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.40853798485196235, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.604888888888889e-05, | |
| "loss": 0.4158, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.41312829928850126, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.6004444444444444e-05, | |
| "loss": 0.4432, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.41771861372504016, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.5960000000000003e-05, | |
| "loss": 0.4214, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.42230892816157906, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.5915555555555558e-05, | |
| "loss": 0.4437, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.42689924259811796, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.5871111111111114e-05, | |
| "loss": 0.4371, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.43148955703465686, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.582666666666667e-05, | |
| "loss": 0.4232, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.43607987147119576, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.5782222222222224e-05, | |
| "loss": 0.4173, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.44067018590773466, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.573777777777778e-05, | |
| "loss": 0.3937, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.44526050034427356, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.5693333333333334e-05, | |
| "loss": 0.405, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.44985081478081246, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.564888888888889e-05, | |
| "loss": 0.3862, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.4544411292173514, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.5604444444444445e-05, | |
| "loss": 0.4358, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.4590314436538903, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.556e-05, | |
| "loss": 0.3889, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4636217580904292, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.551555555555556e-05, | |
| "loss": 0.4039, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.4682120725269681, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.5471111111111114e-05, | |
| "loss": 0.438, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.472802386963507, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.542666666666667e-05, | |
| "loss": 0.424, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.4773927014000459, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.5382222222222224e-05, | |
| "loss": 0.3888, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.4819830158365848, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.533777777777778e-05, | |
| "loss": 0.4235, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.4865733302731237, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.5293333333333335e-05, | |
| "loss": 0.4036, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.4911636447096626, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.524888888888889e-05, | |
| "loss": 0.3917, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.4957539591462015, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.5204444444444445e-05, | |
| "loss": 0.41, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5003442735827405, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 1.516e-05, | |
| "loss": 0.4206, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5049345880192794, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.5115555555555557e-05, | |
| "loss": 0.4108, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5095249024558183, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.5071111111111113e-05, | |
| "loss": 0.3925, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5141152168923572, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.5026666666666668e-05, | |
| "loss": 0.4286, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5187055313288961, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.4982222222222223e-05, | |
| "loss": 0.3988, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.523295845765435, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.493777777777778e-05, | |
| "loss": 0.4194, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.5278861602019739, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.4893333333333335e-05, | |
| "loss": 0.4053, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.5324764746385128, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 1.484888888888889e-05, | |
| "loss": 0.3844, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.5370667890750517, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.4804444444444446e-05, | |
| "loss": 0.3897, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.5416571035115906, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.4760000000000001e-05, | |
| "loss": 0.3732, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.5462474179481295, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.4715555555555556e-05, | |
| "loss": 0.3839, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.5508377323846684, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.4671111111111111e-05, | |
| "loss": 0.3831, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5554280468212073, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.4626666666666667e-05, | |
| "loss": 0.3954, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.5600183612577462, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.4582222222222224e-05, | |
| "loss": 0.3804, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.5646086756942851, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.4537777777777779e-05, | |
| "loss": 0.412, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.569198990130824, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.4493333333333334e-05, | |
| "loss": 0.3704, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.5737893045673629, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.444888888888889e-05, | |
| "loss": 0.4001, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5783796190039018, | |
| "grad_norm": 2.625, | |
| "learning_rate": 1.4404444444444445e-05, | |
| "loss": 0.3877, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.5829699334404407, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.4360000000000001e-05, | |
| "loss": 0.3715, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.5875602478769796, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.4315555555555557e-05, | |
| "loss": 0.375, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.5921505623135185, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.4271111111111114e-05, | |
| "loss": 0.4191, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.5967408767500574, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.4226666666666669e-05, | |
| "loss": 0.3581, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6013311911865963, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.4182222222222224e-05, | |
| "loss": 0.4052, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.6059215056231352, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.413777777777778e-05, | |
| "loss": 0.36, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.6105118200596741, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.4093333333333334e-05, | |
| "loss": 0.3925, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.615102134496213, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.404888888888889e-05, | |
| "loss": 0.3638, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.6196924489327519, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.4004444444444445e-05, | |
| "loss": 0.3791, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.6242827633692908, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.396e-05, | |
| "loss": 0.3698, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.6288730778058297, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.3915555555555557e-05, | |
| "loss": 0.3521, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.6334633922423686, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.3871111111111112e-05, | |
| "loss": 0.3822, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.6380537066789075, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.3826666666666668e-05, | |
| "loss": 0.3564, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.6426440211154464, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 1.3782222222222223e-05, | |
| "loss": 0.3876, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.6472343355519853, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.3737777777777778e-05, | |
| "loss": 0.4142, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.6518246499885242, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.3693333333333333e-05, | |
| "loss": 0.3698, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.6564149644250631, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.3648888888888888e-05, | |
| "loss": 0.3479, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.661005278861602, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.3604444444444445e-05, | |
| "loss": 0.3649, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.6655955932981409, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.3560000000000002e-05, | |
| "loss": 0.3677, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.6701859077346798, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.3515555555555558e-05, | |
| "loss": 0.3579, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.6747762221712187, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.3471111111111113e-05, | |
| "loss": 0.3829, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.6793665366077576, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.3426666666666668e-05, | |
| "loss": 0.3927, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.6839568510442965, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.3382222222222223e-05, | |
| "loss": 0.365, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.6885471654808354, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.3337777777777778e-05, | |
| "loss": 0.3987, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6931374799173743, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.3293333333333334e-05, | |
| "loss": 0.3839, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.6977277943539132, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.3248888888888889e-05, | |
| "loss": 0.3668, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.7023181087904522, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.3204444444444446e-05, | |
| "loss": 0.3912, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.7069084232269911, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.3160000000000001e-05, | |
| "loss": 0.3737, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.71149873766353, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.3115555555555556e-05, | |
| "loss": 0.3388, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.7160890521000689, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.3071111111111112e-05, | |
| "loss": 0.3966, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.7206793665366078, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.3026666666666667e-05, | |
| "loss": 0.3868, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.7252696809731467, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.2982222222222222e-05, | |
| "loss": 0.3778, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.7298599954096856, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.2937777777777777e-05, | |
| "loss": 0.3543, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.7344503098462245, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.2893333333333336e-05, | |
| "loss": 0.338, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.7390406242827634, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.2848888888888891e-05, | |
| "loss": 0.4152, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.7436309387193023, | |
| "grad_norm": 2.5, | |
| "learning_rate": 1.2804444444444446e-05, | |
| "loss": 0.3633, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.7482212531558412, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 1.2760000000000001e-05, | |
| "loss": 0.38, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.7528115675923801, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2715555555555557e-05, | |
| "loss": 0.3948, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.757401882028919, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 1.2671111111111112e-05, | |
| "loss": 0.3908, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.7619921964654579, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 1.2626666666666667e-05, | |
| "loss": 0.3543, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.7665825109019968, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2582222222222222e-05, | |
| "loss": 0.3683, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.7711728253385357, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.253777777777778e-05, | |
| "loss": 0.3427, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.7757631397750746, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.2493333333333335e-05, | |
| "loss": 0.3741, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.7803534542116135, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.244888888888889e-05, | |
| "loss": 0.3203, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.7849437686481524, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.2404444444444445e-05, | |
| "loss": 0.337, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.7895340830846913, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.236e-05, | |
| "loss": 0.3734, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.7941243975212302, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.2315555555555555e-05, | |
| "loss": 0.3164, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.7987147119577691, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.227111111111111e-05, | |
| "loss": 0.3596, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.803305026394308, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.2226666666666666e-05, | |
| "loss": 0.3463, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.8078953408308469, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.2182222222222225e-05, | |
| "loss": 0.3292, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.8124856552673858, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.213777777777778e-05, | |
| "loss": 0.3378, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.8170759697039247, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.2093333333333335e-05, | |
| "loss": 0.3668, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.8216662841404636, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.204888888888889e-05, | |
| "loss": 0.3307, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.8262565985770025, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.2004444444444445e-05, | |
| "loss": 0.3472, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.8308469130135414, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.196e-05, | |
| "loss": 0.3581, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.8354372274500803, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.1915555555555556e-05, | |
| "loss": 0.3667, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.8400275418866192, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.1871111111111111e-05, | |
| "loss": 0.3608, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.8446178563231581, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.1826666666666668e-05, | |
| "loss": 0.3794, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.849208170759697, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.1782222222222223e-05, | |
| "loss": 0.3756, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.8537984851962359, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.1737777777777779e-05, | |
| "loss": 0.3656, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.8583887996327748, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1693333333333334e-05, | |
| "loss": 0.3204, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.8629791140693137, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.1648888888888889e-05, | |
| "loss": 0.3786, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.8675694285058526, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.1604444444444444e-05, | |
| "loss": 0.4005, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.8721597429423915, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.156e-05, | |
| "loss": 0.3343, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.8767500573789304, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.1515555555555558e-05, | |
| "loss": 0.3558, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.8813403718154693, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.1471111111111113e-05, | |
| "loss": 0.3644, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.8859306862520082, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.1426666666666669e-05, | |
| "loss": 0.3624, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.8905210006885471, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.1382222222222224e-05, | |
| "loss": 0.3468, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.895111315125086, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.1337777777777779e-05, | |
| "loss": 0.3467, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.8997016295616249, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.1293333333333334e-05, | |
| "loss": 0.3787, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.9042919439981639, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.124888888888889e-05, | |
| "loss": 0.3425, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.9088822584347028, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1204444444444445e-05, | |
| "loss": 0.3394, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.9134725728712417, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.1160000000000002e-05, | |
| "loss": 0.3141, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.9180628873077806, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.1115555555555557e-05, | |
| "loss": 0.3508, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9226532017443195, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.1071111111111112e-05, | |
| "loss": 0.3005, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.9272435161808584, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.1026666666666667e-05, | |
| "loss": 0.3277, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.9318338306173973, | |
| "grad_norm": 2.375, | |
| "learning_rate": 1.0982222222222222e-05, | |
| "loss": 0.326, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.9364241450539362, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.0937777777777778e-05, | |
| "loss": 0.3708, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.9410144594904751, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.0893333333333333e-05, | |
| "loss": 0.3328, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.945604773927014, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.0848888888888888e-05, | |
| "loss": 0.3576, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.9501950883635529, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.0804444444444447e-05, | |
| "loss": 0.2926, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.9547854028000918, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.0760000000000002e-05, | |
| "loss": 0.3302, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.9593757172366307, | |
| "grad_norm": 2.375, | |
| "learning_rate": 1.0715555555555557e-05, | |
| "loss": 0.3317, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.9639660316731696, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.0671111111111112e-05, | |
| "loss": 0.3591, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.9685563461097085, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.0626666666666668e-05, | |
| "loss": 0.3708, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.9731466605462474, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 1.0582222222222223e-05, | |
| "loss": 0.3532, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.9777369749827863, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 1.0537777777777778e-05, | |
| "loss": 0.3749, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.9823272894193252, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.0493333333333333e-05, | |
| "loss": 0.3334, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.9869176038558641, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 1.044888888888889e-05, | |
| "loss": 0.3396, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.991507918292403, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.0404444444444446e-05, | |
| "loss": 0.3528, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.9960982327289419, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.036e-05, | |
| "loss": 0.348, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.000459031443654, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.0315555555555556e-05, | |
| "loss": 0.358, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.0050493458801928, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.0271111111111111e-05, | |
| "loss": 0.2952, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.0096396603167317, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.0226666666666666e-05, | |
| "loss": 0.3138, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.0142299747532706, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.0182222222222222e-05, | |
| "loss": 0.2763, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.0188202891898095, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.013777777777778e-05, | |
| "loss": 0.2877, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.0234106036263484, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 1.0093333333333336e-05, | |
| "loss": 0.3352, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.0280009180628873, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.004888888888889e-05, | |
| "loss": 0.2946, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.0325912324994262, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.0004444444444446e-05, | |
| "loss": 0.3117, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.037181546935965, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.2875, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.041771861372504, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 9.915555555555556e-06, | |
| "loss": 0.3025, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.046362175809043, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 9.871111111111112e-06, | |
| "loss": 0.2851, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.0509524902455818, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 9.826666666666667e-06, | |
| "loss": 0.2921, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.0555428046821207, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 9.782222222222222e-06, | |
| "loss": 0.2863, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.0601331191186596, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 9.737777777777779e-06, | |
| "loss": 0.3047, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.0647234335551985, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.693333333333334e-06, | |
| "loss": 0.3034, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.0693137479917374, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 9.64888888888889e-06, | |
| "loss": 0.3071, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.0739040624282763, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 9.604444444444445e-06, | |
| "loss": 0.2733, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.0784943768648152, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 9.56e-06, | |
| "loss": 0.2914, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.083084691301354, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.515555555555557e-06, | |
| "loss": 0.3074, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.087675005737893, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 9.471111111111112e-06, | |
| "loss": 0.3069, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.092265320174432, | |
| "grad_norm": 1.875, | |
| "learning_rate": 9.426666666666667e-06, | |
| "loss": 0.3117, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.0968556346109708, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 9.382222222222223e-06, | |
| "loss": 0.3178, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.1014459490475097, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.33777777777778e-06, | |
| "loss": 0.3124, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.1060362634840486, | |
| "grad_norm": 1.75, | |
| "learning_rate": 9.293333333333335e-06, | |
| "loss": 0.2702, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.1106265779205875, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 9.24888888888889e-06, | |
| "loss": 0.2875, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.1152168923571264, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 9.204444444444445e-06, | |
| "loss": 0.2989, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.1198072067936653, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.16e-06, | |
| "loss": 0.2875, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.1243975212302042, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.115555555555556e-06, | |
| "loss": 0.3044, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.128987835666743, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 9.07111111111111e-06, | |
| "loss": 0.2912, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.133578150103282, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.026666666666666e-06, | |
| "loss": 0.2751, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.138168464539821, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.982222222222223e-06, | |
| "loss": 0.2746, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.1427587789763598, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 8.937777777777778e-06, | |
| "loss": 0.2821, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.1473490934128987, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.893333333333333e-06, | |
| "loss": 0.317, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.1519394078494376, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 8.848888888888889e-06, | |
| "loss": 0.3086, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.1565297222859765, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 8.804444444444446e-06, | |
| "loss": 0.3134, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.1611200367225154, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 8.76e-06, | |
| "loss": 0.3023, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.1657103511590543, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 8.715555555555556e-06, | |
| "loss": 0.2862, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.1703006655955932, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 8.671111111111113e-06, | |
| "loss": 0.2734, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.174890980032132, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 8.626666666666668e-06, | |
| "loss": 0.3035, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.1794812944686712, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 8.582222222222223e-06, | |
| "loss": 0.2929, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.1840716089052101, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.537777777777779e-06, | |
| "loss": 0.2805, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.188661923341749, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 8.493333333333334e-06, | |
| "loss": 0.287, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.193252237778288, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 8.448888888888889e-06, | |
| "loss": 0.3009, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.1978425522148268, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 8.404444444444444e-06, | |
| "loss": 0.29, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.2024328666513657, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.36e-06, | |
| "loss": 0.2578, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.2070231810879046, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 8.315555555555557e-06, | |
| "loss": 0.2527, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.2116134955244435, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.271111111111112e-06, | |
| "loss": 0.277, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.2162038099609824, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 8.226666666666667e-06, | |
| "loss": 0.2937, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.2207941243975213, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 8.182222222222222e-06, | |
| "loss": 0.2902, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.2253844388340602, | |
| "grad_norm": 2.25, | |
| "learning_rate": 8.137777777777779e-06, | |
| "loss": 0.2891, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.2299747532705991, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 8.093333333333334e-06, | |
| "loss": 0.2994, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 1.234565067707138, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 8.04888888888889e-06, | |
| "loss": 0.2704, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 1.239155382143677, | |
| "grad_norm": 2.5, | |
| "learning_rate": 8.004444444444445e-06, | |
| "loss": 0.2715, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.2437456965802158, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 0.3063, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 1.2483360110167547, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 7.915555555555557e-06, | |
| "loss": 0.2981, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 1.2529263254532936, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 7.871111111111112e-06, | |
| "loss": 0.2597, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 1.2575166398898325, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.826666666666667e-06, | |
| "loss": 0.2594, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 1.2621069543263714, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 7.782222222222223e-06, | |
| "loss": 0.2842, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.2666972687629103, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.737777777777778e-06, | |
| "loss": 0.2955, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 1.2712875831994492, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 7.693333333333333e-06, | |
| "loss": 0.2963, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 1.2758778976359881, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 7.648888888888888e-06, | |
| "loss": 0.3119, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 1.280468212072527, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 7.604444444444445e-06, | |
| "loss": 0.272, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 1.285058526509066, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.5600000000000005e-06, | |
| "loss": 0.2884, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.2896488409456048, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 7.515555555555556e-06, | |
| "loss": 0.3103, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 1.2942391553821437, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 7.471111111111111e-06, | |
| "loss": 0.2716, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 1.2988294698186826, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 7.426666666666668e-06, | |
| "loss": 0.2636, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 1.3034197842552215, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 7.382222222222223e-06, | |
| "loss": 0.2621, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 1.3080100986917604, | |
| "grad_norm": 2.0, | |
| "learning_rate": 7.337777777777778e-06, | |
| "loss": 0.2793, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.3126004131282993, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 7.2933333333333335e-06, | |
| "loss": 0.3049, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 1.3171907275648382, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 7.24888888888889e-06, | |
| "loss": 0.3035, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 1.3217810420013771, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.204444444444445e-06, | |
| "loss": 0.306, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 1.326371356437916, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 7.16e-06, | |
| "loss": 0.2964, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 1.330961670874455, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 7.115555555555557e-06, | |
| "loss": 0.2747, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.3355519853109938, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 7.071111111111112e-06, | |
| "loss": 0.2819, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 1.3401422997475327, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 7.0266666666666674e-06, | |
| "loss": 0.302, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 1.3447326141840716, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 6.982222222222223e-06, | |
| "loss": 0.2876, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 1.3493229286206105, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 6.937777777777779e-06, | |
| "loss": 0.2868, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.3539132430571494, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 6.893333333333334e-06, | |
| "loss": 0.3028, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.3585035574936883, | |
| "grad_norm": 1.75, | |
| "learning_rate": 6.848888888888889e-06, | |
| "loss": 0.2412, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 1.3630938719302272, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 6.8044444444444444e-06, | |
| "loss": 0.2717, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 1.3676841863667661, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.760000000000001e-06, | |
| "loss": 0.3104, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 1.372274500803305, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 6.7155555555555566e-06, | |
| "loss": 0.2707, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 1.376864815239844, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 6.671111111111112e-06, | |
| "loss": 0.2663, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.3814551296763828, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 6.626666666666667e-06, | |
| "loss": 0.2791, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 1.3860454441129217, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 6.582222222222223e-06, | |
| "loss": 0.3027, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 1.3906357585494606, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 6.537777777777778e-06, | |
| "loss": 0.2836, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 1.3952260729859995, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 6.4933333333333336e-06, | |
| "loss": 0.2509, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 1.3998163874225384, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 6.448888888888889e-06, | |
| "loss": 0.2916, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.4044067018590773, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 6.404444444444446e-06, | |
| "loss": 0.3131, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 1.4089970162956162, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 6.360000000000001e-06, | |
| "loss": 0.2802, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 1.4135873307321551, | |
| "grad_norm": 2.25, | |
| "learning_rate": 6.315555555555556e-06, | |
| "loss": 0.281, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 1.418177645168694, | |
| "grad_norm": 2.0, | |
| "learning_rate": 6.271111111111111e-06, | |
| "loss": 0.2605, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 1.422767959605233, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 6.2266666666666675e-06, | |
| "loss": 0.2643, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.4273582740417718, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 6.182222222222223e-06, | |
| "loss": 0.2783, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 1.4319485884783107, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 6.137777777777778e-06, | |
| "loss": 0.3186, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 1.4365389029148496, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 6.093333333333333e-06, | |
| "loss": 0.2773, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 1.4411292173513885, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 6.04888888888889e-06, | |
| "loss": 0.2642, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 1.4457195317879274, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.004444444444445e-06, | |
| "loss": 0.2483, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 1.4503098462244663, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 0.2518, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 1.4549001606610052, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 5.915555555555556e-06, | |
| "loss": 0.2773, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 1.4594904750975441, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 5.871111111111112e-06, | |
| "loss": 0.2938, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 1.464080789534083, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 5.826666666666667e-06, | |
| "loss": 0.2725, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 1.468671103970622, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.782222222222222e-06, | |
| "loss": 0.2776, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.4732614184071608, | |
| "grad_norm": 2.375, | |
| "learning_rate": 5.737777777777778e-06, | |
| "loss": 0.2751, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 1.4778517328436998, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 5.6933333333333344e-06, | |
| "loss": 0.3097, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 1.4824420472802387, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 5.64888888888889e-06, | |
| "loss": 0.2948, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 1.4870323617167776, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.604444444444445e-06, | |
| "loss": 0.2748, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 1.4916226761533165, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 5.560000000000001e-06, | |
| "loss": 0.2916, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 1.4962129905898554, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 5.515555555555556e-06, | |
| "loss": 0.2835, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 1.5008033050263943, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.4711111111111114e-06, | |
| "loss": 0.2733, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 1.5053936194629332, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 5.426666666666667e-06, | |
| "loss": 0.2662, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 1.509983933899472, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 5.382222222222223e-06, | |
| "loss": 0.2976, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 1.514574248336011, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 5.337777777777779e-06, | |
| "loss": 0.3028, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.5191645627725499, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.293333333333334e-06, | |
| "loss": 0.2738, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 1.5237548772090888, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 5.248888888888889e-06, | |
| "loss": 0.2685, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 1.5283451916456277, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 5.204444444444445e-06, | |
| "loss": 0.2644, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 1.5329355060821666, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 5.1600000000000006e-06, | |
| "loss": 0.2839, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 1.5375258205187055, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.115555555555556e-06, | |
| "loss": 0.2954, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 1.5421161349552444, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 5.071111111111111e-06, | |
| "loss": 0.2565, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 1.5467064493917833, | |
| "grad_norm": 2.5, | |
| "learning_rate": 5.026666666666667e-06, | |
| "loss": 0.2875, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 1.5512967638283222, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.982222222222222e-06, | |
| "loss": 0.2957, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 1.555887078264861, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.937777777777778e-06, | |
| "loss": 0.2848, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 1.5604773927014, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 4.893333333333334e-06, | |
| "loss": 0.2673, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.5650677071379389, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.848888888888889e-06, | |
| "loss": 0.2899, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 1.5696580215744778, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.804444444444445e-06, | |
| "loss": 0.259, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 1.5742483360110167, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 4.76e-06, | |
| "loss": 0.2894, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 1.5788386504475556, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.715555555555556e-06, | |
| "loss": 0.2805, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 1.5834289648840945, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.6711111111111115e-06, | |
| "loss": 0.2924, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.5880192793206334, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 4.626666666666667e-06, | |
| "loss": 0.2903, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 1.5926095937571723, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 4.582222222222223e-06, | |
| "loss": 0.2842, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 1.5971999081937112, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 4.537777777777778e-06, | |
| "loss": 0.299, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 1.60179022263025, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 4.493333333333333e-06, | |
| "loss": 0.2871, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 1.606380537066789, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.448888888888889e-06, | |
| "loss": 0.288, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6109708515033279, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 4.404444444444445e-06, | |
| "loss": 0.2686, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 1.6155611659398668, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.360000000000001e-06, | |
| "loss": 0.267, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 1.6201514803764057, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 4.315555555555556e-06, | |
| "loss": 0.2832, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 1.6247417948129446, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 4.271111111111111e-06, | |
| "loss": 0.2552, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 1.6293321092494835, | |
| "grad_norm": 2.25, | |
| "learning_rate": 4.226666666666667e-06, | |
| "loss": 0.2728, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.6339224236860224, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.182222222222222e-06, | |
| "loss": 0.2701, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 1.6385127381225613, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 4.1377777777777784e-06, | |
| "loss": 0.2717, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 1.6431030525591002, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.093333333333334e-06, | |
| "loss": 0.2659, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 1.647693366995639, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 4.04888888888889e-06, | |
| "loss": 0.2699, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 1.652283681432178, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.004444444444445e-06, | |
| "loss": 0.2894, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.6568739958687169, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 3.96e-06, | |
| "loss": 0.2702, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 1.6614643103052558, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 3.9155555555555554e-06, | |
| "loss": 0.2807, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 1.6660546247417947, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.8711111111111115e-06, | |
| "loss": 0.2831, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 1.6706449391783336, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 3.826666666666667e-06, | |
| "loss": 0.2691, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 1.6752352536148725, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 3.782222222222223e-06, | |
| "loss": 0.2811, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.6798255680514114, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 3.737777777777778e-06, | |
| "loss": 0.2775, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 1.6844158824879503, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.6933333333333337e-06, | |
| "loss": 0.3138, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 1.6890061969244892, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 3.648888888888889e-06, | |
| "loss": 0.2453, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 1.693596511361028, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 3.604444444444445e-06, | |
| "loss": 0.2495, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 1.698186825797567, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 3.5600000000000002e-06, | |
| "loss": 0.3126, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.702777140234106, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 3.515555555555556e-06, | |
| "loss": 0.2937, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 1.707367454670645, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 3.471111111111111e-06, | |
| "loss": 0.2764, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 1.711957769107184, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.426666666666667e-06, | |
| "loss": 0.2783, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 1.7165480835437228, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 3.3822222222222224e-06, | |
| "loss": 0.2514, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 1.7211383979802617, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 3.337777777777778e-06, | |
| "loss": 0.3022, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.7257287124168006, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.2933333333333333e-06, | |
| "loss": 0.275, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 1.7303190268533395, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 3.2488888888888894e-06, | |
| "loss": 0.2586, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 1.7349093412898784, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.2044444444444446e-06, | |
| "loss": 0.2941, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 1.7394996557264173, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 3.1600000000000002e-06, | |
| "loss": 0.2618, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 1.7440899701629562, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.1155555555555555e-06, | |
| "loss": 0.2761, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.748680284599495, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 3.0711111111111115e-06, | |
| "loss": 0.2831, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 1.753270599036034, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.0266666666666668e-06, | |
| "loss": 0.2872, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 1.757860913472573, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 2.9822222222222224e-06, | |
| "loss": 0.273, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 1.7624512279091118, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 2.937777777777778e-06, | |
| "loss": 0.2611, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 1.7670415423456507, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 2.8933333333333337e-06, | |
| "loss": 0.2725, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.7716318567821896, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 2.8488888888888894e-06, | |
| "loss": 0.2715, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 1.7762221712187285, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 2.8044444444444446e-06, | |
| "loss": 0.3087, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 1.7808124856552674, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 2.7600000000000003e-06, | |
| "loss": 0.2967, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 1.7854028000918063, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 2.715555555555556e-06, | |
| "loss": 0.2605, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 1.7899931145283452, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 2.6711111111111116e-06, | |
| "loss": 0.2592, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.794583428964884, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 2.6266666666666668e-06, | |
| "loss": 0.2519, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 1.799173743401423, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 2.5822222222222224e-06, | |
| "loss": 0.2945, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 1.803764057837962, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.537777777777778e-06, | |
| "loss": 0.297, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 1.8083543722745008, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.4933333333333333e-06, | |
| "loss": 0.3222, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 1.8129446867110397, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 2.448888888888889e-06, | |
| "loss": 0.249, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.8175350011475786, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 2.4044444444444446e-06, | |
| "loss": 0.2789, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 1.8221253155841175, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.3600000000000003e-06, | |
| "loss": 0.2713, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 1.8267156300206564, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.3155555555555555e-06, | |
| "loss": 0.2693, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 1.8313059444571953, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 2.2711111111111116e-06, | |
| "loss": 0.2679, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 1.8358962588937342, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.226666666666667e-06, | |
| "loss": 0.2764, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 4500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.5989514663232225e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |