| { | |
| "best_metric": 1.6357628107070923, | |
| "best_model_checkpoint": "/mnt/users/n3thakur/research_new/miracl-unanswerable/models/Meta-Llama-3-8B-Instruct-nomiracl-sft/checkpoint-600", | |
| "epoch": 1.0, | |
| "eval_steps": 200, | |
| "global_step": 671, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0014903129657228018, | |
| "grad_norm": 1.2864242608707517, | |
| "learning_rate": 1.4705882352941178e-07, | |
| "loss": 1.9819, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007451564828614009, | |
| "grad_norm": 1.3601881935430797, | |
| "learning_rate": 7.352941176470589e-07, | |
| "loss": 2.0209, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014903129657228018, | |
| "grad_norm": 1.308324783880395, | |
| "learning_rate": 1.4705882352941177e-06, | |
| "loss": 2.0789, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.022354694485842028, | |
| "grad_norm": 1.3402040033505984, | |
| "learning_rate": 2.2058823529411767e-06, | |
| "loss": 2.1346, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.029806259314456036, | |
| "grad_norm": 1.252511058906224, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 1.9887, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037257824143070044, | |
| "grad_norm": 1.1571538128626155, | |
| "learning_rate": 3.6764705882352946e-06, | |
| "loss": 1.9678, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.044709388971684055, | |
| "grad_norm": 0.9688203213594095, | |
| "learning_rate": 4.411764705882353e-06, | |
| "loss": 1.9569, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05216095380029806, | |
| "grad_norm": 0.9113616550249095, | |
| "learning_rate": 5.147058823529411e-06, | |
| "loss": 1.9236, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05961251862891207, | |
| "grad_norm": 0.8080169675485035, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 1.9358, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06706408345752608, | |
| "grad_norm": 0.6008229695631025, | |
| "learning_rate": 6.61764705882353e-06, | |
| "loss": 1.8434, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07451564828614009, | |
| "grad_norm": 0.5413811186895102, | |
| "learning_rate": 7.352941176470589e-06, | |
| "loss": 1.8786, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08196721311475409, | |
| "grad_norm": 0.4866905959375867, | |
| "learning_rate": 8.088235294117648e-06, | |
| "loss": 1.7887, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08941877794336811, | |
| "grad_norm": 0.3991956099552151, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 1.8388, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09687034277198212, | |
| "grad_norm": 0.4089753043909763, | |
| "learning_rate": 9.558823529411766e-06, | |
| "loss": 1.7719, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.10432190760059612, | |
| "grad_norm": 0.3567858319466126, | |
| "learning_rate": 9.999728567909404e-06, | |
| "loss": 1.7757, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11177347242921014, | |
| "grad_norm": 0.3733731735267369, | |
| "learning_rate": 9.996675295326345e-06, | |
| "loss": 1.7357, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11922503725782414, | |
| "grad_norm": 0.4120045550242149, | |
| "learning_rate": 9.99023153874608e-06, | |
| "loss": 1.76, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12667660208643816, | |
| "grad_norm": 0.47838141094777215, | |
| "learning_rate": 9.980401670566705e-06, | |
| "loss": 1.8124, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.13412816691505217, | |
| "grad_norm": 0.34098601847525956, | |
| "learning_rate": 9.967192360825558e-06, | |
| "loss": 1.7454, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14157973174366617, | |
| "grad_norm": 0.3091360382279767, | |
| "learning_rate": 9.950612572673255e-06, | |
| "loss": 1.7528, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.14903129657228018, | |
| "grad_norm": 0.2937982462814676, | |
| "learning_rate": 9.930673556291789e-06, | |
| "loss": 1.6999, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15648286140089418, | |
| "grad_norm": 0.3033053874366656, | |
| "learning_rate": 9.907388841260723e-06, | |
| "loss": 1.6328, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.16393442622950818, | |
| "grad_norm": 0.3147533435412934, | |
| "learning_rate": 9.880774227376727e-06, | |
| "loss": 1.7202, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17138599105812222, | |
| "grad_norm": 0.30138135147702194, | |
| "learning_rate": 9.850847773932656e-06, | |
| "loss": 1.7147, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17883755588673622, | |
| "grad_norm": 0.2821814099944737, | |
| "learning_rate": 9.817629787463456e-06, | |
| "loss": 1.7141, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18628912071535023, | |
| "grad_norm": 0.278421347158755, | |
| "learning_rate": 9.781142807967205e-06, | |
| "loss": 1.7812, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.19374068554396423, | |
| "grad_norm": 0.2745120712204819, | |
| "learning_rate": 9.741411593610635e-06, | |
| "loss": 1.6429, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.20119225037257824, | |
| "grad_norm": 0.2915048113702178, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 1.6622, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.20864381520119224, | |
| "grad_norm": 0.29052352322277714, | |
| "learning_rate": 9.652326481535434e-06, | |
| "loss": 1.7323, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21609538002980627, | |
| "grad_norm": 0.30795018674862495, | |
| "learning_rate": 9.603033032340875e-06, | |
| "loss": 1.7064, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.22354694485842028, | |
| "grad_norm": 0.27173782182479533, | |
| "learning_rate": 9.550616204316922e-06, | |
| "loss": 1.7384, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23099850968703428, | |
| "grad_norm": 0.3106086974094058, | |
| "learning_rate": 9.495111564797073e-06, | |
| "loss": 1.7267, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.23845007451564829, | |
| "grad_norm": 0.3021509489363488, | |
| "learning_rate": 9.43655677634312e-06, | |
| "loss": 1.6479, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2459016393442623, | |
| "grad_norm": 0.29504453527495433, | |
| "learning_rate": 9.374991571189292e-06, | |
| "loss": 1.6919, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2533532041728763, | |
| "grad_norm": 0.3337725680370752, | |
| "learning_rate": 9.310457724282034e-06, | |
| "loss": 1.7475, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2608047690014903, | |
| "grad_norm": 0.271616871396071, | |
| "learning_rate": 9.242999024933694e-06, | |
| "loss": 1.6249, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.26825633383010433, | |
| "grad_norm": 0.27469314451092647, | |
| "learning_rate": 9.172661247109382e-06, | |
| "loss": 1.6289, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2757078986587183, | |
| "grad_norm": 0.32247904554795664, | |
| "learning_rate": 9.099492118367123e-06, | |
| "loss": 1.6759, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.28315946348733234, | |
| "grad_norm": 0.32346400909431516, | |
| "learning_rate": 9.023541287472435e-06, | |
| "loss": 1.7517, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2906110283159464, | |
| "grad_norm": 0.28010968189921137, | |
| "learning_rate": 8.944860290709245e-06, | |
| "loss": 1.6495, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.29806259314456035, | |
| "grad_norm": 0.3073056244750479, | |
| "learning_rate": 8.863502516910058e-06, | |
| "loss": 1.6576, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.29806259314456035, | |
| "eval_loss": 1.6655718088150024, | |
| "eval_runtime": 581.6685, | |
| "eval_samples_per_second": 4.102, | |
| "eval_steps_per_second": 0.258, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3055141579731744, | |
| "grad_norm": 0.2982943497092564, | |
| "learning_rate": 8.779523171229061e-06, | |
| "loss": 1.6927, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.31296572280178836, | |
| "grad_norm": 0.3097195272484475, | |
| "learning_rate": 8.692979237682785e-06, | |
| "loss": 1.6619, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3204172876304024, | |
| "grad_norm": 0.34367552210427565, | |
| "learning_rate": 8.603929440483714e-06, | |
| "loss": 1.684, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.32786885245901637, | |
| "grad_norm": 0.3023959348087803, | |
| "learning_rate": 8.51243420419308e-06, | |
| "loss": 1.5742, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3353204172876304, | |
| "grad_norm": 0.2908694884413669, | |
| "learning_rate": 8.418555612719912e-06, | |
| "loss": 1.6211, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.34277198211624443, | |
| "grad_norm": 0.3265065297068578, | |
| "learning_rate": 8.32235736719411e-06, | |
| "loss": 1.6614, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3502235469448584, | |
| "grad_norm": 0.29528676002450155, | |
| "learning_rate": 8.223904742742182e-06, | |
| "loss": 1.6757, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.35767511177347244, | |
| "grad_norm": 0.30070470151226314, | |
| "learning_rate": 8.123264544194934e-06, | |
| "loss": 1.6443, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3651266766020864, | |
| "grad_norm": 0.3090720148911324, | |
| "learning_rate": 8.02050506075718e-06, | |
| "loss": 1.6689, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.37257824143070045, | |
| "grad_norm": 0.29743017001679256, | |
| "learning_rate": 7.91569601967025e-06, | |
| "loss": 1.6278, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38002980625931443, | |
| "grad_norm": 0.297851200782397, | |
| "learning_rate": 7.808908538898703e-06, | |
| "loss": 1.6155, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.38748137108792846, | |
| "grad_norm": 0.3435439482120138, | |
| "learning_rate": 7.70021507887338e-06, | |
| "loss": 1.5611, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3949329359165425, | |
| "grad_norm": 0.32733272749783104, | |
| "learning_rate": 7.5896893933235135e-06, | |
| "loss": 1.6055, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.40238450074515647, | |
| "grad_norm": 0.3056544624446995, | |
| "learning_rate": 7.477406479231299e-06, | |
| "loss": 1.678, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4098360655737705, | |
| "grad_norm": 0.31368790810083264, | |
| "learning_rate": 7.363442525942827e-06, | |
| "loss": 1.5912, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4172876304023845, | |
| "grad_norm": 0.2993808857251287, | |
| "learning_rate": 7.247874863469964e-06, | |
| "loss": 1.6412, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4247391952309985, | |
| "grad_norm": 0.32056752037833786, | |
| "learning_rate": 7.1307819100182275e-06, | |
| "loss": 1.6949, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.43219076005961254, | |
| "grad_norm": 0.3361900057942298, | |
| "learning_rate": 7.01224311877627e-06, | |
| "loss": 1.6513, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4396423248882265, | |
| "grad_norm": 0.3373143541448011, | |
| "learning_rate": 6.892338924003068e-06, | |
| "loss": 1.6699, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.44709388971684055, | |
| "grad_norm": 0.30913581638324406, | |
| "learning_rate": 6.771150686449436e-06, | |
| "loss": 1.5889, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.3324140439223907, | |
| "learning_rate": 6.648760638150833e-06, | |
| "loss": 1.6559, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.46199701937406856, | |
| "grad_norm": 0.3143801425022107, | |
| "learning_rate": 6.525251826628991e-06, | |
| "loss": 1.6297, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.46944858420268254, | |
| "grad_norm": 0.32690734523122295, | |
| "learning_rate": 6.400708058540182e-06, | |
| "loss": 1.6407, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.47690014903129657, | |
| "grad_norm": 0.31268911619063705, | |
| "learning_rate": 6.275213842808383e-06, | |
| "loss": 1.6245, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4843517138599106, | |
| "grad_norm": 0.36149919350714527, | |
| "learning_rate": 6.148854333281905e-06, | |
| "loss": 1.6476, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4918032786885246, | |
| "grad_norm": 0.3590585354206088, | |
| "learning_rate": 6.021715270952435e-06, | |
| "loss": 1.6452, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4992548435171386, | |
| "grad_norm": 0.3039096750446457, | |
| "learning_rate": 5.893882925775648e-06, | |
| "loss": 1.6415, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5067064083457526, | |
| "grad_norm": 0.31141048527235143, | |
| "learning_rate": 5.765444038132901e-06, | |
| "loss": 1.6571, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5141579731743666, | |
| "grad_norm": 0.3173169163707792, | |
| "learning_rate": 5.636485759973729e-06, | |
| "loss": 1.5806, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5216095380029806, | |
| "grad_norm": 0.3363924031410881, | |
| "learning_rate": 5.507095595679059e-06, | |
| "loss": 1.6871, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5290611028315947, | |
| "grad_norm": 0.3307680686760043, | |
| "learning_rate": 5.377361342685287e-06, | |
| "loss": 1.6933, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5365126676602087, | |
| "grad_norm": 0.32876995727805525, | |
| "learning_rate": 5.2473710319095054e-06, | |
| "loss": 1.6201, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5439642324888226, | |
| "grad_norm": 0.3452442083785589, | |
| "learning_rate": 5.117212868016303e-06, | |
| "loss": 1.7146, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5514157973174366, | |
| "grad_norm": 0.3659924599969553, | |
| "learning_rate": 4.9869751695666615e-06, | |
| "loss": 1.6215, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5588673621460507, | |
| "grad_norm": 0.3233317882165812, | |
| "learning_rate": 4.856746309089582e-06, | |
| "loss": 1.6315, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5663189269746647, | |
| "grad_norm": 0.30217314493419994, | |
| "learning_rate": 4.726614653117071e-06, | |
| "loss": 1.6945, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5737704918032787, | |
| "grad_norm": 0.35844882665530153, | |
| "learning_rate": 4.596668502223214e-06, | |
| "loss": 1.6924, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5812220566318927, | |
| "grad_norm": 0.34041157047813575, | |
| "learning_rate": 4.466996031108004e-06, | |
| "loss": 1.629, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5886736214605067, | |
| "grad_norm": 0.314658238073181, | |
| "learning_rate": 4.337685228766561e-06, | |
| "loss": 1.6797, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5961251862891207, | |
| "grad_norm": 0.32647573861445145, | |
| "learning_rate": 4.208823838784387e-06, | |
| "loss": 1.6447, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5961251862891207, | |
| "eval_loss": 1.6408501863479614, | |
| "eval_runtime": 580.0543, | |
| "eval_samples_per_second": 4.113, | |
| "eval_steps_per_second": 0.259, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6035767511177347, | |
| "grad_norm": 0.37909958848060366, | |
| "learning_rate": 4.0804992997991335e-06, | |
| "loss": 1.6574, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6110283159463488, | |
| "grad_norm": 0.3790204664396362, | |
| "learning_rate": 3.952798686169279e-06, | |
| "loss": 1.7235, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6184798807749627, | |
| "grad_norm": 0.3703026855695823, | |
| "learning_rate": 3.825808648890005e-06, | |
| "loss": 1.675, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6259314456035767, | |
| "grad_norm": 0.3426042362005175, | |
| "learning_rate": 3.699615356796342e-06, | |
| "loss": 1.5951, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6333830104321908, | |
| "grad_norm": 0.3472991803298781, | |
| "learning_rate": 3.5743044380934655e-06, | |
| "loss": 1.7072, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6408345752608048, | |
| "grad_norm": 0.3645949602782747, | |
| "learning_rate": 3.449960922253858e-06, | |
| "loss": 1.6207, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6482861400894188, | |
| "grad_norm": 0.3314569171608062, | |
| "learning_rate": 3.326669182320736e-06, | |
| "loss": 1.6281, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6557377049180327, | |
| "grad_norm": 0.31811370472714595, | |
| "learning_rate": 3.2045128776568783e-06, | |
| "loss": 1.6572, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6631892697466468, | |
| "grad_norm": 0.3377222335896338, | |
| "learning_rate": 3.0835748971777413e-06, | |
| "loss": 1.6651, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6706408345752608, | |
| "grad_norm": 0.331361220355902, | |
| "learning_rate": 2.9639373031073525e-06, | |
| "loss": 1.6878, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6780923994038748, | |
| "grad_norm": 0.30006578695394276, | |
| "learning_rate": 2.8456812752951483e-06, | |
| "loss": 1.6887, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6855439642324889, | |
| "grad_norm": 0.34387137915082105, | |
| "learning_rate": 2.728887056131553e-06, | |
| "loss": 1.6273, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6929955290611028, | |
| "grad_norm": 0.3426777102093575, | |
| "learning_rate": 2.6136338960996665e-06, | |
| "loss": 1.6867, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7004470938897168, | |
| "grad_norm": 0.3047030104059201, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 1.6514, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7078986587183308, | |
| "grad_norm": 0.3194123549278598, | |
| "learning_rate": 2.3880624738847837e-06, | |
| "loss": 1.6494, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7153502235469449, | |
| "grad_norm": 0.28354706856772705, | |
| "learning_rate": 2.277897272737787e-06, | |
| "loss": 1.5821, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7228017883755589, | |
| "grad_norm": 0.36693040405683097, | |
| "learning_rate": 2.1695791489352346e-06, | |
| "loss": 1.6627, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7302533532041728, | |
| "grad_norm": 0.3268289701429299, | |
| "learning_rate": 2.063181601522722e-06, | |
| "loss": 1.543, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7377049180327869, | |
| "grad_norm": 0.337638547503993, | |
| "learning_rate": 1.9587768263425886e-06, | |
| "loss": 1.6527, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7451564828614009, | |
| "grad_norm": 0.2879745114712397, | |
| "learning_rate": 1.856435667045577e-06, | |
| "loss": 1.5879, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7526080476900149, | |
| "grad_norm": 0.3081841268702886, | |
| "learning_rate": 1.7562275670200041e-06, | |
| "loss": 1.5423, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7600596125186289, | |
| "grad_norm": 0.3237369195299089, | |
| "learning_rate": 1.6582205222711051e-06, | |
| "loss": 1.6172, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.767511177347243, | |
| "grad_norm": 0.3484298930902378, | |
| "learning_rate": 1.5624810352824709e-06, | |
| "loss": 1.6456, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7749627421758569, | |
| "grad_norm": 0.3160389330113336, | |
| "learning_rate": 1.4690740698909223e-06, | |
| "loss": 1.5929, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7824143070044709, | |
| "grad_norm": 0.36108945655051267, | |
| "learning_rate": 1.3780630072054313e-06, | |
| "loss": 1.6718, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.789865871833085, | |
| "grad_norm": 0.32927194714240243, | |
| "learning_rate": 1.289509602599996e-06, | |
| "loss": 1.6058, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.797317436661699, | |
| "grad_norm": 0.3769043230835878, | |
| "learning_rate": 1.203473943809651e-06, | |
| "loss": 1.6295, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8047690014903129, | |
| "grad_norm": 0.33941548710215963, | |
| "learning_rate": 1.1200144101580634e-06, | |
| "loss": 1.6792, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.812220566318927, | |
| "grad_norm": 0.3706526425603311, | |
| "learning_rate": 1.0391876329443534e-06, | |
| "loss": 1.6146, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.819672131147541, | |
| "grad_norm": 0.35398755740858545, | |
| "learning_rate": 9.610484570160444e-07, | |
| "loss": 1.7103, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.827123695976155, | |
| "grad_norm": 0.35561387409815454, | |
| "learning_rate": 8.856499035541972e-07, | |
| "loss": 1.6163, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.834575260804769, | |
| "grad_norm": 0.348602818005502, | |
| "learning_rate": 8.130431340959982e-07, | |
| "loss": 1.6807, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.842026825633383, | |
| "grad_norm": 0.3675149781602246, | |
| "learning_rate": 7.432774158191946e-07, | |
| "loss": 1.6201, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.849478390461997, | |
| "grad_norm": 0.35625514274912495, | |
| "learning_rate": 6.764000881119631e-07, | |
| "loss": 1.5743, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.856929955290611, | |
| "grad_norm": 0.3186585069526685, | |
| "learning_rate": 6.12456530450844e-07, | |
| "loss": 1.6272, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8643815201192251, | |
| "grad_norm": 0.3347624139888847, | |
| "learning_rate": 5.514901316086058e-07, | |
| "loss": 1.6818, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8718330849478391, | |
| "grad_norm": 0.3821665104781894, | |
| "learning_rate": 4.935422602128698e-07, | |
| "loss": 1.7344, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.879284649776453, | |
| "grad_norm": 0.31160320897543975, | |
| "learning_rate": 4.386522366755169e-07, | |
| "loss": 1.5668, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.886736214605067, | |
| "grad_norm": 0.32657433748759934, | |
| "learning_rate": 3.868573065118936e-07, | |
| "loss": 1.6732, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8941877794336811, | |
| "grad_norm": 0.34252424286072203, | |
| "learning_rate": 3.381926150679543e-07, | |
| "loss": 1.6245, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8941877794336811, | |
| "eval_loss": 1.6357628107070923, | |
| "eval_runtime": 580.6663, | |
| "eval_samples_per_second": 4.109, | |
| "eval_steps_per_second": 0.258, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9016393442622951, | |
| "grad_norm": 0.3432161875223241, | |
| "learning_rate": 2.9269118367244385e-07, | |
| "loss": 1.6999, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.3458124106576635, | |
| "learning_rate": 2.5038388723034935e-07, | |
| "loss": 1.6037, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9165424739195231, | |
| "grad_norm": 0.34778709529194757, | |
| "learning_rate": 2.1129943327279522e-07, | |
| "loss": 1.7079, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9239940387481371, | |
| "grad_norm": 0.3605158717479098, | |
| "learning_rate": 1.7546434247760147e-07, | |
| "loss": 1.6404, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9314456035767511, | |
| "grad_norm": 0.29865645458342877, | |
| "learning_rate": 1.429029306737345e-07, | |
| "loss": 1.6086, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9388971684053651, | |
| "grad_norm": 0.3290659895478686, | |
| "learning_rate": 1.1363729234184828e-07, | |
| "loss": 1.596, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9463487332339792, | |
| "grad_norm": 0.32311961294020886, | |
| "learning_rate": 8.768728562211948e-08, | |
| "loss": 1.5774, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9538002980625931, | |
| "grad_norm": 0.3628568702485729, | |
| "learning_rate": 6.507051883954618e-08, | |
| "loss": 1.6421, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9612518628912071, | |
| "grad_norm": 0.35678638324651946, | |
| "learning_rate": 4.580233855585425e-08, | |
| "loss": 1.5815, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9687034277198212, | |
| "grad_norm": 0.32862520535733186, | |
| "learning_rate": 2.989581915611994e-08, | |
| "loss": 1.5756, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9761549925484352, | |
| "grad_norm": 0.36012152823713983, | |
| "learning_rate": 1.7361753977169214e-08, | |
| "loss": 1.6405, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9836065573770492, | |
| "grad_norm": 0.3640507189017552, | |
| "learning_rate": 8.208647983782847e-09, | |
| "loss": 1.62, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9910581222056631, | |
| "grad_norm": 0.33586265256557224, | |
| "learning_rate": 2.442711997670544e-09, | |
| "loss": 1.6375, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.9985096870342772, | |
| "grad_norm": 0.3189391910196833, | |
| "learning_rate": 6.785848312707011e-11, | |
| "loss": 1.6954, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 671, | |
| "total_flos": 919616052789248.0, | |
| "train_loss": 1.6831368206154926, | |
| "train_runtime": 18035.2779, | |
| "train_samples_per_second": 1.191, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 671, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 919616052789248.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |