| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 3000000, | |
| "global_step": 2225, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0044943820224719105, | |
| "grad_norm": 21.028866393310857, | |
| "learning_rate": 4.484304932735426e-07, | |
| "loss": 1.8157, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008988764044943821, | |
| "grad_norm": 18.149495621801602, | |
| "learning_rate": 8.968609865470852e-07, | |
| "loss": 1.8095, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01348314606741573, | |
| "grad_norm": 7.110153316235121, | |
| "learning_rate": 1.345291479820628e-06, | |
| "loss": 1.4583, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.017977528089887642, | |
| "grad_norm": 3.8403679133270168, | |
| "learning_rate": 1.7937219730941704e-06, | |
| "loss": 1.2703, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02247191011235955, | |
| "grad_norm": 3.1058777831353708, | |
| "learning_rate": 2.242152466367713e-06, | |
| "loss": 1.165, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02696629213483146, | |
| "grad_norm": 3.3684279642867847, | |
| "learning_rate": 2.690582959641256e-06, | |
| "loss": 1.0807, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03146067415730337, | |
| "grad_norm": 3.157236299420834, | |
| "learning_rate": 3.1390134529147986e-06, | |
| "loss": 1.0534, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.035955056179775284, | |
| "grad_norm": 3.5078357528479485, | |
| "learning_rate": 3.587443946188341e-06, | |
| "loss": 1.0051, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04044943820224719, | |
| "grad_norm": 3.4693923350197293, | |
| "learning_rate": 4.0358744394618836e-06, | |
| "loss": 0.9734, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0449438202247191, | |
| "grad_norm": 3.1682184152102453, | |
| "learning_rate": 4.484304932735426e-06, | |
| "loss": 0.9764, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04943820224719101, | |
| "grad_norm": 3.4044270322944357, | |
| "learning_rate": 4.932735426008969e-06, | |
| "loss": 0.9358, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05393258426966292, | |
| "grad_norm": 3.629261207410094, | |
| "learning_rate": 5.381165919282512e-06, | |
| "loss": 0.975, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.058426966292134834, | |
| "grad_norm": 2.7702327333157615, | |
| "learning_rate": 5.8295964125560544e-06, | |
| "loss": 0.9692, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06292134831460675, | |
| "grad_norm": 3.5659770341558144, | |
| "learning_rate": 6.278026905829597e-06, | |
| "loss": 0.9665, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06741573033707865, | |
| "grad_norm": 3.0158598967423944, | |
| "learning_rate": 6.72645739910314e-06, | |
| "loss": 0.9409, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07191011235955057, | |
| "grad_norm": 3.6116843848118427, | |
| "learning_rate": 7.174887892376682e-06, | |
| "loss": 0.9369, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.07640449438202247, | |
| "grad_norm": 3.397427112776282, | |
| "learning_rate": 7.6233183856502244e-06, | |
| "loss": 0.9143, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08089887640449438, | |
| "grad_norm": 3.1019973911170857, | |
| "learning_rate": 8.071748878923767e-06, | |
| "loss": 0.9262, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.0853932584269663, | |
| "grad_norm": 3.28653951390302, | |
| "learning_rate": 8.52017937219731e-06, | |
| "loss": 0.9151, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.0898876404494382, | |
| "grad_norm": 2.9229201820388826, | |
| "learning_rate": 8.968609865470853e-06, | |
| "loss": 0.9093, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09438202247191012, | |
| "grad_norm": 3.113906403159331, | |
| "learning_rate": 9.417040358744395e-06, | |
| "loss": 0.9199, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.09887640449438202, | |
| "grad_norm": 3.2663540679465624, | |
| "learning_rate": 9.865470852017938e-06, | |
| "loss": 0.9305, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.10337078651685393, | |
| "grad_norm": 3.051806061363382, | |
| "learning_rate": 9.999698350006063e-06, | |
| "loss": 0.9105, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.10786516853932585, | |
| "grad_norm": 3.2656274646398753, | |
| "learning_rate": 9.998220968267148e-06, | |
| "loss": 0.9234, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.11235955056179775, | |
| "grad_norm": 3.0943126322747556, | |
| "learning_rate": 9.995512813021446e-06, | |
| "loss": 0.9278, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.11685393258426967, | |
| "grad_norm": 3.1197182560800294, | |
| "learning_rate": 9.991574551131379e-06, | |
| "loss": 0.9055, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.12134831460674157, | |
| "grad_norm": 2.788031040285689, | |
| "learning_rate": 9.986407152363672e-06, | |
| "loss": 0.9106, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.1258426966292135, | |
| "grad_norm": 3.4226340314131485, | |
| "learning_rate": 9.980011889150569e-06, | |
| "loss": 0.9111, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1303370786516854, | |
| "grad_norm": 3.3566995644063504, | |
| "learning_rate": 9.97239033627649e-06, | |
| "loss": 0.915, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1348314606741573, | |
| "grad_norm": 3.4076936734585543, | |
| "learning_rate": 9.96354437049027e-06, | |
| "loss": 0.8995, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1393258426966292, | |
| "grad_norm": 3.209280832093268, | |
| "learning_rate": 9.953476170043011e-06, | |
| "loss": 0.8969, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.14382022471910114, | |
| "grad_norm": 3.009956911416, | |
| "learning_rate": 9.942188214151703e-06, | |
| "loss": 0.8791, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.14831460674157304, | |
| "grad_norm": 3.318098809004364, | |
| "learning_rate": 9.929683282388753e-06, | |
| "loss": 0.9135, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.15280898876404495, | |
| "grad_norm": 3.316896081979771, | |
| "learning_rate": 9.915964453997516e-06, | |
| "loss": 0.8809, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.15730337078651685, | |
| "grad_norm": 2.570617534902647, | |
| "learning_rate": 9.901035107134067e-06, | |
| "loss": 0.8975, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.16179775280898875, | |
| "grad_norm": 2.5338700266262766, | |
| "learning_rate": 9.884898918035354e-06, | |
| "loss": 0.8997, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.1662921348314607, | |
| "grad_norm": 2.775240802278996, | |
| "learning_rate": 9.86755986011395e-06, | |
| "loss": 0.8984, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1707865168539326, | |
| "grad_norm": 2.822348053524811, | |
| "learning_rate": 9.849022202979635e-06, | |
| "loss": 0.8993, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.1752808988764045, | |
| "grad_norm": 3.055973621744187, | |
| "learning_rate": 9.829290511388033e-06, | |
| "loss": 0.8883, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.1797752808988764, | |
| "grad_norm": 3.1358574509096155, | |
| "learning_rate": 9.808369644116578e-06, | |
| "loss": 0.8855, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1842696629213483, | |
| "grad_norm": 2.7698413336711285, | |
| "learning_rate": 9.78626475276808e-06, | |
| "loss": 0.9028, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.18876404494382024, | |
| "grad_norm": 2.693223915430037, | |
| "learning_rate": 9.762981280502178e-06, | |
| "loss": 0.8918, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.19325842696629214, | |
| "grad_norm": 3.008257840198355, | |
| "learning_rate": 9.738524960695003e-06, | |
| "loss": 0.8804, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.19775280898876405, | |
| "grad_norm": 2.9264871287794336, | |
| "learning_rate": 9.712901815527387e-06, | |
| "loss": 0.8781, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.20224719101123595, | |
| "grad_norm": 3.121215777703405, | |
| "learning_rate": 9.686118154501938e-06, | |
| "loss": 0.8856, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.20674157303370785, | |
| "grad_norm": 3.0456191477056738, | |
| "learning_rate": 9.658180572889372e-06, | |
| "loss": 0.9005, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.21123595505617979, | |
| "grad_norm": 2.6847965693331837, | |
| "learning_rate": 9.62909595010449e-06, | |
| "loss": 0.8902, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2157303370786517, | |
| "grad_norm": 3.16590640260948, | |
| "learning_rate": 9.598871448012166e-06, | |
| "loss": 0.8872, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2202247191011236, | |
| "grad_norm": 2.917831346210603, | |
| "learning_rate": 9.567514509163791e-06, | |
| "loss": 0.8791, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.2247191011235955, | |
| "grad_norm": 2.395460178627161, | |
| "learning_rate": 9.535032854964608e-06, | |
| "loss": 0.8556, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2292134831460674, | |
| "grad_norm": 2.8755215373997465, | |
| "learning_rate": 9.501434483772371e-06, | |
| "loss": 0.8879, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.23370786516853934, | |
| "grad_norm": 2.7839596342446016, | |
| "learning_rate": 9.466727668927817e-06, | |
| "loss": 0.8921, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.23820224719101124, | |
| "grad_norm": 2.8101352476646113, | |
| "learning_rate": 9.43092095671741e-06, | |
| "loss": 0.8719, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.24269662921348314, | |
| "grad_norm": 2.8841472788021756, | |
| "learning_rate": 9.39402316426889e-06, | |
| "loss": 0.8736, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.24719101123595505, | |
| "grad_norm": 2.8954194572601444, | |
| "learning_rate": 9.356043377380126e-06, | |
| "loss": 0.8987, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.251685393258427, | |
| "grad_norm": 3.554378286149293, | |
| "learning_rate": 9.316990948281799e-06, | |
| "loss": 0.8514, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.25617977528089886, | |
| "grad_norm": 2.3570345682703353, | |
| "learning_rate": 9.27687549333449e-06, | |
| "loss": 0.8627, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2606741573033708, | |
| "grad_norm": 2.8727271733313953, | |
| "learning_rate": 9.235706890660735e-06, | |
| "loss": 0.8372, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2651685393258427, | |
| "grad_norm": 2.838666672391731, | |
| "learning_rate": 9.193495277712593e-06, | |
| "loss": 0.8639, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2696629213483146, | |
| "grad_norm": 2.7577407799514613, | |
| "learning_rate": 9.150251048775403e-06, | |
| "loss": 0.8675, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.27415730337078653, | |
| "grad_norm": 3.5011471176346243, | |
| "learning_rate": 9.105984852408247e-06, | |
| "loss": 0.9128, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.2786516853932584, | |
| "grad_norm": 2.4709369938165353, | |
| "learning_rate": 9.060707588821835e-06, | |
| "loss": 0.8481, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.28314606741573034, | |
| "grad_norm": 2.687018184877083, | |
| "learning_rate": 9.014430407194413e-06, | |
| "loss": 0.8665, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2876404494382023, | |
| "grad_norm": 2.8113496819372195, | |
| "learning_rate": 8.967164702926359e-06, | |
| "loss": 0.8705, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.29213483146067415, | |
| "grad_norm": 3.098175777433656, | |
| "learning_rate": 8.918922114834156e-06, | |
| "loss": 0.8594, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.2966292134831461, | |
| "grad_norm": 2.5305855672874586, | |
| "learning_rate": 8.869714522284424e-06, | |
| "loss": 0.9048, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.30112359550561796, | |
| "grad_norm": 3.1823485712315587, | |
| "learning_rate": 8.819554042268715e-06, | |
| "loss": 0.8581, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.3056179775280899, | |
| "grad_norm": 3.141603475434689, | |
| "learning_rate": 8.768453026419793e-06, | |
| "loss": 0.8924, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3101123595505618, | |
| "grad_norm": 2.8298252741410206, | |
| "learning_rate": 8.71642405797015e-06, | |
| "loss": 0.8442, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.3146067415730337, | |
| "grad_norm": 2.4651915483445284, | |
| "learning_rate": 8.663479948653472e-06, | |
| "loss": 0.8492, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.31910112359550563, | |
| "grad_norm": 2.7122168547316394, | |
| "learning_rate": 8.609633735549852e-06, | |
| "loss": 0.8333, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3235955056179775, | |
| "grad_norm": 2.9065395439540427, | |
| "learning_rate": 8.554898677875509e-06, | |
| "loss": 0.8361, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.32808988764044944, | |
| "grad_norm": 2.541328327536001, | |
| "learning_rate": 8.49928825371781e-06, | |
| "loss": 0.8614, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.3325842696629214, | |
| "grad_norm": 2.8452183525450985, | |
| "learning_rate": 8.442816156716386e-06, | |
| "loss": 0.8398, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.33707865168539325, | |
| "grad_norm": 2.6414675162195773, | |
| "learning_rate": 8.385496292691192e-06, | |
| "loss": 0.8525, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3415730337078652, | |
| "grad_norm": 2.780280291371649, | |
| "learning_rate": 8.32734277621829e-06, | |
| "loss": 0.8707, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.34606741573033706, | |
| "grad_norm": 2.601178957016528, | |
| "learning_rate": 8.26836992715426e-06, | |
| "loss": 0.8806, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.350561797752809, | |
| "grad_norm": 2.962305315436321, | |
| "learning_rate": 8.208592267110033e-06, | |
| "loss": 0.8614, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.3550561797752809, | |
| "grad_norm": 2.793835722937421, | |
| "learning_rate": 8.148024515875057e-06, | |
| "loss": 0.8598, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.3595505617977528, | |
| "grad_norm": 2.6570500425410786, | |
| "learning_rate": 8.08668158779267e-06, | |
| "loss": 0.8723, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.36404494382022473, | |
| "grad_norm": 3.006706054601333, | |
| "learning_rate": 8.024578588087548e-06, | |
| "loss": 0.8451, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.3685393258426966, | |
| "grad_norm": 3.194741458838549, | |
| "learning_rate": 7.961730809146162e-06, | |
| "loss": 0.8445, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.37303370786516854, | |
| "grad_norm": 2.6452523909592873, | |
| "learning_rate": 7.898153726751155e-06, | |
| "loss": 0.8523, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.3775280898876405, | |
| "grad_norm": 2.5344015025695534, | |
| "learning_rate": 7.833862996270534e-06, | |
| "loss": 0.8274, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.38202247191011235, | |
| "grad_norm": 2.7320847056126456, | |
| "learning_rate": 7.768874448802665e-06, | |
| "loss": 0.8425, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.3865168539325843, | |
| "grad_norm": 2.6938059359159077, | |
| "learning_rate": 7.703204087277989e-06, | |
| "loss": 0.8226, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.39101123595505616, | |
| "grad_norm": 2.8434749905260035, | |
| "learning_rate": 7.636868082518413e-06, | |
| "loss": 0.8647, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.3955056179775281, | |
| "grad_norm": 2.6787607584691937, | |
| "learning_rate": 7.5698827692553785e-06, | |
| "loss": 0.8819, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.3474923030984285, | |
| "learning_rate": 7.502264642107559e-06, | |
| "loss": 0.828, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.4044943820224719, | |
| "grad_norm": 2.608043188896359, | |
| "learning_rate": 7.434030351519183e-06, | |
| "loss": 0.8424, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.40898876404494383, | |
| "grad_norm": 2.31822219138967, | |
| "learning_rate": 7.365196699659989e-06, | |
| "loss": 0.8627, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.4134831460674157, | |
| "grad_norm": 2.503238393535594, | |
| "learning_rate": 7.295780636287828e-06, | |
| "loss": 0.8374, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.41797752808988764, | |
| "grad_norm": 2.987882942685997, | |
| "learning_rate": 7.2257992545749045e-06, | |
| "loss": 0.8397, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.42247191011235957, | |
| "grad_norm": 3.0751658643780675, | |
| "learning_rate": 7.155269786898728e-06, | |
| "loss": 0.8419, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.42696629213483145, | |
| "grad_norm": 3.0944989985866798, | |
| "learning_rate": 7.084209600598757e-06, | |
| "loss": 0.847, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.4314606741573034, | |
| "grad_norm": 2.7171358523992026, | |
| "learning_rate": 7.012636193699838e-06, | |
| "loss": 0.8442, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.43595505617977526, | |
| "grad_norm": 2.9449405004821823, | |
| "learning_rate": 6.9405671906034465e-06, | |
| "loss": 0.7978, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.4404494382022472, | |
| "grad_norm": 2.5697814837358024, | |
| "learning_rate": 6.868020337747812e-06, | |
| "loss": 0.8415, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.4449438202247191, | |
| "grad_norm": 2.7123982325774976, | |
| "learning_rate": 6.7950134992379935e-06, | |
| "loss": 0.8368, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.449438202247191, | |
| "grad_norm": 2.7457304327214547, | |
| "learning_rate": 6.721564652446987e-06, | |
| "loss": 0.8244, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.45393258426966293, | |
| "grad_norm": 2.5727374251550574, | |
| "learning_rate": 6.647691883588928e-06, | |
| "loss": 0.8536, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.4584269662921348, | |
| "grad_norm": 3.436622027549439, | |
| "learning_rate": 6.573413383265505e-06, | |
| "loss": 0.8282, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.46292134831460674, | |
| "grad_norm": 2.554581349417234, | |
| "learning_rate": 6.4987474419866635e-06, | |
| "loss": 0.8785, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.46741573033707867, | |
| "grad_norm": 2.572629251373284, | |
| "learning_rate": 6.423712445666703e-06, | |
| "loss": 0.8404, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.47191011235955055, | |
| "grad_norm": 2.6110629987997296, | |
| "learning_rate": 6.348326871096895e-06, | |
| "loss": 0.8335, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.4764044943820225, | |
| "grad_norm": 2.579522074041865, | |
| "learning_rate": 6.272609281395708e-06, | |
| "loss": 0.814, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.48089887640449436, | |
| "grad_norm": 2.604436458085443, | |
| "learning_rate": 6.1965783214377895e-06, | |
| "loss": 0.8356, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.4853932584269663, | |
| "grad_norm": 3.544439618834328, | |
| "learning_rate": 6.120252713262801e-06, | |
| "loss": 0.8145, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.4898876404494382, | |
| "grad_norm": 2.5461314452599044, | |
| "learning_rate": 6.043651251465266e-06, | |
| "loss": 0.8149, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.4943820224719101, | |
| "grad_norm": 2.684504023450718, | |
| "learning_rate": 5.9667927985665385e-06, | |
| "loss": 0.8336, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.49887640449438203, | |
| "grad_norm": 2.624225359715667, | |
| "learning_rate": 5.889696280370063e-06, | |
| "loss": 0.8193, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.503370786516854, | |
| "grad_norm": 2.6410937801065564, | |
| "learning_rate": 5.812380681301031e-06, | |
| "loss": 0.8303, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5078651685393258, | |
| "grad_norm": 2.5558280002336327, | |
| "learning_rate": 5.734865039731612e-06, | |
| "loss": 0.8522, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.5123595505617977, | |
| "grad_norm": 3.205793969640316, | |
| "learning_rate": 5.657168443292909e-06, | |
| "loss": 0.8246, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.5168539325842697, | |
| "grad_norm": 2.5584552525736317, | |
| "learning_rate": 5.57931002417477e-06, | |
| "loss": 0.8527, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.5213483146067416, | |
| "grad_norm": 2.4712332329098974, | |
| "learning_rate": 5.501308954414632e-06, | |
| "loss": 0.8416, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.5258426966292135, | |
| "grad_norm": 2.650810844110697, | |
| "learning_rate": 5.42318444117656e-06, | |
| "loss": 0.8456, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.5303370786516854, | |
| "grad_norm": 2.326300138355385, | |
| "learning_rate": 5.3449557220216245e-06, | |
| "loss": 0.8504, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.5348314606741573, | |
| "grad_norm": 2.2706355883101503, | |
| "learning_rate": 5.2666420601708036e-06, | |
| "loss": 0.8422, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.5393258426966292, | |
| "grad_norm": 3.108873698113913, | |
| "learning_rate": 5.188262739761562e-06, | |
| "loss": 0.8173, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5438202247191011, | |
| "grad_norm": 3.0375290855665686, | |
| "learning_rate": 5.109837061099274e-06, | |
| "loss": 0.8352, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.5483146067415731, | |
| "grad_norm": 3.015315272637417, | |
| "learning_rate": 5.0313843359046865e-06, | |
| "loss": 0.7971, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.5528089887640449, | |
| "grad_norm": 2.8103059408990525, | |
| "learning_rate": 4.9529238825585425e-06, | |
| "loss": 0.8014, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.5573033707865168, | |
| "grad_norm": 2.968627530793131, | |
| "learning_rate": 4.874475021344583e-06, | |
| "loss": 0.8299, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.5617977528089888, | |
| "grad_norm": 3.1399655089271343, | |
| "learning_rate": 4.796057069692073e-06, | |
| "loss": 0.8472, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5662921348314607, | |
| "grad_norm": 2.664176882967019, | |
| "learning_rate": 4.717689337419043e-06, | |
| "loss": 0.8451, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.5707865168539326, | |
| "grad_norm": 2.7440588116901243, | |
| "learning_rate": 4.639391121977378e-06, | |
| "loss": 0.8319, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.5752808988764045, | |
| "grad_norm": 2.8756872704696206, | |
| "learning_rate": 4.561181703700986e-06, | |
| "loss": 0.8137, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.5797752808988764, | |
| "grad_norm": 2.20156098848189, | |
| "learning_rate": 4.4830803410581506e-06, | |
| "loss": 0.8266, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.5842696629213483, | |
| "grad_norm": 2.215449378595676, | |
| "learning_rate": 4.405106265909284e-06, | |
| "loss": 0.7933, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.5887640449438202, | |
| "grad_norm": 2.7246034503185173, | |
| "learning_rate": 4.327278678771237e-06, | |
| "loss": 0.8249, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.5932584269662922, | |
| "grad_norm": 2.7460406192905746, | |
| "learning_rate": 4.2496167440893075e-06, | |
| "loss": 0.8177, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.597752808988764, | |
| "grad_norm": 2.968842648457339, | |
| "learning_rate": 4.172139585518156e-06, | |
| "loss": 0.8424, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.6022471910112359, | |
| "grad_norm": 2.504248019881339, | |
| "learning_rate": 4.094866281212735e-06, | |
| "loss": 0.8365, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.6067415730337079, | |
| "grad_norm": 2.587216532975083, | |
| "learning_rate": 4.017815859130462e-06, | |
| "loss": 0.8157, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.6112359550561798, | |
| "grad_norm": 2.5776922439224057, | |
| "learning_rate": 3.941007292345711e-06, | |
| "loss": 0.776, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.6157303370786517, | |
| "grad_norm": 2.1324124549933514, | |
| "learning_rate": 3.864459494377854e-06, | |
| "loss": 0.8218, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.6202247191011236, | |
| "grad_norm": 2.644470734779867, | |
| "learning_rate": 3.7881913145339387e-06, | |
| "loss": 0.8259, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.6247191011235955, | |
| "grad_norm": 2.9101594862192623, | |
| "learning_rate": 3.712221533267206e-06, | |
| "loss": 0.8208, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.6292134831460674, | |
| "grad_norm": 2.7469407604097555, | |
| "learning_rate": 3.6365688575525315e-06, | |
| "loss": 0.8049, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.6337078651685393, | |
| "grad_norm": 2.533667628455592, | |
| "learning_rate": 3.5612519162799954e-06, | |
| "loss": 0.8253, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.6382022471910113, | |
| "grad_norm": 2.7113902576435214, | |
| "learning_rate": 3.48628925566764e-06, | |
| "loss": 0.84, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.6426966292134831, | |
| "grad_norm": 2.9783117236592673, | |
| "learning_rate": 3.4116993346946286e-06, | |
| "loss": 0.8184, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.647191011235955, | |
| "grad_norm": 2.3877175660365118, | |
| "learning_rate": 3.337500520555843e-06, | |
| "loss": 0.806, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.651685393258427, | |
| "grad_norm": 2.2716341939562716, | |
| "learning_rate": 3.2637110841391202e-06, | |
| "loss": 0.8045, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.6561797752808989, | |
| "grad_norm": 2.4122921966517463, | |
| "learning_rate": 3.190349195526172e-06, | |
| "loss": 0.8254, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.6606741573033708, | |
| "grad_norm": 2.899653247745689, | |
| "learning_rate": 3.117432919518353e-06, | |
| "loss": 0.8059, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.6651685393258427, | |
| "grad_norm": 2.7343035364639743, | |
| "learning_rate": 3.0449802111883365e-06, | |
| "loss": 0.8199, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.6696629213483146, | |
| "grad_norm": 2.526014584573272, | |
| "learning_rate": 2.973008911458816e-06, | |
| "loss": 0.7953, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.6741573033707865, | |
| "grad_norm": 2.620122537254342, | |
| "learning_rate": 2.901536742709323e-06, | |
| "loss": 0.8445, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6786516853932584, | |
| "grad_norm": 2.503228611506985, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.8116, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.6831460674157304, | |
| "grad_norm": 2.6396612931098886, | |
| "learning_rate": 2.760160068798934e-06, | |
| "loss": 0.8182, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.6876404494382022, | |
| "grad_norm": 2.817133442950779, | |
| "learning_rate": 2.6902903765576426e-06, | |
| "loss": 0.8029, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.6921348314606741, | |
| "grad_norm": 2.5166671771316613, | |
| "learning_rate": 2.620989432563167e-06, | |
| "loss": 0.8179, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.6966292134831461, | |
| "grad_norm": 2.623144140751482, | |
| "learning_rate": 2.552274301640454e-06, | |
| "loss": 0.8165, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.701123595505618, | |
| "grad_norm": 2.4289501142037433, | |
| "learning_rate": 2.484161904362481e-06, | |
| "loss": 0.7876, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.7056179775280899, | |
| "grad_norm": 2.929065269125912, | |
| "learning_rate": 2.416669012883698e-06, | |
| "loss": 0.8179, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.7101123595505618, | |
| "grad_norm": 2.236454595662634, | |
| "learning_rate": 2.3498122468100114e-06, | |
| "loss": 0.8066, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.7146067415730337, | |
| "grad_norm": 2.289287136198235, | |
| "learning_rate": 2.283608069106336e-06, | |
| "loss": 0.8043, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.7191011235955056, | |
| "grad_norm": 2.356768634878435, | |
| "learning_rate": 2.2180727820427156e-06, | |
| "loss": 0.7739, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.7235955056179775, | |
| "grad_norm": 2.905123153518728, | |
| "learning_rate": 2.153222523180007e-06, | |
| "loss": 0.8094, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.7280898876404495, | |
| "grad_norm": 2.2771620513777346, | |
| "learning_rate": 2.089073261396148e-06, | |
| "loss": 0.8106, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.7325842696629213, | |
| "grad_norm": 2.410875702229844, | |
| "learning_rate": 2.0256407929539203e-06, | |
| "loss": 0.7578, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.7370786516853932, | |
| "grad_norm": 2.296833553163406, | |
| "learning_rate": 1.962940737611264e-06, | |
| "loss": 0.7917, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.7415730337078652, | |
| "grad_norm": 3.1435922132540184, | |
| "learning_rate": 1.9009885347750102e-06, | |
| "loss": 0.8191, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.7460674157303371, | |
| "grad_norm": 2.649409366248491, | |
| "learning_rate": 1.8397994396990687e-06, | |
| "loss": 0.8069, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.750561797752809, | |
| "grad_norm": 2.423649988581548, | |
| "learning_rate": 1.7793885197279164e-06, | |
| "loss": 0.8193, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.755056179775281, | |
| "grad_norm": 2.479988768714348, | |
| "learning_rate": 1.719770650586397e-06, | |
| "loss": 0.7967, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.7595505617977528, | |
| "grad_norm": 2.244972982243665, | |
| "learning_rate": 1.6609605127166784e-06, | |
| "loss": 0.7885, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.7640449438202247, | |
| "grad_norm": 2.1802352668743277, | |
| "learning_rate": 1.6029725876633085e-06, | |
| "loss": 0.7924, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.7685393258426966, | |
| "grad_norm": 2.701959045692566, | |
| "learning_rate": 1.5458211545072537e-06, | |
| "loss": 0.809, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.7730337078651686, | |
| "grad_norm": 2.647399701909666, | |
| "learning_rate": 1.489520286349767e-06, | |
| "loss": 0.7813, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.7775280898876404, | |
| "grad_norm": 3.3231840138362876, | |
| "learning_rate": 1.4340838468470198e-06, | |
| "loss": 0.7778, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.7820224719101123, | |
| "grad_norm": 2.3425746647284833, | |
| "learning_rate": 1.3795254867962548e-06, | |
| "loss": 0.8023, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.7865168539325843, | |
| "grad_norm": 2.5597817340601625, | |
| "learning_rate": 1.3258586407744117e-06, | |
| "loss": 0.8104, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.7910112359550562, | |
| "grad_norm": 2.517875514542219, | |
| "learning_rate": 1.273096523829939e-06, | |
| "loss": 0.7894, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.7955056179775281, | |
| "grad_norm": 2.4199435511686507, | |
| "learning_rate": 1.2212521282287093e-06, | |
| "loss": 0.814, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.5842013891324784, | |
| "learning_rate": 1.1703382202547619e-06, | |
| "loss": 0.8325, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.8044943820224719, | |
| "grad_norm": 2.9426361073785334, | |
| "learning_rate": 1.1203673370667111e-06, | |
| "loss": 0.8239, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.8089887640449438, | |
| "grad_norm": 2.841556764864622, | |
| "learning_rate": 1.0713517836105624e-06, | |
| "loss": 0.8129, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.8134831460674158, | |
| "grad_norm": 2.827206278816824, | |
| "learning_rate": 1.0233036295897147e-06, | |
| "loss": 0.7546, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.8179775280898877, | |
| "grad_norm": 2.475358629778951, | |
| "learning_rate": 9.762347064928918e-07, | |
| "loss": 0.7787, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.8224719101123595, | |
| "grad_norm": 2.6449064423582573, | |
| "learning_rate": 9.301566046807247e-07, | |
| "loss": 0.816, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.8269662921348314, | |
| "grad_norm": 2.8617687301277215, | |
| "learning_rate": 8.850806705317183e-07, | |
| "loss": 0.7975, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.8314606741573034, | |
| "grad_norm": 2.6088401680292916, | |
| "learning_rate": 8.410180036482895e-07, | |
| "loss": 0.8093, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.8359550561797753, | |
| "grad_norm": 2.602084546554059, | |
| "learning_rate": 7.979794541235792e-07, | |
| "loss": 0.776, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.8404494382022472, | |
| "grad_norm": 2.344373295351233, | |
| "learning_rate": 7.559756198696971e-07, | |
| "loss": 0.8251, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.8449438202247191, | |
| "grad_norm": 2.4323771087778607, | |
| "learning_rate": 7.150168440080652e-07, | |
| "loss": 0.7999, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.849438202247191, | |
| "grad_norm": 2.3773975505911933, | |
| "learning_rate": 6.75113212322514e-07, | |
| "loss": 0.78, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.8539325842696629, | |
| "grad_norm": 2.1579711929276937, | |
| "learning_rate": 6.36274550775719e-07, | |
| "loss": 0.7707, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.8584269662921349, | |
| "grad_norm": 2.7293834324024586, | |
| "learning_rate": 5.985104230896515e-07, | |
| "loss": 0.8189, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.8629213483146068, | |
| "grad_norm": 2.685801601082177, | |
| "learning_rate": 5.618301283905703e-07, | |
| "loss": 0.7541, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.8674157303370786, | |
| "grad_norm": 3.116629925764096, | |
| "learning_rate": 5.262426989191921e-07, | |
| "loss": 0.8066, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.8719101123595505, | |
| "grad_norm": 3.2051350133529324, | |
| "learning_rate": 4.917568978065684e-07, | |
| "loss": 0.8207, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.8764044943820225, | |
| "grad_norm": 2.2836460078075675, | |
| "learning_rate": 4.5838121691622995e-07, | |
| "loss": 0.7935, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.8808988764044944, | |
| "grad_norm": 2.7205905760599887, | |
| "learning_rate": 4.261238747531382e-07, | |
| "loss": 0.7772, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.8853932584269663, | |
| "grad_norm": 2.0574139164637004, | |
| "learning_rate": 3.949928144399301e-07, | |
| "loss": 0.7907, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.8898876404494382, | |
| "grad_norm": 2.8112752890836084, | |
| "learning_rate": 3.649957017609979e-07, | |
| "loss": 0.7942, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.8943820224719101, | |
| "grad_norm": 2.4763979006513615, | |
| "learning_rate": 3.361399232748358e-07, | |
| "loss": 0.7781, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.898876404494382, | |
| "grad_norm": 2.376939861217481, | |
| "learning_rate": 3.0843258449516966e-07, | |
| "loss": 0.7835, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.903370786516854, | |
| "grad_norm": 2.2478850058039725, | |
| "learning_rate": 2.8188050814126675e-07, | |
| "loss": 0.8014, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.9078651685393259, | |
| "grad_norm": 2.923436127261158, | |
| "learning_rate": 2.564902324579022e-07, | |
| "loss": 0.7951, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.9123595505617977, | |
| "grad_norm": 2.7437515079644283, | |
| "learning_rate": 2.32268009605362e-07, | |
| "loss": 0.7875, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.9168539325842696, | |
| "grad_norm": 2.5936965409935233, | |
| "learning_rate": 2.092198041198995e-07, | |
| "loss": 0.8139, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.9213483146067416, | |
| "grad_norm": 2.518280793652536, | |
| "learning_rate": 1.873512914450104e-07, | |
| "loss": 0.7903, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.9258426966292135, | |
| "grad_norm": 2.4421720063174868, | |
| "learning_rate": 1.666678565339025e-07, | |
| "loss": 0.7862, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.9303370786516854, | |
| "grad_norm": 2.2383431847411903, | |
| "learning_rate": 1.471745925234902e-07, | |
| "loss": 0.7941, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.9348314606741573, | |
| "grad_norm": 2.8799057324357245, | |
| "learning_rate": 1.2887629948024727e-07, | |
| "loss": 0.8059, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.9393258426966292, | |
| "grad_norm": 2.250643175569056, | |
| "learning_rate": 1.1177748321823223e-07, | |
| "loss": 0.8312, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.9438202247191011, | |
| "grad_norm": 2.512971359525962, | |
| "learning_rate": 9.588235418955926e-08, | |
| "loss": 0.8094, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.9483146067415731, | |
| "grad_norm": 2.3088708858588802, | |
| "learning_rate": 8.119482644761423e-08, | |
| "loss": 0.7921, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.952808988764045, | |
| "grad_norm": 2.3516337137858714, | |
| "learning_rate": 6.771851668324226e-08, | |
| "loss": 0.8052, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.9573033707865168, | |
| "grad_norm": 2.7956368487770566, | |
| "learning_rate": 5.545674333417073e-08, | |
| "loss": 0.8082, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.9617977528089887, | |
| "grad_norm": 2.805865942089696, | |
| "learning_rate": 4.4412525767866256e-08, | |
| "loss": 0.7951, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.9662921348314607, | |
| "grad_norm": 2.4160854525275015, | |
| "learning_rate": 3.458858353804107e-08, | |
| "loss": 0.7968, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.9707865168539326, | |
| "grad_norm": 2.6385371261261867, | |
| "learning_rate": 2.59873357149798e-08, | |
| "loss": 0.8013, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.9752808988764045, | |
| "grad_norm": 2.638712402614864, | |
| "learning_rate": 1.8610900289867673e-08, | |
| "loss": 0.7815, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.9797752808988764, | |
| "grad_norm": 2.4261506557100887, | |
| "learning_rate": 1.2461093653246547e-08, | |
| "loss": 0.7871, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.9842696629213483, | |
| "grad_norm": 2.564929645903007, | |
| "learning_rate": 7.539430147745497e-09, | |
| "loss": 0.7924, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.9887640449438202, | |
| "grad_norm": 2.6033526258950683, | |
| "learning_rate": 3.847121695183531e-09, | |
| "loss": 0.7667, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.9932584269662922, | |
| "grad_norm": 2.7502692590510867, | |
| "learning_rate": 1.3850774981466474e-09, | |
| "loss": 0.7814, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.9977528089887641, | |
| "grad_norm": 2.4346251881242487, | |
| "learning_rate": 1.5390381609858839e-10, | |
| "loss": 0.803, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 2225, | |
| "total_flos": 3591464654209024.0, | |
| "train_loss": 0.8581783727581581, | |
| "train_runtime": 10043.9825, | |
| "train_samples_per_second": 28.354, | |
| "train_steps_per_second": 0.222 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2225, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3591464654209024.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |