| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 6024, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 41.35739895402257, | |
| "learning_rate": 7.193423539345941e-06, | |
| "loss": 0.5141, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 10.443694874625423, | |
| "learning_rate": 9.358859796204429e-06, | |
| "loss": 0.4195, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.720021591145773, | |
| "learning_rate": 1.0625558804168632e-05, | |
| "loss": 0.2851, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 7.59250257002781, | |
| "learning_rate": 1.1524296053062918e-05, | |
| "loss": 0.2174, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 11.521550428114184, | |
| "learning_rate": 1.2221410821833392e-05, | |
| "loss": 0.1817, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.027998448017936, | |
| "learning_rate": 1.2790995061027121e-05, | |
| "loss": 0.1886, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.929816146115093, | |
| "learning_rate": 1.3272571673439616e-05, | |
| "loss": 0.1553, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.3881515971871345, | |
| "learning_rate": 1.3689732309921406e-05, | |
| "loss": 0.129, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 5.441118636680569, | |
| "learning_rate": 1.4057694068991321e-05, | |
| "loss": 0.1433, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.085213496930297, | |
| "learning_rate": 1.4386847078691883e-05, | |
| "loss": 0.1092, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 4.713361075579101, | |
| "learning_rate": 1.4684602194465794e-05, | |
| "loss": 0.1231, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.8259500358417924, | |
| "learning_rate": 1.495643131788561e-05, | |
| "loss": 0.0697, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.67920682727699, | |
| "learning_rate": 1.5206489871327869e-05, | |
| "loss": 0.084, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 8.189491553913532, | |
| "learning_rate": 1.54380079302981e-05, | |
| "loss": 0.1023, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 12.000369384166694, | |
| "learning_rate": 1.5653546086656083e-05, | |
| "loss": 0.0972, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.3726942012239953, | |
| "learning_rate": 1.5855168566779895e-05, | |
| "loss": 0.1036, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.0242117206977044, | |
| "learning_rate": 1.604456377435124e-05, | |
| "loss": 0.1081, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.2261934744219967, | |
| "learning_rate": 1.6223130325849813e-05, | |
| "loss": 0.082, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 4.70421073268978, | |
| "learning_rate": 1.6392039793463407e-05, | |
| "loss": 0.0744, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 12.064244277626111, | |
| "learning_rate": 1.6552283335550368e-05, | |
| "loss": 0.0934, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.150990470911233, | |
| "learning_rate": 1.67047069382623e-05, | |
| "loss": 0.0737, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 4.074258308547766, | |
| "learning_rate": 1.6850038451324284e-05, | |
| "loss": 0.0841, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.6764310841274237, | |
| "learning_rate": 1.6988908609137504e-05, | |
| "loss": 0.0821, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 3.1391283710987063, | |
| "learning_rate": 1.71218675747441e-05, | |
| "loss": 0.0747, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 3.1103728720325754, | |
| "learning_rate": 1.7249398104320845e-05, | |
| "loss": 0.0907, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 5.354173276825422, | |
| "learning_rate": 1.7371926128186358e-05, | |
| "loss": 0.0765, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.8677808267225005, | |
| "learning_rate": 1.7489829333814013e-05, | |
| "loss": 0.0777, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.6124301773304874, | |
| "learning_rate": 1.760344418715659e-05, | |
| "loss": 0.0897, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.915238130503912, | |
| "learning_rate": 1.7713071721324668e-05, | |
| "loss": 0.0866, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 4.061775729833908, | |
| "learning_rate": 1.781898234351457e-05, | |
| "loss": 0.0804, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.5648704857906242, | |
| "learning_rate": 1.7921419853452233e-05, | |
| "loss": 0.0834, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.225837991191409, | |
| "learning_rate": 1.8020604823638384e-05, | |
| "loss": 0.0646, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 4.041622096074851, | |
| "learning_rate": 1.811673745928848e-05, | |
| "loss": 0.0636, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.21100008000919218, | |
| "learning_rate": 1.821000003120973e-05, | |
| "loss": 0.0656, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.2923304434569443, | |
| "learning_rate": 1.8300558955927067e-05, | |
| "loss": 0.0756, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.6524729986322013, | |
| "learning_rate": 1.83885665827083e-05, | |
| "loss": 0.0466, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.1793660025907284, | |
| "learning_rate": 1.847416273569235e-05, | |
| "loss": 0.0704, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 4.363503091034068, | |
| "learning_rate": 1.8557476050321896e-05, | |
| "loss": 0.0712, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 3.0942807753734476, | |
| "learning_rate": 1.863862513615056e-05, | |
| "loss": 0.0831, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.789230871814937, | |
| "learning_rate": 1.8717719592408857e-05, | |
| "loss": 0.078, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 3.4361028432853677, | |
| "learning_rate": 1.879486089815082e-05, | |
| "loss": 0.0663, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 3.690439378170989, | |
| "learning_rate": 1.8870143195120794e-05, | |
| "loss": 0.0738, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.159404934400987, | |
| "learning_rate": 1.8943653978491198e-05, | |
| "loss": 0.0768, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.8042070009727055, | |
| "learning_rate": 1.901547470818277e-05, | |
| "loss": 0.0777, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.70758347995601, | |
| "learning_rate": 1.9085681351478775e-05, | |
| "loss": 0.05, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.2113560599237037, | |
| "learning_rate": 1.9154344865995993e-05, | |
| "loss": 0.0715, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.5816071085486527, | |
| "learning_rate": 1.9221531630710657e-05, | |
| "loss": 0.0688, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.4447841193953872, | |
| "learning_rate": 1.9287303831602588e-05, | |
| "loss": 0.0659, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.2668557011358101, | |
| "learning_rate": 1.9351719807533285e-05, | |
| "loss": 0.0515, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.0353893651728865, | |
| "learning_rate": 1.9414834361179333e-05, | |
| "loss": 0.0687, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 2.005863891715338, | |
| "learning_rate": 1.947669903917393e-05, | |
| "loss": 0.0459, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.4330303529877615, | |
| "learning_rate": 1.9537362385044847e-05, | |
| "loss": 0.0557, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.8683959370360874, | |
| "learning_rate": 1.959687016805845e-05, | |
| "loss": 0.0656, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.5963373312962135, | |
| "learning_rate": 1.9655265590672502e-05, | |
| "loss": 0.0519, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.824110595633705, | |
| "learning_rate": 1.9712589476953243e-05, | |
| "loss": 0.0557, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.024299349665376, | |
| "learning_rate": 1.976888044401508e-05, | |
| "loss": 0.0617, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.0490238284158413, | |
| "learning_rate": 1.98241750582861e-05, | |
| "loss": 0.0529, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.2602675746455845, | |
| "learning_rate": 1.9878507978183157e-05, | |
| "loss": 0.0617, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.7353304044327778, | |
| "learning_rate": 1.9931912084590654e-05, | |
| "loss": 0.0464, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.043608410331547, | |
| "learning_rate": 1.998441860037306e-05, | |
| "loss": 0.0716, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.9522735357022469, | |
| "learning_rate": 1.997786386275595e-05, | |
| "loss": 0.0496, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.3740017908881452, | |
| "learning_rate": 1.9940970300682533e-05, | |
| "loss": 0.0686, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.1533525092583308, | |
| "learning_rate": 1.9904076738609114e-05, | |
| "loss": 0.0562, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.564671647601224, | |
| "learning_rate": 1.9867183176535695e-05, | |
| "loss": 0.0553, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.8996693997352794, | |
| "learning_rate": 1.9830289614462276e-05, | |
| "loss": 0.0524, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.9340857910489522, | |
| "learning_rate": 1.979339605238886e-05, | |
| "loss": 0.0534, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.892682744806536, | |
| "learning_rate": 1.9756502490315442e-05, | |
| "loss": 0.0625, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.2063853192894494, | |
| "learning_rate": 1.9719608928242023e-05, | |
| "loss": 0.0519, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.984312822464147, | |
| "learning_rate": 1.9682715366168604e-05, | |
| "loss": 0.0494, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.6705446030280595, | |
| "learning_rate": 1.9645821804095185e-05, | |
| "loss": 0.059, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.7422620817223426, | |
| "learning_rate": 1.960892824202177e-05, | |
| "loss": 0.0454, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.3182961229762868, | |
| "learning_rate": 1.957203467994835e-05, | |
| "loss": 0.0581, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.330360819481426, | |
| "learning_rate": 1.9535141117874932e-05, | |
| "loss": 0.0575, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.9840741868184866, | |
| "learning_rate": 1.9498247555801517e-05, | |
| "loss": 0.063, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.6018064731760029, | |
| "learning_rate": 1.9461353993728094e-05, | |
| "loss": 0.0519, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.391561342963203, | |
| "learning_rate": 1.9424460431654675e-05, | |
| "loss": 0.0572, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.0319629863193043, | |
| "learning_rate": 1.938756686958126e-05, | |
| "loss": 0.0619, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.7398111190442345, | |
| "learning_rate": 1.935067330750784e-05, | |
| "loss": 0.05, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.675180749962208, | |
| "learning_rate": 1.9313779745434422e-05, | |
| "loss": 0.0654, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.486943722740635, | |
| "learning_rate": 1.9276886183361007e-05, | |
| "loss": 0.0463, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.1444156835752686, | |
| "learning_rate": 1.9239992621287588e-05, | |
| "loss": 0.0598, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.1260777714975718, | |
| "learning_rate": 1.920309905921417e-05, | |
| "loss": 0.039, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.328313971146618, | |
| "learning_rate": 1.916620549714075e-05, | |
| "loss": 0.0591, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.8805517771854091, | |
| "learning_rate": 1.912931193506733e-05, | |
| "loss": 0.0392, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.9704226543415952, | |
| "learning_rate": 1.9092418372993916e-05, | |
| "loss": 0.056, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.4328262810995938, | |
| "learning_rate": 1.9055524810920497e-05, | |
| "loss": 0.0602, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 2.6133255890647167, | |
| "learning_rate": 1.9018631248847078e-05, | |
| "loss": 0.0507, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.4574352859865667, | |
| "learning_rate": 1.898173768677366e-05, | |
| "loss": 0.0667, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.9116869451062006, | |
| "learning_rate": 1.894484412470024e-05, | |
| "loss": 0.0533, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.8152499611885836, | |
| "learning_rate": 1.890795056262682e-05, | |
| "loss": 0.0534, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.9440345318883696, | |
| "learning_rate": 1.8871057000553406e-05, | |
| "loss": 0.0419, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 2.7588896552457074, | |
| "learning_rate": 1.8834163438479987e-05, | |
| "loss": 0.0418, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.8886319208504987, | |
| "learning_rate": 1.8797269876406568e-05, | |
| "loss": 0.0397, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.8820615671771377, | |
| "learning_rate": 1.876037631433315e-05, | |
| "loss": 0.0454, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 1.2868214865586614, | |
| "learning_rate": 1.872348275225973e-05, | |
| "loss": 0.0373, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.016876778284122, | |
| "learning_rate": 1.8686589190186315e-05, | |
| "loss": 0.0619, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.8992764743601639, | |
| "learning_rate": 1.8649695628112896e-05, | |
| "loss": 0.0403, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.42416507318131186, | |
| "learning_rate": 1.8612802066039477e-05, | |
| "loss": 0.067, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.8601423331120188, | |
| "learning_rate": 1.8575908503966062e-05, | |
| "loss": 0.066, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.3893250379275068, | |
| "learning_rate": 1.8539014941892643e-05, | |
| "loss": 0.0437, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 2.3240219930504376, | |
| "learning_rate": 1.850212137981922e-05, | |
| "loss": 0.0488, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.8849896953543195, | |
| "learning_rate": 1.8465227817745805e-05, | |
| "loss": 0.0472, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.8655212477880712, | |
| "learning_rate": 1.8428334255672386e-05, | |
| "loss": 0.0546, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.2016665548799828, | |
| "learning_rate": 1.8391440693598967e-05, | |
| "loss": 0.0956, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.492659569010543, | |
| "learning_rate": 1.8354547131525552e-05, | |
| "loss": 0.0449, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.4224837643985645, | |
| "learning_rate": 1.8317653569452133e-05, | |
| "loss": 0.0545, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.8405573730480247, | |
| "learning_rate": 1.8280760007378714e-05, | |
| "loss": 0.0328, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.3988661237020132, | |
| "learning_rate": 1.8243866445305295e-05, | |
| "loss": 0.0451, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5573396543240564, | |
| "learning_rate": 1.8206972883231876e-05, | |
| "loss": 0.0502, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.7855566721285819, | |
| "learning_rate": 1.817007932115846e-05, | |
| "loss": 0.0341, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.497005175064917, | |
| "learning_rate": 1.8133185759085042e-05, | |
| "loss": 0.0513, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.650975953086994, | |
| "learning_rate": 1.8096292197011623e-05, | |
| "loss": 0.0473, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.8418054071334755, | |
| "learning_rate": 1.8059398634938204e-05, | |
| "loss": 0.0444, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.33651471077315, | |
| "learning_rate": 1.8022505072864785e-05, | |
| "loss": 0.0476, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.1529220423121023, | |
| "learning_rate": 1.7985611510791367e-05, | |
| "loss": 0.0379, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.6912013035706749, | |
| "learning_rate": 1.794871794871795e-05, | |
| "loss": 0.0526, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 2.806793867054328, | |
| "learning_rate": 1.7911824386644532e-05, | |
| "loss": 0.047, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 3.323409487413899, | |
| "learning_rate": 1.7874930824571113e-05, | |
| "loss": 0.0455, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.0877783561592873, | |
| "learning_rate": 1.7838037262497695e-05, | |
| "loss": 0.0555, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.9158165193439103, | |
| "learning_rate": 1.7801143700424276e-05, | |
| "loss": 0.0458, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.0863638939536142, | |
| "learning_rate": 1.776425013835086e-05, | |
| "loss": 0.0462, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 2.3398031954044978, | |
| "learning_rate": 1.772735657627744e-05, | |
| "loss": 0.032, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.9272917966743286, | |
| "learning_rate": 1.7690463014204022e-05, | |
| "loss": 0.0464, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.4608044079217914, | |
| "learning_rate": 1.7653569452130607e-05, | |
| "loss": 0.0544, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.0799767984416713, | |
| "learning_rate": 1.7616675890057188e-05, | |
| "loss": 0.0629, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.438242481318271, | |
| "learning_rate": 1.757978232798377e-05, | |
| "loss": 0.0586, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.22666963149041416, | |
| "learning_rate": 1.754288876591035e-05, | |
| "loss": 0.0419, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.6906855856145777, | |
| "learning_rate": 1.750599520383693e-05, | |
| "loss": 0.0417, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.8417294291584847, | |
| "learning_rate": 1.7469101641763513e-05, | |
| "loss": 0.0496, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 2.384648750559231, | |
| "learning_rate": 1.7432208079690097e-05, | |
| "loss": 0.037, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.0969098360012755, | |
| "learning_rate": 1.7395314517616678e-05, | |
| "loss": 0.0392, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.265947825753472, | |
| "learning_rate": 1.735842095554326e-05, | |
| "loss": 0.0434, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.0043877066833349, | |
| "learning_rate": 1.732152739346984e-05, | |
| "loss": 0.0423, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.3834891081373628, | |
| "learning_rate": 1.728463383139642e-05, | |
| "loss": 0.0423, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.8917610552257311, | |
| "learning_rate": 1.7247740269323006e-05, | |
| "loss": 0.0491, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.7777935423085822, | |
| "learning_rate": 1.7210846707249587e-05, | |
| "loss": 0.0465, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 2.882686457165806, | |
| "learning_rate": 1.717395314517617e-05, | |
| "loss": 0.0366, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.4070944899475468, | |
| "learning_rate": 1.713705958310275e-05, | |
| "loss": 0.0364, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.368387839772382, | |
| "learning_rate": 1.710016602102933e-05, | |
| "loss": 0.0302, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.2202288141019488, | |
| "learning_rate": 1.7063272458955912e-05, | |
| "loss": 0.0455, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.7419694841079256, | |
| "learning_rate": 1.7026378896882496e-05, | |
| "loss": 0.0402, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.5792094366611027, | |
| "learning_rate": 1.6989485334809077e-05, | |
| "loss": 0.0364, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.0092569099059323, | |
| "learning_rate": 1.695259177273566e-05, | |
| "loss": 0.0335, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.5374731387141227, | |
| "learning_rate": 1.691569821066224e-05, | |
| "loss": 0.0428, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.6268956058587998, | |
| "learning_rate": 1.687880464858882e-05, | |
| "loss": 0.0483, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.6628348966888566, | |
| "learning_rate": 1.6841911086515402e-05, | |
| "loss": 0.0487, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 8.037103442054311, | |
| "learning_rate": 1.6805017524441987e-05, | |
| "loss": 0.0679, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.620270154467724, | |
| "learning_rate": 1.6768123962368568e-05, | |
| "loss": 0.0437, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.0436875513091504, | |
| "learning_rate": 1.6731230400295152e-05, | |
| "loss": 0.0346, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8374706362800531, | |
| "learning_rate": 1.6694336838221733e-05, | |
| "loss": 0.0517, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.1939983249441695, | |
| "learning_rate": 1.6657443276148314e-05, | |
| "loss": 0.0393, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.0650702970882078, | |
| "learning_rate": 1.6620549714074896e-05, | |
| "loss": 0.0433, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.9588706576483126, | |
| "learning_rate": 1.6583656152001477e-05, | |
| "loss": 0.0381, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.4412233810414918, | |
| "learning_rate": 1.6546762589928058e-05, | |
| "loss": 0.0346, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.9533871774762268, | |
| "learning_rate": 1.6509869027854642e-05, | |
| "loss": 0.0239, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.44748765523495393, | |
| "learning_rate": 1.6472975465781223e-05, | |
| "loss": 0.0424, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.6542278196418172, | |
| "learning_rate": 1.6436081903707805e-05, | |
| "loss": 0.0322, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.774821420829752, | |
| "learning_rate": 1.6399188341634386e-05, | |
| "loss": 0.0337, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 3.4366984039627053, | |
| "learning_rate": 1.6362294779560967e-05, | |
| "loss": 0.0402, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.9409785268524657, | |
| "learning_rate": 1.6325401217487548e-05, | |
| "loss": 0.0351, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.3755255908395096, | |
| "learning_rate": 1.6288507655414133e-05, | |
| "loss": 0.0287, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.9105883051718862, | |
| "learning_rate": 1.6251614093340714e-05, | |
| "loss": 0.0377, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.669451044858242, | |
| "learning_rate": 1.6214720531267295e-05, | |
| "loss": 0.0397, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.0173498911812129, | |
| "learning_rate": 1.6177826969193876e-05, | |
| "loss": 0.0362, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.443726695479527, | |
| "learning_rate": 1.6140933407120457e-05, | |
| "loss": 0.0388, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.48614710000952543, | |
| "learning_rate": 1.610403984504704e-05, | |
| "loss": 0.044, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.8955875806318787, | |
| "learning_rate": 1.6067146282973623e-05, | |
| "loss": 0.0361, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.7136011214906065, | |
| "learning_rate": 1.6030252720900204e-05, | |
| "loss": 0.0399, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.342383821315822, | |
| "learning_rate": 1.599335915882679e-05, | |
| "loss": 0.0398, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.9154366056765233, | |
| "learning_rate": 1.5956465596753366e-05, | |
| "loss": 0.0351, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 3.380571358597728, | |
| "learning_rate": 1.5919572034679947e-05, | |
| "loss": 0.0355, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 1.4160947580259224, | |
| "learning_rate": 1.5882678472606532e-05, | |
| "loss": 0.0333, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.6737638581076882, | |
| "learning_rate": 1.5845784910533113e-05, | |
| "loss": 0.0296, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.206337590570484, | |
| "learning_rate": 1.5808891348459694e-05, | |
| "loss": 0.0308, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.7350933133142535, | |
| "learning_rate": 1.577199778638628e-05, | |
| "loss": 0.0335, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 1.3603619905771789, | |
| "learning_rate": 1.573510422431286e-05, | |
| "loss": 0.0254, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.6337146921122443, | |
| "learning_rate": 1.569821066223944e-05, | |
| "loss": 0.0381, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.3833442570741445, | |
| "learning_rate": 1.5661317100166022e-05, | |
| "loss": 0.0295, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.1660396503165016, | |
| "learning_rate": 1.5624423538092603e-05, | |
| "loss": 0.0313, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.085393536126589, | |
| "learning_rate": 1.5587529976019188e-05, | |
| "loss": 0.0294, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.9428955580680065, | |
| "learning_rate": 1.555063641394577e-05, | |
| "loss": 0.0403, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.2824389825772662, | |
| "learning_rate": 1.551374285187235e-05, | |
| "loss": 0.0326, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.071530839035684, | |
| "learning_rate": 1.547684928979893e-05, | |
| "loss": 0.0224, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.7250437295813942, | |
| "learning_rate": 1.5439955727725512e-05, | |
| "loss": 0.0253, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.4676245052117833, | |
| "learning_rate": 1.5403062165652093e-05, | |
| "loss": 0.0245, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.3654237978697314, | |
| "learning_rate": 1.5366168603578678e-05, | |
| "loss": 0.0274, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.8371654653693097, | |
| "learning_rate": 1.532927504150526e-05, | |
| "loss": 0.0408, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.918371136718616, | |
| "learning_rate": 1.529238147943184e-05, | |
| "loss": 0.0252, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.0100840319244568, | |
| "learning_rate": 1.5255487917358423e-05, | |
| "loss": 0.0282, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.2752226504227586, | |
| "learning_rate": 1.5218594355285004e-05, | |
| "loss": 0.0481, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.016628356639124, | |
| "learning_rate": 1.5181700793211587e-05, | |
| "loss": 0.0534, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.2018849311057271, | |
| "learning_rate": 1.5144807231138168e-05, | |
| "loss": 0.0357, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.3543149671289594, | |
| "learning_rate": 1.5107913669064749e-05, | |
| "loss": 0.0354, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.0460241475499934, | |
| "learning_rate": 1.5071020106991332e-05, | |
| "loss": 0.0328, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.7712993931217346, | |
| "learning_rate": 1.5034126544917913e-05, | |
| "loss": 0.0352, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.6403552627247137, | |
| "learning_rate": 1.4997232982844494e-05, | |
| "loss": 0.0374, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.6985534187744373, | |
| "learning_rate": 1.4960339420771077e-05, | |
| "loss": 0.0236, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 2.7973859948867315, | |
| "learning_rate": 1.4923445858697658e-05, | |
| "loss": 0.0333, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.7856751051336177, | |
| "learning_rate": 1.488655229662424e-05, | |
| "loss": 0.0283, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.1090831103953211, | |
| "learning_rate": 1.4849658734550822e-05, | |
| "loss": 0.0424, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.37705371680669364, | |
| "learning_rate": 1.4812765172477403e-05, | |
| "loss": 0.0264, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 2.011193423521568, | |
| "learning_rate": 1.4775871610403986e-05, | |
| "loss": 0.0325, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.00662492673962, | |
| "learning_rate": 1.4738978048330567e-05, | |
| "loss": 0.0325, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.0249189181939053, | |
| "learning_rate": 1.4702084486257148e-05, | |
| "loss": 0.0296, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.294162625472094, | |
| "learning_rate": 1.4665190924183733e-05, | |
| "loss": 0.0362, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.8753798886295676, | |
| "learning_rate": 1.4628297362110312e-05, | |
| "loss": 0.0352, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.32404034083576144, | |
| "learning_rate": 1.4591403800036893e-05, | |
| "loss": 0.0319, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 1.3881694244187606, | |
| "learning_rate": 1.4554510237963478e-05, | |
| "loss": 0.0383, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.9106958124280001, | |
| "learning_rate": 1.4517616675890059e-05, | |
| "loss": 0.0366, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 1.3920757998129725, | |
| "learning_rate": 1.448072311381664e-05, | |
| "loss": 0.0359, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.5378193320802556, | |
| "learning_rate": 1.4443829551743223e-05, | |
| "loss": 0.0375, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.6327275888367767, | |
| "learning_rate": 1.4406935989669804e-05, | |
| "loss": 0.0491, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.457863332305797, | |
| "learning_rate": 1.4370042427596385e-05, | |
| "loss": 0.0379, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.19454376571103232, | |
| "learning_rate": 1.4333148865522968e-05, | |
| "loss": 0.0357, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.7421971370673671, | |
| "learning_rate": 1.429625530344955e-05, | |
| "loss": 0.0418, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.0937417741981473, | |
| "learning_rate": 1.4259361741376132e-05, | |
| "loss": 0.0248, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.7544900086855553, | |
| "learning_rate": 1.4222468179302713e-05, | |
| "loss": 0.0242, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.10357082442667653, | |
| "learning_rate": 1.4185574617229294e-05, | |
| "loss": 0.0277, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.429204704240934, | |
| "learning_rate": 1.4148681055155877e-05, | |
| "loss": 0.0253, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.7367754983146269, | |
| "learning_rate": 1.4111787493082458e-05, | |
| "loss": 0.0337, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.6428519364176438, | |
| "learning_rate": 1.407489393100904e-05, | |
| "loss": 0.0331, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.7848514705508334, | |
| "learning_rate": 1.4038000368935622e-05, | |
| "loss": 0.031, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.2604412780544694, | |
| "learning_rate": 1.4001106806862203e-05, | |
| "loss": 0.0381, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.0154534415085918, | |
| "learning_rate": 1.3964213244788784e-05, | |
| "loss": 0.0294, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.2166332545112624, | |
| "learning_rate": 1.3927319682715367e-05, | |
| "loss": 0.0275, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.8758369938501386, | |
| "learning_rate": 1.3890426120641948e-05, | |
| "loss": 0.0359, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.9645035178583596, | |
| "learning_rate": 1.385353255856853e-05, | |
| "loss": 0.0257, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.24217577106008117, | |
| "learning_rate": 1.3816638996495112e-05, | |
| "loss": 0.0188, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.8656287403232453, | |
| "learning_rate": 1.3779745434421693e-05, | |
| "loss": 0.035, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 1.5535880950760363, | |
| "learning_rate": 1.3742851872348278e-05, | |
| "loss": 0.0325, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.6356266461691455, | |
| "learning_rate": 1.3705958310274859e-05, | |
| "loss": 0.0292, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.3067413428110854, | |
| "learning_rate": 1.3669064748201439e-05, | |
| "loss": 0.0206, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 1.3327319855425117, | |
| "learning_rate": 1.3632171186128023e-05, | |
| "loss": 0.023, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 2.308037550483557, | |
| "learning_rate": 1.3595277624054604e-05, | |
| "loss": 0.0283, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.447528110026965, | |
| "learning_rate": 1.3558384061981185e-05, | |
| "loss": 0.0344, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.4781377553893515, | |
| "learning_rate": 1.3521490499907768e-05, | |
| "loss": 0.0338, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.2213361639825704, | |
| "learning_rate": 1.348459693783435e-05, | |
| "loss": 0.0213, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.806579624483452, | |
| "learning_rate": 1.344770337576093e-05, | |
| "loss": 0.0308, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.36305855519021074, | |
| "learning_rate": 1.3410809813687513e-05, | |
| "loss": 0.0307, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.7963102033556697, | |
| "learning_rate": 1.3373916251614094e-05, | |
| "loss": 0.0438, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.8741626138535608, | |
| "learning_rate": 1.3337022689540676e-05, | |
| "loss": 0.0282, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.8183941890622122, | |
| "learning_rate": 1.3300129127467258e-05, | |
| "loss": 0.03, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 1.315171338874376, | |
| "learning_rate": 1.326323556539384e-05, | |
| "loss": 0.0347, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 1.268446896432008, | |
| "learning_rate": 1.3226342003320422e-05, | |
| "loss": 0.0376, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.913219454382149, | |
| "learning_rate": 1.3189448441247003e-05, | |
| "loss": 0.0358, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.1043754536185766, | |
| "learning_rate": 1.3152554879173585e-05, | |
| "loss": 0.0284, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 1.203565789266508, | |
| "learning_rate": 1.3115661317100167e-05, | |
| "loss": 0.0292, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.9568074416295906, | |
| "learning_rate": 1.3078767755026749e-05, | |
| "loss": 0.0262, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.2859744392048369, | |
| "learning_rate": 1.304187419295333e-05, | |
| "loss": 0.0295, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 1.077880610534017, | |
| "learning_rate": 1.3004980630879912e-05, | |
| "loss": 0.0288, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.9333475338900054, | |
| "learning_rate": 1.2968087068806494e-05, | |
| "loss": 0.0355, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.904061045593678, | |
| "learning_rate": 1.2931193506733075e-05, | |
| "loss": 0.0251, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.58158119774247, | |
| "learning_rate": 1.2894299944659658e-05, | |
| "loss": 0.0233, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.7277481905024581, | |
| "learning_rate": 1.2857406382586239e-05, | |
| "loss": 0.0298, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 1.8262592151991606, | |
| "learning_rate": 1.2820512820512823e-05, | |
| "loss": 0.0276, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.6465268769367779, | |
| "learning_rate": 1.2783619258439404e-05, | |
| "loss": 0.027, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.6081326854882737, | |
| "learning_rate": 1.2746725696365985e-05, | |
| "loss": 0.0337, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.7068771700231296, | |
| "learning_rate": 1.2709832134292568e-05, | |
| "loss": 0.0236, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.8049539476685674, | |
| "learning_rate": 1.267293857221915e-05, | |
| "loss": 0.0282, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.553245442957324, | |
| "learning_rate": 1.263604501014573e-05, | |
| "loss": 0.0226, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.8295872212818136, | |
| "learning_rate": 1.2599151448072313e-05, | |
| "loss": 0.0232, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.9798348798123743, | |
| "learning_rate": 1.2562257885998895e-05, | |
| "loss": 0.0456, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.19245304335107, | |
| "learning_rate": 1.2525364323925476e-05, | |
| "loss": 0.0284, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.5998767447680797, | |
| "learning_rate": 1.2488470761852058e-05, | |
| "loss": 0.0307, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.7431018978542536, | |
| "learning_rate": 1.245157719977864e-05, | |
| "loss": 0.0216, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.0864224894613053, | |
| "learning_rate": 1.241468363770522e-05, | |
| "loss": 0.0388, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 2.164074288346446, | |
| "learning_rate": 1.2377790075631804e-05, | |
| "loss": 0.0302, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.6890867086297026, | |
| "learning_rate": 1.2340896513558385e-05, | |
| "loss": 0.0258, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 1.1150440915546074, | |
| "learning_rate": 1.2304002951484968e-05, | |
| "loss": 0.0279, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.7515709240098276, | |
| "learning_rate": 1.2267109389411549e-05, | |
| "loss": 0.0419, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.7795946895425347, | |
| "learning_rate": 1.223021582733813e-05, | |
| "loss": 0.0152, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.7206757082802774, | |
| "learning_rate": 1.2193322265264713e-05, | |
| "loss": 0.039, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.9261417250470075, | |
| "learning_rate": 1.2156428703191294e-05, | |
| "loss": 0.031, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.9738836548814439, | |
| "learning_rate": 1.2119535141117875e-05, | |
| "loss": 0.0318, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 1.6517158198941853, | |
| "learning_rate": 1.2082641579044458e-05, | |
| "loss": 0.0346, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.8931816948527265, | |
| "learning_rate": 1.2045748016971039e-05, | |
| "loss": 0.0278, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.5388621088460506, | |
| "learning_rate": 1.200885445489762e-05, | |
| "loss": 0.0296, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.6080136375748658, | |
| "learning_rate": 1.1971960892824204e-05, | |
| "loss": 0.0314, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.7855564824393652, | |
| "learning_rate": 1.1935067330750784e-05, | |
| "loss": 0.0371, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 1.391898825501054, | |
| "learning_rate": 1.1898173768677365e-05, | |
| "loss": 0.0215, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.023622261403316365, | |
| "learning_rate": 1.186128020660395e-05, | |
| "loss": 0.0289, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 1.2844725300311655, | |
| "learning_rate": 1.182438664453053e-05, | |
| "loss": 0.0272, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.008098058731733223, | |
| "learning_rate": 1.1787493082457114e-05, | |
| "loss": 0.021, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 2.4455783687272565, | |
| "learning_rate": 1.1750599520383695e-05, | |
| "loss": 0.0387, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.5832035368747368, | |
| "learning_rate": 1.1713705958310276e-05, | |
| "loss": 0.0392, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.4288935348488189, | |
| "learning_rate": 1.1676812396236859e-05, | |
| "loss": 0.0243, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.85140142754813, | |
| "learning_rate": 1.163991883416344e-05, | |
| "loss": 0.0288, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.6208450731238502, | |
| "learning_rate": 1.1603025272090021e-05, | |
| "loss": 0.0257, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5506545117162834, | |
| "learning_rate": 1.1566131710016604e-05, | |
| "loss": 0.0257, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.6983830876671564, | |
| "learning_rate": 1.1529238147943185e-05, | |
| "loss": 0.0153, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.5365569446878222, | |
| "learning_rate": 1.1492344585869766e-05, | |
| "loss": 0.0287, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.48404097576321364, | |
| "learning_rate": 1.1455451023796349e-05, | |
| "loss": 0.0189, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.2656124897423078, | |
| "learning_rate": 1.141855746172293e-05, | |
| "loss": 0.0321, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.9816015622085352, | |
| "learning_rate": 1.1381663899649511e-05, | |
| "loss": 0.0392, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 1.6810737744401563, | |
| "learning_rate": 1.1344770337576094e-05, | |
| "loss": 0.022, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.4543583055955254, | |
| "learning_rate": 1.1307876775502675e-05, | |
| "loss": 0.0331, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.4710753680279536, | |
| "learning_rate": 1.1270983213429258e-05, | |
| "loss": 0.0383, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.2478883072438642, | |
| "learning_rate": 1.1234089651355839e-05, | |
| "loss": 0.0316, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 2.2702061081470153, | |
| "learning_rate": 1.119719608928242e-05, | |
| "loss": 0.0282, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.6940800534656072, | |
| "learning_rate": 1.1160302527209005e-05, | |
| "loss": 0.0243, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.576849535693127, | |
| "learning_rate": 1.1123408965135584e-05, | |
| "loss": 0.0209, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.7227915044457328, | |
| "learning_rate": 1.1086515403062165e-05, | |
| "loss": 0.0243, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 9.838138591586603, | |
| "learning_rate": 1.104962184098875e-05, | |
| "loss": 0.0287, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.8776202139061747, | |
| "learning_rate": 1.101272827891533e-05, | |
| "loss": 0.0243, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.3717854528241673, | |
| "learning_rate": 1.097583471684191e-05, | |
| "loss": 0.0229, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 1.5052530178666412, | |
| "learning_rate": 1.0938941154768495e-05, | |
| "loss": 0.0232, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.7935110408300886, | |
| "learning_rate": 1.0902047592695076e-05, | |
| "loss": 0.0223, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 2.5764455237822945, | |
| "learning_rate": 1.0865154030621657e-05, | |
| "loss": 0.017, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.844609705609384, | |
| "learning_rate": 1.082826046854824e-05, | |
| "loss": 0.0247, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 1.3041682780604213, | |
| "learning_rate": 1.0791366906474821e-05, | |
| "loss": 0.0216, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.4847618373381675, | |
| "learning_rate": 1.0754473344401404e-05, | |
| "loss": 0.0229, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.575833133756361, | |
| "learning_rate": 1.0717579782327985e-05, | |
| "loss": 0.0225, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.064228503124154, | |
| "learning_rate": 1.0680686220254566e-05, | |
| "loss": 0.0212, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.6517908414314305, | |
| "learning_rate": 1.0643792658181149e-05, | |
| "loss": 0.0222, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 1.6514888202035902, | |
| "learning_rate": 1.060689909610773e-05, | |
| "loss": 0.0326, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.5296943703361738, | |
| "learning_rate": 1.0570005534034311e-05, | |
| "loss": 0.0292, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.020711807231569, | |
| "learning_rate": 1.0533111971960894e-05, | |
| "loss": 0.0226, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.5315183634882714, | |
| "learning_rate": 1.0496218409887475e-05, | |
| "loss": 0.0198, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 0.7542133397646632, | |
| "learning_rate": 1.0459324847814056e-05, | |
| "loss": 0.0201, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 1.4709649935718534, | |
| "learning_rate": 1.0422431285740639e-05, | |
| "loss": 0.0221, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 1.6579768697093855, | |
| "learning_rate": 1.038553772366722e-05, | |
| "loss": 0.0263, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.6342891371924907, | |
| "learning_rate": 1.0348644161593803e-05, | |
| "loss": 0.0229, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.37353433845779555, | |
| "learning_rate": 1.0311750599520384e-05, | |
| "loss": 0.0201, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.4459249487588341, | |
| "learning_rate": 1.0274857037446965e-05, | |
| "loss": 0.0283, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 1.6151962921561034, | |
| "learning_rate": 1.023796347537355e-05, | |
| "loss": 0.0118, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 1.4262238116120927, | |
| "learning_rate": 1.020106991330013e-05, | |
| "loss": 0.0229, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 1.228045108128222, | |
| "learning_rate": 1.016417635122671e-05, | |
| "loss": 0.0217, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.6530702622761245, | |
| "learning_rate": 1.0127282789153295e-05, | |
| "loss": 0.0207, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 1.9041502597805957, | |
| "learning_rate": 1.0090389227079876e-05, | |
| "loss": 0.0231, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 0.6535669839017003, | |
| "learning_rate": 1.0053495665006457e-05, | |
| "loss": 0.0225, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.3311305874562087, | |
| "learning_rate": 1.001660210293304e-05, | |
| "loss": 0.0203, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.8823147611211436, | |
| "learning_rate": 9.979708540859621e-06, | |
| "loss": 0.0276, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 1.5562568296719255, | |
| "learning_rate": 9.942814978786202e-06, | |
| "loss": 0.0269, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.6094359831984283, | |
| "learning_rate": 9.905921416712785e-06, | |
| "loss": 0.0205, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 2.2948821888190865, | |
| "learning_rate": 9.869027854639366e-06, | |
| "loss": 0.02, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 0.2523552251279377, | |
| "learning_rate": 9.832134292565947e-06, | |
| "loss": 0.0202, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.4590207341781818, | |
| "learning_rate": 9.79524073049253e-06, | |
| "loss": 0.02, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.385641051376586, | |
| "learning_rate": 9.758347168419111e-06, | |
| "loss": 0.0326, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 2.304443346442453, | |
| "learning_rate": 9.721453606345692e-06, | |
| "loss": 0.0469, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.7921276423545808, | |
| "learning_rate": 9.684560044272275e-06, | |
| "loss": 0.0226, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.7313140221633634, | |
| "learning_rate": 9.647666482198858e-06, | |
| "loss": 0.0208, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.8616674537642799, | |
| "learning_rate": 9.61077292012544e-06, | |
| "loss": 0.0228, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 0.3223458276996969, | |
| "learning_rate": 9.57387935805202e-06, | |
| "loss": 0.0204, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 0.25847436736502, | |
| "learning_rate": 9.536985795978603e-06, | |
| "loss": 0.0184, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 1.0102739659378692, | |
| "learning_rate": 9.500092233905184e-06, | |
| "loss": 0.021, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.4614770769943895, | |
| "learning_rate": 9.463198671831765e-06, | |
| "loss": 0.0192, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.7181037817361005, | |
| "learning_rate": 9.426305109758348e-06, | |
| "loss": 0.0219, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 0.8881807080809706, | |
| "learning_rate": 9.38941154768493e-06, | |
| "loss": 0.0149, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.8001361388360979, | |
| "learning_rate": 9.35251798561151e-06, | |
| "loss": 0.0148, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.4596057880898412, | |
| "learning_rate": 9.315624423538093e-06, | |
| "loss": 0.0171, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 0.3649340327954503, | |
| "learning_rate": 9.278730861464676e-06, | |
| "loss": 0.0217, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.3661755964392428, | |
| "learning_rate": 9.241837299391257e-06, | |
| "loss": 0.0262, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 1.1221188633493766, | |
| "learning_rate": 9.204943737317838e-06, | |
| "loss": 0.0254, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.12414790838540636, | |
| "learning_rate": 9.168050175244421e-06, | |
| "loss": 0.0193, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.8194229025263218, | |
| "learning_rate": 9.131156613171002e-06, | |
| "loss": 0.0132, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.7710864622318587, | |
| "learning_rate": 9.094263051097584e-06, | |
| "loss": 0.0229, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 1.2621629718343774, | |
| "learning_rate": 9.057369489024166e-06, | |
| "loss": 0.0214, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.31130549301919014, | |
| "learning_rate": 9.020475926950747e-06, | |
| "loss": 0.024, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 1.086050014018613, | |
| "learning_rate": 8.98358236487733e-06, | |
| "loss": 0.0251, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 0.7029996124364184, | |
| "learning_rate": 8.946688802803911e-06, | |
| "loss": 0.0215, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.3162716307135196, | |
| "learning_rate": 8.909795240730493e-06, | |
| "loss": 0.0198, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.15643570116786681, | |
| "learning_rate": 8.872901678657075e-06, | |
| "loss": 0.019, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.45575568712361675, | |
| "learning_rate": 8.836008116583657e-06, | |
| "loss": 0.0236, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.917784424801474, | |
| "learning_rate": 8.79911455451024e-06, | |
| "loss": 0.0262, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 1.3932007070706611, | |
| "learning_rate": 8.76222099243682e-06, | |
| "loss": 0.0203, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 1.0304061442597747, | |
| "learning_rate": 8.725327430363403e-06, | |
| "loss": 0.0265, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 1.161967395476058, | |
| "learning_rate": 8.688433868289984e-06, | |
| "loss": 0.024, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.43755927706096664, | |
| "learning_rate": 8.651540306216566e-06, | |
| "loss": 0.0249, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.45983444385917566, | |
| "learning_rate": 8.614646744143148e-06, | |
| "loss": 0.0121, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 1.0607636793892183, | |
| "learning_rate": 8.57775318206973e-06, | |
| "loss": 0.023, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 5.484759977473768, | |
| "learning_rate": 8.54085961999631e-06, | |
| "loss": 0.0208, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 0.6873972122341183, | |
| "learning_rate": 8.503966057922893e-06, | |
| "loss": 0.0224, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.9003081708047279, | |
| "learning_rate": 8.467072495849475e-06, | |
| "loss": 0.018, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.8521011818568595, | |
| "learning_rate": 8.430178933776056e-06, | |
| "loss": 0.0306, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 0.6956977848317097, | |
| "learning_rate": 8.393285371702639e-06, | |
| "loss": 0.023, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.5043637468484221, | |
| "learning_rate": 8.356391809629221e-06, | |
| "loss": 0.0193, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.9799158125352796, | |
| "learning_rate": 8.319498247555803e-06, | |
| "loss": 0.0211, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 0.33259982625437207, | |
| "learning_rate": 8.282604685482384e-06, | |
| "loss": 0.0217, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 1.0900206610270098, | |
| "learning_rate": 8.245711123408966e-06, | |
| "loss": 0.0291, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.8493131274550083, | |
| "learning_rate": 8.208817561335548e-06, | |
| "loss": 0.0186, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 0.0120847201082528, | |
| "learning_rate": 8.171923999262129e-06, | |
| "loss": 0.0241, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 1.1886551408595796, | |
| "learning_rate": 8.135030437188712e-06, | |
| "loss": 0.0263, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.7674536373536617, | |
| "learning_rate": 8.098136875115293e-06, | |
| "loss": 0.0162, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 0.9599986258139311, | |
| "learning_rate": 8.061243313041874e-06, | |
| "loss": 0.0311, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.7855164569900942, | |
| "learning_rate": 8.024349750968457e-06, | |
| "loss": 0.02, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.7680612067976029, | |
| "learning_rate": 7.987456188895038e-06, | |
| "loss": 0.0196, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 0.49399032116203606, | |
| "learning_rate": 7.95056262682162e-06, | |
| "loss": 0.0184, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.8318343668356561, | |
| "learning_rate": 7.913669064748202e-06, | |
| "loss": 0.0146, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.5414520871794223, | |
| "learning_rate": 7.876775502674785e-06, | |
| "loss": 0.0201, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.9871603598974189, | |
| "learning_rate": 7.839881940601366e-06, | |
| "loss": 0.0182, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.9254049832223281, | |
| "learning_rate": 7.802988378527947e-06, | |
| "loss": 0.019, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.04138757868391416, | |
| "learning_rate": 7.76609481645453e-06, | |
| "loss": 0.0127, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 1.0624519786080426, | |
| "learning_rate": 7.72920125438111e-06, | |
| "loss": 0.0137, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 1.545143622725685, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0264, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.7165870463329832, | |
| "learning_rate": 7.655414130234275e-06, | |
| "loss": 0.0169, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 0.6659772755609369, | |
| "learning_rate": 7.618520568160857e-06, | |
| "loss": 0.0292, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.4324913671742698, | |
| "learning_rate": 7.581627006087439e-06, | |
| "loss": 0.0186, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.4677680379599719, | |
| "learning_rate": 7.54473344401402e-06, | |
| "loss": 0.0207, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.5611844573800361, | |
| "learning_rate": 7.507839881940602e-06, | |
| "loss": 0.0214, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 0.7203601548324894, | |
| "learning_rate": 7.470946319867184e-06, | |
| "loss": 0.0208, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 0.7513235652150956, | |
| "learning_rate": 7.434052757793766e-06, | |
| "loss": 0.0297, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.6663228554103, | |
| "learning_rate": 7.397159195720347e-06, | |
| "loss": 0.0203, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 1.025646882506606, | |
| "learning_rate": 7.360265633646929e-06, | |
| "loss": 0.0165, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 1.104810538987063, | |
| "learning_rate": 7.323372071573512e-06, | |
| "loss": 0.025, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 1.1222939795401854, | |
| "learning_rate": 7.286478509500092e-06, | |
| "loss": 0.0184, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.40339534329472254, | |
| "learning_rate": 7.249584947426675e-06, | |
| "loss": 0.0172, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.3827747835522215, | |
| "learning_rate": 7.212691385353257e-06, | |
| "loss": 0.0176, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 0.5976609123895832, | |
| "learning_rate": 7.175797823279839e-06, | |
| "loss": 0.0288, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 1.248750777678682, | |
| "learning_rate": 7.13890426120642e-06, | |
| "loss": 0.018, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 0.7797224751733178, | |
| "learning_rate": 7.102010699133002e-06, | |
| "loss": 0.0202, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 0.7965109498444111, | |
| "learning_rate": 7.065117137059584e-06, | |
| "loss": 0.0193, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.7504440026508671, | |
| "learning_rate": 7.028223574986166e-06, | |
| "loss": 0.0208, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.3650634488062586, | |
| "learning_rate": 6.991330012912747e-06, | |
| "loss": 0.0156, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.4739908339986351, | |
| "learning_rate": 6.954436450839329e-06, | |
| "loss": 0.0253, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.40917205803434015, | |
| "learning_rate": 6.917542888765911e-06, | |
| "loss": 0.0223, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 1.0343715356497936, | |
| "learning_rate": 6.880649326692492e-06, | |
| "loss": 0.0239, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 1.2172584699084175, | |
| "learning_rate": 6.843755764619075e-06, | |
| "loss": 0.0198, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.1649553267102644, | |
| "learning_rate": 6.806862202545657e-06, | |
| "loss": 0.0155, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.750390467725635, | |
| "learning_rate": 6.769968640472239e-06, | |
| "loss": 0.0142, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 0.37624795209764544, | |
| "learning_rate": 6.73307507839882e-06, | |
| "loss": 0.017, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.7620471735392086, | |
| "learning_rate": 6.696181516325402e-06, | |
| "loss": 0.0196, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.699688822688377, | |
| "learning_rate": 6.659287954251984e-06, | |
| "loss": 0.0211, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.6564323081238257, | |
| "learning_rate": 6.622394392178565e-06, | |
| "loss": 0.0175, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 4.47664345340197, | |
| "learning_rate": 6.585500830105147e-06, | |
| "loss": 0.0268, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.5855473045668194, | |
| "learning_rate": 6.548607268031729e-06, | |
| "loss": 0.0197, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.3900950078903462, | |
| "learning_rate": 6.511713705958311e-06, | |
| "loss": 0.0139, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.6231685029419647, | |
| "learning_rate": 6.474820143884892e-06, | |
| "loss": 0.0207, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 1.4647351847789896, | |
| "learning_rate": 6.437926581811474e-06, | |
| "loss": 0.0251, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 0.8968411402674165, | |
| "learning_rate": 6.401033019738057e-06, | |
| "loss": 0.02, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 1.6599408896389465, | |
| "learning_rate": 6.364139457664637e-06, | |
| "loss": 0.0204, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.5235302401701253, | |
| "learning_rate": 6.32724589559122e-06, | |
| "loss": 0.021, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 1.0083480713032584, | |
| "learning_rate": 6.290352333517802e-06, | |
| "loss": 0.0176, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.0120381326163095, | |
| "learning_rate": 6.253458771444384e-06, | |
| "loss": 0.0234, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.3180666902395828, | |
| "learning_rate": 6.216565209370965e-06, | |
| "loss": 0.0227, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.948900773905558, | |
| "learning_rate": 6.179671647297547e-06, | |
| "loss": 0.0191, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.6367559831650437, | |
| "learning_rate": 6.142778085224129e-06, | |
| "loss": 0.0189, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.6717504563338927, | |
| "learning_rate": 6.10588452315071e-06, | |
| "loss": 0.0193, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 0.4138448431041914, | |
| "learning_rate": 6.068990961077292e-06, | |
| "loss": 0.02, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.6941281207519591, | |
| "learning_rate": 6.032097399003874e-06, | |
| "loss": 0.0163, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 1.054337846229134, | |
| "learning_rate": 5.995203836930457e-06, | |
| "loss": 0.0163, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 0.5690430661632323, | |
| "learning_rate": 5.958310274857037e-06, | |
| "loss": 0.0183, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 1.0019959557731468, | |
| "learning_rate": 5.92141671278362e-06, | |
| "loss": 0.0198, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 2.3627897902276356, | |
| "learning_rate": 5.884523150710202e-06, | |
| "loss": 0.0182, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.3794782980567666, | |
| "learning_rate": 5.847629588636783e-06, | |
| "loss": 0.0182, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.3295687746448634, | |
| "learning_rate": 5.810736026563365e-06, | |
| "loss": 0.0195, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.6192695411260121, | |
| "learning_rate": 5.773842464489947e-06, | |
| "loss": 0.0322, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.49045425590494857, | |
| "learning_rate": 5.736948902416529e-06, | |
| "loss": 0.0131, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.5148148095849162, | |
| "learning_rate": 5.70005534034311e-06, | |
| "loss": 0.0082, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 1.1798023900555077, | |
| "learning_rate": 5.663161778269692e-06, | |
| "loss": 0.0146, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 2.0171878876748512, | |
| "learning_rate": 5.626268216196274e-06, | |
| "loss": 0.0177, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 0.48523277809422005, | |
| "learning_rate": 5.589374654122855e-06, | |
| "loss": 0.0179, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.1666734279624884, | |
| "learning_rate": 5.552481092049437e-06, | |
| "loss": 0.0201, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 0.8382732166996802, | |
| "learning_rate": 5.51558752997602e-06, | |
| "loss": 0.0154, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 0.6194132737037249, | |
| "learning_rate": 5.478693967902602e-06, | |
| "loss": 0.0138, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.4187980161723267, | |
| "learning_rate": 5.441800405829183e-06, | |
| "loss": 0.0124, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 1.5698176121313157, | |
| "learning_rate": 5.404906843755765e-06, | |
| "loss": 0.0211, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 1.8056971380362028, | |
| "learning_rate": 5.368013281682347e-06, | |
| "loss": 0.0184, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.8795976081303007, | |
| "learning_rate": 5.331119719608928e-06, | |
| "loss": 0.0202, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.439105477045408, | |
| "learning_rate": 5.29422615753551e-06, | |
| "loss": 0.0202, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.29907687436659397, | |
| "learning_rate": 5.257332595462092e-06, | |
| "loss": 0.0099, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 1.0058773890984, | |
| "learning_rate": 5.220439033388674e-06, | |
| "loss": 0.0154, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 2.328331650132827, | |
| "learning_rate": 5.183545471315255e-06, | |
| "loss": 0.0197, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.6643990281031331, | |
| "learning_rate": 5.146651909241837e-06, | |
| "loss": 0.0169, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 0.8595932076596299, | |
| "learning_rate": 5.10975834716842e-06, | |
| "loss": 0.0223, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 1.1086116904916707, | |
| "learning_rate": 5.0728647850950005e-06, | |
| "loss": 0.0203, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 0.42085198590831047, | |
| "learning_rate": 5.035971223021583e-06, | |
| "loss": 0.0129, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.43355176911444604, | |
| "learning_rate": 4.999077660948165e-06, | |
| "loss": 0.0177, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 0.9661999333427626, | |
| "learning_rate": 4.962184098874746e-06, | |
| "loss": 0.0245, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 0.8742867014694161, | |
| "learning_rate": 4.925290536801328e-06, | |
| "loss": 0.0206, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.7590937058122642, | |
| "learning_rate": 4.88839697472791e-06, | |
| "loss": 0.0116, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.6407424307401028, | |
| "learning_rate": 4.851503412654492e-06, | |
| "loss": 0.0134, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.6194114934908045, | |
| "learning_rate": 4.814609850581074e-06, | |
| "loss": 0.0142, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 0.519041762974111, | |
| "learning_rate": 4.7777162885076554e-06, | |
| "loss": 0.017, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 0.8245823496587308, | |
| "learning_rate": 4.7408227264342374e-06, | |
| "loss": 0.0189, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 0.3531919757658654, | |
| "learning_rate": 4.703929164360819e-06, | |
| "loss": 0.0153, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.6794322400514526, | |
| "learning_rate": 4.667035602287401e-06, | |
| "loss": 0.0142, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.8540708015113813, | |
| "learning_rate": 4.630142040213983e-06, | |
| "loss": 0.0159, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.33219353692785597, | |
| "learning_rate": 4.593248478140565e-06, | |
| "loss": 0.0112, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 1.565286821401098, | |
| "learning_rate": 4.5563549160671465e-06, | |
| "loss": 0.0177, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 0.5652412008151505, | |
| "learning_rate": 4.5194613539937284e-06, | |
| "loss": 0.018, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 0.7094026702996225, | |
| "learning_rate": 4.4825677919203104e-06, | |
| "loss": 0.0112, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 1.2272305946952888, | |
| "learning_rate": 4.4456742298468916e-06, | |
| "loss": 0.0155, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 0.33889031674826015, | |
| "learning_rate": 4.408780667773474e-06, | |
| "loss": 0.0155, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 1.0052030625806332, | |
| "learning_rate": 4.3718871057000555e-06, | |
| "loss": 0.0171, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 0.8759824822777607, | |
| "learning_rate": 4.3349935436266375e-06, | |
| "loss": 0.011, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.9023761495743264, | |
| "learning_rate": 4.2980999815532195e-06, | |
| "loss": 0.0131, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 1.964594375832822, | |
| "learning_rate": 4.261206419479801e-06, | |
| "loss": 0.0187, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 0.20577618910606157, | |
| "learning_rate": 4.224312857406383e-06, | |
| "loss": 0.0138, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.43094266153941596, | |
| "learning_rate": 4.1874192953329646e-06, | |
| "loss": 0.0155, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.7550497919046862, | |
| "learning_rate": 4.1505257332595465e-06, | |
| "loss": 0.019, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 0.39737859899397643, | |
| "learning_rate": 4.1136321711861285e-06, | |
| "loss": 0.0135, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 1.00857408712709, | |
| "learning_rate": 4.0767386091127105e-06, | |
| "loss": 0.0114, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 0.6237895449638233, | |
| "learning_rate": 4.039845047039292e-06, | |
| "loss": 0.0163, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 0.040120689992290896, | |
| "learning_rate": 4.0029514849658744e-06, | |
| "loss": 0.0171, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.42882730907516153, | |
| "learning_rate": 3.966057922892456e-06, | |
| "loss": 0.0191, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.9108726072671286, | |
| "learning_rate": 3.9291643608190376e-06, | |
| "loss": 0.0185, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.5272498718341149, | |
| "learning_rate": 3.8922707987456195e-06, | |
| "loss": 0.0173, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.10708904724548446, | |
| "learning_rate": 3.855377236672201e-06, | |
| "loss": 0.0117, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.34583690447093357, | |
| "learning_rate": 3.818483674598783e-06, | |
| "loss": 0.0137, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 0.5823921518397646, | |
| "learning_rate": 3.7815901125253646e-06, | |
| "loss": 0.0166, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.6663422664501841, | |
| "learning_rate": 3.7446965504519466e-06, | |
| "loss": 0.0157, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.6376232439475752, | |
| "learning_rate": 3.707802988378528e-06, | |
| "loss": 0.0127, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 0.5322081396431619, | |
| "learning_rate": 3.6709094263051097e-06, | |
| "loss": 0.0121, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.25710001666211046, | |
| "learning_rate": 3.634015864231692e-06, | |
| "loss": 0.013, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.37957235167928116, | |
| "learning_rate": 3.5971223021582737e-06, | |
| "loss": 0.017, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 2.5532029064156547, | |
| "learning_rate": 3.5602287400848556e-06, | |
| "loss": 0.0144, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 2.1803077170850034, | |
| "learning_rate": 3.523335178011437e-06, | |
| "loss": 0.023, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 2.16703443584886, | |
| "learning_rate": 3.486441615938019e-06, | |
| "loss": 0.0188, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 0.07551800240220259, | |
| "learning_rate": 3.4495480538646007e-06, | |
| "loss": 0.0113, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.6989331221451501, | |
| "learning_rate": 3.4126544917911823e-06, | |
| "loss": 0.0194, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.6496164107042043, | |
| "learning_rate": 3.3757609297177647e-06, | |
| "loss": 0.0167, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 1.0013063957343065, | |
| "learning_rate": 3.3388673676443462e-06, | |
| "loss": 0.0135, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 1.1140749594913264, | |
| "learning_rate": 3.3019738055709282e-06, | |
| "loss": 0.0185, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 7.6331271025716365, | |
| "learning_rate": 3.2650802434975098e-06, | |
| "loss": 0.0276, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 1.1428692829815614, | |
| "learning_rate": 3.228186681424092e-06, | |
| "loss": 0.0128, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 0.041793369208575376, | |
| "learning_rate": 3.1912931193506737e-06, | |
| "loss": 0.0111, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 0.8486168336406528, | |
| "learning_rate": 3.1543995572772553e-06, | |
| "loss": 0.017, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.6725951925782523, | |
| "learning_rate": 3.1175059952038373e-06, | |
| "loss": 0.0145, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.6049207114447787, | |
| "learning_rate": 3.080612433130419e-06, | |
| "loss": 0.0169, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.27588155289452265, | |
| "learning_rate": 3.043718871057001e-06, | |
| "loss": 0.0092, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 0.4707157560687748, | |
| "learning_rate": 3.0068253089835823e-06, | |
| "loss": 0.018, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 0.35669193880545386, | |
| "learning_rate": 2.9699317469101647e-06, | |
| "loss": 0.0203, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 1.293228077308707, | |
| "learning_rate": 2.9330381848367463e-06, | |
| "loss": 0.0166, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 1.0181838445829068, | |
| "learning_rate": 2.8961446227633283e-06, | |
| "loss": 0.0154, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 0.5127757272780367, | |
| "learning_rate": 2.85925106068991e-06, | |
| "loss": 0.013, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 1.5370851778988082, | |
| "learning_rate": 2.8223574986164914e-06, | |
| "loss": 0.0293, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.4244549945275462, | |
| "learning_rate": 2.7854639365430734e-06, | |
| "loss": 0.0169, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.2734364373322682, | |
| "learning_rate": 2.7485703744696553e-06, | |
| "loss": 0.0116, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 1.1193751543770176, | |
| "learning_rate": 2.7116768123962373e-06, | |
| "loss": 0.0178, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.7704894398479883, | |
| "learning_rate": 2.674783250322819e-06, | |
| "loss": 0.0125, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 3.1008558260983445, | |
| "learning_rate": 2.637889688249401e-06, | |
| "loss": 0.0122, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 0.5179756509312545, | |
| "learning_rate": 2.6009961261759824e-06, | |
| "loss": 0.0184, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 0.7022943888833756, | |
| "learning_rate": 2.564102564102564e-06, | |
| "loss": 0.0076, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.25920563764099924, | |
| "learning_rate": 2.5272090020291464e-06, | |
| "loss": 0.0152, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.8652259016806108, | |
| "learning_rate": 2.490315439955728e-06, | |
| "loss": 0.0135, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.7916837531471141, | |
| "learning_rate": 2.4534218778823095e-06, | |
| "loss": 0.0143, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 0.5874610928577679, | |
| "learning_rate": 2.4165283158088915e-06, | |
| "loss": 0.0134, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 0.4016794628100762, | |
| "learning_rate": 2.3796347537354734e-06, | |
| "loss": 0.0189, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 1.6272336573727344, | |
| "learning_rate": 2.342741191662055e-06, | |
| "loss": 0.015, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 0.8579537864190223, | |
| "learning_rate": 2.305847629588637e-06, | |
| "loss": 0.0137, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 0.5560817121042592, | |
| "learning_rate": 2.268954067515219e-06, | |
| "loss": 0.0126, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.7159023263950224, | |
| "learning_rate": 2.2320605054418005e-06, | |
| "loss": 0.0149, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 4.058676695871013, | |
| "learning_rate": 2.1951669433683825e-06, | |
| "loss": 0.0143, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 1.9762340167157488, | |
| "learning_rate": 2.158273381294964e-06, | |
| "loss": 0.0131, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 0.6880794597389035, | |
| "learning_rate": 2.121379819221546e-06, | |
| "loss": 0.0161, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 1.1166888638088444, | |
| "learning_rate": 2.084486257148128e-06, | |
| "loss": 0.0149, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 0.5254579805784297, | |
| "learning_rate": 2.0475926950747095e-06, | |
| "loss": 0.016, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.45966794670628464, | |
| "learning_rate": 2.0106991330012915e-06, | |
| "loss": 0.0219, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.3169781496673292, | |
| "learning_rate": 1.9738055709278735e-06, | |
| "loss": 0.0181, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.7768794834685826, | |
| "learning_rate": 1.936912008854455e-06, | |
| "loss": 0.0289, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 0.4672959664269277, | |
| "learning_rate": 1.9000184467810368e-06, | |
| "loss": 0.0172, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 1.501176214000934, | |
| "learning_rate": 1.8631248847076186e-06, | |
| "loss": 0.0202, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 0.9123461332321072, | |
| "learning_rate": 1.8262313226342003e-06, | |
| "loss": 0.0186, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 0.5579770825783235, | |
| "learning_rate": 1.7893377605607823e-06, | |
| "loss": 0.0156, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 1.0793157994878555, | |
| "learning_rate": 1.752444198487364e-06, | |
| "loss": 0.0111, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 0.6229119468646244, | |
| "learning_rate": 1.715550636413946e-06, | |
| "loss": 0.0233, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.30488059454238836, | |
| "learning_rate": 1.6786570743405278e-06, | |
| "loss": 0.0122, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.559048874050897, | |
| "learning_rate": 1.6417635122671094e-06, | |
| "loss": 0.0114, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 1.4016309536448794, | |
| "learning_rate": 1.6048699501936912e-06, | |
| "loss": 0.0115, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 0.9407392806274478, | |
| "learning_rate": 1.5679763881202731e-06, | |
| "loss": 0.0139, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 0.5442497182397634, | |
| "learning_rate": 1.531082826046855e-06, | |
| "loss": 0.0101, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 1.1127989142385872, | |
| "learning_rate": 1.4941892639734369e-06, | |
| "loss": 0.0145, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 0.33784353335998774, | |
| "learning_rate": 1.4572957019000186e-06, | |
| "loss": 0.0121, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.46441423784320535, | |
| "learning_rate": 1.4204021398266004e-06, | |
| "loss": 0.0138, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.27429616166915427, | |
| "learning_rate": 1.3835085777531824e-06, | |
| "loss": 0.0115, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.42197086869726136, | |
| "learning_rate": 1.346615015679764e-06, | |
| "loss": 0.0157, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.8479098454856343, | |
| "learning_rate": 1.3097214536063457e-06, | |
| "loss": 0.0174, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 0.5165606880963921, | |
| "learning_rate": 1.2728278915329275e-06, | |
| "loss": 0.0109, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.7486171689925694, | |
| "learning_rate": 1.2359343294595095e-06, | |
| "loss": 0.0157, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.49242728311315703, | |
| "learning_rate": 1.1990407673860912e-06, | |
| "loss": 0.0137, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 1.361331802866714, | |
| "learning_rate": 1.162147205312673e-06, | |
| "loss": 0.0167, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 1.1385403748575034, | |
| "learning_rate": 1.1252536432392548e-06, | |
| "loss": 0.0251, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.08947600380281688, | |
| "learning_rate": 1.0883600811658367e-06, | |
| "loss": 0.0123, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 0.7741530990096059, | |
| "learning_rate": 1.0514665190924185e-06, | |
| "loss": 0.0107, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 0.6521877399219315, | |
| "learning_rate": 1.0145729570190003e-06, | |
| "loss": 0.0154, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 0.3481535622815111, | |
| "learning_rate": 9.77679394945582e-07, | |
| "loss": 0.0115, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 0.5721956281850312, | |
| "learning_rate": 9.407858328721639e-07, | |
| "loss": 0.012, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 1.4702700601709895, | |
| "learning_rate": 9.038922707987458e-07, | |
| "loss": 0.0246, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.939439483987666, | |
| "learning_rate": 8.669987087253274e-07, | |
| "loss": 0.017, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 0.048547956632345776, | |
| "learning_rate": 8.301051466519093e-07, | |
| "loss": 0.0105, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 0.9040698761413287, | |
| "learning_rate": 7.932115845784912e-07, | |
| "loss": 0.0171, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 0.05292960515314995, | |
| "learning_rate": 7.563180225050728e-07, | |
| "loss": 0.0167, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.745288754130905, | |
| "learning_rate": 7.194244604316547e-07, | |
| "loss": 0.0093, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 0.3231742072867581, | |
| "learning_rate": 6.825308983582366e-07, | |
| "loss": 0.0153, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 1.104730096451692, | |
| "learning_rate": 6.456373362848183e-07, | |
| "loss": 0.0172, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 0.27650101996659654, | |
| "learning_rate": 6.087437742114002e-07, | |
| "loss": 0.0121, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.6137572949630508, | |
| "learning_rate": 5.71850212137982e-07, | |
| "loss": 0.0181, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.3947830907180403, | |
| "learning_rate": 5.349566500645638e-07, | |
| "loss": 0.011, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 0.33990532766618414, | |
| "learning_rate": 4.980630879911455e-07, | |
| "loss": 0.014, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.7278388922473019, | |
| "learning_rate": 4.611695259177274e-07, | |
| "loss": 0.0144, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.8133836855766845, | |
| "learning_rate": 4.242759638443092e-07, | |
| "loss": 0.0154, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 0.7134865138510943, | |
| "learning_rate": 3.8738240177089103e-07, | |
| "loss": 0.0261, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 0.5721112977333791, | |
| "learning_rate": 3.504888396974728e-07, | |
| "loss": 0.0158, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 0.7556723035420685, | |
| "learning_rate": 3.1359527762405467e-07, | |
| "loss": 0.0142, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.20174754283725183, | |
| "learning_rate": 2.7670171555063644e-07, | |
| "loss": 0.0168, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.4654226465048818, | |
| "learning_rate": 2.3980815347721825e-07, | |
| "loss": 0.0171, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.5512068706949325, | |
| "learning_rate": 2.0291459140380005e-07, | |
| "loss": 0.0308, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 0.9662650446935767, | |
| "learning_rate": 1.6602102933038187e-07, | |
| "loss": 0.0176, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.357835040985804, | |
| "learning_rate": 1.2912746725696369e-07, | |
| "loss": 0.0179, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.7756082916304456, | |
| "learning_rate": 9.223390518354548e-08, | |
| "loss": 0.0131, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 0.1608432595701506, | |
| "learning_rate": 5.5340343110127284e-08, | |
| "loss": 0.0114, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6872400095655116, | |
| "learning_rate": 1.8446781036709097e-08, | |
| "loss": 0.0184, | |
| "step": 6020 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 6024, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |