| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.15186490098408456, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006074596039363382, | |
| "grad_norm": 2.236969232559204, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2628, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0012149192078726764, | |
| "grad_norm": 1.873207688331604, | |
| "learning_rate": 0.0002, | |
| "loss": 1.6589, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0018223788118090147, | |
| "grad_norm": 1.7144545316696167, | |
| "learning_rate": 0.0002, | |
| "loss": 1.6129, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0024298384157453528, | |
| "grad_norm": 1.6338379383087158, | |
| "learning_rate": 0.0002, | |
| "loss": 1.565, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.003037298019681691, | |
| "grad_norm": 1.733494758605957, | |
| "learning_rate": 0.0002, | |
| "loss": 1.5043, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0036447576236180294, | |
| "grad_norm": 1.6118093729019165, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4686, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.004252217227554368, | |
| "grad_norm": 1.5905429124832153, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4902, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0048596768314907056, | |
| "grad_norm": 1.667108416557312, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4311, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.005467136435427044, | |
| "grad_norm": 1.813370943069458, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3872, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.006074596039363382, | |
| "grad_norm": 1.6819970607757568, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4375, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.006682055643299721, | |
| "grad_norm": 1.6235817670822144, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4136, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.007289515247236059, | |
| "grad_norm": 1.6433429718017578, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3771, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.007896974851172398, | |
| "grad_norm": 1.6093605756759644, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4332, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.008504434455108735, | |
| "grad_norm": 1.63667893409729, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4148, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.009111894059045073, | |
| "grad_norm": 1.5507344007492065, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3764, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.009719353662981411, | |
| "grad_norm": 1.6159253120422363, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3141, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.01032681326691775, | |
| "grad_norm": 1.6511636972427368, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4633, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.010934272870854089, | |
| "grad_norm": 1.8297101259231567, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3747, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.011541732474790426, | |
| "grad_norm": 1.8446108102798462, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3742, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.012149192078726764, | |
| "grad_norm": 1.6419591903686523, | |
| "learning_rate": 0.0002, | |
| "loss": 1.329, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.012756651682663102, | |
| "grad_norm": 1.595116376876831, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3658, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.013364111286599442, | |
| "grad_norm": 1.6085125207901, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3062, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.01397157089053578, | |
| "grad_norm": 1.6018210649490356, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3563, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.014579030494472118, | |
| "grad_norm": 1.5757384300231934, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4252, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.015186490098408455, | |
| "grad_norm": 1.6396926641464233, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3044, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.015793949702344795, | |
| "grad_norm": 1.7595162391662598, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3885, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.01640140930628113, | |
| "grad_norm": 1.5528141260147095, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2728, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.01700886891021747, | |
| "grad_norm": 1.536747932434082, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3512, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.01761632851415381, | |
| "grad_norm": 1.6435034275054932, | |
| "learning_rate": 0.0002, | |
| "loss": 1.338, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.018223788118090146, | |
| "grad_norm": 1.5331226587295532, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3519, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.018831247722026486, | |
| "grad_norm": 1.6326663494110107, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3089, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.019438707325962822, | |
| "grad_norm": 1.567794919013977, | |
| "learning_rate": 0.0002, | |
| "loss": 1.253, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.020046166929899162, | |
| "grad_norm": 1.5589838027954102, | |
| "learning_rate": 0.0002, | |
| "loss": 1.33, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.0206536265338355, | |
| "grad_norm": 1.5770776271820068, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3316, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.021261086137771838, | |
| "grad_norm": 1.4471594095230103, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2864, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.021868545741708177, | |
| "grad_norm": 1.5740830898284912, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2921, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.022476005345644513, | |
| "grad_norm": 1.7635418176651, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2212, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.023083464949580853, | |
| "grad_norm": 1.5588957071304321, | |
| "learning_rate": 0.0002, | |
| "loss": 1.226, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.023690924553517193, | |
| "grad_norm": 1.6329195499420166, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3694, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.02429838415745353, | |
| "grad_norm": 1.5749831199645996, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3151, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.024905843761389868, | |
| "grad_norm": 1.776865839958191, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2971, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.025513303365326204, | |
| "grad_norm": 1.4414085149765015, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3434, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.026120762969262544, | |
| "grad_norm": 1.672532081604004, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2647, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.026728222573198884, | |
| "grad_norm": 1.7707642316818237, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2875, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.02733568217713522, | |
| "grad_norm": 1.510501503944397, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2962, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.02794314178107156, | |
| "grad_norm": 1.5336624383926392, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3686, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.028550601385007895, | |
| "grad_norm": 1.785957932472229, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2887, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.029158060988944235, | |
| "grad_norm": 1.5779095888137817, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3048, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.029765520592880575, | |
| "grad_norm": 1.9468932151794434, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2892, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.03037298019681691, | |
| "grad_norm": 1.5167356729507446, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3858, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03098043980075325, | |
| "grad_norm": 1.3795329332351685, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3023, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.03158789940468959, | |
| "grad_norm": 1.5920099020004272, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3765, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.032195359008625926, | |
| "grad_norm": 1.462296724319458, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2663, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.03280281861256226, | |
| "grad_norm": 1.3863338232040405, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2386, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.033410278216498605, | |
| "grad_norm": 1.8175960779190063, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3238, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.03401773782043494, | |
| "grad_norm": 1.7048077583312988, | |
| "learning_rate": 0.0002, | |
| "loss": 1.257, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.03462519742437128, | |
| "grad_norm": 1.5479084253311157, | |
| "learning_rate": 0.0002, | |
| "loss": 1.23, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.03523265702830762, | |
| "grad_norm": 1.515726089477539, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2893, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.03584011663224396, | |
| "grad_norm": 1.7023965120315552, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2783, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.03644757623618029, | |
| "grad_norm": 1.7180956602096558, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2578, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.03705503584011663, | |
| "grad_norm": 1.3848283290863037, | |
| "learning_rate": 0.0002, | |
| "loss": 1.349, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.03766249544405297, | |
| "grad_norm": 1.4944247007369995, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2718, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.03826995504798931, | |
| "grad_norm": 1.581346869468689, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2798, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.038877414651925644, | |
| "grad_norm": 1.490548849105835, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2878, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.03948487425586199, | |
| "grad_norm": 1.4963841438293457, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2777, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.040092333859798324, | |
| "grad_norm": 1.729665994644165, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2781, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.04069979346373466, | |
| "grad_norm": 1.5069007873535156, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2978, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.041307253067671, | |
| "grad_norm": 4.421722412109375, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2765, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.04191471267160734, | |
| "grad_norm": 1.6383577585220337, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3364, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.042522172275543675, | |
| "grad_norm": 1.9025250673294067, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3184, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.04312963187948001, | |
| "grad_norm": 1.4485440254211426, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2749, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.043737091483416354, | |
| "grad_norm": 1.3731114864349365, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2254, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.04434455108735269, | |
| "grad_norm": 2.1447882652282715, | |
| "learning_rate": 0.0002, | |
| "loss": 1.227, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.04495201069128903, | |
| "grad_norm": 1.4844560623168945, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2656, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.04555947029522537, | |
| "grad_norm": 1.6620945930480957, | |
| "learning_rate": 0.0002, | |
| "loss": 1.299, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.046166929899161706, | |
| "grad_norm": 1.4194166660308838, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2961, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.04677438950309804, | |
| "grad_norm": 1.4977781772613525, | |
| "learning_rate": 0.0002, | |
| "loss": 1.324, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.047381849107034385, | |
| "grad_norm": 1.5294965505599976, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2344, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.04798930871097072, | |
| "grad_norm": 1.6185721158981323, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3811, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.04859676831490706, | |
| "grad_norm": 1.8349323272705078, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2636, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.0492042279188434, | |
| "grad_norm": 1.4457494020462036, | |
| "learning_rate": 0.0002, | |
| "loss": 1.278, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.049811687522779736, | |
| "grad_norm": 1.3572221994400024, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2306, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.05041914712671607, | |
| "grad_norm": 1.3937549591064453, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3023, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.05102660673065241, | |
| "grad_norm": 1.450095295906067, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2393, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.05163406633458875, | |
| "grad_norm": 1.4043591022491455, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2574, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.05224152593852509, | |
| "grad_norm": 1.4549307823181152, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2938, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.052848985542461424, | |
| "grad_norm": 1.4602893590927124, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1861, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.05345644514639777, | |
| "grad_norm": 1.867261290550232, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3015, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.0540639047503341, | |
| "grad_norm": 1.6120606660842896, | |
| "learning_rate": 0.0002, | |
| "loss": 1.268, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.05467136435427044, | |
| "grad_norm": 1.6458678245544434, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2661, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.05527882395820678, | |
| "grad_norm": 1.729551076889038, | |
| "learning_rate": 0.0002, | |
| "loss": 1.28, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.05588628356214312, | |
| "grad_norm": 1.5503125190734863, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2612, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.056493743166079455, | |
| "grad_norm": 1.5638302564620972, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3259, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.05710120277001579, | |
| "grad_norm": 1.4959323406219482, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2489, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.057708662373952134, | |
| "grad_norm": 1.5417566299438477, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2675, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.05831612197788847, | |
| "grad_norm": 1.5263869762420654, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3438, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.058923581581824806, | |
| "grad_norm": 1.4479107856750488, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2868, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.05953104118576115, | |
| "grad_norm": 1.678945541381836, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2204, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.060138500789697485, | |
| "grad_norm": 1.431054949760437, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2624, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.06074596039363382, | |
| "grad_norm": 1.6529697179794312, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2867, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.061353419997570165, | |
| "grad_norm": 1.4838560819625854, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2556, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.0619608796015065, | |
| "grad_norm": 1.303206443786621, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1895, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.06256833920544284, | |
| "grad_norm": 1.6381781101226807, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2975, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.06317579880937918, | |
| "grad_norm": 1.589048981666565, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2776, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.06378325841331552, | |
| "grad_norm": 1.71570885181427, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2703, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.06439071801725185, | |
| "grad_norm": 1.416674017906189, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2465, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.06499817762118819, | |
| "grad_norm": 1.773748517036438, | |
| "learning_rate": 0.0002, | |
| "loss": 1.339, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.06560563722512452, | |
| "grad_norm": 1.5272996425628662, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2973, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.06621309682906086, | |
| "grad_norm": 1.3926866054534912, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3862, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.06682055643299721, | |
| "grad_norm": 1.576254963874817, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2522, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.06742801603693355, | |
| "grad_norm": 1.6619774103164673, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3027, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.06803547564086988, | |
| "grad_norm": 1.4636040925979614, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3649, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.06864293524480622, | |
| "grad_norm": 1.6307445764541626, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2553, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.06925039484874256, | |
| "grad_norm": 1.531379222869873, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2578, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.06985785445267889, | |
| "grad_norm": 1.5244723558425903, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2745, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.07046531405661524, | |
| "grad_norm": 1.7863515615463257, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2835, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.07107277366055158, | |
| "grad_norm": 1.5718461275100708, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2352, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.07168023326448791, | |
| "grad_norm": 1.5710781812667847, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2357, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.07228769286842425, | |
| "grad_norm": 1.4836217164993286, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2811, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.07289515247236059, | |
| "grad_norm": 1.5769175291061401, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3826, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.07350261207629692, | |
| "grad_norm": 1.552350640296936, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2512, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.07411007168023326, | |
| "grad_norm": 1.4635241031646729, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3651, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.07471753128416961, | |
| "grad_norm": 1.6063421964645386, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3135, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.07532499088810594, | |
| "grad_norm": 1.3370627164840698, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3172, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.07593245049204228, | |
| "grad_norm": 1.691219449043274, | |
| "learning_rate": 0.0002, | |
| "loss": 1.114, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.07653991009597862, | |
| "grad_norm": 1.4442918300628662, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2884, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.07714736969991495, | |
| "grad_norm": 1.3976494073867798, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2585, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.07775482930385129, | |
| "grad_norm": 1.357978105545044, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2449, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.07836228890778764, | |
| "grad_norm": 1.4768654108047485, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1673, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.07896974851172398, | |
| "grad_norm": 1.3699668645858765, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3029, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.07957720811566031, | |
| "grad_norm": 1.494268774986267, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2156, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.08018466771959665, | |
| "grad_norm": 1.4966034889221191, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2387, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.08079212732353298, | |
| "grad_norm": 1.5984331369400024, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2559, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.08139958692746932, | |
| "grad_norm": 1.6565812826156616, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2035, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.08200704653140566, | |
| "grad_norm": 1.3882875442504883, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2266, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.082614506135342, | |
| "grad_norm": 1.4133611917495728, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2908, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.08322196573927834, | |
| "grad_norm": 1.5866031646728516, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2541, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.08382942534321468, | |
| "grad_norm": 1.7129892110824585, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2653, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.08443688494715101, | |
| "grad_norm": 1.821727991104126, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2413, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.08504434455108735, | |
| "grad_norm": 1.5216853618621826, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2449, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.08565180415502369, | |
| "grad_norm": 1.3088600635528564, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1995, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.08625926375896002, | |
| "grad_norm": 1.467633843421936, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2901, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.08686672336289637, | |
| "grad_norm": 1.467429518699646, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2593, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.08747418296683271, | |
| "grad_norm": 1.5163699388504028, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3415, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.08808164257076904, | |
| "grad_norm": 1.4762097597122192, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2328, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.08868910217470538, | |
| "grad_norm": 1.3106896877288818, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2625, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.08929656177864172, | |
| "grad_norm": 1.7591036558151245, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2654, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.08990402138257805, | |
| "grad_norm": 1.425759196281433, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2492, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.0905114809865144, | |
| "grad_norm": 1.4507300853729248, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3173, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.09111894059045074, | |
| "grad_norm": 1.564005970954895, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2037, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.09172640019438708, | |
| "grad_norm": 1.4253226518630981, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2143, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.09233385979832341, | |
| "grad_norm": 1.4880207777023315, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2818, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.09294131940225975, | |
| "grad_norm": 1.439846396446228, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2082, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.09354877900619608, | |
| "grad_norm": 1.5116918087005615, | |
| "learning_rate": 0.0002, | |
| "loss": 1.279, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.09415623861013242, | |
| "grad_norm": 1.2679647207260132, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2541, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.09476369821406877, | |
| "grad_norm": 1.4966439008712769, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2426, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.0953711578180051, | |
| "grad_norm": 1.795647144317627, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2582, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.09597861742194144, | |
| "grad_norm": 1.485668659210205, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2277, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.09658607702587778, | |
| "grad_norm": 1.4666286706924438, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1659, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.09719353662981411, | |
| "grad_norm": 1.4714016914367676, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2541, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.09780099623375045, | |
| "grad_norm": 1.351139783859253, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2047, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.0984084558376868, | |
| "grad_norm": 1.4304152727127075, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2028, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.09901591544162314, | |
| "grad_norm": 1.3749319314956665, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2509, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.09962337504555947, | |
| "grad_norm": 1.5823308229446411, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2278, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.10023083464949581, | |
| "grad_norm": 1.3146113157272339, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2092, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.10083829425343215, | |
| "grad_norm": 1.5478814840316772, | |
| "learning_rate": 0.0002, | |
| "loss": 1.284, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.10144575385736848, | |
| "grad_norm": 1.6172051429748535, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2521, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.10205321346130482, | |
| "grad_norm": 1.52104914188385, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2477, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.10266067306524117, | |
| "grad_norm": 1.5709279775619507, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2424, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.1032681326691775, | |
| "grad_norm": 1.4448177814483643, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2296, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.10387559227311384, | |
| "grad_norm": 1.3035105466842651, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2177, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.10448305187705018, | |
| "grad_norm": 1.6021480560302734, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2391, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.10509051148098651, | |
| "grad_norm": 1.5413893461227417, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2627, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.10569797108492285, | |
| "grad_norm": 1.582407832145691, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2344, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.10630543068885918, | |
| "grad_norm": 1.4765949249267578, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2268, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.10691289029279553, | |
| "grad_norm": 1.6818124055862427, | |
| "learning_rate": 0.0002, | |
| "loss": 1.229, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.10752034989673187, | |
| "grad_norm": 1.570087194442749, | |
| "learning_rate": 0.0002, | |
| "loss": 1.322, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.1081278095006682, | |
| "grad_norm": 1.5548206567764282, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1993, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.10873526910460454, | |
| "grad_norm": 1.4159990549087524, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2136, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.10934272870854088, | |
| "grad_norm": 1.4259227514266968, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2826, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.10995018831247722, | |
| "grad_norm": 1.721754550933838, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2958, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.11055764791641357, | |
| "grad_norm": 1.5920616388320923, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2152, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.1111651075203499, | |
| "grad_norm": 1.4243106842041016, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2273, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.11177256712428624, | |
| "grad_norm": 1.5096100568771362, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2232, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.11238002672822257, | |
| "grad_norm": 1.573041558265686, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2303, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.11298748633215891, | |
| "grad_norm": 1.546109676361084, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2366, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.11359494593609525, | |
| "grad_norm": 1.5966582298278809, | |
| "learning_rate": 0.0002, | |
| "loss": 1.174, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.11420240554003158, | |
| "grad_norm": 1.5600007772445679, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1639, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.11480986514396793, | |
| "grad_norm": 1.7233026027679443, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2161, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.11541732474790427, | |
| "grad_norm": 1.5901647806167603, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2153, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.1160247843518406, | |
| "grad_norm": 1.376562237739563, | |
| "learning_rate": 0.0002, | |
| "loss": 1.266, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.11663224395577694, | |
| "grad_norm": 1.656231164932251, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2664, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.11723970355971328, | |
| "grad_norm": 1.356429100036621, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2244, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.11784716316364961, | |
| "grad_norm": 1.5584162473678589, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2545, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.11845462276758595, | |
| "grad_norm": 1.5809051990509033, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2847, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.1190620823715223, | |
| "grad_norm": 1.6206623315811157, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1984, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.11966954197545863, | |
| "grad_norm": 1.5814268589019775, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2822, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.12027700157939497, | |
| "grad_norm": 1.6698721647262573, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2771, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.12088446118333131, | |
| "grad_norm": 1.7568098306655884, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2153, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.12149192078726764, | |
| "grad_norm": 1.5525120496749878, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1899, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.12209938039120398, | |
| "grad_norm": 1.451067328453064, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2241, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.12270683999514033, | |
| "grad_norm": 1.6643682718276978, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2726, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.12331429959907667, | |
| "grad_norm": 1.7143374681472778, | |
| "learning_rate": 0.0002, | |
| "loss": 1.277, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.123921759203013, | |
| "grad_norm": 1.7562227249145508, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2066, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.12452921880694934, | |
| "grad_norm": 1.5043004751205444, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1832, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.1251366784108857, | |
| "grad_norm": 1.7817741632461548, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1517, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.125744138014822, | |
| "grad_norm": 1.7085005044937134, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2548, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.12635159761875836, | |
| "grad_norm": 1.3320608139038086, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1689, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.12695905722269468, | |
| "grad_norm": 1.2270746231079102, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3156, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.12756651682663103, | |
| "grad_norm": 1.4981920719146729, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2243, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.12817397643056735, | |
| "grad_norm": 1.4143863916397095, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2765, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.1287814360345037, | |
| "grad_norm": 1.4701218605041504, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1679, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.12938889563844005, | |
| "grad_norm": 1.3249510526657104, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2313, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.12999635524237638, | |
| "grad_norm": 1.3991800546646118, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2078, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.13060381484631273, | |
| "grad_norm": 1.5967034101486206, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1447, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.13121127445024905, | |
| "grad_norm": 1.5604697465896606, | |
| "learning_rate": 0.0002, | |
| "loss": 1.247, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.1318187340541854, | |
| "grad_norm": 1.3492066860198975, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2734, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.13242619365812172, | |
| "grad_norm": 1.6497987508773804, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2067, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.13303365326205807, | |
| "grad_norm": 1.4936901330947876, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2062, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.13364111286599442, | |
| "grad_norm": 1.4576996564865112, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2225, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.13424857246993074, | |
| "grad_norm": 1.3753255605697632, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2431, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.1348560320738671, | |
| "grad_norm": 1.563539743423462, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2793, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.13546349167780342, | |
| "grad_norm": 1.4935153722763062, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2361, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.13607095128173977, | |
| "grad_norm": 1.307816505432129, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2314, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.13667841088567612, | |
| "grad_norm": 1.535267949104309, | |
| "learning_rate": 0.0002, | |
| "loss": 1.224, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.13728587048961244, | |
| "grad_norm": 1.3963550329208374, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2803, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.1378933300935488, | |
| "grad_norm": 1.527891755104065, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1724, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.1385007896974851, | |
| "grad_norm": 1.4339513778686523, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2773, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.13910824930142146, | |
| "grad_norm": 1.9729173183441162, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2676, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.13971570890535778, | |
| "grad_norm": 1.4557331800460815, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1851, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.14032316850929413, | |
| "grad_norm": 1.4285987615585327, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2804, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.14093062811323048, | |
| "grad_norm": 1.2934836149215698, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2107, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.1415380877171668, | |
| "grad_norm": 1.3619877099990845, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1618, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.14214554732110316, | |
| "grad_norm": 1.4161388874053955, | |
| "learning_rate": 0.0002, | |
| "loss": 1.228, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.14275300692503948, | |
| "grad_norm": 1.3923609256744385, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2526, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.14336046652897583, | |
| "grad_norm": 1.3667365312576294, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2691, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.14396792613291215, | |
| "grad_norm": 1.420930027961731, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2337, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.1445753857368485, | |
| "grad_norm": 1.5273494720458984, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2606, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.14518284534078485, | |
| "grad_norm": 1.5331289768218994, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1693, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.14579030494472117, | |
| "grad_norm": 1.8542994260787964, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2367, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.14639776454865752, | |
| "grad_norm": 1.3973726034164429, | |
| "learning_rate": 0.0002, | |
| "loss": 1.27, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.14700522415259384, | |
| "grad_norm": 1.5454223155975342, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1725, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.1476126837565302, | |
| "grad_norm": 1.3396209478378296, | |
| "learning_rate": 0.0002, | |
| "loss": 1.243, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.14822014336046652, | |
| "grad_norm": 1.4007655382156372, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2339, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.14882760296440287, | |
| "grad_norm": 1.459521770477295, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2409, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.14943506256833922, | |
| "grad_norm": 1.4330353736877441, | |
| "learning_rate": 0.0002, | |
| "loss": 1.265, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.15004252217227554, | |
| "grad_norm": 1.3534977436065674, | |
| "learning_rate": 0.0002, | |
| "loss": 1.1958, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.1506499817762119, | |
| "grad_norm": 1.6496937274932861, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2301, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.1512574413801482, | |
| "grad_norm": 1.6388399600982666, | |
| "learning_rate": 0.0002, | |
| "loss": 1.2952, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.15186490098408456, | |
| "grad_norm": 1.3567882776260376, | |
| "learning_rate": 0.0002, | |
| "loss": 1.202, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 49386, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3843779431581532e+18, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |