| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.462025316455696, | |
| "eval_steps": 500, | |
| "global_step": 3063, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02109704641350211, | |
| "grad_norm": 0.47141513228416443, | |
| "learning_rate": 1e-05, | |
| "loss": 3.0389, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04219409282700422, | |
| "grad_norm": 0.701974093914032, | |
| "learning_rate": 2e-05, | |
| "loss": 3.2294, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06329113924050633, | |
| "grad_norm": 0.5129281878471375, | |
| "learning_rate": 1.9999467080038847e-05, | |
| "loss": 2.8774, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08438818565400844, | |
| "grad_norm": 0.5039157867431641, | |
| "learning_rate": 1.9997868376956112e-05, | |
| "loss": 2.8694, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10548523206751055, | |
| "grad_norm": 0.6525703072547913, | |
| "learning_rate": 1.9995204061147954e-05, | |
| "loss": 2.549, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12658227848101267, | |
| "grad_norm": 0.7681527137756348, | |
| "learning_rate": 1.9991474416587794e-05, | |
| "loss": 2.4746, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14767932489451477, | |
| "grad_norm": 0.909417450428009, | |
| "learning_rate": 1.9986679840796036e-05, | |
| "loss": 2.2414, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16877637130801687, | |
| "grad_norm": 1.059488296508789, | |
| "learning_rate": 1.998082084479771e-05, | |
| "loss": 2.1423, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.189873417721519, | |
| "grad_norm": 0.5037703514099121, | |
| "learning_rate": 1.9973898053067996e-05, | |
| "loss": 1.9704, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2109704641350211, | |
| "grad_norm": 1.0699340105056763, | |
| "learning_rate": 1.996591220346568e-05, | |
| "loss": 1.9176, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2320675105485232, | |
| "grad_norm": 0.5377259850502014, | |
| "learning_rate": 1.995686414715449e-05, | |
| "loss": 1.6842, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25316455696202533, | |
| "grad_norm": 0.4896199107170105, | |
| "learning_rate": 1.994675484851239e-05, | |
| "loss": 1.8138, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2742616033755274, | |
| "grad_norm": 0.85350102186203, | |
| "learning_rate": 1.9935585385028786e-05, | |
| "loss": 1.6322, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29535864978902954, | |
| "grad_norm": 0.7110411524772644, | |
| "learning_rate": 1.992335694718969e-05, | |
| "loss": 1.6378, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31645569620253167, | |
| "grad_norm": 7.485868453979492, | |
| "learning_rate": 1.9910070838350828e-05, | |
| "loss": 1.5367, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33755274261603374, | |
| "grad_norm": 0.6181347370147705, | |
| "learning_rate": 1.9895728474598713e-05, | |
| "loss": 1.5012, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35864978902953587, | |
| "grad_norm": 0.609489917755127, | |
| "learning_rate": 1.9880331384599743e-05, | |
| "loss": 1.4693, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.379746835443038, | |
| "grad_norm": 0.4934925138950348, | |
| "learning_rate": 1.986388120943723e-05, | |
| "loss": 1.411, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4008438818565401, | |
| "grad_norm": 0.5047251582145691, | |
| "learning_rate": 1.9846379702436518e-05, | |
| "loss": 1.389, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4219409282700422, | |
| "grad_norm": 0.558677077293396, | |
| "learning_rate": 1.9827828728978096e-05, | |
| "loss": 1.3247, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4430379746835443, | |
| "grad_norm": 0.6455519199371338, | |
| "learning_rate": 1.980823026629877e-05, | |
| "loss": 1.4037, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4641350210970464, | |
| "grad_norm": 0.54356849193573, | |
| "learning_rate": 1.978758640328094e-05, | |
| "loss": 1.2992, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48523206751054854, | |
| "grad_norm": 0.569212794303894, | |
| "learning_rate": 1.9765899340229934e-05, | |
| "loss": 1.3177, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 0.525545060634613, | |
| "learning_rate": 1.9743171388639518e-05, | |
| "loss": 1.2712, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5274261603375527, | |
| "grad_norm": 0.53231281042099, | |
| "learning_rate": 1.9719404970945506e-05, | |
| "loss": 1.249, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5485232067510548, | |
| "grad_norm": 0.8335980176925659, | |
| "learning_rate": 1.9694602620267575e-05, | |
| "loss": 1.2184, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.569620253164557, | |
| "grad_norm": 0.48649874329566956, | |
| "learning_rate": 1.9668766980139283e-05, | |
| "loss": 1.2538, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5907172995780591, | |
| "grad_norm": 0.5631985068321228, | |
| "learning_rate": 1.9641900804226286e-05, | |
| "loss": 1.2088, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6118143459915611, | |
| "grad_norm": 0.5892934799194336, | |
| "learning_rate": 1.961400695603288e-05, | |
| "loss": 1.189, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6329113924050633, | |
| "grad_norm": 0.9515223503112793, | |
| "learning_rate": 1.9585088408596758e-05, | |
| "loss": 1.2768, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6540084388185654, | |
| "grad_norm": 0.4848034679889679, | |
| "learning_rate": 1.9555148244172155e-05, | |
| "loss": 1.1582, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6751054852320675, | |
| "grad_norm": 0.8658956289291382, | |
| "learning_rate": 1.9524189653901326e-05, | |
| "loss": 1.1797, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6962025316455697, | |
| "grad_norm": 0.5689557194709778, | |
| "learning_rate": 1.9492215937474416e-05, | |
| "loss": 1.2036, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7172995780590717, | |
| "grad_norm": 0.47713547945022583, | |
| "learning_rate": 1.9459230502777765e-05, | |
| "loss": 1.1822, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7383966244725738, | |
| "grad_norm": 0.5391554832458496, | |
| "learning_rate": 1.9425236865530693e-05, | |
| "loss": 1.1276, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.4558831751346588, | |
| "learning_rate": 1.9390238648910765e-05, | |
| "loss": 1.1413, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7805907172995781, | |
| "grad_norm": 0.6209776401519775, | |
| "learning_rate": 1.935423958316763e-05, | |
| "loss": 1.1602, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8016877637130801, | |
| "grad_norm": 3.7902824878692627, | |
| "learning_rate": 1.9317243505225432e-05, | |
| "loss": 1.1335, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8227848101265823, | |
| "grad_norm": 0.46277958154678345, | |
| "learning_rate": 1.9279254358273857e-05, | |
| "loss": 1.0967, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8438818565400844, | |
| "grad_norm": 0.5270448327064514, | |
| "learning_rate": 1.9240276191347843e-05, | |
| "loss": 1.1255, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8649789029535865, | |
| "grad_norm": 0.5431308150291443, | |
| "learning_rate": 1.9200313158896034e-05, | |
| "loss": 1.1804, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8860759493670886, | |
| "grad_norm": 0.5265894532203674, | |
| "learning_rate": 1.915936952033797e-05, | |
| "loss": 1.1411, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9071729957805907, | |
| "grad_norm": 0.4602664113044739, | |
| "learning_rate": 1.9117449639610105e-05, | |
| "loss": 1.1545, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9282700421940928, | |
| "grad_norm": 0.5221014618873596, | |
| "learning_rate": 1.9074557984700685e-05, | |
| "loss": 1.098, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9493670886075949, | |
| "grad_norm": 0.45506101846694946, | |
| "learning_rate": 1.9030699127173518e-05, | |
| "loss": 1.1323, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9704641350210971, | |
| "grad_norm": 0.6054735779762268, | |
| "learning_rate": 1.8985877741680735e-05, | |
| "loss": 1.1156, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9915611814345991, | |
| "grad_norm": 0.5910170078277588, | |
| "learning_rate": 1.8940098605464544e-05, | |
| "loss": 1.0949, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0126582278481013, | |
| "grad_norm": 0.5487635135650635, | |
| "learning_rate": 1.8893366597848036e-05, | |
| "loss": 1.1163, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0337552742616034, | |
| "grad_norm": 0.78420090675354, | |
| "learning_rate": 1.884568669971515e-05, | |
| "loss": 1.1012, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0548523206751055, | |
| "grad_norm": 0.5282578468322754, | |
| "learning_rate": 1.8797063992979786e-05, | |
| "loss": 1.0687, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0759493670886076, | |
| "grad_norm": 0.5505881309509277, | |
| "learning_rate": 1.8747503660044127e-05, | |
| "loss": 1.1008, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0970464135021096, | |
| "grad_norm": 0.5666527152061462, | |
| "learning_rate": 1.869701098324633e-05, | |
| "loss": 1.071, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1181434599156117, | |
| "grad_norm": 0.507770836353302, | |
| "learning_rate": 1.864559134429745e-05, | |
| "loss": 1.0876, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.139240506329114, | |
| "grad_norm": 0.5918306112289429, | |
| "learning_rate": 1.85932502237079e-05, | |
| "loss": 1.0926, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.160337552742616, | |
| "grad_norm": 0.5296492576599121, | |
| "learning_rate": 1.853999320020326e-05, | |
| "loss": 1.0804, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1814345991561181, | |
| "grad_norm": 0.583659827709198, | |
| "learning_rate": 1.8485825950129715e-05, | |
| "loss": 1.1073, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2025316455696202, | |
| "grad_norm": 0.46049240231513977, | |
| "learning_rate": 1.8430754246849024e-05, | |
| "loss": 1.0384, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2236286919831223, | |
| "grad_norm": 0.49372121691703796, | |
| "learning_rate": 1.8374783960123178e-05, | |
| "loss": 1.0382, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2447257383966246, | |
| "grad_norm": 1.3448891639709473, | |
| "learning_rate": 1.831792105548879e-05, | |
| "loss": 1.0804, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2658227848101267, | |
| "grad_norm": 0.4508301019668579, | |
| "learning_rate": 1.826017159362124e-05, | |
| "loss": 1.0698, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2869198312236287, | |
| "grad_norm": 0.6035999655723572, | |
| "learning_rate": 1.8201541729688727e-05, | |
| "loss": 1.0308, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3080168776371308, | |
| "grad_norm": 0.6147104501724243, | |
| "learning_rate": 1.8142037712696213e-05, | |
| "loss": 1.0251, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3291139240506329, | |
| "grad_norm": 0.6955983638763428, | |
| "learning_rate": 1.8081665884819376e-05, | |
| "loss": 1.0428, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.350210970464135, | |
| "grad_norm": 0.7107847929000854, | |
| "learning_rate": 1.8020432680728655e-05, | |
| "loss": 1.0337, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.371308016877637, | |
| "grad_norm": 0.6034618020057678, | |
| "learning_rate": 1.79583446269034e-05, | |
| "loss": 1.0401, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.3924050632911391, | |
| "grad_norm": 0.49049898982048035, | |
| "learning_rate": 1.789540834093626e-05, | |
| "loss": 1.0474, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.4135021097046414, | |
| "grad_norm": 1.1223995685577393, | |
| "learning_rate": 1.7831630530827843e-05, | |
| "loss": 1.0551, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.4345991561181435, | |
| "grad_norm": 0.9353031516075134, | |
| "learning_rate": 1.7767017994271764e-05, | |
| "loss": 1.0567, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 0.4723357856273651, | |
| "learning_rate": 1.7701577617930126e-05, | |
| "loss": 1.0528, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4767932489451476, | |
| "grad_norm": 0.5055142641067505, | |
| "learning_rate": 1.763531637669949e-05, | |
| "loss": 1.0344, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.49789029535865, | |
| "grad_norm": 0.49486976861953735, | |
| "learning_rate": 1.7568241332967474e-05, | |
| "loss": 1.0346, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 0.7445654273033142, | |
| "learning_rate": 1.7500359635860018e-05, | |
| "loss": 0.9996, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.540084388185654, | |
| "grad_norm": 0.5551081299781799, | |
| "learning_rate": 1.7431678520479402e-05, | |
| "loss": 1.0206, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5611814345991561, | |
| "grad_norm": 0.6225650310516357, | |
| "learning_rate": 1.7362205307133095e-05, | |
| "loss": 1.0851, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5822784810126582, | |
| "grad_norm": 0.4787660539150238, | |
| "learning_rate": 1.729194740055352e-05, | |
| "loss": 1.0411, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6033755274261603, | |
| "grad_norm": 0.5873054265975952, | |
| "learning_rate": 1.7220912289108858e-05, | |
| "loss": 1.0371, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.6244725738396624, | |
| "grad_norm": 0.6177183389663696, | |
| "learning_rate": 1.7149107544004868e-05, | |
| "loss": 1.0862, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.6455696202531644, | |
| "grad_norm": 0.6516501903533936, | |
| "learning_rate": 1.7076540818477946e-05, | |
| "loss": 1.0608, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.7590875625610352, | |
| "learning_rate": 1.7003219846979403e-05, | |
| "loss": 1.0223, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.6877637130801688, | |
| "grad_norm": 0.5621942281723022, | |
| "learning_rate": 1.6929152444351095e-05, | |
| "loss": 1.0259, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.6706472635269165, | |
| "learning_rate": 1.6854346504992486e-05, | |
| "loss": 1.0054, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.729957805907173, | |
| "grad_norm": 0.7586253881454468, | |
| "learning_rate": 1.677881000201924e-05, | |
| "loss": 1.0763, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7510548523206753, | |
| "grad_norm": 0.6050952076911926, | |
| "learning_rate": 1.6702550986413396e-05, | |
| "loss": 0.9916, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7721518987341773, | |
| "grad_norm": 0.5705825686454773, | |
| "learning_rate": 1.6625577586165288e-05, | |
| "loss": 1.0412, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.7932489451476794, | |
| "grad_norm": 0.4899527132511139, | |
| "learning_rate": 1.6547898005407207e-05, | |
| "loss": 1.0354, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8143459915611815, | |
| "grad_norm": 0.6274574995040894, | |
| "learning_rate": 1.6469520523538986e-05, | |
| "loss": 0.9917, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.8354430379746836, | |
| "grad_norm": 0.490570992231369, | |
| "learning_rate": 1.639045349434554e-05, | |
| "loss": 1.0623, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8565400843881856, | |
| "grad_norm": 0.49156588315963745, | |
| "learning_rate": 1.6310705345106505e-05, | |
| "loss": 1.0722, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.8776371308016877, | |
| "grad_norm": 0.5059338212013245, | |
| "learning_rate": 1.6230284575697986e-05, | |
| "loss": 1.0307, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.8987341772151898, | |
| "grad_norm": 0.5933522582054138, | |
| "learning_rate": 1.614919975768665e-05, | |
| "loss": 1.0517, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.9198312236286919, | |
| "grad_norm": 0.5222599506378174, | |
| "learning_rate": 1.6067459533416118e-05, | |
| "loss": 0.987, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.9409282700421941, | |
| "grad_norm": 0.6938186883926392, | |
| "learning_rate": 1.5985072615085806e-05, | |
| "loss": 1.0169, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 0.5320377945899963, | |
| "learning_rate": 1.5902047783822385e-05, | |
| "loss": 1.0421, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9831223628691983, | |
| "grad_norm": 0.5304467678070068, | |
| "learning_rate": 1.5818393888743823e-05, | |
| "loss": 1.0214, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.0042194092827006, | |
| "grad_norm": 0.5151244401931763, | |
| "learning_rate": 1.5734119846016224e-05, | |
| "loss": 1.0074, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.0253164556962027, | |
| "grad_norm": 0.5229523777961731, | |
| "learning_rate": 1.5649234637903503e-05, | |
| "loss": 0.9872, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.0464135021097047, | |
| "grad_norm": 0.5985708832740784, | |
| "learning_rate": 1.5563747311810023e-05, | |
| "loss": 0.9669, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.067510548523207, | |
| "grad_norm": 0.522693395614624, | |
| "learning_rate": 1.547766697931628e-05, | |
| "loss": 0.9778, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.088607594936709, | |
| "grad_norm": 0.5560949444770813, | |
| "learning_rate": 1.5391002815207772e-05, | |
| "loss": 1.0171, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.109704641350211, | |
| "grad_norm": 0.6248462200164795, | |
| "learning_rate": 1.5303764056497086e-05, | |
| "loss": 0.9818, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.130801687763713, | |
| "grad_norm": 0.4498722553253174, | |
| "learning_rate": 1.5215960001439402e-05, | |
| "loss": 1.0226, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.151898734177215, | |
| "grad_norm": 1.4349712133407593, | |
| "learning_rate": 1.512760000854145e-05, | |
| "loss": 1.0547, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.172995780590717, | |
| "grad_norm": 0.5725253224372864, | |
| "learning_rate": 1.503869349556402e-05, | |
| "loss": 1.0465, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.1940928270042193, | |
| "grad_norm": 0.706743597984314, | |
| "learning_rate": 1.4949249938518203e-05, | |
| "loss": 0.9972, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.2151898734177213, | |
| "grad_norm": 0.5506371259689331, | |
| "learning_rate": 1.4859278870655387e-05, | |
| "loss": 0.9572, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.2362869198312234, | |
| "grad_norm": 0.5830360054969788, | |
| "learning_rate": 1.476878988145117e-05, | |
| "loss": 0.9966, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.257383966244726, | |
| "grad_norm": 0.564736008644104, | |
| "learning_rate": 1.4677792615583277e-05, | |
| "loss": 0.9768, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.278481012658228, | |
| "grad_norm": 0.5060303807258606, | |
| "learning_rate": 1.4586296771903582e-05, | |
| "loss": 0.9935, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.29957805907173, | |
| "grad_norm": 0.499325692653656, | |
| "learning_rate": 1.4494312102404382e-05, | |
| "loss": 0.9918, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.320675105485232, | |
| "grad_norm": 0.5975970029830933, | |
| "learning_rate": 1.440184841117897e-05, | |
| "loss": 0.9811, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.3417721518987342, | |
| "grad_norm": 0.6109151244163513, | |
| "learning_rate": 1.43089155533767e-05, | |
| "loss": 1.0358, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.3628691983122363, | |
| "grad_norm": 0.5909796953201294, | |
| "learning_rate": 1.4215523434152562e-05, | |
| "loss": 1.0327, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.3839662447257384, | |
| "grad_norm": 0.5238848924636841, | |
| "learning_rate": 1.4121682007611467e-05, | |
| "loss": 1.003, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 0.6874323487281799, | |
| "learning_rate": 1.4027401275747292e-05, | |
| "loss": 0.9948, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.4261603375527425, | |
| "grad_norm": 0.5528364181518555, | |
| "learning_rate": 1.3932691287376829e-05, | |
| "loss": 1.0084, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.4472573839662446, | |
| "grad_norm": 0.5873010754585266, | |
| "learning_rate": 1.3837562137068746e-05, | |
| "loss": 0.9936, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.4683544303797467, | |
| "grad_norm": 0.5948253273963928, | |
| "learning_rate": 1.3742023964067657e-05, | |
| "loss": 1.0179, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.489451476793249, | |
| "grad_norm": 2.4981517791748047, | |
| "learning_rate": 1.3646086951213451e-05, | |
| "loss": 0.9609, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.510548523206751, | |
| "grad_norm": 0.8416944146156311, | |
| "learning_rate": 1.3549761323855963e-05, | |
| "loss": 0.9583, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.5316455696202533, | |
| "grad_norm": 0.5092337131500244, | |
| "learning_rate": 1.3453057348765109e-05, | |
| "loss": 0.9639, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.5527426160337554, | |
| "grad_norm": 0.6320915222167969, | |
| "learning_rate": 1.335598533303662e-05, | |
| "loss": 1.0155, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.5738396624472575, | |
| "grad_norm": 0.5696821808815002, | |
| "learning_rate": 1.3258555622993465e-05, | |
| "loss": 0.9251, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.5949367088607596, | |
| "grad_norm": 0.5761372447013855, | |
| "learning_rate": 1.3160778603083104e-05, | |
| "loss": 1.0202, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.6160337552742616, | |
| "grad_norm": 0.5222246050834656, | |
| "learning_rate": 1.3062664694770668e-05, | |
| "loss": 0.9805, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.6371308016877637, | |
| "grad_norm": 0.7842735052108765, | |
| "learning_rate": 1.2964224355428194e-05, | |
| "loss": 1.0286, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.5675404667854309, | |
| "learning_rate": 1.2865468077220052e-05, | |
| "loss": 1.0586, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.679324894514768, | |
| "grad_norm": 0.5645381212234497, | |
| "learning_rate": 1.2766406385984625e-05, | |
| "loss": 0.9967, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.70042194092827, | |
| "grad_norm": 0.5045151114463806, | |
| "learning_rate": 1.2667049840112446e-05, | |
| "loss": 1.0062, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.721518987341772, | |
| "grad_norm": 0.615481972694397, | |
| "learning_rate": 1.2567409029420828e-05, | |
| "loss": 1.0349, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.742616033755274, | |
| "grad_norm": 0.5723300576210022, | |
| "learning_rate": 1.2467494574025166e-05, | |
| "loss": 0.9948, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7637130801687766, | |
| "grad_norm": 0.5182732939720154, | |
| "learning_rate": 1.2367317123206989e-05, | |
| "loss": 0.9959, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.7848101265822782, | |
| "grad_norm": 0.5043283104896545, | |
| "learning_rate": 1.2266887354278949e-05, | |
| "loss": 0.9774, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.8059071729957807, | |
| "grad_norm": 0.609108030796051, | |
| "learning_rate": 1.2166215971446749e-05, | |
| "loss": 0.9801, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.827004219409283, | |
| "grad_norm": 0.6657718420028687, | |
| "learning_rate": 1.2065313704668276e-05, | |
| "loss": 0.9891, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.848101265822785, | |
| "grad_norm": 0.49846649169921875, | |
| "learning_rate": 1.1964191308509949e-05, | |
| "loss": 0.9791, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.869198312236287, | |
| "grad_norm": 0.7247899174690247, | |
| "learning_rate": 1.1862859561000452e-05, | |
| "loss": 0.969, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.890295358649789, | |
| "grad_norm": 0.542321503162384, | |
| "learning_rate": 1.1761329262481981e-05, | |
| "loss": 0.9655, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.5614388585090637, | |
| "learning_rate": 1.165961123445908e-05, | |
| "loss": 0.9902, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.932489451476793, | |
| "grad_norm": 0.5627961754798889, | |
| "learning_rate": 1.1557716318445259e-05, | |
| "loss": 0.9746, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.9535864978902953, | |
| "grad_norm": 0.6578301787376404, | |
| "learning_rate": 1.1455655374807452e-05, | |
| "loss": 0.9727, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9746835443037973, | |
| "grad_norm": 0.6406633257865906, | |
| "learning_rate": 1.1353439281608486e-05, | |
| "loss": 0.9755, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.9957805907173, | |
| "grad_norm": 0.5521556735038757, | |
| "learning_rate": 1.1251078933447643e-05, | |
| "loss": 0.9525, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.0168776371308015, | |
| "grad_norm": 0.5983248353004456, | |
| "learning_rate": 1.1148585240299477e-05, | |
| "loss": 0.9821, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.037974683544304, | |
| "grad_norm": 0.7607985734939575, | |
| "learning_rate": 1.1045969126350982e-05, | |
| "loss": 0.9958, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.059071729957806, | |
| "grad_norm": 0.5998318791389465, | |
| "learning_rate": 1.094324152883725e-05, | |
| "loss": 1.0006, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.080168776371308, | |
| "grad_norm": 0.5375860929489136, | |
| "learning_rate": 1.0840413396875733e-05, | |
| "loss": 0.9893, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.1012658227848102, | |
| "grad_norm": 0.5799211263656616, | |
| "learning_rate": 1.0737495690299256e-05, | |
| "loss": 0.9625, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.1223628691983123, | |
| "grad_norm": 0.5588505864143372, | |
| "learning_rate": 1.063449937848785e-05, | |
| "loss": 0.9497, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.1434599156118144, | |
| "grad_norm": 0.6329061388969421, | |
| "learning_rate": 1.053143543919962e-05, | |
| "loss": 0.9664, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.1645569620253164, | |
| "grad_norm": 0.6371050477027893, | |
| "learning_rate": 1.0428314857400666e-05, | |
| "loss": 0.9833, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.1856540084388185, | |
| "grad_norm": 0.7161750197410583, | |
| "learning_rate": 1.0325148624094274e-05, | |
| "loss": 0.9926, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.2067510548523206, | |
| "grad_norm": 0.7472956776618958, | |
| "learning_rate": 1.0221947735149463e-05, | |
| "loss": 0.963, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.2278481012658227, | |
| "grad_norm": 0.5231541395187378, | |
| "learning_rate": 1.0118723190128969e-05, | |
| "loss": 1.0121, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.2489451476793247, | |
| "grad_norm": 0.5988788604736328, | |
| "learning_rate": 1.0015485991116904e-05, | |
| "loss": 0.9946, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.270042194092827, | |
| "grad_norm": 0.5369338393211365, | |
| "learning_rate": 9.91224714154608e-06, | |
| "loss": 0.9719, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.291139240506329, | |
| "grad_norm": 0.7602208852767944, | |
| "learning_rate": 9.809017645025242e-06, | |
| "loss": 1.0155, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.3122362869198314, | |
| "grad_norm": 1.0432655811309814, | |
| "learning_rate": 9.705808504166241e-06, | |
| "loss": 0.9271, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.6576730608940125, | |
| "learning_rate": 9.602630719411347e-06, | |
| "loss": 0.9607, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 0.6119015216827393, | |
| "learning_rate": 9.499495287860767e-06, | |
| "loss": 0.9346, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.3755274261603376, | |
| "grad_norm": 0.5240073800086975, | |
| "learning_rate": 9.39641320210054e-06, | |
| "loss": 0.9865, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.3966244725738397, | |
| "grad_norm": 0.5481302738189697, | |
| "learning_rate": 9.29339544903089e-06, | |
| "loss": 0.9459, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.4177215189873418, | |
| "grad_norm": 0.5367913246154785, | |
| "learning_rate": 9.190453008695211e-06, | |
| "loss": 0.9328, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.438818565400844, | |
| "grad_norm": 0.5393932461738586, | |
| "learning_rate": 9.087596853109763e-06, | |
| "loss": 0.9799, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.459915611814346, | |
| "grad_norm": 0.5419475436210632, | |
| "learning_rate": 8.984837945094236e-06, | |
| "loss": 0.946, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.481012658227848, | |
| "grad_norm": 0.4995289742946625, | |
| "learning_rate": 8.882187237103283e-06, | |
| "loss": 0.9512, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.50210970464135, | |
| "grad_norm": 0.6618465781211853, | |
| "learning_rate": 8.779655670059167e-06, | |
| "loss": 0.9976, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.523206751054852, | |
| "grad_norm": 0.5813521146774292, | |
| "learning_rate": 8.677254172185632e-06, | |
| "loss": 0.9626, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 3.5443037974683547, | |
| "grad_norm": 0.49877235293388367, | |
| "learning_rate": 8.574993657843135e-06, | |
| "loss": 0.9186, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.5654008438818563, | |
| "grad_norm": 0.6298680305480957, | |
| "learning_rate": 8.472885026365537e-06, | |
| "loss": 0.9909, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.586497890295359, | |
| "grad_norm": 0.8999304175376892, | |
| "learning_rate": 8.37093916089843e-06, | |
| "loss": 0.9434, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.607594936708861, | |
| "grad_norm": 0.6086742281913757, | |
| "learning_rate": 8.269166927239136e-06, | |
| "loss": 1.0081, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.628691983122363, | |
| "grad_norm": 0.5396959185600281, | |
| "learning_rate": 8.16757917267863e-06, | |
| "loss": 1.0028, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.649789029535865, | |
| "grad_norm": 0.48324915766716003, | |
| "learning_rate": 8.066186724845344e-06, | |
| "loss": 0.9769, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.670886075949367, | |
| "grad_norm": 0.5358965396881104, | |
| "learning_rate": 7.965000390551155e-06, | |
| "loss": 0.9697, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.691983122362869, | |
| "grad_norm": 0.6094380021095276, | |
| "learning_rate": 7.86403095463953e-06, | |
| "loss": 1.0159, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.7130801687763713, | |
| "grad_norm": 1.5224417448043823, | |
| "learning_rate": 7.763289178836044e-06, | |
| "loss": 0.9389, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.7341772151898733, | |
| "grad_norm": 0.5542327761650085, | |
| "learning_rate": 7.662785800601342e-06, | |
| "loss": 0.8919, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.7552742616033754, | |
| "grad_norm": 0.5416964888572693, | |
| "learning_rate": 7.562531531986713e-06, | |
| "loss": 0.9842, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.7763713080168775, | |
| "grad_norm": 0.5305083394050598, | |
| "learning_rate": 7.462537058492343e-06, | |
| "loss": 0.95, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.7974683544303796, | |
| "grad_norm": 0.7290380001068115, | |
| "learning_rate": 7.362813037928419e-06, | |
| "loss": 0.9207, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.818565400843882, | |
| "grad_norm": 0.6322868466377258, | |
| "learning_rate": 7.263370099279173e-06, | |
| "loss": 0.9681, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.8396624472573837, | |
| "grad_norm": 0.5915969014167786, | |
| "learning_rate": 7.1642188415700074e-06, | |
| "loss": 0.9749, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.5843455195426941, | |
| "learning_rate": 7.065369832737796e-06, | |
| "loss": 0.972, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.8818565400843883, | |
| "grad_norm": 0.5843347907066345, | |
| "learning_rate": 6.966833608504542e-06, | |
| "loss": 1.0092, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.9029535864978904, | |
| "grad_norm": 0.5809078216552734, | |
| "learning_rate": 6.868620671254389e-06, | |
| "loss": 0.9749, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.9240506329113924, | |
| "grad_norm": 0.6204249262809753, | |
| "learning_rate": 6.770741488914291e-06, | |
| "loss": 0.9546, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.9451476793248945, | |
| "grad_norm": 0.6413384675979614, | |
| "learning_rate": 6.673206493838247e-06, | |
| "loss": 0.9374, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.9662447257383966, | |
| "grad_norm": 0.5627954006195068, | |
| "learning_rate": 6.576026081695421e-06, | |
| "loss": 0.9751, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.9873417721518987, | |
| "grad_norm": 0.5923116207122803, | |
| "learning_rate": 6.479210610362103e-06, | |
| "loss": 0.9937, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.008438818565401, | |
| "grad_norm": 0.5560117363929749, | |
| "learning_rate": 6.382770398817741e-06, | |
| "loss": 0.9403, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.029535864978903, | |
| "grad_norm": 0.6586416363716125, | |
| "learning_rate": 6.286715726045089e-06, | |
| "loss": 0.9965, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.050632911392405, | |
| "grad_norm": 0.4938979148864746, | |
| "learning_rate": 6.191056829934647e-06, | |
| "loss": 0.9334, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.071729957805907, | |
| "grad_norm": 0.5240390300750732, | |
| "learning_rate": 6.095803906193452e-06, | |
| "loss": 0.9452, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.0928270042194095, | |
| "grad_norm": 0.8655621409416199, | |
| "learning_rate": 6.000967107258388e-06, | |
| "loss": 0.9959, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.113924050632911, | |
| "grad_norm": 0.6976946592330933, | |
| "learning_rate": 5.906556541214099e-06, | |
| "loss": 0.9801, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.135021097046414, | |
| "grad_norm": 0.5815954804420471, | |
| "learning_rate": 5.812582270715621e-06, | |
| "loss": 0.9402, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.156118143459915, | |
| "grad_norm": 0.5947538614273071, | |
| "learning_rate": 5.7190543119158704e-06, | |
| "loss": 0.9987, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.177215189873418, | |
| "grad_norm": 0.6958327889442444, | |
| "learning_rate": 5.6259826333980835e-06, | |
| "loss": 0.9857, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.198312236286919, | |
| "grad_norm": 0.9180216193199158, | |
| "learning_rate": 5.533377155113313e-06, | |
| "loss": 1.007, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.219409282700422, | |
| "grad_norm": 0.5390021204948425, | |
| "learning_rate": 5.441247747323148e-06, | |
| "loss": 0.9497, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.2405063291139244, | |
| "grad_norm": 0.9804661273956299, | |
| "learning_rate": 5.3496042295476665e-06, | |
| "loss": 0.9817, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.261603375527426, | |
| "grad_norm": 0.5353003144264221, | |
| "learning_rate": 5.258456369518856e-06, | |
| "loss": 0.9536, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.282700421940929, | |
| "grad_norm": 0.7157588005065918, | |
| "learning_rate": 5.167813882139518e-06, | |
| "loss": 0.9536, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.6708821058273315, | |
| "learning_rate": 5.077686428447835e-06, | |
| "loss": 0.9559, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.324894514767933, | |
| "grad_norm": 0.6380380392074585, | |
| "learning_rate": 4.9880836145876145e-06, | |
| "loss": 0.9588, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.345991561181434, | |
| "grad_norm": 0.6591257452964783, | |
| "learning_rate": 4.899014990784485e-06, | |
| "loss": 1.0262, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.367088607594937, | |
| "grad_norm": 0.5861677527427673, | |
| "learning_rate": 4.810490050327952e-06, | |
| "loss": 0.9388, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 4.3881856540084385, | |
| "grad_norm": 0.9419888257980347, | |
| "learning_rate": 4.7225182285595785e-06, | |
| "loss": 0.9411, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 4.409282700421941, | |
| "grad_norm": 0.5278356075286865, | |
| "learning_rate": 4.63510890186733e-06, | |
| "loss": 0.9645, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 4.430379746835443, | |
| "grad_norm": 0.7032003998756409, | |
| "learning_rate": 4.548271386686211e-06, | |
| "loss": 0.9719, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.451476793248945, | |
| "grad_norm": 0.6221036314964294, | |
| "learning_rate": 4.4620149385052615e-06, | |
| "loss": 0.9467, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 4.472573839662447, | |
| "grad_norm": 0.5741263031959534, | |
| "learning_rate": 4.3763487508810845e-06, | |
| "loss": 0.9745, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 4.493670886075949, | |
| "grad_norm": 0.5083219408988953, | |
| "learning_rate": 4.291281954457952e-06, | |
| "loss": 0.917, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 4.514767932489452, | |
| "grad_norm": 0.6289077401161194, | |
| "learning_rate": 4.206823615994641e-06, | |
| "loss": 0.9575, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 4.5358649789029535, | |
| "grad_norm": 0.5753020644187927, | |
| "learning_rate": 4.1229827373980366e-06, | |
| "loss": 0.9767, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 4.556962025316456, | |
| "grad_norm": 0.6226878762245178, | |
| "learning_rate": 4.039768254763695e-06, | |
| "loss": 0.9485, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 4.578059071729958, | |
| "grad_norm": 0.6323665380477905, | |
| "learning_rate": 3.957189037423381e-06, | |
| "loss": 1.0249, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 4.59915611814346, | |
| "grad_norm": 0.5639895796775818, | |
| "learning_rate": 3.8752538869997655e-06, | |
| "loss": 0.9755, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 4.620253164556962, | |
| "grad_norm": 0.5681629180908203, | |
| "learning_rate": 3.7939715364682704e-06, | |
| "loss": 0.9597, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 4.641350210970464, | |
| "grad_norm": 1.1015621423721313, | |
| "learning_rate": 3.7133506492263237e-06, | |
| "loss": 0.9353, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.662447257383966, | |
| "grad_norm": 0.5930614471435547, | |
| "learning_rate": 3.633399818169939e-06, | |
| "loss": 0.9802, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 4.6835443037974684, | |
| "grad_norm": 0.7959696650505066, | |
| "learning_rate": 3.5541275647778804e-06, | |
| "loss": 0.9717, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 4.70464135021097, | |
| "grad_norm": 0.5339838266372681, | |
| "learning_rate": 3.475542338203377e-06, | |
| "loss": 0.9193, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 4.725738396624473, | |
| "grad_norm": 0.5793105959892273, | |
| "learning_rate": 3.397652514373616e-06, | |
| "loss": 0.9267, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 4.746835443037975, | |
| "grad_norm": 0.5588264465332031, | |
| "learning_rate": 3.3204663950969717e-06, | |
| "loss": 0.9833, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 4.767932489451477, | |
| "grad_norm": 0.5399629473686218, | |
| "learning_rate": 3.2439922071781814e-06, | |
| "loss": 0.92, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 4.789029535864979, | |
| "grad_norm": 0.599014163017273, | |
| "learning_rate": 3.1682381015414942e-06, | |
| "loss": 0.93, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.5534657835960388, | |
| "learning_rate": 3.093212152361923e-06, | |
| "loss": 0.9384, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 4.831223628691983, | |
| "grad_norm": 0.4902456998825073, | |
| "learning_rate": 3.0189223562046476e-06, | |
| "loss": 0.965, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 4.852320675105485, | |
| "grad_norm": 0.525508463382721, | |
| "learning_rate": 2.945376631172726e-06, | |
| "loss": 0.9382, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.8734177215189876, | |
| "grad_norm": 0.5271111130714417, | |
| "learning_rate": 2.872582816063141e-06, | |
| "loss": 0.8921, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.894514767932489, | |
| "grad_norm": 0.5723162889480591, | |
| "learning_rate": 2.8005486695313234e-06, | |
| "loss": 0.9736, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.915611814345992, | |
| "grad_norm": 0.6840436458587646, | |
| "learning_rate": 2.7292818692641778e-06, | |
| "loss": 0.9456, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.936708860759493, | |
| "grad_norm": 0.5373286604881287, | |
| "learning_rate": 2.6587900111617983e-06, | |
| "loss": 0.9295, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 4.957805907172996, | |
| "grad_norm": 0.6412104964256287, | |
| "learning_rate": 2.5890806085278352e-06, | |
| "loss": 0.8981, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 4.978902953586498, | |
| "grad_norm": 0.8979920148849487, | |
| "learning_rate": 2.5201610912687267e-06, | |
| "loss": 0.9915, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.023783802986145, | |
| "learning_rate": 2.452038805101753e-06, | |
| "loss": 0.9515, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 5.0210970464135025, | |
| "grad_norm": 0.526311993598938, | |
| "learning_rate": 2.384721010772143e-06, | |
| "loss": 0.985, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 5.042194092827004, | |
| "grad_norm": 0.6119584441184998, | |
| "learning_rate": 2.3182148832791596e-06, | |
| "loss": 0.963, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 5.063291139240507, | |
| "grad_norm": 0.5512840747833252, | |
| "learning_rate": 2.252527511111381e-06, | |
| "loss": 0.995, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.084388185654008, | |
| "grad_norm": 0.655735194683075, | |
| "learning_rate": 2.187665895491169e-06, | |
| "loss": 0.977, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 5.105485232067511, | |
| "grad_norm": 0.4825851619243622, | |
| "learning_rate": 2.1236369496284637e-06, | |
| "loss": 0.9355, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 5.1265822784810124, | |
| "grad_norm": 0.5352940559387207, | |
| "learning_rate": 2.0604474979839304e-06, | |
| "loss": 0.96, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 5.147679324894515, | |
| "grad_norm": 0.50180584192276, | |
| "learning_rate": 1.998104275541595e-06, | |
| "loss": 0.9304, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 5.168776371308017, | |
| "grad_norm": 0.6091214418411255, | |
| "learning_rate": 1.936613927090988e-06, | |
| "loss": 0.9572, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 5.189873417721519, | |
| "grad_norm": 0.5810332298278809, | |
| "learning_rate": 1.87598300651894e-06, | |
| "loss": 0.9476, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 5.210970464135021, | |
| "grad_norm": 0.8323220014572144, | |
| "learning_rate": 1.8162179761110078e-06, | |
| "loss": 0.942, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 5.232067510548523, | |
| "grad_norm": 0.47791892290115356, | |
| "learning_rate": 1.7573252058627367e-06, | |
| "loss": 0.9274, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.549113392829895, | |
| "learning_rate": 1.6993109728006873e-06, | |
| "loss": 0.9421, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 5.274261603375527, | |
| "grad_norm": 0.5429459810256958, | |
| "learning_rate": 1.6421814603134346e-06, | |
| "loss": 0.9327, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.29535864978903, | |
| "grad_norm": 0.5613409876823425, | |
| "learning_rate": 1.5859427574924812e-06, | |
| "loss": 0.9333, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 5.3164556962025316, | |
| "grad_norm": 0.5223398208618164, | |
| "learning_rate": 1.530600858483301e-06, | |
| "loss": 0.9675, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 5.337552742616034, | |
| "grad_norm": 0.9334096908569336, | |
| "learning_rate": 1.4761616618464247e-06, | |
| "loss": 0.9793, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 5.358649789029536, | |
| "grad_norm": 0.7707008123397827, | |
| "learning_rate": 1.4226309699287644e-06, | |
| "loss": 0.9329, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 5.379746835443038, | |
| "grad_norm": 0.7877485752105713, | |
| "learning_rate": 1.3700144882451693e-06, | |
| "loss": 0.9336, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 5.40084388185654, | |
| "grad_norm": 0.6202056407928467, | |
| "learning_rate": 1.3183178248703187e-06, | |
| "loss": 0.9396, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 5.421940928270042, | |
| "grad_norm": 0.6117923259735107, | |
| "learning_rate": 1.2675464898409772e-06, | |
| "loss": 0.9758, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 5.443037974683544, | |
| "grad_norm": 0.6278054714202881, | |
| "learning_rate": 1.2177058945687248e-06, | |
| "loss": 0.9744, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 5.4641350210970465, | |
| "grad_norm": 0.6328499913215637, | |
| "learning_rate": 1.1688013512631768e-06, | |
| "loss": 1.0033, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 5.485232067510548, | |
| "grad_norm": 0.6842343211174011, | |
| "learning_rate": 1.120838072365803e-06, | |
| "loss": 0.9515, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.506329113924051, | |
| "grad_norm": 0.7078588008880615, | |
| "learning_rate": 1.0738211699943458e-06, | |
| "loss": 0.9005, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 5.527426160337553, | |
| "grad_norm": 0.6245320439338684, | |
| "learning_rate": 1.0277556553979605e-06, | |
| "loss": 0.9987, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 5.548523206751055, | |
| "grad_norm": 0.6007537245750427, | |
| "learning_rate": 9.826464384230983e-07, | |
| "loss": 0.95, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 5.569620253164557, | |
| "grad_norm": 0.5233466625213623, | |
| "learning_rate": 9.38498326990196e-07, | |
| "loss": 0.9524, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 5.590717299578059, | |
| "grad_norm": 0.9564990401268005, | |
| "learning_rate": 8.953160265812088e-07, | |
| "loss": 0.9375, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 5.6118143459915615, | |
| "grad_norm": 0.5528080463409424, | |
| "learning_rate": 8.531041397381179e-07, | |
| "loss": 0.9601, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 5.632911392405063, | |
| "grad_norm": 0.5790895223617554, | |
| "learning_rate": 8.118671655723375e-07, | |
| "loss": 0.9187, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 5.654008438818566, | |
| "grad_norm": 0.609240710735321, | |
| "learning_rate": 7.716094992852041e-07, | |
| "loss": 0.9405, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 5.675105485232067, | |
| "grad_norm": 0.6695911288261414, | |
| "learning_rate": 7.323354316995057e-07, | |
| "loss": 0.942, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 5.69620253164557, | |
| "grad_norm": 0.6779336333274841, | |
| "learning_rate": 6.940491488021595e-07, | |
| "loss": 0.9464, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.717299578059071, | |
| "grad_norm": 0.5404800176620483, | |
| "learning_rate": 6.567547312980438e-07, | |
| "loss": 0.9421, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 5.738396624472574, | |
| "grad_norm": 0.6370323896408081, | |
| "learning_rate": 6.204561541750643e-07, | |
| "loss": 0.9601, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.5309698581695557, | |
| "learning_rate": 5.851572862804811e-07, | |
| "loss": 0.9586, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 5.780590717299578, | |
| "grad_norm": 0.5793449878692627, | |
| "learning_rate": 5.508618899085583e-07, | |
| "loss": 0.9522, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 5.80168776371308, | |
| "grad_norm": 0.6610496640205383, | |
| "learning_rate": 5.175736203995541e-07, | |
| "loss": 0.9673, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 5.822784810126582, | |
| "grad_norm": 0.707855224609375, | |
| "learning_rate": 4.852960257501293e-07, | |
| "loss": 0.9816, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 5.843881856540085, | |
| "grad_norm": 0.5480538606643677, | |
| "learning_rate": 4.5403254623517914e-07, | |
| "loss": 0.9561, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 5.864978902953586, | |
| "grad_norm": 0.6886956095695496, | |
| "learning_rate": 4.2378651404116566e-07, | |
| "loss": 0.9979, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 5.886075949367089, | |
| "grad_norm": 0.5038341879844666, | |
| "learning_rate": 3.9456115291094275e-07, | |
| "loss": 0.9212, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 5.9071729957805905, | |
| "grad_norm": 0.6231849789619446, | |
| "learning_rate": 3.6635957780017716e-07, | |
| "loss": 0.9102, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.928270042194093, | |
| "grad_norm": 0.9208401441574097, | |
| "learning_rate": 3.3918479454533524e-07, | |
| "loss": 0.9473, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 5.949367088607595, | |
| "grad_norm": 0.5422254800796509, | |
| "learning_rate": 3.130396995432983e-07, | |
| "loss": 0.903, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 5.970464135021097, | |
| "grad_norm": 0.7246553301811218, | |
| "learning_rate": 2.879270794426714e-07, | |
| "loss": 0.9435, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 5.991561181434599, | |
| "grad_norm": 0.5847915410995483, | |
| "learning_rate": 2.638496108467614e-07, | |
| "loss": 0.9915, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 6.012658227848101, | |
| "grad_norm": 0.5698049068450928, | |
| "learning_rate": 2.4080986002829267e-07, | |
| "loss": 1.0059, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 6.033755274261603, | |
| "grad_norm": 0.5052976608276367, | |
| "learning_rate": 2.188102826558869e-07, | |
| "loss": 0.9318, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 6.0548523206751055, | |
| "grad_norm": 0.6817172765731812, | |
| "learning_rate": 1.9785322353232938e-07, | |
| "loss": 0.9258, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 6.075949367088608, | |
| "grad_norm": 0.53616863489151, | |
| "learning_rate": 1.779409163446455e-07, | |
| "loss": 0.9498, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 6.09704641350211, | |
| "grad_norm": 0.5920806527137756, | |
| "learning_rate": 1.5907548342603107e-07, | |
| "loss": 0.9657, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 6.118143459915612, | |
| "grad_norm": 0.8010777235031128, | |
| "learning_rate": 1.4125893552964122e-07, | |
| "loss": 0.9843, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 6.139240506329114, | |
| "grad_norm": 0.5737888813018799, | |
| "learning_rate": 1.2449317161427942e-07, | |
| "loss": 0.9208, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 6.160337552742616, | |
| "grad_norm": 0.5839754343032837, | |
| "learning_rate": 1.0877997864199519e-07, | |
| "loss": 0.9246, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 6.181434599156118, | |
| "grad_norm": 0.6490985751152039, | |
| "learning_rate": 9.412103138762618e-08, | |
| "loss": 0.9679, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.5477576851844788, | |
| "learning_rate": 8.051789226029338e-08, | |
| "loss": 0.9961, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 6.223628691983122, | |
| "grad_norm": 0.6516560912132263, | |
| "learning_rate": 6.797201113687313e-08, | |
| "loss": 0.9696, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 6.244725738396625, | |
| "grad_norm": 0.5474268794059753, | |
| "learning_rate": 5.648472520745851e-08, | |
| "loss": 0.9495, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 6.265822784810126, | |
| "grad_norm": 0.49119266867637634, | |
| "learning_rate": 4.605725883284562e-08, | |
| "loss": 0.9818, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 6.286919831223629, | |
| "grad_norm": 0.711467444896698, | |
| "learning_rate": 3.669072341403013e-08, | |
| "loss": 0.96, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 6.308016877637131, | |
| "grad_norm": 0.6409249305725098, | |
| "learning_rate": 2.8386117273749846e-08, | |
| "loss": 0.9632, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 6.329113924050633, | |
| "grad_norm": 0.7857047915458679, | |
| "learning_rate": 2.1144325550079837e-08, | |
| "loss": 0.9326, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.350210970464135, | |
| "grad_norm": 0.6604498028755188, | |
| "learning_rate": 1.4966120102094527e-08, | |
| "loss": 0.9718, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 6.371308016877637, | |
| "grad_norm": 0.5844841003417969, | |
| "learning_rate": 9.852159427594655e-09, | |
| "loss": 0.9742, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 6.3924050632911396, | |
| "grad_norm": 0.6354609727859497, | |
| "learning_rate": 5.802988592925607e-09, | |
| "loss": 0.9509, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 6.413502109704641, | |
| "grad_norm": 0.6316894888877869, | |
| "learning_rate": 2.8190391748794678e-09, | |
| "loss": 0.9592, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 6.434599156118144, | |
| "grad_norm": 0.5979645848274231, | |
| "learning_rate": 9.006292146984674e-10, | |
| "loss": 0.9484, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 6.455696202531645, | |
| "grad_norm": 0.6670655608177185, | |
| "learning_rate": 4.796318417432666e-11, | |
| "loss": 0.9456, | |
| "step": 3060 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3063, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.3676544022452634e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |