| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 1000, | |
| "global_step": 1317, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007593014426727411, | |
| "eval_accuracy": 0.14927598308927809, | |
| "eval_loss": 7.863234519958496, | |
| "eval_runtime": 368.109, | |
| "eval_samples_per_second": 192.592, | |
| "eval_steps_per_second": 0.752, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007593014426727411, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1e-05, | |
| "loss": 7.8621, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015186028853454821, | |
| "grad_norm": 9.625, | |
| "learning_rate": 2e-05, | |
| "loss": 7.8394, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.022779043280182234, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 3e-05, | |
| "loss": 7.7726, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.030372057706909643, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 4e-05, | |
| "loss": 7.6201, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.037965072133637055, | |
| "grad_norm": 8.375, | |
| "learning_rate": 5e-05, | |
| "loss": 7.3344, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04555808656036447, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 6e-05, | |
| "loss": 7.0089, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05315110098709187, | |
| "grad_norm": 5.0, | |
| "learning_rate": 7e-05, | |
| "loss": 6.5946, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.060744115413819286, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8e-05, | |
| "loss": 6.0679, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0683371298405467, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 9e-05, | |
| "loss": 5.6963, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07593014426727411, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.0001, | |
| "loss": 5.4276, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08352315869400152, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 9.998334155418425e-05, | |
| "loss": 5.2216, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.09111617312072894, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.993337731688963e-05, | |
| "loss": 5.0464, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09870918754745633, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.985014058117778e-05, | |
| "loss": 4.9161, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10630220197418375, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.973368681083473e-05, | |
| "loss": 4.8156, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11389521640091116, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.958409360341343e-05, | |
| "loss": 4.735, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12148823082763857, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 9.940146063852748e-05, | |
| "loss": 4.6663, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.12908124525436598, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 9.918590961143088e-05, | |
| "loss": 4.6082, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1366742596810934, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.893758415192783e-05, | |
| "loss": 4.5533, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1442672741078208, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 9.865664972866682e-05, | |
| "loss": 4.5055, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.15186028853454822, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 9.834329353888254e-05, | |
| "loss": 4.4668, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15945330296127563, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 9.799772438365936e-05, | |
| "loss": 4.4333, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.16704631738800305, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.762017252879917e-05, | |
| "loss": 4.4072, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.17463933181473046, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 9.721088955138664e-05, | |
| "loss": 4.3759, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.18223234624145787, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.67701481721539e-05, | |
| "loss": 4.3542, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.18982536066818528, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 9.629824207375632e-05, | |
| "loss": 4.3318, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.19741837509491267, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 9.57954857050807e-05, | |
| "loss": 4.3142, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.20501138952164008, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 9.52622140717161e-05, | |
| "loss": 4.2948, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2126044039483675, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 9.46987825127269e-05, | |
| "loss": 4.282, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2201974183750949, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 9.410556646387694e-05, | |
| "loss": 4.2621, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.22779043280182232, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 9.348296120746251e-05, | |
| "loss": 4.2483, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.23538344722854973, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 9.283138160892074e-05, | |
| "loss": 4.2307, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.24297646165527714, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 9.215126184038911e-05, | |
| "loss": 4.2256, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2505694760820046, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 9.144305509140012e-05, | |
| "loss": 4.2123, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.25816249050873197, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 9.070723326690396e-05, | |
| "loss": 4.1984, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.26575550493545935, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 8.994428667282032e-05, | |
| "loss": 4.1908, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2733485193621868, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 8.915472368932916e-05, | |
| "loss": 4.1878, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2809415337889142, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 8.83390704321176e-05, | |
| "loss": 4.1766, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2885345482156416, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 8.749787040180923e-05, | |
| "loss": 4.1719, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.296127562642369, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 8.663168412180907e-05, | |
| "loss": 4.1615, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.30372057706909644, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 8.574108876480557e-05, | |
| "loss": 4.1505, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3113135914958238, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 8.482667776817868e-05, | |
| "loss": 4.1422, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.31890660592255127, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 8.388906043857001e-05, | |
| "loss": 4.1427, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.32649962034927865, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 8.292886154587884e-05, | |
| "loss": 4.14, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3340926347760061, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 8.194672090695419e-05, | |
| "loss": 4.127, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3416856492027335, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 8.094329295926075e-05, | |
| "loss": 4.1251, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3492786636294609, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 7.991924632480238e-05, | |
| "loss": 4.1234, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3568716780561883, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 7.887526336459401e-05, | |
| "loss": 4.1151, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.36446469248291574, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 7.781203972397866e-05, | |
| "loss": 4.1102, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.3720577069096431, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 7.673028386909259e-05, | |
| "loss": 4.1107, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.37965072133637057, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 7.563071661478758e-05, | |
| "loss": 4.1017, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.38724373576309795, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 7.451407064432471e-05, | |
| "loss": 4.096, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.39483675018982534, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 7.338109002115976e-05, | |
| "loss": 4.0982, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4024297646165528, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 7.223252969314557e-05, | |
| "loss": 4.0925, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.41002277904328016, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 7.106915498948175e-05, | |
| "loss": 4.0917, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4176157934700076, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 6.989174111074688e-05, | |
| "loss": 4.0865, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.425208807896735, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 6.870107261235304e-05, | |
| "loss": 4.0831, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.4328018223234624, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 6.749794288176681e-05, | |
| "loss": 4.0833, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4403948367501898, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 6.628315360984526e-05, | |
| "loss": 4.0813, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.44798785117691725, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 6.505751425663892e-05, | |
| "loss": 4.0757, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.45558086560364464, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 6.382184151201804e-05, | |
| "loss": 4.0782, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4631738800303721, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 6.257695875148106e-05, | |
| "loss": 4.0763, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.47076689445709946, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 6.132369548750851e-05, | |
| "loss": 4.0728, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.4783599088838269, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 6.0062886816827494e-05, | |
| "loss": 4.0696, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.4859529233105543, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 5.8795372863954924e-05, | |
| "loss": 4.0634, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.4935459377372817, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 5.7521998221391004e-05, | |
| "loss": 4.0658, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5011389521640092, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 5.6243611386835205e-05, | |
| "loss": 4.0631, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5087319665907365, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 5.49610641978001e-05, | |
| "loss": 4.0639, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5163249810174639, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 5.36752112639999e-05, | |
| "loss": 4.0631, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5239179954441914, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 5.2386909397891595e-05, | |
| "loss": 4.0592, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5315110098709187, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 5.109701704374841e-05, | |
| "loss": 4.0625, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5391040242976461, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 4.980639370564593e-05, | |
| "loss": 4.0632, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.5466970387243736, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 4.851589937474202e-05, | |
| "loss": 4.0557, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.554290053151101, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 4.722639395623215e-05, | |
| "loss": 4.0654, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.5618830675778284, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 4.593873669636204e-05, | |
| "loss": 4.0542, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 4.465378560987944e-05, | |
| "loss": 4.0584, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5770690964312832, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 4.337239690830632e-05, | |
| "loss": 4.058, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.5846621108580107, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 4.209542442941284e-05, | |
| "loss": 4.0574, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.592255125284738, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 4.082371906827295e-05, | |
| "loss": 4.0569, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5998481397114654, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 3.9558128210280706e-05, | |
| "loss": 4.0512, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6074411541381929, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.829949516650543e-05, | |
| "loss": 4.054, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6150341685649203, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 3.704865861176159e-05, | |
| "loss": 4.0514, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6226271829916477, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 3.5806452025768056e-05, | |
| "loss": 4.0564, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6302201974183751, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 3.457370313776897e-05, | |
| "loss": 4.0529, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6378132118451025, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 3.335123337498662e-05, | |
| "loss": 4.055, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.6454062262718299, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.213985731527313e-05, | |
| "loss": 4.053, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6529992406985573, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 3.0940382144326666e-05, | |
| "loss": 4.0494, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.6605922551252847, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 2.9753607117832903e-05, | |
| "loss": 4.0488, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6681852695520122, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 2.858032302889084e-05, | |
| "loss": 4.0548, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6757782839787395, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 2.7421311681077344e-05, | |
| "loss": 4.0507, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.683371298405467, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 2.627734536750191e-05, | |
| "loss": 4.0524, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6909643128321944, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 2.5149186356198562e-05, | |
| "loss": 4.0494, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.6985573272589218, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 2.4037586382197685e-05, | |
| "loss": 4.0516, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7061503416856492, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 2.2943286146616788e-05, | |
| "loss": 4.0507, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7137433561123766, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 2.1867014823103044e-05, | |
| "loss": 4.0507, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.721336370539104, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 2.080948957195749e-05, | |
| "loss": 4.0506, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7289293849658315, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 1.9771415062263965e-05, | |
| "loss": 4.0502, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.7365223993925588, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 1.875348300234131e-05, | |
| "loss": 4.0503, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.7441154138192863, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 1.7756371678832145e-05, | |
| "loss": 4.049, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.7517084282460137, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 1.6780745504734617e-05, | |
| "loss": 4.0465, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.7593014426727411, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 1.58272545766791e-05, | |
| "loss": 4.0503, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7593014426727411, | |
| "eval_accuracy": 0.30700041923136506, | |
| "eval_loss": 4.048835277557373, | |
| "eval_runtime": 369.3769, | |
| "eval_samples_per_second": 191.931, | |
| "eval_steps_per_second": 0.75, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7668944570994685, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 1.4896534241743988e-05, | |
| "loss": 4.0501, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.7744874715261959, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 1.398920467410002e-05, | |
| "loss": 4.0469, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.7820804859529233, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 1.3105870461764847e-05, | |
| "loss": 4.0512, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.7896735003796507, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 1.2247120203742945e-05, | |
| "loss": 4.0463, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.7972665148063781, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 1.141352611782005e-05, | |
| "loss": 4.0457, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8048595292331056, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 1.0605643659272668e-05, | |
| "loss": 4.0511, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.812452543659833, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 9.824011150747226e-06, | |
| "loss": 4.0533, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.8200455580865603, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 9.069149423555378e-06, | |
| "loss": 4.0541, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.8276385725132878, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 8.341561470624365e-06, | |
| "loss": 4.0464, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.8352315869400152, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 7.641732111333793e-06, | |
| "loss": 4.0518, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8428246013667426, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 6.970127668462129e-06, | |
| "loss": 4.0492, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.85041761579347, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 6.3271956574582296e-06, | |
| "loss": 4.0489, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.8580106302201974, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 5.7133644882448376e-06, | |
| "loss": 4.0462, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.8656036446469249, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 5.12904317975283e-06, | |
| "loss": 4.0466, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.8731966590736523, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 4.574621087376501e-06, | |
| "loss": 4.0496, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.8807896735003796, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 4.050467643531214e-06, | |
| "loss": 4.0478, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.8883826879271071, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 3.556932111486727e-06, | |
| "loss": 4.0524, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.8959757023538345, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 3.0943433526397838e-06, | |
| "loss": 4.0526, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.9035687167805618, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 2.663009607381345e-06, | |
| "loss": 4.0484, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 2.2632182897043773e-06, | |
| "loss": 4.0489, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9187547456340167, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 1.8952357956889988e-06, | |
| "loss": 4.0477, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.9263477600607442, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.559307325992726e-06, | |
| "loss": 4.0498, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.9339407744874715, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 1.2556567224639714e-06, | |
| "loss": 4.0451, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.9415337889141989, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 9.844863189877596e-07, | |
| "loss": 4.0513, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.9491268033409264, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 7.459768066630291e-07, | |
| "loss": 4.0487, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.9567198177676538, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 5.402871134012666e-07, | |
| "loss": 4.0509, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.9643128321943811, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 3.6755429802690087e-07, | |
| "loss": 4.048, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.9719058466211086, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 2.2789345894972945e-07, | |
| "loss": 4.0524, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.979498861047836, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 1.213976574705733e-07, | |
| "loss": 4.0501, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.9870918754745635, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 4.813785577097996e-08, | |
| "loss": 4.0455, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.9946848899012908, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 8.162869628425274e-09, | |
| "loss": 4.0489, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1317, | |
| "total_flos": 1.0857282238836376e+19, | |
| "train_loss": 4.36365500488513, | |
| "train_runtime": 5657.2998, | |
| "train_samples_per_second": 238.343, | |
| "train_steps_per_second": 0.233 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1317, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0857282238836376e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |