diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,15209 @@ +{ + "best_global_step": 18888, + "best_metric": 0.00379174, + "best_model_checkpoint": "/ssd/yuky/vlnckpt/v3-20260427-211707/checkpoint-18888", + "epoch": 1.0, + "eval_steps": 3000, + "global_step": 18888, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 5.294401831863034e-05, + "grad_norm": 40.624629974365234, + "learning_rate": 3.527336860670194e-08, + "loss": 1.610304832458496, + "step": 1, + "token_acc": 0.8053193263298316 + }, + { + "epoch": 0.0005294401831863034, + "grad_norm": 39.16465759277344, + "learning_rate": 3.527336860670194e-07, + "loss": 1.5792715284559462, + "step": 10, + "token_acc": 0.8077929064331825 + }, + { + "epoch": 0.0010588803663726068, + "grad_norm": 27.180404663085938, + "learning_rate": 7.054673721340388e-07, + "loss": 1.3994749069213868, + "step": 20, + "token_acc": 0.8149931166826582 + }, + { + "epoch": 0.0015883205495589102, + "grad_norm": 9.793509483337402, + "learning_rate": 1.0582010582010582e-06, + "loss": 0.8898136138916015, + "step": 30, + "token_acc": 0.8414840641604178 + }, + { + "epoch": 0.0021177607327452137, + "grad_norm": 4.886561393737793, + "learning_rate": 1.4109347442680776e-06, + "loss": 0.4104780197143555, + "step": 40, + "token_acc": 0.9024746362301428 + }, + { + "epoch": 0.002647200915931517, + "grad_norm": 1.9797357320785522, + "learning_rate": 1.763668430335097e-06, + "loss": 0.16789214611053466, + "step": 50, + "token_acc": 0.9540350701569982 + }, + { + "epoch": 0.0031766410991178203, + "grad_norm": 1.3802133798599243, + "learning_rate": 2.1164021164021164e-06, + "loss": 0.06390936374664306, + "step": 60, + "token_acc": 0.9774533928348 + }, + { + "epoch": 0.0037060812823041236, + "grad_norm": 1.091347098350525, + "learning_rate": 2.469135802469136e-06, + "loss": 0.0429516077041626, + "step": 70, + "token_acc": 0.9827495599803138 + }, + { + "epoch": 0.004235521465490427, + "grad_norm": 0.49583902955055237, + "learning_rate": 2.8218694885361552e-06, + "loss": 0.030794396996498108, + "step": 80, + "token_acc": 0.9855282803259682 + }, + { + "epoch": 0.00476496164867673, + "grad_norm": 0.6311731934547424, + "learning_rate": 3.1746031746031746e-06, + "loss": 0.025531718134880067, + "step": 90, + "token_acc": 0.9881108320748893 + }, + { + "epoch": 0.005294401831863034, + "grad_norm": 0.3630579710006714, + "learning_rate": 3.527336860670194e-06, + "loss": 0.02389259934425354, + "step": 100, + "token_acc": 0.9889622145652047 + }, + { + "epoch": 0.005823842015049337, + "grad_norm": 0.5834900736808777, + "learning_rate": 3.8800705467372134e-06, + "loss": 0.023002691566944122, + "step": 110, + "token_acc": 0.9900490036481421 + }, + { + "epoch": 0.006353282198235641, + "grad_norm": 0.5861024856567383, + "learning_rate": 4.232804232804233e-06, + "loss": 0.02275083214044571, + "step": 120, + "token_acc": 0.9895634328046918 + }, + { + "epoch": 0.006882722381421944, + "grad_norm": 0.5614109039306641, + "learning_rate": 4.585537918871252e-06, + "loss": 0.022587382793426515, + "step": 130, + "token_acc": 0.9898183153484723 + }, + { + "epoch": 0.007412162564608247, + "grad_norm": 0.34563350677490234, + "learning_rate": 4.938271604938272e-06, + "loss": 0.0221064954996109, + "step": 140, + "token_acc": 0.9898917449250221 + }, + { + "epoch": 0.007941602747794551, + "grad_norm": 0.4929299056529999, + "learning_rate": 5.291005291005291e-06, + "loss": 0.02250640094280243, + "step": 150, + "token_acc": 0.9897106806973037 + }, + { + "epoch": 0.008471042930980855, + "grad_norm": 0.30376073718070984, + "learning_rate": 5.6437389770723105e-06, + "loss": 0.02110200822353363, + "step": 160, + "token_acc": 0.9903955273698264 + }, + { + "epoch": 0.009000483114167157, + "grad_norm": 0.39431607723236084, + "learning_rate": 5.99647266313933e-06, + "loss": 0.02064204663038254, + "step": 170, + "token_acc": 0.9909323510046921 + }, + { + "epoch": 0.00952992329735346, + "grad_norm": 0.4077151417732239, + "learning_rate": 6.349206349206349e-06, + "loss": 0.020060770213603973, + "step": 180, + "token_acc": 0.9910823030849059 + }, + { + "epoch": 0.010059363480539764, + "grad_norm": 0.4160288870334625, + "learning_rate": 6.701940035273369e-06, + "loss": 0.02049023360013962, + "step": 190, + "token_acc": 0.9910893071686023 + }, + { + "epoch": 0.010588803663726068, + "grad_norm": 0.4125887155532837, + "learning_rate": 7.054673721340388e-06, + "loss": 0.02052287757396698, + "step": 200, + "token_acc": 0.9904836570780406 + }, + { + "epoch": 0.011118243846912372, + "grad_norm": 0.3472023010253906, + "learning_rate": 7.4074074074074075e-06, + "loss": 0.020289844274520873, + "step": 210, + "token_acc": 0.9911398107823998 + }, + { + "epoch": 0.011647684030098674, + "grad_norm": 0.23152336478233337, + "learning_rate": 7.760141093474427e-06, + "loss": 0.01985529363155365, + "step": 220, + "token_acc": 0.9909851308865594 + }, + { + "epoch": 0.012177124213284977, + "grad_norm": 0.2731718420982361, + "learning_rate": 8.112874779541447e-06, + "loss": 0.01922576129436493, + "step": 230, + "token_acc": 0.991420352691041 + }, + { + "epoch": 0.012706564396471281, + "grad_norm": 0.26822414994239807, + "learning_rate": 8.465608465608466e-06, + "loss": 0.018826961517333984, + "step": 240, + "token_acc": 0.9914661817211118 + }, + { + "epoch": 0.013236004579657585, + "grad_norm": 0.43043428659439087, + "learning_rate": 8.818342151675486e-06, + "loss": 0.02907015085220337, + "step": 250, + "token_acc": 0.9904069136955905 + }, + { + "epoch": 0.013765444762843889, + "grad_norm": 0.29103559255599976, + "learning_rate": 9.171075837742504e-06, + "loss": 0.021755072474479675, + "step": 260, + "token_acc": 0.9901904308696705 + }, + { + "epoch": 0.01429488494603019, + "grad_norm": 0.49291878938674927, + "learning_rate": 9.523809523809525e-06, + "loss": 0.020091281831264497, + "step": 270, + "token_acc": 0.9909787361885369 + }, + { + "epoch": 0.014824325129216494, + "grad_norm": 1.144215703010559, + "learning_rate": 9.876543209876543e-06, + "loss": 0.02563856840133667, + "step": 280, + "token_acc": 0.9900806727456264 + }, + { + "epoch": 0.015353765312402798, + "grad_norm": 0.21126340329647064, + "learning_rate": 1.0229276895943564e-05, + "loss": 0.01882001757621765, + "step": 290, + "token_acc": 0.9916805046771084 + }, + { + "epoch": 0.015883205495589102, + "grad_norm": 0.25845855474472046, + "learning_rate": 1.0582010582010582e-05, + "loss": 0.01925322413444519, + "step": 300, + "token_acc": 0.9910653952999474 + }, + { + "epoch": 0.016412645678775404, + "grad_norm": 0.2962028980255127, + "learning_rate": 1.0934744268077602e-05, + "loss": 0.018808317184448243, + "step": 310, + "token_acc": 0.9915867589427166 + }, + { + "epoch": 0.01694208586196171, + "grad_norm": 0.38271716237068176, + "learning_rate": 1.1287477954144621e-05, + "loss": 0.018583652377128602, + "step": 320, + "token_acc": 0.9915624843517885 + }, + { + "epoch": 0.01747152604514801, + "grad_norm": 0.25600460171699524, + "learning_rate": 1.1640211640211641e-05, + "loss": 0.0189639613032341, + "step": 330, + "token_acc": 0.9915119982657128 + }, + { + "epoch": 0.018000966228334313, + "grad_norm": 0.307173490524292, + "learning_rate": 1.199294532627866e-05, + "loss": 0.01903609186410904, + "step": 340, + "token_acc": 0.9915718148135786 + }, + { + "epoch": 0.01853040641152062, + "grad_norm": 0.19155730307102203, + "learning_rate": 1.234567901234568e-05, + "loss": 0.019222331047058106, + "step": 350, + "token_acc": 0.9912315796492632 + }, + { + "epoch": 0.01905984659470692, + "grad_norm": 0.21802888810634613, + "learning_rate": 1.2698412698412699e-05, + "loss": 0.019385236501693725, + "step": 360, + "token_acc": 0.9910702374599787 + }, + { + "epoch": 0.019589286777893226, + "grad_norm": 0.20818020403385162, + "learning_rate": 1.3051146384479719e-05, + "loss": 0.019593854248523713, + "step": 370, + "token_acc": 0.9912240454480993 + }, + { + "epoch": 0.02011872696107953, + "grad_norm": 0.42564448714256287, + "learning_rate": 1.3403880070546737e-05, + "loss": 0.019342853128910063, + "step": 380, + "token_acc": 0.9914328657314629 + }, + { + "epoch": 0.02064816714426583, + "grad_norm": 0.1828458309173584, + "learning_rate": 1.3756613756613758e-05, + "loss": 0.01897099018096924, + "step": 390, + "token_acc": 0.9913612674588285 + }, + { + "epoch": 0.021177607327452136, + "grad_norm": 0.31627896428108215, + "learning_rate": 1.4109347442680776e-05, + "loss": 0.019131506979465484, + "step": 400, + "token_acc": 0.9913008632348767 + }, + { + "epoch": 0.021707047510638438, + "grad_norm": 0.2053859978914261, + "learning_rate": 1.4462081128747796e-05, + "loss": 0.01921783685684204, + "step": 410, + "token_acc": 0.9912911459984317 + }, + { + "epoch": 0.022236487693824743, + "grad_norm": 0.3623993992805481, + "learning_rate": 1.4814814814814815e-05, + "loss": 0.01962321698665619, + "step": 420, + "token_acc": 0.9913244408852407 + }, + { + "epoch": 0.022765927877011045, + "grad_norm": 0.3896368741989136, + "learning_rate": 1.5167548500881835e-05, + "loss": 0.01966039091348648, + "step": 430, + "token_acc": 0.9911938766141082 + }, + { + "epoch": 0.023295368060197347, + "grad_norm": 0.35099607706069946, + "learning_rate": 1.5520282186948854e-05, + "loss": 0.019052752852439882, + "step": 440, + "token_acc": 0.9913014469788582 + }, + { + "epoch": 0.023824808243383653, + "grad_norm": 0.334000825881958, + "learning_rate": 1.5873015873015872e-05, + "loss": 0.01981915384531021, + "step": 450, + "token_acc": 0.9909396085242835 + }, + { + "epoch": 0.024354248426569955, + "grad_norm": 0.35724470019340515, + "learning_rate": 1.6225749559082894e-05, + "loss": 0.02018025517463684, + "step": 460, + "token_acc": 0.991064724433098 + }, + { + "epoch": 0.02488368860975626, + "grad_norm": 0.29710838198661804, + "learning_rate": 1.6578483245149913e-05, + "loss": 0.019318394362926483, + "step": 470, + "token_acc": 0.9913819688117286 + }, + { + "epoch": 0.025413128792942562, + "grad_norm": 0.46417856216430664, + "learning_rate": 1.693121693121693e-05, + "loss": 0.01900869607925415, + "step": 480, + "token_acc": 0.9914532052564873 + }, + { + "epoch": 0.025942568976128864, + "grad_norm": 0.21939238905906677, + "learning_rate": 1.728395061728395e-05, + "loss": 0.01896592676639557, + "step": 490, + "token_acc": 0.9912252872027785 + }, + { + "epoch": 0.02647200915931517, + "grad_norm": 0.9802926182746887, + "learning_rate": 1.7636684303350972e-05, + "loss": 0.019215902686119078, + "step": 500, + "token_acc": 0.9911182646840521 + }, + { + "epoch": 0.027001449342501472, + "grad_norm": 0.2899588346481323, + "learning_rate": 1.798941798941799e-05, + "loss": 0.03785286247730255, + "step": 510, + "token_acc": 0.9902278772314542 + }, + { + "epoch": 0.027530889525687777, + "grad_norm": 0.3563251197338104, + "learning_rate": 1.834215167548501e-05, + "loss": 0.018962323665618896, + "step": 520, + "token_acc": 0.9917635375898125 + }, + { + "epoch": 0.02806032970887408, + "grad_norm": 0.2315487414598465, + "learning_rate": 1.8694885361552028e-05, + "loss": 0.017761531472206115, + "step": 530, + "token_acc": 0.9915845717148105 + }, + { + "epoch": 0.02858976989206038, + "grad_norm": 0.29373639822006226, + "learning_rate": 1.904761904761905e-05, + "loss": 0.018248121440410613, + "step": 540, + "token_acc": 0.9920141525572236 + }, + { + "epoch": 0.029119210075246687, + "grad_norm": 2.8977622985839844, + "learning_rate": 1.9400352733686068e-05, + "loss": 0.019626700878143312, + "step": 550, + "token_acc": 0.9914989822137685 + }, + { + "epoch": 0.02964865025843299, + "grad_norm": 0.3122707009315491, + "learning_rate": 1.9753086419753087e-05, + "loss": 0.039229866862297055, + "step": 560, + "token_acc": 0.9887721990974383 + }, + { + "epoch": 0.030178090441619294, + "grad_norm": 0.1385003626346588, + "learning_rate": 1.999999867683532e-05, + "loss": 0.01779875159263611, + "step": 570, + "token_acc": 0.9922677832643801 + }, + { + "epoch": 0.030707530624805596, + "grad_norm": 0.23507219552993774, + "learning_rate": 1.999997515391738e-05, + "loss": 0.016664466261863707, + "step": 580, + "token_acc": 0.9922150092618861 + }, + { + "epoch": 0.0312369708079919, + "grad_norm": 0.3166334331035614, + "learning_rate": 1.9999922227419454e-05, + "loss": 0.0165520042181015, + "step": 590, + "token_acc": 0.9926574272626388 + }, + { + "epoch": 0.031766410991178204, + "grad_norm": 0.3929966390132904, + "learning_rate": 1.999983989749716e-05, + "loss": 0.01776043176651001, + "step": 600, + "token_acc": 0.9920499194981355 + }, + { + "epoch": 0.03229585117436451, + "grad_norm": 0.19735302031040192, + "learning_rate": 1.9999728164392586e-05, + "loss": 0.017202050983905794, + "step": 610, + "token_acc": 0.9922943990382609 + }, + { + "epoch": 0.03282529135755081, + "grad_norm": 0.4107261002063751, + "learning_rate": 1.999958702843426e-05, + "loss": 0.01816055029630661, + "step": 620, + "token_acc": 0.9917638125119954 + }, + { + "epoch": 0.03335473154073711, + "grad_norm": 0.24938711524009705, + "learning_rate": 1.9999416490037177e-05, + "loss": 0.01815718412399292, + "step": 630, + "token_acc": 0.9918322056381976 + }, + { + "epoch": 0.03388417172392342, + "grad_norm": 0.3341120779514313, + "learning_rate": 1.9999216549702787e-05, + "loss": 0.018011400103569032, + "step": 640, + "token_acc": 0.9919566795437669 + }, + { + "epoch": 0.03441361190710972, + "grad_norm": 0.36312881112098694, + "learning_rate": 1.999898720801898e-05, + "loss": 0.018776944279670714, + "step": 650, + "token_acc": 0.9917009676638308 + }, + { + "epoch": 0.03494305209029602, + "grad_norm": 0.2795328199863434, + "learning_rate": 1.9998728465660105e-05, + "loss": 0.018062490224838256, + "step": 660, + "token_acc": 0.9916811843059457 + }, + { + "epoch": 0.03547249227348233, + "grad_norm": 0.38559475541114807, + "learning_rate": 1.999844032338696e-05, + "loss": 0.01781948357820511, + "step": 670, + "token_acc": 0.9917864476386037 + }, + { + "epoch": 0.03600193245666863, + "grad_norm": 0.23458664119243622, + "learning_rate": 1.9998122782046795e-05, + "loss": 0.01758660078048706, + "step": 680, + "token_acc": 0.9922643634998123 + }, + { + "epoch": 0.03653137263985493, + "grad_norm": 0.4144543707370758, + "learning_rate": 1.9997775842573287e-05, + "loss": 0.01761503517627716, + "step": 690, + "token_acc": 0.992129332410642 + }, + { + "epoch": 0.03706081282304124, + "grad_norm": 0.2508455216884613, + "learning_rate": 1.9997399505986576e-05, + "loss": 0.017394205927848815, + "step": 700, + "token_acc": 0.9923904881101376 + }, + { + "epoch": 0.03759025300622754, + "grad_norm": 0.27183935046195984, + "learning_rate": 1.999699377339322e-05, + "loss": 0.01798824518918991, + "step": 710, + "token_acc": 0.9917480998914223 + }, + { + "epoch": 0.03811969318941384, + "grad_norm": 0.13980643451213837, + "learning_rate": 1.9996558645986224e-05, + "loss": 0.018337398767471313, + "step": 720, + "token_acc": 0.9919505522004605 + }, + { + "epoch": 0.03864913337260015, + "grad_norm": 0.2429850548505783, + "learning_rate": 1.999609412504502e-05, + "loss": 0.018147243559360503, + "step": 730, + "token_acc": 0.991636160472311 + }, + { + "epoch": 0.03917857355578645, + "grad_norm": 0.49824216961860657, + "learning_rate": 1.9995600211935472e-05, + "loss": 0.01676984578371048, + "step": 740, + "token_acc": 0.992411417897379 + }, + { + "epoch": 0.03970801373897275, + "grad_norm": 0.23630906641483307, + "learning_rate": 1.9995076908109866e-05, + "loss": 0.017143572866916656, + "step": 750, + "token_acc": 0.9923325295977706 + }, + { + "epoch": 0.04023745392215906, + "grad_norm": 0.3479686677455902, + "learning_rate": 1.9994524215106902e-05, + "loss": 0.016651883721351624, + "step": 760, + "token_acc": 0.9926608109010754 + }, + { + "epoch": 0.04076689410534536, + "grad_norm": 0.26318269968032837, + "learning_rate": 1.9993942134551702e-05, + "loss": 0.017795515060424805, + "step": 770, + "token_acc": 0.9919672349938273 + }, + { + "epoch": 0.04129633428853166, + "grad_norm": 0.19690784811973572, + "learning_rate": 1.99933306681558e-05, + "loss": 0.01817816197872162, + "step": 780, + "token_acc": 0.9918298880886611 + }, + { + "epoch": 0.041825774471717966, + "grad_norm": 0.39895081520080566, + "learning_rate": 1.999268981771712e-05, + "loss": 0.01750715374946594, + "step": 790, + "token_acc": 0.9921704798711217 + }, + { + "epoch": 0.04235521465490427, + "grad_norm": 0.31517377495765686, + "learning_rate": 1.9992019585120005e-05, + "loss": 0.01943216472864151, + "step": 800, + "token_acc": 0.9916292500542472 + }, + { + "epoch": 0.04288465483809058, + "grad_norm": 0.2869214713573456, + "learning_rate": 1.9991319972335176e-05, + "loss": 0.016813355684280395, + "step": 810, + "token_acc": 0.9922077055280238 + }, + { + "epoch": 0.043414095021276876, + "grad_norm": 0.36023807525634766, + "learning_rate": 1.999059098141976e-05, + "loss": 0.01655134856700897, + "step": 820, + "token_acc": 0.9925307813502947 + }, + { + "epoch": 0.04394353520446318, + "grad_norm": 0.19860194623470306, + "learning_rate": 1.9989832614517247e-05, + "loss": 0.017387935519218446, + "step": 830, + "token_acc": 0.992331505912 + }, + { + "epoch": 0.04447297538764949, + "grad_norm": 0.29435139894485474, + "learning_rate": 1.998904487385752e-05, + "loss": 0.01737343370914459, + "step": 840, + "token_acc": 0.992257763094225 + }, + { + "epoch": 0.045002415570835785, + "grad_norm": 0.27239280939102173, + "learning_rate": 1.998822776175682e-05, + "loss": 0.0170346200466156, + "step": 850, + "token_acc": 0.9919921842382138 + }, + { + "epoch": 0.04553185575402209, + "grad_norm": 0.3314070999622345, + "learning_rate": 1.9987381280617762e-05, + "loss": 0.01667698919773102, + "step": 860, + "token_acc": 0.9925007090542051 + }, + { + "epoch": 0.046061295937208396, + "grad_norm": 0.4412219822406769, + "learning_rate": 1.99865054329293e-05, + "loss": 0.017632770538330077, + "step": 870, + "token_acc": 0.99207527653114 + }, + { + "epoch": 0.046590736120394695, + "grad_norm": 0.31852784752845764, + "learning_rate": 1.998560022126675e-05, + "loss": 0.01677253991365433, + "step": 880, + "token_acc": 0.9921393225798376 + }, + { + "epoch": 0.047120176303581, + "grad_norm": 0.19955475628376007, + "learning_rate": 1.9984665648291767e-05, + "loss": 0.016786649823188782, + "step": 890, + "token_acc": 0.9923666669447481 + }, + { + "epoch": 0.047649616486767306, + "grad_norm": 0.22426548600196838, + "learning_rate": 1.9983701716752333e-05, + "loss": 0.016336151957511903, + "step": 900, + "token_acc": 0.9928172785744676 + }, + { + "epoch": 0.048179056669953604, + "grad_norm": 0.2601512372493744, + "learning_rate": 1.9982708429482763e-05, + "loss": 0.017291010916233064, + "step": 910, + "token_acc": 0.9923563477361104 + }, + { + "epoch": 0.04870849685313991, + "grad_norm": 0.28832370042800903, + "learning_rate": 1.9981685789403683e-05, + "loss": 0.01724453568458557, + "step": 920, + "token_acc": 0.9921842416962782 + }, + { + "epoch": 0.049237937036326215, + "grad_norm": 0.26290687918663025, + "learning_rate": 1.9980633799522032e-05, + "loss": 0.017322060465812684, + "step": 930, + "token_acc": 0.9922752867570386 + }, + { + "epoch": 0.04976737721951252, + "grad_norm": 0.2839105427265167, + "learning_rate": 1.9979552462931034e-05, + "loss": 0.016471859812736512, + "step": 940, + "token_acc": 0.9926858763187524 + }, + { + "epoch": 0.05029681740269882, + "grad_norm": 0.2301226705312729, + "learning_rate": 1.9978441782810225e-05, + "loss": 0.016872580349445342, + "step": 950, + "token_acc": 0.9923663485283322 + }, + { + "epoch": 0.050826257585885125, + "grad_norm": 0.20936915278434753, + "learning_rate": 1.9977301762425407e-05, + "loss": 0.016406601667404173, + "step": 960, + "token_acc": 0.9925556863039651 + }, + { + "epoch": 0.05135569776907143, + "grad_norm": 0.29504403471946716, + "learning_rate": 1.9976132405128646e-05, + "loss": 0.016918857395648957, + "step": 970, + "token_acc": 0.9923744368429834 + }, + { + "epoch": 0.05188513795225773, + "grad_norm": 0.2577407956123352, + "learning_rate": 1.99749337143583e-05, + "loss": 0.016857327520847322, + "step": 980, + "token_acc": 0.9926246673174313 + }, + { + "epoch": 0.052414578135444034, + "grad_norm": 0.27821770310401917, + "learning_rate": 1.9973705693638935e-05, + "loss": 0.017641085386276244, + "step": 990, + "token_acc": 0.9919225634178905 + }, + { + "epoch": 0.05294401831863034, + "grad_norm": 0.2685238718986511, + "learning_rate": 1.99724483465814e-05, + "loss": 0.01667487621307373, + "step": 1000, + "token_acc": 0.9927382601873028 + }, + { + "epoch": 0.05347345850181664, + "grad_norm": 0.1889924854040146, + "learning_rate": 1.9971161676882744e-05, + "loss": 0.016096821427345274, + "step": 1010, + "token_acc": 0.9930279881242285 + }, + { + "epoch": 0.054002898685002944, + "grad_norm": 0.2063930779695511, + "learning_rate": 1.996984568832625e-05, + "loss": 0.016250564157962798, + "step": 1020, + "token_acc": 0.9927046901997446 + }, + { + "epoch": 0.05453233886818925, + "grad_norm": 0.339690625667572, + "learning_rate": 1.9968500384781403e-05, + "loss": 0.0167578786611557, + "step": 1030, + "token_acc": 0.9920583279109731 + }, + { + "epoch": 0.055061779051375555, + "grad_norm": 0.17077094316482544, + "learning_rate": 1.9967125770203887e-05, + "loss": 0.01681506335735321, + "step": 1040, + "token_acc": 0.9923818900922851 + }, + { + "epoch": 0.05559121923456185, + "grad_norm": 0.18625879287719727, + "learning_rate": 1.996572184863557e-05, + "loss": 0.016425205767154692, + "step": 1050, + "token_acc": 0.9928563679304331 + }, + { + "epoch": 0.05612065941774816, + "grad_norm": 0.3343530297279358, + "learning_rate": 1.9964288624204496e-05, + "loss": 0.016851165890693666, + "step": 1060, + "token_acc": 0.9925785712498854 + }, + { + "epoch": 0.056650099600934464, + "grad_norm": 0.1621190756559372, + "learning_rate": 1.9962826101124867e-05, + "loss": 0.016009426116943358, + "step": 1070, + "token_acc": 0.9926658934844098 + }, + { + "epoch": 0.05717953978412076, + "grad_norm": 0.18228836357593536, + "learning_rate": 1.9961334283697038e-05, + "loss": 0.01679401695728302, + "step": 1080, + "token_acc": 0.9921864576384256 + }, + { + "epoch": 0.05770897996730707, + "grad_norm": 0.2137574553489685, + "learning_rate": 1.9959813176307496e-05, + "loss": 0.016052186489105225, + "step": 1090, + "token_acc": 0.9926035379465218 + }, + { + "epoch": 0.058238420150493374, + "grad_norm": 0.349079966545105, + "learning_rate": 1.995826278342885e-05, + "loss": 0.016752496361732483, + "step": 1100, + "token_acc": 0.9922574401161384 + }, + { + "epoch": 0.05876786033367967, + "grad_norm": 0.20915967226028442, + "learning_rate": 1.9956683109619824e-05, + "loss": 0.01655551791191101, + "step": 1110, + "token_acc": 0.9927447902751804 + }, + { + "epoch": 0.05929730051686598, + "grad_norm": 0.2501552700996399, + "learning_rate": 1.9955074159525236e-05, + "loss": 0.016435810923576356, + "step": 1120, + "token_acc": 0.9926656487020952 + }, + { + "epoch": 0.05982674070005228, + "grad_norm": 0.2083655595779419, + "learning_rate": 1.9953435937875988e-05, + "loss": 0.015829847753047945, + "step": 1130, + "token_acc": 0.9929595109240329 + }, + { + "epoch": 0.06035618088323859, + "grad_norm": 0.1325041949748993, + "learning_rate": 1.9951768449489045e-05, + "loss": 0.015565156936645508, + "step": 1140, + "token_acc": 0.9929556329926472 + }, + { + "epoch": 0.06088562106642489, + "grad_norm": 0.2056419998407364, + "learning_rate": 1.9950071699267438e-05, + "loss": 0.01673245131969452, + "step": 1150, + "token_acc": 0.9925984646194926 + }, + { + "epoch": 0.06141506124961119, + "grad_norm": 0.2566123604774475, + "learning_rate": 1.9948345692200227e-05, + "loss": 0.016523544490337373, + "step": 1160, + "token_acc": 0.9925987734156619 + }, + { + "epoch": 0.0619445014327975, + "grad_norm": 0.2760147154331207, + "learning_rate": 1.9946590433362512e-05, + "loss": 0.01607349216938019, + "step": 1170, + "token_acc": 0.9928859754470243 + }, + { + "epoch": 0.0624739416159838, + "grad_norm": 0.19814899563789368, + "learning_rate": 1.9944805927915384e-05, + "loss": 0.015280820429325104, + "step": 1180, + "token_acc": 0.9933356687685584 + }, + { + "epoch": 0.0630033817991701, + "grad_norm": 0.1616675853729248, + "learning_rate": 1.9942992181105948e-05, + "loss": 0.016144189238548278, + "step": 1190, + "token_acc": 0.9928970996490234 + }, + { + "epoch": 0.06353282198235641, + "grad_norm": 0.08978190273046494, + "learning_rate": 1.994114919826728e-05, + "loss": 0.01823064237833023, + "step": 1200, + "token_acc": 0.992564466327297 + }, + { + "epoch": 0.06406226216554271, + "grad_norm": 0.19705332815647125, + "learning_rate": 1.9939276984818423e-05, + "loss": 0.015870603919029235, + "step": 1210, + "token_acc": 0.9929304732493114 + }, + { + "epoch": 0.06459170234872902, + "grad_norm": 4.296022891998291, + "learning_rate": 1.9937375546264367e-05, + "loss": 0.032708361744880676, + "step": 1220, + "token_acc": 0.991791860261426 + }, + { + "epoch": 0.06512114253191531, + "grad_norm": 0.19522494077682495, + "learning_rate": 1.9935444888196044e-05, + "loss": 0.015410324931144715, + "step": 1230, + "token_acc": 0.9932358607805033 + }, + { + "epoch": 0.06565058271510162, + "grad_norm": 0.1904059797525406, + "learning_rate": 1.993348501629028e-05, + "loss": 0.014347922801971436, + "step": 1240, + "token_acc": 0.9934757761073243 + }, + { + "epoch": 0.06618002289828792, + "grad_norm": 0.21662530303001404, + "learning_rate": 1.9931495936309827e-05, + "loss": 0.015134570002555848, + "step": 1250, + "token_acc": 0.9930633717682565 + }, + { + "epoch": 0.06670946308147423, + "grad_norm": 0.15513503551483154, + "learning_rate": 1.99294776541033e-05, + "loss": 0.015029802918434143, + "step": 1260, + "token_acc": 0.9932858477347034 + }, + { + "epoch": 0.06723890326466053, + "grad_norm": 0.18613627552986145, + "learning_rate": 1.992743017560519e-05, + "loss": 0.014833299815654755, + "step": 1270, + "token_acc": 0.9933416215404384 + }, + { + "epoch": 0.06776834344784684, + "grad_norm": 0.13888917863368988, + "learning_rate": 1.9925353506835827e-05, + "loss": 0.01528899371623993, + "step": 1280, + "token_acc": 0.993001743308282 + }, + { + "epoch": 0.06829778363103314, + "grad_norm": 0.2606760561466217, + "learning_rate": 1.992324765390138e-05, + "loss": 0.015293537080287934, + "step": 1290, + "token_acc": 0.9930779110303071 + }, + { + "epoch": 0.06882722381421943, + "grad_norm": 0.16797654330730438, + "learning_rate": 1.9921112622993825e-05, + "loss": 0.015485766530036926, + "step": 1300, + "token_acc": 0.9929817826772704 + }, + { + "epoch": 0.06935666399740574, + "grad_norm": 0.18898263573646545, + "learning_rate": 1.991894842039093e-05, + "loss": 0.014706256985664367, + "step": 1310, + "token_acc": 0.9933361134278565 + }, + { + "epoch": 0.06988610418059205, + "grad_norm": 0.2777235507965088, + "learning_rate": 1.991675505245624e-05, + "loss": 0.015264616906642913, + "step": 1320, + "token_acc": 0.993325936246006 + }, + { + "epoch": 0.07041554436377835, + "grad_norm": 0.24200356006622314, + "learning_rate": 1.9914532525639067e-05, + "loss": 0.015748044848442076, + "step": 1330, + "token_acc": 0.9928487387244553 + }, + { + "epoch": 0.07094498454696466, + "grad_norm": 0.21135491132736206, + "learning_rate": 1.9912280846474445e-05, + "loss": 0.014889243245124816, + "step": 1340, + "token_acc": 0.9933887057055804 + }, + { + "epoch": 0.07147442473015096, + "grad_norm": 0.12637273967266083, + "learning_rate": 1.9910000021583132e-05, + "loss": 0.015416422486305236, + "step": 1350, + "token_acc": 0.9933259919244503 + }, + { + "epoch": 0.07200386491333725, + "grad_norm": 0.1456395387649536, + "learning_rate": 1.9907690057671592e-05, + "loss": 0.015557226538658143, + "step": 1360, + "token_acc": 0.9929674399979979 + }, + { + "epoch": 0.07253330509652356, + "grad_norm": 0.12905369699001312, + "learning_rate": 1.9905350961531958e-05, + "loss": 0.014846476912498473, + "step": 1370, + "token_acc": 0.9931621650920189 + }, + { + "epoch": 0.07306274527970986, + "grad_norm": 0.12199676781892776, + "learning_rate": 1.9902982740042027e-05, + "loss": 0.015367740392684936, + "step": 1380, + "token_acc": 0.9934052357419527 + }, + { + "epoch": 0.07359218546289617, + "grad_norm": 0.16988356411457062, + "learning_rate": 1.990058540016524e-05, + "loss": 0.0151124969124794, + "step": 1390, + "token_acc": 0.9931626212175537 + }, + { + "epoch": 0.07412162564608248, + "grad_norm": 0.17141181230545044, + "learning_rate": 1.9898158948950645e-05, + "loss": 0.015593770146369933, + "step": 1400, + "token_acc": 0.9929320070763377 + }, + { + "epoch": 0.07465106582926878, + "grad_norm": 0.1844063699245453, + "learning_rate": 1.9895703393532898e-05, + "loss": 0.01625789701938629, + "step": 1410, + "token_acc": 0.9930574098798398 + }, + { + "epoch": 0.07518050601245509, + "grad_norm": 0.23952791094779968, + "learning_rate": 1.9893218741132233e-05, + "loss": 0.016111481189727783, + "step": 1420, + "token_acc": 0.9926685850118854 + }, + { + "epoch": 0.07570994619564138, + "grad_norm": 0.3312993049621582, + "learning_rate": 1.9890704999054433e-05, + "loss": 0.015351589024066924, + "step": 1430, + "token_acc": 0.9931177162496976 + }, + { + "epoch": 0.07623938637882768, + "grad_norm": 0.18104445934295654, + "learning_rate": 1.9888162174690814e-05, + "loss": 0.01543954312801361, + "step": 1440, + "token_acc": 0.9931117400996919 + }, + { + "epoch": 0.07676882656201399, + "grad_norm": 0.25670966506004333, + "learning_rate": 1.9885590275518216e-05, + "loss": 0.014887252449989319, + "step": 1450, + "token_acc": 0.9932421158017687 + }, + { + "epoch": 0.0772982667452003, + "grad_norm": 0.20747993886470795, + "learning_rate": 1.9882989309098954e-05, + "loss": 0.014649266004562378, + "step": 1460, + "token_acc": 0.9931799801569104 + }, + { + "epoch": 0.0778277069283866, + "grad_norm": 0.1586734503507614, + "learning_rate": 1.9880359283080828e-05, + "loss": 0.01536465585231781, + "step": 1470, + "token_acc": 0.9933120408282395 + }, + { + "epoch": 0.0783571471115729, + "grad_norm": 0.20884263515472412, + "learning_rate": 1.9877700205197067e-05, + "loss": 0.01486780196428299, + "step": 1480, + "token_acc": 0.9935166797389984 + }, + { + "epoch": 0.0788865872947592, + "grad_norm": 0.2085837572813034, + "learning_rate": 1.987501208326634e-05, + "loss": 0.015686462819576263, + "step": 1490, + "token_acc": 0.9931257143333861 + }, + { + "epoch": 0.0794160274779455, + "grad_norm": 0.17206107079982758, + "learning_rate": 1.9872294925192695e-05, + "loss": 0.014749321341514587, + "step": 1500, + "token_acc": 0.9933413993074388 + }, + { + "epoch": 0.07994546766113181, + "grad_norm": 0.31142544746398926, + "learning_rate": 1.986954873896558e-05, + "loss": 0.015649592876434325, + "step": 1510, + "token_acc": 0.9931577717699677 + }, + { + "epoch": 0.08047490784431811, + "grad_norm": 0.2183057963848114, + "learning_rate": 1.9866773532659786e-05, + "loss": 0.015441799163818359, + "step": 1520, + "token_acc": 0.9930077515498927 + }, + { + "epoch": 0.08100434802750442, + "grad_norm": 0.2614285945892334, + "learning_rate": 1.986396931443543e-05, + "loss": 0.014922884106636048, + "step": 1530, + "token_acc": 0.9935867498415452 + }, + { + "epoch": 0.08153378821069072, + "grad_norm": 0.21387344598770142, + "learning_rate": 1.9861136092537933e-05, + "loss": 0.01545359194278717, + "step": 1540, + "token_acc": 0.9929254334006307 + }, + { + "epoch": 0.08206322839387703, + "grad_norm": 0.19913002848625183, + "learning_rate": 1.9858273875298012e-05, + "loss": 0.01475047767162323, + "step": 1550, + "token_acc": 0.9933782566634419 + }, + { + "epoch": 0.08259266857706332, + "grad_norm": 0.25380730628967285, + "learning_rate": 1.9855382671131625e-05, + "loss": 0.014745526015758514, + "step": 1560, + "token_acc": 0.99333511286098 + }, + { + "epoch": 0.08312210876024963, + "grad_norm": 0.15402960777282715, + "learning_rate": 1.9852462488539967e-05, + "loss": 0.01432792842388153, + "step": 1570, + "token_acc": 0.9933849413570464 + }, + { + "epoch": 0.08365154894343593, + "grad_norm": 0.20179462432861328, + "learning_rate": 1.9849513336109446e-05, + "loss": 0.01479189395904541, + "step": 1580, + "token_acc": 0.9936367882042899 + }, + { + "epoch": 0.08418098912662224, + "grad_norm": 0.15874245762825012, + "learning_rate": 1.9846535222511648e-05, + "loss": 0.014734578132629395, + "step": 1590, + "token_acc": 0.993673474497759 + }, + { + "epoch": 0.08471042930980854, + "grad_norm": 0.1751202642917633, + "learning_rate": 1.9843528156503314e-05, + "loss": 0.014660277962684631, + "step": 1600, + "token_acc": 0.9937424387801927 + }, + { + "epoch": 0.08523986949299485, + "grad_norm": 0.23757734894752502, + "learning_rate": 1.984049214692631e-05, + "loss": 0.014745841920375823, + "step": 1610, + "token_acc": 0.993356007936111 + }, + { + "epoch": 0.08576930967618115, + "grad_norm": 0.2505788505077362, + "learning_rate": 1.9837427202707623e-05, + "loss": 0.015558727085590363, + "step": 1620, + "token_acc": 0.9929686718046241 + }, + { + "epoch": 0.08629874985936745, + "grad_norm": 0.16628512740135193, + "learning_rate": 1.9834333332859304e-05, + "loss": 0.015397495031356812, + "step": 1630, + "token_acc": 0.992912656209565 + }, + { + "epoch": 0.08682819004255375, + "grad_norm": 0.1155611127614975, + "learning_rate": 1.983121054647846e-05, + "loss": 0.014987055957317353, + "step": 1640, + "token_acc": 0.993581825465714 + }, + { + "epoch": 0.08735763022574006, + "grad_norm": 0.18505799770355225, + "learning_rate": 1.9828058852747224e-05, + "loss": 0.014484158158302307, + "step": 1650, + "token_acc": 0.9933281561829371 + }, + { + "epoch": 0.08788707040892636, + "grad_norm": 0.2567390501499176, + "learning_rate": 1.9824878260932728e-05, + "loss": 0.015050376951694488, + "step": 1660, + "token_acc": 0.9930978650953947 + }, + { + "epoch": 0.08841651059211267, + "grad_norm": 0.12895241379737854, + "learning_rate": 1.9821668780387073e-05, + "loss": 0.014717523753643037, + "step": 1670, + "token_acc": 0.9934380791771719 + }, + { + "epoch": 0.08894595077529897, + "grad_norm": 0.18911804258823395, + "learning_rate": 1.9818430420547307e-05, + "loss": 0.015375326573848724, + "step": 1680, + "token_acc": 0.9930386878677852 + }, + { + "epoch": 0.08947539095848527, + "grad_norm": 0.2011958211660385, + "learning_rate": 1.9815163190935384e-05, + "loss": 0.015271103382110596, + "step": 1690, + "token_acc": 0.993287077624427 + }, + { + "epoch": 0.09000483114167157, + "grad_norm": 0.1693992167711258, + "learning_rate": 1.9811867101158162e-05, + "loss": 0.015348242223262787, + "step": 1700, + "token_acc": 0.9932380538952149 + }, + { + "epoch": 0.09053427132485788, + "grad_norm": 0.24018199741840363, + "learning_rate": 1.9808542160907342e-05, + "loss": 0.015835171937942503, + "step": 1710, + "token_acc": 0.9928925371640223 + }, + { + "epoch": 0.09106371150804418, + "grad_norm": 0.12215518206357956, + "learning_rate": 1.9805188379959466e-05, + "loss": 0.014699850976467133, + "step": 1720, + "token_acc": 0.9934657431361095 + }, + { + "epoch": 0.09159315169123049, + "grad_norm": 0.1710125356912613, + "learning_rate": 1.980180576817587e-05, + "loss": 0.014636707305908204, + "step": 1730, + "token_acc": 0.9933871478843876 + }, + { + "epoch": 0.09212259187441679, + "grad_norm": 0.3093157708644867, + "learning_rate": 1.9798394335502676e-05, + "loss": 0.01454448103904724, + "step": 1740, + "token_acc": 0.9936185121540233 + }, + { + "epoch": 0.0926520320576031, + "grad_norm": 0.10877680778503418, + "learning_rate": 1.979495409197074e-05, + "loss": 0.015331670641899109, + "step": 1750, + "token_acc": 0.9932189535669305 + }, + { + "epoch": 0.09318147224078939, + "grad_norm": 0.13713012635707855, + "learning_rate": 1.979148504769563e-05, + "loss": 0.014645929634571075, + "step": 1760, + "token_acc": 0.9935109346589487 + }, + { + "epoch": 0.0937109124239757, + "grad_norm": 0.15919052064418793, + "learning_rate": 1.978798721287761e-05, + "loss": 0.014620964229106904, + "step": 1770, + "token_acc": 0.9934340063407309 + }, + { + "epoch": 0.094240352607162, + "grad_norm": 0.08753754198551178, + "learning_rate": 1.9784460597801587e-05, + "loss": 0.015422160923480987, + "step": 1780, + "token_acc": 0.9928837201541721 + }, + { + "epoch": 0.0947697927903483, + "grad_norm": 0.16888871788978577, + "learning_rate": 1.9780905212837102e-05, + "loss": 0.014722776412963868, + "step": 1790, + "token_acc": 0.9937094325140579 + }, + { + "epoch": 0.09529923297353461, + "grad_norm": 0.11716227978467941, + "learning_rate": 1.9777321068438282e-05, + "loss": 0.014879959821701049, + "step": 1800, + "token_acc": 0.9934569611590526 + }, + { + "epoch": 0.09582867315672092, + "grad_norm": 0.12512709200382233, + "learning_rate": 1.9773708175143823e-05, + "loss": 0.014694008231163024, + "step": 1810, + "token_acc": 0.9934850974732856 + }, + { + "epoch": 0.09635811333990721, + "grad_norm": 0.2145359218120575, + "learning_rate": 1.9770066543576946e-05, + "loss": 0.014114877581596375, + "step": 1820, + "token_acc": 0.9938457433057865 + }, + { + "epoch": 0.09688755352309351, + "grad_norm": 0.15780009329319, + "learning_rate": 1.976639618444539e-05, + "loss": 0.014264130592346191, + "step": 1830, + "token_acc": 0.9934067399621576 + }, + { + "epoch": 0.09741699370627982, + "grad_norm": 0.12755165994167328, + "learning_rate": 1.9762697108541338e-05, + "loss": 0.014090609550476075, + "step": 1840, + "token_acc": 0.993715783148618 + }, + { + "epoch": 0.09794643388946613, + "grad_norm": 0.14389929175376892, + "learning_rate": 1.975896932674143e-05, + "loss": 0.014824743568897247, + "step": 1850, + "token_acc": 0.993267201174694 + }, + { + "epoch": 0.09847587407265243, + "grad_norm": 0.13695280253887177, + "learning_rate": 1.975521285000671e-05, + "loss": 0.014638055860996247, + "step": 1860, + "token_acc": 0.9932954270303668 + }, + { + "epoch": 0.09900531425583874, + "grad_norm": 0.09902188926935196, + "learning_rate": 1.9751427689382588e-05, + "loss": 0.014058363437652589, + "step": 1870, + "token_acc": 0.9938034277136066 + }, + { + "epoch": 0.09953475443902504, + "grad_norm": 0.1432419866323471, + "learning_rate": 1.974761385599882e-05, + "loss": 0.014428834617137908, + "step": 1880, + "token_acc": 0.9936818208374717 + }, + { + "epoch": 0.10006419462221133, + "grad_norm": 0.20492440462112427, + "learning_rate": 1.9743771361069467e-05, + "loss": 0.014150190353393554, + "step": 1890, + "token_acc": 0.993652091223036 + }, + { + "epoch": 0.10059363480539764, + "grad_norm": 0.1718825399875641, + "learning_rate": 1.973990021589287e-05, + "loss": 0.014242845773696899, + "step": 1900, + "token_acc": 0.993646134731962 + }, + { + "epoch": 0.10112307498858394, + "grad_norm": 0.20733670890331268, + "learning_rate": 1.973600043185161e-05, + "loss": 0.014840033650398255, + "step": 1910, + "token_acc": 0.9933642722401216 + }, + { + "epoch": 0.10165251517177025, + "grad_norm": 0.13032479584217072, + "learning_rate": 1.9732072020412475e-05, + "loss": 0.01504979133605957, + "step": 1920, + "token_acc": 0.9934881394088465 + }, + { + "epoch": 0.10218195535495656, + "grad_norm": 0.16003693640232086, + "learning_rate": 1.9728114993126432e-05, + "loss": 0.014648252725601196, + "step": 1930, + "token_acc": 0.9934719636700586 + }, + { + "epoch": 0.10271139553814286, + "grad_norm": 0.20438160002231598, + "learning_rate": 1.9724129361628586e-05, + "loss": 0.014987148344516754, + "step": 1940, + "token_acc": 0.9934909958776307 + }, + { + "epoch": 0.10324083572132917, + "grad_norm": 0.15040352940559387, + "learning_rate": 1.972011513763815e-05, + "loss": 0.014513610303401947, + "step": 1950, + "token_acc": 0.9934434434434435 + }, + { + "epoch": 0.10377027590451546, + "grad_norm": 0.21010926365852356, + "learning_rate": 1.971607233295841e-05, + "loss": 0.014066994190216064, + "step": 1960, + "token_acc": 0.9935120710503274 + }, + { + "epoch": 0.10429971608770176, + "grad_norm": 0.1677207201719284, + "learning_rate": 1.9712000959476683e-05, + "loss": 0.014186528325080872, + "step": 1970, + "token_acc": 0.9936180862601152 + }, + { + "epoch": 0.10482915627088807, + "grad_norm": 0.21293596923351288, + "learning_rate": 1.9707901029164297e-05, + "loss": 0.014313945174217224, + "step": 1980, + "token_acc": 0.9938200045036405 + }, + { + "epoch": 0.10535859645407437, + "grad_norm": 0.13706688582897186, + "learning_rate": 1.970377255407655e-05, + "loss": 0.01445833146572113, + "step": 1990, + "token_acc": 0.9935971187034165 + }, + { + "epoch": 0.10588803663726068, + "grad_norm": 0.1516571044921875, + "learning_rate": 1.969961554635266e-05, + "loss": 0.014031200110912323, + "step": 2000, + "token_acc": 0.9937349839829152 + }, + { + "epoch": 0.10641747682044699, + "grad_norm": 0.17179790139198303, + "learning_rate": 1.9695430018215752e-05, + "loss": 0.013821715116500854, + "step": 2010, + "token_acc": 0.9938317975427351 + }, + { + "epoch": 0.10694691700363328, + "grad_norm": 0.12614597380161285, + "learning_rate": 1.969121598197281e-05, + "loss": 0.01368074119091034, + "step": 2020, + "token_acc": 0.9938387400681991 + }, + { + "epoch": 0.10747635718681958, + "grad_norm": 0.1423429548740387, + "learning_rate": 1.9686973450014635e-05, + "loss": 0.014041239023208618, + "step": 2030, + "token_acc": 0.9939541012041763 + }, + { + "epoch": 0.10800579737000589, + "grad_norm": 0.1439235657453537, + "learning_rate": 1.968270243481582e-05, + "loss": 0.013630753755569458, + "step": 2040, + "token_acc": 0.9939198556800909 + }, + { + "epoch": 0.1085352375531922, + "grad_norm": 0.19314269721508026, + "learning_rate": 1.9678402948934717e-05, + "loss": 0.014843356609344483, + "step": 2050, + "token_acc": 0.9935561416981069 + }, + { + "epoch": 0.1090646777363785, + "grad_norm": 0.16090211272239685, + "learning_rate": 1.9674075005013378e-05, + "loss": 0.014360342919826508, + "step": 2060, + "token_acc": 0.9934506378327869 + }, + { + "epoch": 0.1095941179195648, + "grad_norm": 0.13298411667346954, + "learning_rate": 1.9669718615777542e-05, + "loss": 0.014340427517890931, + "step": 2070, + "token_acc": 0.993610916491509 + }, + { + "epoch": 0.11012355810275111, + "grad_norm": 0.09223160147666931, + "learning_rate": 1.9665333794036586e-05, + "loss": 0.01421847939491272, + "step": 2080, + "token_acc": 0.9937143143343684 + }, + { + "epoch": 0.1106529982859374, + "grad_norm": 0.09074156731367111, + "learning_rate": 1.9660920552683484e-05, + "loss": 0.014728325605392455, + "step": 2090, + "token_acc": 0.993379526561106 + }, + { + "epoch": 0.1111824384691237, + "grad_norm": 0.20380668342113495, + "learning_rate": 1.9656478904694783e-05, + "loss": 0.014153891801834106, + "step": 2100, + "token_acc": 0.9938824904022701 + }, + { + "epoch": 0.11171187865231001, + "grad_norm": 0.1178550198674202, + "learning_rate": 1.9652008863130544e-05, + "loss": 0.01406555324792862, + "step": 2110, + "token_acc": 0.993663594470046 + }, + { + "epoch": 0.11224131883549632, + "grad_norm": 0.1450866460800171, + "learning_rate": 1.9647510441134327e-05, + "loss": 0.01425243616104126, + "step": 2120, + "token_acc": 0.9935185725844796 + }, + { + "epoch": 0.11277075901868262, + "grad_norm": 0.1353432536125183, + "learning_rate": 1.964298365193314e-05, + "loss": 0.014033317565917969, + "step": 2130, + "token_acc": 0.9938276753690883 + }, + { + "epoch": 0.11330019920186893, + "grad_norm": 0.11431889981031418, + "learning_rate": 1.9638428508837382e-05, + "loss": 0.013508771359920502, + "step": 2140, + "token_acc": 0.9938711183749708 + }, + { + "epoch": 0.11382963938505523, + "grad_norm": 0.12013133615255356, + "learning_rate": 1.9633845025240855e-05, + "loss": 0.013967834413051605, + "step": 2150, + "token_acc": 0.9938416348873887 + }, + { + "epoch": 0.11435907956824153, + "grad_norm": 0.1236162781715393, + "learning_rate": 1.9629233214620668e-05, + "loss": 0.013715249300003052, + "step": 2160, + "token_acc": 0.9937885609471402 + }, + { + "epoch": 0.11488851975142783, + "grad_norm": 0.09243787825107574, + "learning_rate": 1.9624593090537223e-05, + "loss": 0.013384032249450683, + "step": 2170, + "token_acc": 0.9941388337696553 + }, + { + "epoch": 0.11541795993461414, + "grad_norm": 0.1580934077501297, + "learning_rate": 1.9619924666634188e-05, + "loss": 0.013232611119747162, + "step": 2180, + "token_acc": 0.9941723289841177 + }, + { + "epoch": 0.11594740011780044, + "grad_norm": 0.19918663799762726, + "learning_rate": 1.9615227956638433e-05, + "loss": 0.013959816098213196, + "step": 2190, + "token_acc": 0.9940662972359461 + }, + { + "epoch": 0.11647684030098675, + "grad_norm": 0.15594160556793213, + "learning_rate": 1.961050297436e-05, + "loss": 0.014273026585578918, + "step": 2200, + "token_acc": 0.9934810734109595 + }, + { + "epoch": 0.11700628048417305, + "grad_norm": 0.1488836705684662, + "learning_rate": 1.9605749733692063e-05, + "loss": 0.01350422203540802, + "step": 2210, + "token_acc": 0.9938529546686684 + }, + { + "epoch": 0.11753572066735934, + "grad_norm": 0.0939706340432167, + "learning_rate": 1.9600968248610885e-05, + "loss": 0.014030486345291138, + "step": 2220, + "token_acc": 0.993677537742931 + }, + { + "epoch": 0.11806516085054565, + "grad_norm": 0.14121605455875397, + "learning_rate": 1.959615853317578e-05, + "loss": 0.013287697732448579, + "step": 2230, + "token_acc": 0.9944368916653461 + }, + { + "epoch": 0.11859460103373196, + "grad_norm": 0.17856624722480774, + "learning_rate": 1.9591320601529068e-05, + "loss": 0.01357399970293045, + "step": 2240, + "token_acc": 0.9939437923871969 + }, + { + "epoch": 0.11912404121691826, + "grad_norm": 0.10048573464155197, + "learning_rate": 1.958645446789604e-05, + "loss": 0.01314356029033661, + "step": 2250, + "token_acc": 0.9943610276943611 + }, + { + "epoch": 0.11965348140010457, + "grad_norm": 0.19854819774627686, + "learning_rate": 1.9581560146584896e-05, + "loss": 0.013932856917381286, + "step": 2260, + "token_acc": 0.9936940093088434 + }, + { + "epoch": 0.12018292158329087, + "grad_norm": 0.25018957257270813, + "learning_rate": 1.9576637651986738e-05, + "loss": 0.013639228045940399, + "step": 2270, + "token_acc": 0.99406255993729 + }, + { + "epoch": 0.12071236176647718, + "grad_norm": 0.1087864562869072, + "learning_rate": 1.9571686998575493e-05, + "loss": 0.013747358322143554, + "step": 2280, + "token_acc": 0.9939612317752644 + }, + { + "epoch": 0.12124180194966347, + "grad_norm": 0.12415312975645065, + "learning_rate": 1.9566708200907896e-05, + "loss": 0.013799256086349488, + "step": 2290, + "token_acc": 0.9937187711146888 + }, + { + "epoch": 0.12177124213284977, + "grad_norm": 0.14945298433303833, + "learning_rate": 1.9561701273623425e-05, + "loss": 0.014375174045562744, + "step": 2300, + "token_acc": 0.9935965402950434 + }, + { + "epoch": 0.12230068231603608, + "grad_norm": 0.1431857794523239, + "learning_rate": 1.9556666231444278e-05, + "loss": 0.014219081401824952, + "step": 2310, + "token_acc": 0.9936443858741889 + }, + { + "epoch": 0.12283012249922239, + "grad_norm": 0.11537466943264008, + "learning_rate": 1.9551603089175316e-05, + "loss": 0.013759464025497437, + "step": 2320, + "token_acc": 0.9940598354774658 + }, + { + "epoch": 0.12335956268240869, + "grad_norm": 0.2179916650056839, + "learning_rate": 1.9546511861704032e-05, + "loss": 0.014386409521102905, + "step": 2330, + "token_acc": 0.9935441897707937 + }, + { + "epoch": 0.123889002865595, + "grad_norm": 0.08233068883419037, + "learning_rate": 1.954139256400049e-05, + "loss": 0.012992560863494873, + "step": 2340, + "token_acc": 0.9941480924952 + }, + { + "epoch": 0.12441844304878129, + "grad_norm": 0.1415328085422516, + "learning_rate": 1.9536245211117298e-05, + "loss": 0.01337493360042572, + "step": 2350, + "token_acc": 0.9939491562202674 + }, + { + "epoch": 0.1249478832319676, + "grad_norm": 0.10844320803880692, + "learning_rate": 1.9531069818189553e-05, + "loss": 0.01349758803844452, + "step": 2360, + "token_acc": 0.9941084676174342 + }, + { + "epoch": 0.1254773234151539, + "grad_norm": 0.21574494242668152, + "learning_rate": 1.9525866400434798e-05, + "loss": 0.013676650822162628, + "step": 2370, + "token_acc": 0.994159463338117 + }, + { + "epoch": 0.1260067635983402, + "grad_norm": 0.12667444348335266, + "learning_rate": 1.952063497315299e-05, + "loss": 0.013422438502311706, + "step": 2380, + "token_acc": 0.9940165733407883 + }, + { + "epoch": 0.1265362037815265, + "grad_norm": 0.1800079643726349, + "learning_rate": 1.9515375551726424e-05, + "loss": 0.013396120071411133, + "step": 2390, + "token_acc": 0.9939185638133692 + }, + { + "epoch": 0.12706564396471282, + "grad_norm": 0.15108215808868408, + "learning_rate": 1.951008815161973e-05, + "loss": 0.013133403658866883, + "step": 2400, + "token_acc": 0.9942942942942943 + }, + { + "epoch": 0.1275950841478991, + "grad_norm": 0.11209385097026825, + "learning_rate": 1.9504772788379792e-05, + "loss": 0.013791799545288086, + "step": 2410, + "token_acc": 0.9939343373242668 + }, + { + "epoch": 0.12812452433108543, + "grad_norm": 0.21388082206249237, + "learning_rate": 1.9499429477635724e-05, + "loss": 0.013684123754501343, + "step": 2420, + "token_acc": 0.9939735906381985 + }, + { + "epoch": 0.12865396451427172, + "grad_norm": 0.15899330377578735, + "learning_rate": 1.9494058235098804e-05, + "loss": 0.01334327757358551, + "step": 2430, + "token_acc": 0.9941101387371003 + }, + { + "epoch": 0.12918340469745804, + "grad_norm": 0.17888228595256805, + "learning_rate": 1.9488659076562454e-05, + "loss": 0.01430957019329071, + "step": 2440, + "token_acc": 0.9934907786030209 + }, + { + "epoch": 0.12971284488064433, + "grad_norm": 0.10245105624198914, + "learning_rate": 1.948323201790217e-05, + "loss": 0.013035660982131958, + "step": 2450, + "token_acc": 0.9943026834944654 + }, + { + "epoch": 0.13024228506383062, + "grad_norm": 0.13153426349163055, + "learning_rate": 1.9477777075075487e-05, + "loss": 0.013706900179386139, + "step": 2460, + "token_acc": 0.9939764062004639 + }, + { + "epoch": 0.13077172524701694, + "grad_norm": 0.11353341490030289, + "learning_rate": 1.9472294264121926e-05, + "loss": 0.01354885995388031, + "step": 2470, + "token_acc": 0.9941109248183646 + }, + { + "epoch": 0.13130116543020323, + "grad_norm": 0.17793892323970795, + "learning_rate": 1.946678360116296e-05, + "loss": 0.013161835074424744, + "step": 2480, + "token_acc": 0.9941296154231013 + }, + { + "epoch": 0.13183060561338955, + "grad_norm": 0.12724965810775757, + "learning_rate": 1.9461245102401946e-05, + "loss": 0.013097435235977173, + "step": 2490, + "token_acc": 0.9943968936578849 + }, + { + "epoch": 0.13236004579657584, + "grad_norm": 0.1362168788909912, + "learning_rate": 1.94556787841241e-05, + "loss": 0.013753545284271241, + "step": 2500, + "token_acc": 0.993732265064263 + }, + { + "epoch": 0.13288948597976216, + "grad_norm": 0.18147434294223785, + "learning_rate": 1.9450084662696425e-05, + "loss": 0.013303205370903015, + "step": 2510, + "token_acc": 0.9937209709563637 + }, + { + "epoch": 0.13341892616294845, + "grad_norm": 0.2115943729877472, + "learning_rate": 1.9444462754567682e-05, + "loss": 0.013561353087425232, + "step": 2520, + "token_acc": 0.9939848996788053 + }, + { + "epoch": 0.13394836634613475, + "grad_norm": 0.13511165976524353, + "learning_rate": 1.943881307626834e-05, + "loss": 0.013393791019916534, + "step": 2530, + "token_acc": 0.9937148389897 + }, + { + "epoch": 0.13447780652932106, + "grad_norm": 0.10359222441911697, + "learning_rate": 1.943313564441051e-05, + "loss": 0.012736321985721588, + "step": 2540, + "token_acc": 0.9943694633055838 + }, + { + "epoch": 0.13500724671250736, + "grad_norm": 0.07508662343025208, + "learning_rate": 1.9427430475687928e-05, + "loss": 0.01375180184841156, + "step": 2550, + "token_acc": 0.9939321097395064 + }, + { + "epoch": 0.13553668689569368, + "grad_norm": 0.2055548131465912, + "learning_rate": 1.9421697586875857e-05, + "loss": 0.013411034643650056, + "step": 2560, + "token_acc": 0.9940528818083243 + }, + { + "epoch": 0.13606612707887997, + "grad_norm": 0.12977179884910583, + "learning_rate": 1.9415936994831096e-05, + "loss": 0.013373830914497375, + "step": 2570, + "token_acc": 0.9939856523189856 + }, + { + "epoch": 0.1365955672620663, + "grad_norm": 0.1589537262916565, + "learning_rate": 1.9410148716491886e-05, + "loss": 0.013372272253036499, + "step": 2580, + "token_acc": 0.9940783986655546 + }, + { + "epoch": 0.13712500744525258, + "grad_norm": 0.24759438633918762, + "learning_rate": 1.9404332768877882e-05, + "loss": 0.013328668475151063, + "step": 2590, + "token_acc": 0.9941038129232411 + }, + { + "epoch": 0.13765444762843887, + "grad_norm": 0.08139269798994064, + "learning_rate": 1.9398489169090088e-05, + "loss": 0.013412672281265258, + "step": 2600, + "token_acc": 0.9939152136352634 + }, + { + "epoch": 0.1381838878116252, + "grad_norm": 0.1369147002696991, + "learning_rate": 1.939261793431083e-05, + "loss": 0.013577595353126526, + "step": 2610, + "token_acc": 0.993859349057627 + }, + { + "epoch": 0.13871332799481148, + "grad_norm": 0.06886211037635803, + "learning_rate": 1.938671908180368e-05, + "loss": 0.012858301401138306, + "step": 2620, + "token_acc": 0.9944881758447016 + }, + { + "epoch": 0.1392427681779978, + "grad_norm": 0.14026515185832977, + "learning_rate": 1.9380792628913422e-05, + "loss": 0.01358620822429657, + "step": 2630, + "token_acc": 0.9939697149395719 + }, + { + "epoch": 0.1397722083611841, + "grad_norm": 0.1343849152326584, + "learning_rate": 1.9374838593065988e-05, + "loss": 0.013179725408554077, + "step": 2640, + "token_acc": 0.9941789190135853 + }, + { + "epoch": 0.14030164854437038, + "grad_norm": 0.1271725296974182, + "learning_rate": 1.936885699176842e-05, + "loss": 0.013085122406482696, + "step": 2650, + "token_acc": 0.9941663467476757 + }, + { + "epoch": 0.1408310887275567, + "grad_norm": 0.12383197247982025, + "learning_rate": 1.9362847842608812e-05, + "loss": 0.013596424460411071, + "step": 2660, + "token_acc": 0.9941654702681045 + }, + { + "epoch": 0.141360528910743, + "grad_norm": 0.11162195354700089, + "learning_rate": 1.9356811163256257e-05, + "loss": 0.013854682445526123, + "step": 2670, + "token_acc": 0.9939947954894242 + }, + { + "epoch": 0.1418899690939293, + "grad_norm": 0.19413498044013977, + "learning_rate": 1.93507469714608e-05, + "loss": 0.013428667187690735, + "step": 2680, + "token_acc": 0.9941157322783384 + }, + { + "epoch": 0.1424194092771156, + "grad_norm": 0.15126074850559235, + "learning_rate": 1.934465528505337e-05, + "loss": 0.012888062000274658, + "step": 2690, + "token_acc": 0.9943407565690628 + }, + { + "epoch": 0.14294884946030192, + "grad_norm": 0.14599479734897614, + "learning_rate": 1.9338536121945764e-05, + "loss": 0.012997153401374816, + "step": 2700, + "token_acc": 0.9942284756336583 + }, + { + "epoch": 0.14347828964348822, + "grad_norm": 0.09968029707670212, + "learning_rate": 1.933238950013055e-05, + "loss": 0.01259482353925705, + "step": 2710, + "token_acc": 0.9945284089013445 + }, + { + "epoch": 0.1440077298266745, + "grad_norm": 0.13861827552318573, + "learning_rate": 1.9326215437681042e-05, + "loss": 0.013284245133399963, + "step": 2720, + "token_acc": 0.9939440954931058 + }, + { + "epoch": 0.14453717000986083, + "grad_norm": 0.13162416219711304, + "learning_rate": 1.932001395275124e-05, + "loss": 0.013209289312362671, + "step": 2730, + "token_acc": 0.9942290050871487 + }, + { + "epoch": 0.14506661019304712, + "grad_norm": 0.1251908391714096, + "learning_rate": 1.9313785063575773e-05, + "loss": 0.013006813824176788, + "step": 2740, + "token_acc": 0.9942074952007345 + }, + { + "epoch": 0.14559605037623344, + "grad_norm": 0.10537020862102509, + "learning_rate": 1.9307528788469856e-05, + "loss": 0.0123038187623024, + "step": 2750, + "token_acc": 0.9945867044791058 + }, + { + "epoch": 0.14612549055941973, + "grad_norm": 0.06798841059207916, + "learning_rate": 1.9301245145829214e-05, + "loss": 0.013278274238109589, + "step": 2760, + "token_acc": 0.9942989741492283 + }, + { + "epoch": 0.14665493074260605, + "grad_norm": 0.08897854387760162, + "learning_rate": 1.9294934154130062e-05, + "loss": 0.012781307101249695, + "step": 2770, + "token_acc": 0.9943242521367521 + }, + { + "epoch": 0.14718437092579234, + "grad_norm": 0.08661210536956787, + "learning_rate": 1.9288595831929015e-05, + "loss": 0.013171061873435974, + "step": 2780, + "token_acc": 0.9943280868136359 + }, + { + "epoch": 0.14771381110897863, + "grad_norm": 0.09554250538349152, + "learning_rate": 1.9282230197863056e-05, + "loss": 0.013314065337181092, + "step": 2790, + "token_acc": 0.9939943446745686 + }, + { + "epoch": 0.14824325129216495, + "grad_norm": 0.07586398720741272, + "learning_rate": 1.9275837270649474e-05, + "loss": 0.012647213041782379, + "step": 2800, + "token_acc": 0.9940925171884387 + }, + { + "epoch": 0.14877269147535124, + "grad_norm": 0.12106398493051529, + "learning_rate": 1.9269417069085815e-05, + "loss": 0.013267773389816283, + "step": 2810, + "token_acc": 0.9942382576357678 + }, + { + "epoch": 0.14930213165853756, + "grad_norm": 0.1541309505701065, + "learning_rate": 1.9262969612049813e-05, + "loss": 0.013338148593902588, + "step": 2820, + "token_acc": 0.99438352277469 + }, + { + "epoch": 0.14983157184172385, + "grad_norm": 0.17331618070602417, + "learning_rate": 1.9256494918499348e-05, + "loss": 0.01329217255115509, + "step": 2830, + "token_acc": 0.9941176470588236 + }, + { + "epoch": 0.15036101202491017, + "grad_norm": 0.07473639398813248, + "learning_rate": 1.924999300747238e-05, + "loss": 0.01338455080986023, + "step": 2840, + "token_acc": 0.9941346282205313 + }, + { + "epoch": 0.15089045220809646, + "grad_norm": 0.1250215321779251, + "learning_rate": 1.924346389808691e-05, + "loss": 0.013010719418525695, + "step": 2850, + "token_acc": 0.99431278039994 + }, + { + "epoch": 0.15141989239128276, + "grad_norm": 0.12465380877256393, + "learning_rate": 1.9236907609540902e-05, + "loss": 0.01326519250869751, + "step": 2860, + "token_acc": 0.9940869855302115 + }, + { + "epoch": 0.15194933257446908, + "grad_norm": 0.11480288207530975, + "learning_rate": 1.923032416111224e-05, + "loss": 0.013441666960716248, + "step": 2870, + "token_acc": 0.9939989998333055 + }, + { + "epoch": 0.15247877275765537, + "grad_norm": 0.17930011451244354, + "learning_rate": 1.9223713572158663e-05, + "loss": 0.013769230246543885, + "step": 2880, + "token_acc": 0.9936761663218314 + }, + { + "epoch": 0.1530082129408417, + "grad_norm": 0.10309790819883347, + "learning_rate": 1.921707586211772e-05, + "loss": 0.012541404366493225, + "step": 2890, + "token_acc": 0.9943302622253721 + }, + { + "epoch": 0.15353765312402798, + "grad_norm": 0.14423204958438873, + "learning_rate": 1.9210411050506706e-05, + "loss": 0.012900611758232117, + "step": 2900, + "token_acc": 0.994368993593166 + }, + { + "epoch": 0.1540670933072143, + "grad_norm": 0.22388316690921783, + "learning_rate": 1.9203719156922597e-05, + "loss": 0.013069161772727966, + "step": 2910, + "token_acc": 0.9941418474043043 + }, + { + "epoch": 0.1545965334904006, + "grad_norm": 0.11884718388319016, + "learning_rate": 1.9197000201042012e-05, + "loss": 0.013088618218898774, + "step": 2920, + "token_acc": 0.9941278526624849 + }, + { + "epoch": 0.15512597367358688, + "grad_norm": 0.09179212898015976, + "learning_rate": 1.919025420262113e-05, + "loss": 0.012425411492586136, + "step": 2930, + "token_acc": 0.9945263087629124 + }, + { + "epoch": 0.1556554138567732, + "grad_norm": 0.1196509450674057, + "learning_rate": 1.9183481181495645e-05, + "loss": 0.013234367966651917, + "step": 2940, + "token_acc": 0.994100614971254 + }, + { + "epoch": 0.1561848540399595, + "grad_norm": 0.2017110139131546, + "learning_rate": 1.9176681157580726e-05, + "loss": 0.012923815846443176, + "step": 2950, + "token_acc": 0.994383754551742 + }, + { + "epoch": 0.1567142942231458, + "grad_norm": 0.12554901838302612, + "learning_rate": 1.9169854150870922e-05, + "loss": 0.012743508815765381, + "step": 2960, + "token_acc": 0.994592700145196 + }, + { + "epoch": 0.1572437344063321, + "grad_norm": 0.09087054431438446, + "learning_rate": 1.9163000181440125e-05, + "loss": 0.012246143817901612, + "step": 2970, + "token_acc": 0.994500863673156 + }, + { + "epoch": 0.1577731745895184, + "grad_norm": 0.10294390469789505, + "learning_rate": 1.9156119269441506e-05, + "loss": 0.013268329203128815, + "step": 2980, + "token_acc": 0.9942980222569145 + }, + { + "epoch": 0.15830261477270471, + "grad_norm": 0.0954526737332344, + "learning_rate": 1.914921143510747e-05, + "loss": 0.012406562268733979, + "step": 2990, + "token_acc": 0.9948628994596037 + }, + { + "epoch": 0.158832054955891, + "grad_norm": 0.15816745162010193, + "learning_rate": 1.9142276698749566e-05, + "loss": 0.012695509195327758, + "step": 3000, + "token_acc": 0.994343211854225 + }, + { + "epoch": 0.158832054955891, + "eval_loss": 0.013998457230627537, + "eval_runtime": 710.6197, + "eval_samples_per_second": 16.856, + "eval_steps_per_second": 1.054, + "eval_token_acc": 0.9937939609299604, + "step": 3000 + }, + { + "epoch": 0.15936149513907732, + "grad_norm": 0.19043275713920593, + "learning_rate": 1.913531508075845e-05, + "loss": 0.012796023488044738, + "step": 3010, + "token_acc": 0.9941346282205313 + }, + { + "epoch": 0.15989093532226362, + "grad_norm": 0.11168229579925537, + "learning_rate": 1.912832660160383e-05, + "loss": 0.01276124119758606, + "step": 3020, + "token_acc": 0.994217196756964 + }, + { + "epoch": 0.16042037550544994, + "grad_norm": 0.09534970670938492, + "learning_rate": 1.912131128183438e-05, + "loss": 0.012796524167060851, + "step": 3030, + "token_acc": 0.9941739631240245 + }, + { + "epoch": 0.16094981568863623, + "grad_norm": 0.13286903500556946, + "learning_rate": 1.9114269142077704e-05, + "loss": 0.012587401270866393, + "step": 3040, + "token_acc": 0.994142580602096 + }, + { + "epoch": 0.16147925587182252, + "grad_norm": 0.1017802357673645, + "learning_rate": 1.9107200203040275e-05, + "loss": 0.012643876671791076, + "step": 3050, + "token_acc": 0.994469331064341 + }, + { + "epoch": 0.16200869605500884, + "grad_norm": 0.13110768795013428, + "learning_rate": 1.9100104485507343e-05, + "loss": 0.012140783667564391, + "step": 3060, + "token_acc": 0.9946416946266713 + }, + { + "epoch": 0.16253813623819513, + "grad_norm": 0.13625939190387726, + "learning_rate": 1.9092982010342915e-05, + "loss": 0.012839323282241822, + "step": 3070, + "token_acc": 0.9943047738169689 + }, + { + "epoch": 0.16306757642138145, + "grad_norm": 0.14268894493579865, + "learning_rate": 1.9085832798489672e-05, + "loss": 0.012316041439771653, + "step": 3080, + "token_acc": 0.9947938393765956 + }, + { + "epoch": 0.16359701660456774, + "grad_norm": 0.14032742381095886, + "learning_rate": 1.9078656870968906e-05, + "loss": 0.01237119287252426, + "step": 3090, + "token_acc": 0.9943713413718918 + }, + { + "epoch": 0.16412645678775406, + "grad_norm": 0.10425359755754471, + "learning_rate": 1.9071454248880465e-05, + "loss": 0.012520423531532288, + "step": 3100, + "token_acc": 0.9942204727036178 + }, + { + "epoch": 0.16465589697094035, + "grad_norm": 0.10179140418767929, + "learning_rate": 1.906422495340269e-05, + "loss": 0.012277722358703613, + "step": 3110, + "token_acc": 0.9946107384789935 + }, + { + "epoch": 0.16518533715412664, + "grad_norm": 0.13637907803058624, + "learning_rate": 1.9056969005792353e-05, + "loss": 0.01288021206855774, + "step": 3120, + "token_acc": 0.9940242701430503 + }, + { + "epoch": 0.16571477733731296, + "grad_norm": 0.08888508379459381, + "learning_rate": 1.9049686427384585e-05, + "loss": 0.012345415353775025, + "step": 3130, + "token_acc": 0.9945585952496203 + }, + { + "epoch": 0.16624421752049925, + "grad_norm": 0.11450297385454178, + "learning_rate": 1.904237723959283e-05, + "loss": 0.011784133315086365, + "step": 3140, + "token_acc": 0.994668958152573 + }, + { + "epoch": 0.16677365770368557, + "grad_norm": 0.11359819024801254, + "learning_rate": 1.9035041463908773e-05, + "loss": 0.01293257474899292, + "step": 3150, + "token_acc": 0.994237056311716 + }, + { + "epoch": 0.16730309788687187, + "grad_norm": 0.08328907936811447, + "learning_rate": 1.9027679121902268e-05, + "loss": 0.012201464176177979, + "step": 3160, + "token_acc": 0.9945529316572268 + }, + { + "epoch": 0.16783253807005818, + "grad_norm": 0.09557446092367172, + "learning_rate": 1.9020290235221294e-05, + "loss": 0.011904217302799225, + "step": 3170, + "token_acc": 0.9948701694095273 + }, + { + "epoch": 0.16836197825324448, + "grad_norm": 0.10489526391029358, + "learning_rate": 1.9012874825591878e-05, + "loss": 0.012524682283401489, + "step": 3180, + "token_acc": 0.9942999732941648 + }, + { + "epoch": 0.16889141843643077, + "grad_norm": 0.16704648733139038, + "learning_rate": 1.9005432914818033e-05, + "loss": 0.012450940907001495, + "step": 3190, + "token_acc": 0.9945268567805237 + }, + { + "epoch": 0.1694208586196171, + "grad_norm": 0.09519923478364944, + "learning_rate": 1.899796452478169e-05, + "loss": 0.012829813361167907, + "step": 3200, + "token_acc": 0.9942624112883722 + }, + { + "epoch": 0.16995029880280338, + "grad_norm": 0.07067445665597916, + "learning_rate": 1.8990469677442652e-05, + "loss": 0.011842641234397887, + "step": 3210, + "token_acc": 0.9948512967722554 + }, + { + "epoch": 0.1704797389859897, + "grad_norm": 0.187923401594162, + "learning_rate": 1.89829483948385e-05, + "loss": 0.012599028646945953, + "step": 3220, + "token_acc": 0.9943785551051727 + }, + { + "epoch": 0.171009179169176, + "grad_norm": 0.06628217548131943, + "learning_rate": 1.897540069908456e-05, + "loss": 0.013080981373786927, + "step": 3230, + "token_acc": 0.994133257669326 + }, + { + "epoch": 0.1715386193523623, + "grad_norm": 0.08535142987966537, + "learning_rate": 1.896782661237381e-05, + "loss": 0.011299634724855423, + "step": 3240, + "token_acc": 0.9947859746890355 + }, + { + "epoch": 0.1720680595355486, + "grad_norm": 0.10513030737638474, + "learning_rate": 1.896022615697683e-05, + "loss": 0.012114519625902176, + "step": 3250, + "token_acc": 0.994624777771286 + }, + { + "epoch": 0.1725974997187349, + "grad_norm": 0.1872827112674713, + "learning_rate": 1.895259935524174e-05, + "loss": 0.012511488795280457, + "step": 3260, + "token_acc": 0.9943856780565937 + }, + { + "epoch": 0.1731269399019212, + "grad_norm": 0.11911717057228088, + "learning_rate": 1.8944946229594123e-05, + "loss": 0.012278057634830475, + "step": 3270, + "token_acc": 0.9945116356660272 + }, + { + "epoch": 0.1736563800851075, + "grad_norm": 0.11644499003887177, + "learning_rate": 1.893726680253696e-05, + "loss": 0.012293926626443862, + "step": 3280, + "token_acc": 0.9944869806001768 + }, + { + "epoch": 0.17418582026829382, + "grad_norm": 0.12327589839696884, + "learning_rate": 1.892956109665058e-05, + "loss": 0.01279902011156082, + "step": 3290, + "token_acc": 0.9942726422655623 + }, + { + "epoch": 0.17471526045148011, + "grad_norm": 0.08826106041669846, + "learning_rate": 1.892182913459256e-05, + "loss": 0.012657384574413299, + "step": 3300, + "token_acc": 0.9941048271946159 + }, + { + "epoch": 0.1752447006346664, + "grad_norm": 0.140497088432312, + "learning_rate": 1.8914070939097704e-05, + "loss": 0.013047289848327637, + "step": 3310, + "token_acc": 0.994324915291015 + }, + { + "epoch": 0.17577414081785273, + "grad_norm": 0.09164322167634964, + "learning_rate": 1.890628653297794e-05, + "loss": 0.0130844846367836, + "step": 3320, + "token_acc": 0.9940701572951244 + }, + { + "epoch": 0.17630358100103902, + "grad_norm": 0.11808131635189056, + "learning_rate": 1.8898475939122258e-05, + "loss": 0.011877813935279846, + "step": 3330, + "token_acc": 0.9948154950743029 + }, + { + "epoch": 0.17683302118422534, + "grad_norm": 0.09616479277610779, + "learning_rate": 1.8890639180496665e-05, + "loss": 0.012178563326597214, + "step": 3340, + "token_acc": 0.9947634359209921 + }, + { + "epoch": 0.17736246136741163, + "grad_norm": 0.10518739372491837, + "learning_rate": 1.8882776280144095e-05, + "loss": 0.011829140037298203, + "step": 3350, + "token_acc": 0.9947613843959325 + }, + { + "epoch": 0.17789190155059795, + "grad_norm": 0.1052684560418129, + "learning_rate": 1.887488726118434e-05, + "loss": 0.011936575174331665, + "step": 3360, + "token_acc": 0.9948351675858789 + }, + { + "epoch": 0.17842134173378424, + "grad_norm": 0.06065749749541283, + "learning_rate": 1.8866972146814002e-05, + "loss": 0.012907716631889343, + "step": 3370, + "token_acc": 0.9942253932490508 + }, + { + "epoch": 0.17895078191697053, + "grad_norm": 0.11386685073375702, + "learning_rate": 1.8859030960306408e-05, + "loss": 0.012151119858026504, + "step": 3380, + "token_acc": 0.9948038733246036 + }, + { + "epoch": 0.17948022210015685, + "grad_norm": 0.09328997880220413, + "learning_rate": 1.8851063725011548e-05, + "loss": 0.01197451800107956, + "step": 3390, + "token_acc": 0.9947377638415158 + }, + { + "epoch": 0.18000966228334314, + "grad_norm": 0.08318711072206497, + "learning_rate": 1.8843070464356e-05, + "loss": 0.011765899509191513, + "step": 3400, + "token_acc": 0.9947043174407258 + }, + { + "epoch": 0.18053910246652946, + "grad_norm": 0.1237230896949768, + "learning_rate": 1.883505120184288e-05, + "loss": 0.01192016452550888, + "step": 3410, + "token_acc": 0.9946761461306096 + }, + { + "epoch": 0.18106854264971575, + "grad_norm": 0.1543959081172943, + "learning_rate": 1.8827005961051742e-05, + "loss": 0.012184982001781464, + "step": 3420, + "token_acc": 0.9948194308882048 + }, + { + "epoch": 0.18159798283290207, + "grad_norm": 0.15891771018505096, + "learning_rate": 1.8818934765638533e-05, + "loss": 0.01259923428297043, + "step": 3430, + "token_acc": 0.9944762906038532 + }, + { + "epoch": 0.18212742301608836, + "grad_norm": 0.12234760820865631, + "learning_rate": 1.8810837639335516e-05, + "loss": 0.011260591447353363, + "step": 3440, + "token_acc": 0.9950743851329916 + }, + { + "epoch": 0.18265686319927465, + "grad_norm": 0.10251070559024811, + "learning_rate": 1.88027146059512e-05, + "loss": 0.01242552101612091, + "step": 3450, + "token_acc": 0.9945639920839701 + }, + { + "epoch": 0.18318630338246097, + "grad_norm": 0.0978703647851944, + "learning_rate": 1.8794565689370275e-05, + "loss": 0.012132106721401215, + "step": 3460, + "token_acc": 0.9946652195692102 + }, + { + "epoch": 0.18371574356564727, + "grad_norm": 0.09851015359163284, + "learning_rate": 1.878639091355353e-05, + "loss": 0.012237283587455749, + "step": 3470, + "token_acc": 0.9946375554591854 + }, + { + "epoch": 0.18424518374883359, + "grad_norm": 0.11065611988306046, + "learning_rate": 1.877819030253779e-05, + "loss": 0.012428236752748489, + "step": 3480, + "token_acc": 0.9945029737327228 + }, + { + "epoch": 0.18477462393201988, + "grad_norm": 0.1257503181695938, + "learning_rate": 1.8769963880435848e-05, + "loss": 0.01242423877120018, + "step": 3490, + "token_acc": 0.9944367060629061 + }, + { + "epoch": 0.1853040641152062, + "grad_norm": 0.10965922474861145, + "learning_rate": 1.8761711671436394e-05, + "loss": 0.011719176173210144, + "step": 3500, + "token_acc": 0.9950370759619315 + }, + { + "epoch": 0.1858335042983925, + "grad_norm": 0.08124595135450363, + "learning_rate": 1.8753433699803933e-05, + "loss": 0.012090445309877396, + "step": 3510, + "token_acc": 0.9946009095840114 + }, + { + "epoch": 0.18636294448157878, + "grad_norm": 0.0652400329709053, + "learning_rate": 1.8745129989878722e-05, + "loss": 0.01181778609752655, + "step": 3520, + "token_acc": 0.9946521387273591 + }, + { + "epoch": 0.1868923846647651, + "grad_norm": 0.1268422156572342, + "learning_rate": 1.873680056607671e-05, + "loss": 0.012106741964817046, + "step": 3530, + "token_acc": 0.9946039715436644 + }, + { + "epoch": 0.1874218248479514, + "grad_norm": 0.08397526293992996, + "learning_rate": 1.872844545288944e-05, + "loss": 0.011762569844722747, + "step": 3540, + "token_acc": 0.9947066976138395 + }, + { + "epoch": 0.1879512650311377, + "grad_norm": 0.11669668555259705, + "learning_rate": 1.8720064674884e-05, + "loss": 0.01177578866481781, + "step": 3550, + "token_acc": 0.9947763684913218 + }, + { + "epoch": 0.188480705214324, + "grad_norm": 0.09324853122234344, + "learning_rate": 1.8711658256702938e-05, + "loss": 0.011640041321516036, + "step": 3560, + "token_acc": 0.9948346935028956 + }, + { + "epoch": 0.18901014539751032, + "grad_norm": 0.07459814846515656, + "learning_rate": 1.8703226223064194e-05, + "loss": 0.011547382920980453, + "step": 3570, + "token_acc": 0.9949809074385954 + }, + { + "epoch": 0.1895395855806966, + "grad_norm": 0.15576159954071045, + "learning_rate": 1.8694768598761026e-05, + "loss": 0.011832021176815033, + "step": 3580, + "token_acc": 0.994645135998532 + }, + { + "epoch": 0.1900690257638829, + "grad_norm": 0.10806694626808167, + "learning_rate": 1.868628540866194e-05, + "loss": 0.01166851669549942, + "step": 3590, + "token_acc": 0.9947599856483683 + }, + { + "epoch": 0.19059846594706922, + "grad_norm": 0.11775384843349457, + "learning_rate": 1.8677776677710614e-05, + "loss": 0.012158556282520295, + "step": 3600, + "token_acc": 0.9945954060951809 + }, + { + "epoch": 0.19112790613025551, + "grad_norm": 0.13991916179656982, + "learning_rate": 1.8669242430925826e-05, + "loss": 0.012168093025684357, + "step": 3610, + "token_acc": 0.9945336037321716 + }, + { + "epoch": 0.19165734631344183, + "grad_norm": 0.10819360613822937, + "learning_rate": 1.866068269340138e-05, + "loss": 0.011927696317434311, + "step": 3620, + "token_acc": 0.9948190851221812 + }, + { + "epoch": 0.19218678649662813, + "grad_norm": 0.08585803210735321, + "learning_rate": 1.865209749030603e-05, + "loss": 0.011996359378099442, + "step": 3630, + "token_acc": 0.9948748340998823 + }, + { + "epoch": 0.19271622667981442, + "grad_norm": 0.07425203919410706, + "learning_rate": 1.8643486846883406e-05, + "loss": 0.011524158716201782, + "step": 3640, + "token_acc": 0.9949186059124399 + }, + { + "epoch": 0.19324566686300074, + "grad_norm": 0.09276238083839417, + "learning_rate": 1.863485078845195e-05, + "loss": 0.011590300500392914, + "step": 3650, + "token_acc": 0.9950921883998698 + }, + { + "epoch": 0.19377510704618703, + "grad_norm": 0.0752921774983406, + "learning_rate": 1.8626189340404826e-05, + "loss": 0.010996472090482712, + "step": 3660, + "token_acc": 0.9952514875612342 + }, + { + "epoch": 0.19430454722937335, + "grad_norm": 0.12285684794187546, + "learning_rate": 1.8617502528209856e-05, + "loss": 0.011212416738271714, + "step": 3670, + "token_acc": 0.9951209748042135 + }, + { + "epoch": 0.19483398741255964, + "grad_norm": 0.11246007680892944, + "learning_rate": 1.860879037740944e-05, + "loss": 0.011637929081916808, + "step": 3680, + "token_acc": 0.9947763684913218 + }, + { + "epoch": 0.19536342759574596, + "grad_norm": 0.0950104221701622, + "learning_rate": 1.8600052913620488e-05, + "loss": 0.012218789011240006, + "step": 3690, + "token_acc": 0.994696331629377 + }, + { + "epoch": 0.19589286777893225, + "grad_norm": 0.16720029711723328, + "learning_rate": 1.8591290162534332e-05, + "loss": 0.011792376637458801, + "step": 3700, + "token_acc": 0.9948702977729585 + }, + { + "epoch": 0.19642230796211854, + "grad_norm": 0.07429894804954529, + "learning_rate": 1.8582502149916662e-05, + "loss": 0.011602994054555893, + "step": 3710, + "token_acc": 0.9946431844310779 + }, + { + "epoch": 0.19695174814530486, + "grad_norm": 0.1059260293841362, + "learning_rate": 1.857368890160744e-05, + "loss": 0.011660879105329513, + "step": 3720, + "token_acc": 0.9949768031774641 + }, + { + "epoch": 0.19748118832849115, + "grad_norm": 0.08341298997402191, + "learning_rate": 1.856485044352085e-05, + "loss": 0.011923395097255707, + "step": 3730, + "token_acc": 0.9946794759533996 + }, + { + "epoch": 0.19801062851167747, + "grad_norm": 0.0997161865234375, + "learning_rate": 1.8555986801645173e-05, + "loss": 0.011528775095939636, + "step": 3740, + "token_acc": 0.9950846623104591 + }, + { + "epoch": 0.19854006869486376, + "grad_norm": 0.06876150518655777, + "learning_rate": 1.8547098002042757e-05, + "loss": 0.011438699811697007, + "step": 3750, + "token_acc": 0.9950196876668447 + }, + { + "epoch": 0.19906950887805008, + "grad_norm": 0.08931242674589157, + "learning_rate": 1.8538184070849926e-05, + "loss": 0.010738776624202728, + "step": 3760, + "token_acc": 0.9954380931729855 + }, + { + "epoch": 0.19959894906123637, + "grad_norm": 0.07668065279722214, + "learning_rate": 1.8529245034276888e-05, + "loss": 0.011637356877326966, + "step": 3770, + "token_acc": 0.9948586547369651 + }, + { + "epoch": 0.20012838924442267, + "grad_norm": 0.08759427815675735, + "learning_rate": 1.852028091860768e-05, + "loss": 0.011832761764526366, + "step": 3780, + "token_acc": 0.9946024860265288 + }, + { + "epoch": 0.20065782942760899, + "grad_norm": 0.09793692082166672, + "learning_rate": 1.8511291750200072e-05, + "loss": 0.011665716022253036, + "step": 3790, + "token_acc": 0.9948768440024031 + }, + { + "epoch": 0.20118726961079528, + "grad_norm": 0.09731903672218323, + "learning_rate": 1.850227755548551e-05, + "loss": 0.011719143390655518, + "step": 3800, + "token_acc": 0.9948033965600654 + }, + { + "epoch": 0.2017167097939816, + "grad_norm": 0.10248780250549316, + "learning_rate": 1.8493238360969015e-05, + "loss": 0.011086869239807128, + "step": 3810, + "token_acc": 0.9951927490631702 + }, + { + "epoch": 0.2022461499771679, + "grad_norm": 0.1289559304714203, + "learning_rate": 1.8484174193229117e-05, + "loss": 0.011635150760412216, + "step": 3820, + "token_acc": 0.9948837791595376 + }, + { + "epoch": 0.2027755901603542, + "grad_norm": 0.11350011825561523, + "learning_rate": 1.847508507891779e-05, + "loss": 0.011909595131874085, + "step": 3830, + "token_acc": 0.9949163701215904 + }, + { + "epoch": 0.2033050303435405, + "grad_norm": 0.14403733611106873, + "learning_rate": 1.846597104476034e-05, + "loss": 0.011130518466234206, + "step": 3840, + "token_acc": 0.9950444249781004 + }, + { + "epoch": 0.2038344705267268, + "grad_norm": 0.08904826641082764, + "learning_rate": 1.8456832117555366e-05, + "loss": 0.011928752064704895, + "step": 3850, + "token_acc": 0.9947931008586377 + }, + { + "epoch": 0.2043639107099131, + "grad_norm": 0.08891967684030533, + "learning_rate": 1.8447668324174642e-05, + "loss": 0.011090496182441711, + "step": 3860, + "token_acc": 0.9952177469161562 + }, + { + "epoch": 0.2048933508930994, + "grad_norm": 0.10339909046888351, + "learning_rate": 1.843847969156308e-05, + "loss": 0.011843350529670716, + "step": 3870, + "token_acc": 0.9947618213514168 + }, + { + "epoch": 0.20542279107628572, + "grad_norm": 0.12172644585371017, + "learning_rate": 1.842926624673861e-05, + "loss": 0.011696116626262664, + "step": 3880, + "token_acc": 0.9949231808064557 + }, + { + "epoch": 0.205952231259472, + "grad_norm": 0.0876343622803688, + "learning_rate": 1.8420028016792123e-05, + "loss": 0.011564520001411439, + "step": 3890, + "token_acc": 0.9950306830309499 + }, + { + "epoch": 0.20648167144265833, + "grad_norm": 0.09616302698850632, + "learning_rate": 1.84107650288874e-05, + "loss": 0.01125875934958458, + "step": 3900, + "token_acc": 0.9951109220000167 + }, + { + "epoch": 0.20701111162584462, + "grad_norm": 0.09002236276865005, + "learning_rate": 1.8401477310260996e-05, + "loss": 0.011833015084266662, + "step": 3910, + "token_acc": 0.9947115545026859 + }, + { + "epoch": 0.20754055180903092, + "grad_norm": 0.0981248989701271, + "learning_rate": 1.8392164888222206e-05, + "loss": 0.011576320230960845, + "step": 3920, + "token_acc": 0.9950244187502609 + }, + { + "epoch": 0.20806999199221723, + "grad_norm": 0.12523253262043, + "learning_rate": 1.8382827790152947e-05, + "loss": 0.011327150464057922, + "step": 3930, + "token_acc": 0.994967871150797 + }, + { + "epoch": 0.20859943217540353, + "grad_norm": 0.09090828895568848, + "learning_rate": 1.83734660435077e-05, + "loss": 0.011326855421066285, + "step": 3940, + "token_acc": 0.9949678827616659 + }, + { + "epoch": 0.20912887235858985, + "grad_norm": 0.0869918018579483, + "learning_rate": 1.8364079675813425e-05, + "loss": 0.012431064248085022, + "step": 3950, + "token_acc": 0.9948205111096283 + }, + { + "epoch": 0.20965831254177614, + "grad_norm": 0.09381060302257538, + "learning_rate": 1.8354668714669465e-05, + "loss": 0.012398632615804673, + "step": 3960, + "token_acc": 0.9946282301519778 + }, + { + "epoch": 0.21018775272496246, + "grad_norm": 0.08381836116313934, + "learning_rate": 1.8345233187747487e-05, + "loss": 0.011205083131790161, + "step": 3970, + "token_acc": 0.9949327144622165 + }, + { + "epoch": 0.21071719290814875, + "grad_norm": 0.13143491744995117, + "learning_rate": 1.8335773122791386e-05, + "loss": 0.011301149427890778, + "step": 3980, + "token_acc": 0.9949980408993523 + }, + { + "epoch": 0.21124663309133504, + "grad_norm": 0.10211317986249924, + "learning_rate": 1.8326288547617214e-05, + "loss": 0.011369569599628449, + "step": 3990, + "token_acc": 0.9950860565812636 + }, + { + "epoch": 0.21177607327452136, + "grad_norm": 0.11691654473543167, + "learning_rate": 1.8316779490113073e-05, + "loss": 0.011469640582799912, + "step": 4000, + "token_acc": 0.9947591985245642 + }, + { + "epoch": 0.21230551345770765, + "grad_norm": 0.06719864159822464, + "learning_rate": 1.8307245978239082e-05, + "loss": 0.010724474489688874, + "step": 4010, + "token_acc": 0.9953807927693564 + }, + { + "epoch": 0.21283495364089397, + "grad_norm": 0.1969636231660843, + "learning_rate": 1.829768804002724e-05, + "loss": 0.011339026689529418, + "step": 4020, + "token_acc": 0.9949516021361816 + }, + { + "epoch": 0.21336439382408026, + "grad_norm": 0.09002166986465454, + "learning_rate": 1.8288105703581377e-05, + "loss": 0.010930276662111282, + "step": 4030, + "token_acc": 0.9951293556404397 + }, + { + "epoch": 0.21389383400726655, + "grad_norm": 0.12502668797969818, + "learning_rate": 1.8278498997077064e-05, + "loss": 0.011602598428726196, + "step": 4040, + "token_acc": 0.9948935318069555 + }, + { + "epoch": 0.21442327419045287, + "grad_norm": 0.08773364871740341, + "learning_rate": 1.8268867948761527e-05, + "loss": 0.011349610984325409, + "step": 4050, + "token_acc": 0.9951465621481883 + }, + { + "epoch": 0.21495271437363916, + "grad_norm": 0.07717062532901764, + "learning_rate": 1.8259212586953564e-05, + "loss": 0.011223162710666656, + "step": 4060, + "token_acc": 0.9952716901831282 + }, + { + "epoch": 0.21548215455682548, + "grad_norm": 0.09828871488571167, + "learning_rate": 1.824953294004347e-05, + "loss": 0.011029057204723358, + "step": 4070, + "token_acc": 0.9949945357014741 + }, + { + "epoch": 0.21601159474001178, + "grad_norm": 0.09407025575637817, + "learning_rate": 1.8239829036492945e-05, + "loss": 0.011374135315418244, + "step": 4080, + "token_acc": 0.9950498355510292 + }, + { + "epoch": 0.2165410349231981, + "grad_norm": 0.06426401436328888, + "learning_rate": 1.8230100904835004e-05, + "loss": 0.011681027710437775, + "step": 4090, + "token_acc": 0.9948374505846441 + }, + { + "epoch": 0.2170704751063844, + "grad_norm": 0.1273181438446045, + "learning_rate": 1.822034857367391e-05, + "loss": 0.011212971061468124, + "step": 4100, + "token_acc": 0.9953115484403807 + }, + { + "epoch": 0.21759991528957068, + "grad_norm": 0.15100979804992676, + "learning_rate": 1.821057207168508e-05, + "loss": 0.011004224419593811, + "step": 4110, + "token_acc": 0.9952808557826192 + }, + { + "epoch": 0.218129355472757, + "grad_norm": 0.09646367281675339, + "learning_rate": 1.8200771427615e-05, + "loss": 0.011242683231830596, + "step": 4120, + "token_acc": 0.9950736448348418 + }, + { + "epoch": 0.2186587956559433, + "grad_norm": 0.13345639407634735, + "learning_rate": 1.819094667028115e-05, + "loss": 0.01109175980091095, + "step": 4130, + "token_acc": 0.995146440723197 + }, + { + "epoch": 0.2191882358391296, + "grad_norm": 0.18397724628448486, + "learning_rate": 1.818109782857189e-05, + "loss": 0.011379151046276093, + "step": 4140, + "token_acc": 0.9951878169202029 + }, + { + "epoch": 0.2197176760223159, + "grad_norm": 0.07246195524930954, + "learning_rate": 1.817122493144643e-05, + "loss": 0.011480104923248292, + "step": 4150, + "token_acc": 0.9951282179621936 + }, + { + "epoch": 0.22024711620550222, + "grad_norm": 0.09845490753650665, + "learning_rate": 1.8161328007934688e-05, + "loss": 0.010773421823978424, + "step": 4160, + "token_acc": 0.9952570226126457 + }, + { + "epoch": 0.2207765563886885, + "grad_norm": 0.08280066400766373, + "learning_rate": 1.8151407087137233e-05, + "loss": 0.011343392729759216, + "step": 4170, + "token_acc": 0.9951266324529562 + }, + { + "epoch": 0.2213059965718748, + "grad_norm": 0.06328543275594711, + "learning_rate": 1.8141462198225193e-05, + "loss": 0.010863637179136276, + "step": 4180, + "token_acc": 0.9952413948556974 + }, + { + "epoch": 0.22183543675506112, + "grad_norm": 0.0883801132440567, + "learning_rate": 1.813149337044018e-05, + "loss": 0.01121646910905838, + "step": 4190, + "token_acc": 0.9952270091204178 + }, + { + "epoch": 0.2223648769382474, + "grad_norm": 0.09695139527320862, + "learning_rate": 1.8121500633094186e-05, + "loss": 0.011460987478494644, + "step": 4200, + "token_acc": 0.9950843341317466 + }, + { + "epoch": 0.22289431712143373, + "grad_norm": 0.12915395200252533, + "learning_rate": 1.8111484015569513e-05, + "loss": 0.011166107654571534, + "step": 4210, + "token_acc": 0.9951267137862263 + }, + { + "epoch": 0.22342375730462002, + "grad_norm": 0.08524905145168304, + "learning_rate": 1.8101443547318674e-05, + "loss": 0.011791612207889556, + "step": 4220, + "token_acc": 0.994840368688218 + }, + { + "epoch": 0.22395319748780634, + "grad_norm": 0.12845316529273987, + "learning_rate": 1.809137925786431e-05, + "loss": 0.010956542193889618, + "step": 4230, + "token_acc": 0.9951693642583014 + }, + { + "epoch": 0.22448263767099264, + "grad_norm": 0.08242276310920715, + "learning_rate": 1.808129117679911e-05, + "loss": 0.011588747799396514, + "step": 4240, + "token_acc": 0.9949368989131432 + }, + { + "epoch": 0.22501207785417893, + "grad_norm": 0.10287240147590637, + "learning_rate": 1.8071179333785723e-05, + "loss": 0.011173225939273834, + "step": 4250, + "token_acc": 0.994884974508315 + }, + { + "epoch": 0.22554151803736525, + "grad_norm": 0.07787423580884933, + "learning_rate": 1.8061043758556652e-05, + "loss": 0.011547836661338805, + "step": 4260, + "token_acc": 0.9947835812175538 + }, + { + "epoch": 0.22607095822055154, + "grad_norm": 0.08325622230768204, + "learning_rate": 1.80508844809142e-05, + "loss": 0.011148739606142044, + "step": 4270, + "token_acc": 0.9952362677701395 + }, + { + "epoch": 0.22660039840373786, + "grad_norm": 0.09910520166158676, + "learning_rate": 1.804070153073035e-05, + "loss": 0.010768309235572815, + "step": 4280, + "token_acc": 0.995150938514568 + }, + { + "epoch": 0.22712983858692415, + "grad_norm": 0.08337461203336716, + "learning_rate": 1.8030494937946694e-05, + "loss": 0.01144525557756424, + "step": 4290, + "token_acc": 0.994864954984995 + }, + { + "epoch": 0.22765927877011047, + "grad_norm": 0.14330936968326569, + "learning_rate": 1.802026473257435e-05, + "loss": 0.010987612605094909, + "step": 4300, + "token_acc": 0.9953189145889657 + }, + { + "epoch": 0.22818871895329676, + "grad_norm": 0.06947844475507736, + "learning_rate": 1.8010010944693846e-05, + "loss": 0.01124451756477356, + "step": 4310, + "token_acc": 0.994993825718386 + }, + { + "epoch": 0.22871815913648305, + "grad_norm": 0.08575119823217392, + "learning_rate": 1.7999733604455078e-05, + "loss": 0.010849817097187043, + "step": 4320, + "token_acc": 0.9952918391879257 + }, + { + "epoch": 0.22924759931966937, + "grad_norm": 0.06774135679006577, + "learning_rate": 1.7989432742077175e-05, + "loss": 0.01133774071931839, + "step": 4330, + "token_acc": 0.9952449675907001 + }, + { + "epoch": 0.22977703950285566, + "grad_norm": 0.11256641894578934, + "learning_rate": 1.797910838784843e-05, + "loss": 0.011555623263120651, + "step": 4340, + "token_acc": 0.9948785105974793 + }, + { + "epoch": 0.23030647968604198, + "grad_norm": 0.09484875202178955, + "learning_rate": 1.7968760572126226e-05, + "loss": 0.011899186670780182, + "step": 4350, + "token_acc": 0.9947838859632285 + }, + { + "epoch": 0.23083591986922827, + "grad_norm": 0.07960456609725952, + "learning_rate": 1.7958389325336916e-05, + "loss": 0.010491693019866943, + "step": 4360, + "token_acc": 0.9953589315525877 + }, + { + "epoch": 0.23136536005241456, + "grad_norm": 0.06526882201433182, + "learning_rate": 1.7947994677975754e-05, + "loss": 0.010907218605279923, + "step": 4370, + "token_acc": 0.9953357586275949 + }, + { + "epoch": 0.23189480023560088, + "grad_norm": 0.05946412682533264, + "learning_rate": 1.79375766606068e-05, + "loss": 0.010534897446632385, + "step": 4380, + "token_acc": 0.9954273126282939 + }, + { + "epoch": 0.23242424041878718, + "grad_norm": 0.1212170273065567, + "learning_rate": 1.792713530386283e-05, + "loss": 0.011323726922273635, + "step": 4390, + "token_acc": 0.9949686269274415 + }, + { + "epoch": 0.2329536806019735, + "grad_norm": 0.08500424772500992, + "learning_rate": 1.791667063844525e-05, + "loss": 0.010547153651714325, + "step": 4400, + "token_acc": 0.9953937431678029 + }, + { + "epoch": 0.2334831207851598, + "grad_norm": 0.061498209834098816, + "learning_rate": 1.7906182695124e-05, + "loss": 0.010839894413948059, + "step": 4410, + "token_acc": 0.9952850264121972 + }, + { + "epoch": 0.2340125609683461, + "grad_norm": 0.09212823957204819, + "learning_rate": 1.7895671504737463e-05, + "loss": 0.01102587878704071, + "step": 4420, + "token_acc": 0.995049959514846 + }, + { + "epoch": 0.2345420011515324, + "grad_norm": 0.10865923017263412, + "learning_rate": 1.7885137098192376e-05, + "loss": 0.010536351799964904, + "step": 4430, + "token_acc": 0.9953565920553541 + }, + { + "epoch": 0.2350714413347187, + "grad_norm": 0.10068950057029724, + "learning_rate": 1.787457950646375e-05, + "loss": 0.011340776085853576, + "step": 4440, + "token_acc": 0.9949673671735465 + }, + { + "epoch": 0.235600881517905, + "grad_norm": 0.10460250824689865, + "learning_rate": 1.786399876059475e-05, + "loss": 0.010938787460327148, + "step": 4450, + "token_acc": 0.9951771842432434 + }, + { + "epoch": 0.2361303217010913, + "grad_norm": 0.1326463222503662, + "learning_rate": 1.7853394891696643e-05, + "loss": 0.01075664535164833, + "step": 4460, + "token_acc": 0.9954187771723257 + }, + { + "epoch": 0.23665976188427762, + "grad_norm": 0.09301836043596268, + "learning_rate": 1.7842767930948675e-05, + "loss": 0.010936200618743896, + "step": 4470, + "token_acc": 0.9951080241760444 + }, + { + "epoch": 0.2371892020674639, + "grad_norm": 0.05401263386011124, + "learning_rate": 1.7832117909597994e-05, + "loss": 0.011003050208091735, + "step": 4480, + "token_acc": 0.9953702545108737 + }, + { + "epoch": 0.23771864225065023, + "grad_norm": 0.06449698656797409, + "learning_rate": 1.782144485895955e-05, + "loss": 0.010801362246274948, + "step": 4490, + "token_acc": 0.9951924280742169 + }, + { + "epoch": 0.23824808243383652, + "grad_norm": 0.07623039186000824, + "learning_rate": 1.781074881041601e-05, + "loss": 0.011075205355882644, + "step": 4500, + "token_acc": 0.995036330722193 + }, + { + "epoch": 0.2387775226170228, + "grad_norm": 0.11719249933958054, + "learning_rate": 1.7800029795417668e-05, + "loss": 0.010938540846109391, + "step": 4510, + "token_acc": 0.9952295965172718 + }, + { + "epoch": 0.23930696280020913, + "grad_norm": 0.08885107189416885, + "learning_rate": 1.7789287845482342e-05, + "loss": 0.010892553627490998, + "step": 4520, + "token_acc": 0.9955107183685049 + }, + { + "epoch": 0.23983640298339542, + "grad_norm": 0.07599925994873047, + "learning_rate": 1.7778522992195288e-05, + "loss": 0.0111905038356781, + "step": 4530, + "token_acc": 0.9951629986072772 + }, + { + "epoch": 0.24036584316658174, + "grad_norm": 0.1025770753622055, + "learning_rate": 1.7767735267209108e-05, + "loss": 0.010496590286493301, + "step": 4540, + "token_acc": 0.9955779531763645 + }, + { + "epoch": 0.24089528334976804, + "grad_norm": 0.09753808379173279, + "learning_rate": 1.7756924702243658e-05, + "loss": 0.010012496262788773, + "step": 4550, + "token_acc": 0.995616378878461 + }, + { + "epoch": 0.24142472353295436, + "grad_norm": 0.09451748430728912, + "learning_rate": 1.774609132908594e-05, + "loss": 0.010527233779430389, + "step": 4560, + "token_acc": 0.9955353789920806 + }, + { + "epoch": 0.24195416371614065, + "grad_norm": 0.11088404059410095, + "learning_rate": 1.7735235179590035e-05, + "loss": 0.010932166129350662, + "step": 4570, + "token_acc": 0.9951690821256038 + }, + { + "epoch": 0.24248360389932694, + "grad_norm": 0.07473895698785782, + "learning_rate": 1.772435628567699e-05, + "loss": 0.010097935795783997, + "step": 4580, + "token_acc": 0.9955722326454034 + }, + { + "epoch": 0.24301304408251326, + "grad_norm": 0.07963219285011292, + "learning_rate": 1.7713454679334715e-05, + "loss": 0.010204389691352844, + "step": 4590, + "token_acc": 0.9955468085461248 + }, + { + "epoch": 0.24354248426569955, + "grad_norm": 0.07848156988620758, + "learning_rate": 1.770253039261793e-05, + "loss": 0.01066179871559143, + "step": 4600, + "token_acc": 0.9953680134202422 + }, + { + "epoch": 0.24407192444888587, + "grad_norm": 0.0696529671549797, + "learning_rate": 1.7691583457648017e-05, + "loss": 0.010576093196868896, + "step": 4610, + "token_acc": 0.9953768609386474 + }, + { + "epoch": 0.24460136463207216, + "grad_norm": 0.11796292662620544, + "learning_rate": 1.7680613906612974e-05, + "loss": 0.010477733612060548, + "step": 4620, + "token_acc": 0.9955786540755633 + }, + { + "epoch": 0.24513080481525848, + "grad_norm": 0.06933698058128357, + "learning_rate": 1.7669621771767282e-05, + "loss": 0.011219052970409394, + "step": 4630, + "token_acc": 0.9949620485444991 + }, + { + "epoch": 0.24566024499844477, + "grad_norm": 0.08617869019508362, + "learning_rate": 1.7658607085431834e-05, + "loss": 0.011005634069442749, + "step": 4640, + "token_acc": 0.9952268498043175 + }, + { + "epoch": 0.24618968518163106, + "grad_norm": 0.10418684035539627, + "learning_rate": 1.764756987999383e-05, + "loss": 0.010853968560695648, + "step": 4650, + "token_acc": 0.9954378268375883 + }, + { + "epoch": 0.24671912536481738, + "grad_norm": 0.10259804874658585, + "learning_rate": 1.763651018790669e-05, + "loss": 0.010451202094554902, + "step": 4660, + "token_acc": 0.9954725476299663 + }, + { + "epoch": 0.24724856554800367, + "grad_norm": 0.10324206203222275, + "learning_rate": 1.7625428041689948e-05, + "loss": 0.010664881765842437, + "step": 4670, + "token_acc": 0.9951608193233491 + }, + { + "epoch": 0.24777800573119, + "grad_norm": 0.093321792781353, + "learning_rate": 1.7614323473929155e-05, + "loss": 0.010385939478874206, + "step": 4680, + "token_acc": 0.9952861278669103 + }, + { + "epoch": 0.24830744591437628, + "grad_norm": 0.12238208204507828, + "learning_rate": 1.7603196517275808e-05, + "loss": 0.01052580624818802, + "step": 4690, + "token_acc": 0.995435847372066 + }, + { + "epoch": 0.24883688609756258, + "grad_norm": 0.0904463604092598, + "learning_rate": 1.7592047204447216e-05, + "loss": 0.01085222065448761, + "step": 4700, + "token_acc": 0.9952639039439674 + }, + { + "epoch": 0.2493663262807489, + "grad_norm": 0.10422682017087936, + "learning_rate": 1.7580875568226436e-05, + "loss": 0.010660526901483535, + "step": 4710, + "token_acc": 0.9954247547484868 + }, + { + "epoch": 0.2498957664639352, + "grad_norm": 0.12369547039270401, + "learning_rate": 1.756968164146216e-05, + "loss": 0.011555536091327668, + "step": 4720, + "token_acc": 0.9950634990785754 + }, + { + "epoch": 0.2504252066471215, + "grad_norm": 0.08082088083028793, + "learning_rate": 1.7558465457068617e-05, + "loss": 0.01097717359662056, + "step": 4730, + "token_acc": 0.995102578863498 + }, + { + "epoch": 0.2509546468303078, + "grad_norm": 0.0759989470243454, + "learning_rate": 1.754722704802549e-05, + "loss": 0.009982295334339142, + "step": 4740, + "token_acc": 0.9956221178942805 + }, + { + "epoch": 0.2514840870134941, + "grad_norm": 0.08537179231643677, + "learning_rate": 1.7535966447377805e-05, + "loss": 0.010259152948856353, + "step": 4750, + "token_acc": 0.9955501010202208 + }, + { + "epoch": 0.2520135271966804, + "grad_norm": 0.1108422726392746, + "learning_rate": 1.7524683688235844e-05, + "loss": 0.00988430380821228, + "step": 4760, + "token_acc": 0.9957397495539659 + }, + { + "epoch": 0.2525429673798667, + "grad_norm": 4.389804363250732, + "learning_rate": 1.751337880377504e-05, + "loss": 0.01281629353761673, + "step": 4770, + "token_acc": 0.9950690417587919 + }, + { + "epoch": 0.253072407563053, + "grad_norm": 0.08497581630945206, + "learning_rate": 1.7502051827235882e-05, + "loss": 0.010467810928821564, + "step": 4780, + "token_acc": 0.9955544434713708 + }, + { + "epoch": 0.25360184774623934, + "grad_norm": 0.1379275918006897, + "learning_rate": 1.749070279192382e-05, + "loss": 0.010285983234643937, + "step": 4790, + "token_acc": 0.9954808854796348 + }, + { + "epoch": 0.25413128792942563, + "grad_norm": 0.08157876133918762, + "learning_rate": 1.7479331731209163e-05, + "loss": 0.009871353954076767, + "step": 4800, + "token_acc": 0.9958196700793511 + }, + { + "epoch": 0.2546607281126119, + "grad_norm": 0.07739344239234924, + "learning_rate": 1.746793867852698e-05, + "loss": 0.010115179419517516, + "step": 4810, + "token_acc": 0.995535528017691 + }, + { + "epoch": 0.2551901682957982, + "grad_norm": 0.08572009205818176, + "learning_rate": 1.7456523667377013e-05, + "loss": 0.010167035460472106, + "step": 4820, + "token_acc": 0.9956683943013095 + }, + { + "epoch": 0.25571960847898456, + "grad_norm": 0.09203449636697769, + "learning_rate": 1.7445086731323563e-05, + "loss": 0.009985524415969848, + "step": 4830, + "token_acc": 0.9956366124093742 + }, + { + "epoch": 0.25624904866217085, + "grad_norm": 0.1826680600643158, + "learning_rate": 1.74336279039954e-05, + "loss": 0.00958092212677002, + "step": 4840, + "token_acc": 0.9958029120947891 + }, + { + "epoch": 0.25677848884535714, + "grad_norm": 0.09237126260995865, + "learning_rate": 1.742214721908566e-05, + "loss": 0.009882880747318268, + "step": 4850, + "token_acc": 0.9956090556965407 + }, + { + "epoch": 0.25730792902854344, + "grad_norm": 0.19913312792778015, + "learning_rate": 1.7410644710351756e-05, + "loss": 0.0102533258497715, + "step": 4860, + "token_acc": 0.9955115423441763 + }, + { + "epoch": 0.25783736921172973, + "grad_norm": 0.10953837633132935, + "learning_rate": 1.739912041161526e-05, + "loss": 0.010251455754041672, + "step": 4870, + "token_acc": 0.9955061987777528 + }, + { + "epoch": 0.2583668093949161, + "grad_norm": 0.15581069886684418, + "learning_rate": 1.7387574356761826e-05, + "loss": 0.009992486238479615, + "step": 4880, + "token_acc": 0.9956349371948421 + }, + { + "epoch": 0.25889624957810237, + "grad_norm": 0.05655200779438019, + "learning_rate": 1.7376006579741066e-05, + "loss": 0.00950596183538437, + "step": 4890, + "token_acc": 0.9958958274245483 + }, + { + "epoch": 0.25942568976128866, + "grad_norm": 0.09489509463310242, + "learning_rate": 1.736441711456647e-05, + "loss": 0.010028690099716187, + "step": 4900, + "token_acc": 0.9956816419764408 + }, + { + "epoch": 0.25995512994447495, + "grad_norm": 0.08840907365083694, + "learning_rate": 1.735280599531531e-05, + "loss": 0.009550805389881133, + "step": 4910, + "token_acc": 0.9959025969690901 + }, + { + "epoch": 0.26048457012766124, + "grad_norm": 0.07671452313661575, + "learning_rate": 1.7341173256128508e-05, + "loss": 0.009890986233949661, + "step": 4920, + "token_acc": 0.9956684304528535 + }, + { + "epoch": 0.2610140103108476, + "grad_norm": 0.10329990088939667, + "learning_rate": 1.7329518931210573e-05, + "loss": 0.010437099635601044, + "step": 4930, + "token_acc": 0.9954537112731277 + }, + { + "epoch": 0.2615434504940339, + "grad_norm": 0.05995256081223488, + "learning_rate": 1.731784305482947e-05, + "loss": 0.009974733740091325, + "step": 4940, + "token_acc": 0.9956730528737932 + }, + { + "epoch": 0.26207289067722017, + "grad_norm": 0.11443962901830673, + "learning_rate": 1.7306145661316545e-05, + "loss": 0.009995999932289123, + "step": 4950, + "token_acc": 0.9957603071273577 + }, + { + "epoch": 0.26260233086040646, + "grad_norm": 0.08567491173744202, + "learning_rate": 1.7294426785066415e-05, + "loss": 0.010159790515899658, + "step": 4960, + "token_acc": 0.9957355187061346 + }, + { + "epoch": 0.26313177104359275, + "grad_norm": 0.08041193336248398, + "learning_rate": 1.7282686460536843e-05, + "loss": 0.00992010235786438, + "step": 4970, + "token_acc": 0.995691886386027 + }, + { + "epoch": 0.2636612112267791, + "grad_norm": 0.06766381114721298, + "learning_rate": 1.727092472224868e-05, + "loss": 0.010173306614160538, + "step": 4980, + "token_acc": 0.9955562225390394 + }, + { + "epoch": 0.2641906514099654, + "grad_norm": 0.10913963615894318, + "learning_rate": 1.7259141604785737e-05, + "loss": 0.010290446877479553, + "step": 4990, + "token_acc": 0.9955955221141494 + }, + { + "epoch": 0.2647200915931517, + "grad_norm": 0.07953409850597382, + "learning_rate": 1.724733714279468e-05, + "loss": 0.010660607367753983, + "step": 5000, + "token_acc": 0.9954212988832639 + }, + { + "epoch": 0.265249531776338, + "grad_norm": 0.08673790842294693, + "learning_rate": 1.7235511370984935e-05, + "loss": 0.009992580115795135, + "step": 5010, + "token_acc": 0.9960040710090764 + }, + { + "epoch": 0.2657789719595243, + "grad_norm": 0.106283999979496, + "learning_rate": 1.7223664324128596e-05, + "loss": 0.010323873162269593, + "step": 5020, + "token_acc": 0.995570681414379 + }, + { + "epoch": 0.2663084121427106, + "grad_norm": 0.12913547456264496, + "learning_rate": 1.721179603706031e-05, + "loss": 0.009637081623077392, + "step": 5030, + "token_acc": 0.9958964093581885 + }, + { + "epoch": 0.2668378523258969, + "grad_norm": 0.09253279119729996, + "learning_rate": 1.719990654467717e-05, + "loss": 0.010438549518585204, + "step": 5040, + "token_acc": 0.9954425561750859 + }, + { + "epoch": 0.2673672925090832, + "grad_norm": 0.09350863099098206, + "learning_rate": 1.718799588193863e-05, + "loss": 0.010773511230945587, + "step": 5050, + "token_acc": 0.9952622447617777 + }, + { + "epoch": 0.2678967326922695, + "grad_norm": 0.09263801574707031, + "learning_rate": 1.7176064083866397e-05, + "loss": 0.010268905758857727, + "step": 5060, + "token_acc": 0.9955039871876147 + }, + { + "epoch": 0.26842617287545584, + "grad_norm": 0.08650557696819305, + "learning_rate": 1.7164111185544302e-05, + "loss": 0.010438148677349091, + "step": 5070, + "token_acc": 0.9955429057917184 + }, + { + "epoch": 0.26895561305864213, + "grad_norm": 0.0784640684723854, + "learning_rate": 1.7152137222118245e-05, + "loss": 0.010061170160770416, + "step": 5080, + "token_acc": 0.9956363575683749 + }, + { + "epoch": 0.2694850532418284, + "grad_norm": 0.06762394309043884, + "learning_rate": 1.7140142228796047e-05, + "loss": 0.010280284285545348, + "step": 5090, + "token_acc": 0.9954874925973192 + }, + { + "epoch": 0.2700144934250147, + "grad_norm": 0.10271281003952026, + "learning_rate": 1.7128126240847368e-05, + "loss": 0.010046753287315368, + "step": 5100, + "token_acc": 0.995545694171032 + }, + { + "epoch": 0.270543933608201, + "grad_norm": 0.15268613398075104, + "learning_rate": 1.711608929360361e-05, + "loss": 0.010117458552122116, + "step": 5110, + "token_acc": 0.995655436957972 + }, + { + "epoch": 0.27107337379138735, + "grad_norm": 0.07870332151651382, + "learning_rate": 1.7104031422457786e-05, + "loss": 0.01032189503312111, + "step": 5120, + "token_acc": 0.9953214130833639 + }, + { + "epoch": 0.27160281397457364, + "grad_norm": 0.10826383531093597, + "learning_rate": 1.7091952662864448e-05, + "loss": 0.010061499476432801, + "step": 5130, + "token_acc": 0.9954698280536947 + }, + { + "epoch": 0.27213225415775993, + "grad_norm": 0.13327179849147797, + "learning_rate": 1.7079853050339558e-05, + "loss": 0.010724222660064698, + "step": 5140, + "token_acc": 0.9954011668182918 + }, + { + "epoch": 0.2726616943409462, + "grad_norm": 0.11641693115234375, + "learning_rate": 1.70677326204604e-05, + "loss": 0.010362713783979415, + "step": 5150, + "token_acc": 0.9954766991587662 + }, + { + "epoch": 0.2731911345241326, + "grad_norm": 0.080918088555336, + "learning_rate": 1.705559140886546e-05, + "loss": 0.010399863123893738, + "step": 5160, + "token_acc": 0.9954463570856685 + }, + { + "epoch": 0.27372057470731886, + "grad_norm": 0.07317711412906647, + "learning_rate": 1.7043429451254342e-05, + "loss": 0.01009959876537323, + "step": 5170, + "token_acc": 0.9955607847064026 + }, + { + "epoch": 0.27425001489050516, + "grad_norm": 0.11137982457876205, + "learning_rate": 1.7031246783387637e-05, + "loss": 0.00963272824883461, + "step": 5180, + "token_acc": 0.9959059451346619 + }, + { + "epoch": 0.27477945507369145, + "grad_norm": 0.08322513848543167, + "learning_rate": 1.7019043441086846e-05, + "loss": 0.010171153396368027, + "step": 5190, + "token_acc": 0.9955699793098846 + }, + { + "epoch": 0.27530889525687774, + "grad_norm": 0.09878774732351303, + "learning_rate": 1.7006819460234254e-05, + "loss": 0.009505745768547059, + "step": 5200, + "token_acc": 0.9959123746600598 + }, + { + "epoch": 0.2758383354400641, + "grad_norm": 0.09801449626684189, + "learning_rate": 1.6994574876772823e-05, + "loss": 0.010385122895240784, + "step": 5210, + "token_acc": 0.9955388412661352 + }, + { + "epoch": 0.2763677756232504, + "grad_norm": 0.08885489404201508, + "learning_rate": 1.6982309726706106e-05, + "loss": 0.010203893482685088, + "step": 5220, + "token_acc": 0.9956804536357572 + }, + { + "epoch": 0.27689721580643667, + "grad_norm": 0.06610845029354095, + "learning_rate": 1.6970024046098126e-05, + "loss": 0.010080977529287338, + "step": 5230, + "token_acc": 0.995545582702847 + }, + { + "epoch": 0.27742665598962296, + "grad_norm": 0.09676042944192886, + "learning_rate": 1.6957717871073265e-05, + "loss": 0.010139885544776916, + "step": 5240, + "token_acc": 0.9955147586990295 + }, + { + "epoch": 0.27795609617280925, + "grad_norm": 0.1016777753829956, + "learning_rate": 1.6945391237816183e-05, + "loss": 0.010168123245239257, + "step": 5250, + "token_acc": 0.9955729709450165 + }, + { + "epoch": 0.2784855363559956, + "grad_norm": 0.09102777391672134, + "learning_rate": 1.693304418257168e-05, + "loss": 0.009677259624004364, + "step": 5260, + "token_acc": 0.9956712845192123 + }, + { + "epoch": 0.2790149765391819, + "grad_norm": 0.1037028431892395, + "learning_rate": 1.6920676741644606e-05, + "loss": 0.009910434484481812, + "step": 5270, + "token_acc": 0.9957933394541357 + }, + { + "epoch": 0.2795444167223682, + "grad_norm": 0.09485547989606857, + "learning_rate": 1.690828895139976e-05, + "loss": 0.0097355917096138, + "step": 5280, + "token_acc": 0.9956854825249525 + }, + { + "epoch": 0.2800738569055545, + "grad_norm": 0.07754084467887878, + "learning_rate": 1.6895880848261762e-05, + "loss": 0.009660989046096802, + "step": 5290, + "token_acc": 0.9957425494615577 + }, + { + "epoch": 0.28060329708874077, + "grad_norm": 0.08729749172925949, + "learning_rate": 1.688345246871498e-05, + "loss": 0.010247129201889037, + "step": 5300, + "token_acc": 0.9956208397978079 + }, + { + "epoch": 0.2811327372719271, + "grad_norm": 0.06432025134563446, + "learning_rate": 1.687100384930338e-05, + "loss": 0.00970134064555168, + "step": 5310, + "token_acc": 0.9957198995469602 + }, + { + "epoch": 0.2816621774551134, + "grad_norm": 0.11998601257801056, + "learning_rate": 1.685853502663046e-05, + "loss": 0.010519204288721084, + "step": 5320, + "token_acc": 0.995409857955968 + }, + { + "epoch": 0.2821916176382997, + "grad_norm": 0.12320456653833389, + "learning_rate": 1.6846046037359107e-05, + "loss": 0.009626917541027069, + "step": 5330, + "token_acc": 0.9959056720201465 + }, + { + "epoch": 0.282721057821486, + "grad_norm": 0.08224596083164215, + "learning_rate": 1.683353691821151e-05, + "loss": 0.009865088760852814, + "step": 5340, + "token_acc": 0.995929975563173 + }, + { + "epoch": 0.28325049800467234, + "grad_norm": 0.06699672341346741, + "learning_rate": 1.6821007705969056e-05, + "loss": 0.009881328046321868, + "step": 5350, + "token_acc": 0.9956543498206689 + }, + { + "epoch": 0.2837799381878586, + "grad_norm": 0.12560273706912994, + "learning_rate": 1.68084584374722e-05, + "loss": 0.010346464067697524, + "step": 5360, + "token_acc": 0.9956267369949674 + }, + { + "epoch": 0.2843093783710449, + "grad_norm": 0.08353330194950104, + "learning_rate": 1.679588914962038e-05, + "loss": 0.009909333288669586, + "step": 5370, + "token_acc": 0.9957284566546808 + }, + { + "epoch": 0.2848388185542312, + "grad_norm": 0.0698608011007309, + "learning_rate": 1.678329987937189e-05, + "loss": 0.009329697489738465, + "step": 5380, + "token_acc": 0.9960274738572729 + }, + { + "epoch": 0.2853682587374175, + "grad_norm": 0.07612036913633347, + "learning_rate": 1.6770690663743784e-05, + "loss": 0.009307721257209777, + "step": 5390, + "token_acc": 0.9960063697984842 + }, + { + "epoch": 0.28589769892060385, + "grad_norm": 0.12496699392795563, + "learning_rate": 1.6758061539811768e-05, + "loss": 0.010323578119277954, + "step": 5400, + "token_acc": 0.9953011676139446 + }, + { + "epoch": 0.28642713910379014, + "grad_norm": 0.07051340490579605, + "learning_rate": 1.674541254471007e-05, + "loss": 0.009918044507503509, + "step": 5410, + "token_acc": 0.9957061506265685 + }, + { + "epoch": 0.28695657928697643, + "grad_norm": 0.10341159254312515, + "learning_rate": 1.6732743715631365e-05, + "loss": 0.009674561023712159, + "step": 5420, + "token_acc": 0.9957501168926591 + }, + { + "epoch": 0.2874860194701627, + "grad_norm": 0.08817838877439499, + "learning_rate": 1.672005508982663e-05, + "loss": 0.009385728091001511, + "step": 5430, + "token_acc": 0.9958464694990742 + }, + { + "epoch": 0.288015459653349, + "grad_norm": 0.08516643196344376, + "learning_rate": 1.6707346704605073e-05, + "loss": 0.009607335925102234, + "step": 5440, + "token_acc": 0.9960398515986494 + }, + { + "epoch": 0.28854489983653536, + "grad_norm": 0.08707567304372787, + "learning_rate": 1.6694618597333977e-05, + "loss": 0.009947662055492402, + "step": 5450, + "token_acc": 0.9957765406028029 + }, + { + "epoch": 0.28907434001972165, + "grad_norm": 0.07897105067968369, + "learning_rate": 1.668187080543863e-05, + "loss": 0.010012052208185195, + "step": 5460, + "token_acc": 0.9957166832265983 + }, + { + "epoch": 0.28960378020290795, + "grad_norm": 0.07494234293699265, + "learning_rate": 1.6669103366402195e-05, + "loss": 0.009755800664424896, + "step": 5470, + "token_acc": 0.9956864544658128 + }, + { + "epoch": 0.29013322038609424, + "grad_norm": 0.07901187241077423, + "learning_rate": 1.6656316317765612e-05, + "loss": 0.009615787863731384, + "step": 5480, + "token_acc": 0.9958379902246984 + }, + { + "epoch": 0.2906626605692806, + "grad_norm": 0.10357753932476044, + "learning_rate": 1.6643509697127478e-05, + "loss": 0.009917864948511124, + "step": 5490, + "token_acc": 0.9956035338577948 + }, + { + "epoch": 0.2911921007524669, + "grad_norm": 0.09779317677021027, + "learning_rate": 1.6630683542143926e-05, + "loss": 0.00972934439778328, + "step": 5500, + "token_acc": 0.995919729650799 + }, + { + "epoch": 0.29172154093565317, + "grad_norm": 0.0983927994966507, + "learning_rate": 1.6617837890528543e-05, + "loss": 0.009848371148109436, + "step": 5510, + "token_acc": 0.9956534434572227 + }, + { + "epoch": 0.29225098111883946, + "grad_norm": 0.10686007142066956, + "learning_rate": 1.6604972780052236e-05, + "loss": 0.009840734302997589, + "step": 5520, + "token_acc": 0.9957268523927958 + }, + { + "epoch": 0.29278042130202575, + "grad_norm": 0.06463044136762619, + "learning_rate": 1.6592088248543133e-05, + "loss": 0.009312328696250916, + "step": 5530, + "token_acc": 0.9960174663738907 + }, + { + "epoch": 0.2933098614852121, + "grad_norm": 0.06900442391633987, + "learning_rate": 1.6579184333886454e-05, + "loss": 0.010098479688167572, + "step": 5540, + "token_acc": 0.9955735614074575 + }, + { + "epoch": 0.2938393016683984, + "grad_norm": 0.09063943475484848, + "learning_rate": 1.6566261074024436e-05, + "loss": 0.009185981005430222, + "step": 5550, + "token_acc": 0.9958645300071703 + }, + { + "epoch": 0.2943687418515847, + "grad_norm": 0.07650324702262878, + "learning_rate": 1.655331850695617e-05, + "loss": 0.009732353687286376, + "step": 5560, + "token_acc": 0.9958857352683452 + }, + { + "epoch": 0.294898182034771, + "grad_norm": 0.05918656289577484, + "learning_rate": 1.6540356670737542e-05, + "loss": 0.009732091426849365, + "step": 5570, + "token_acc": 0.995771864132565 + }, + { + "epoch": 0.29542762221795726, + "grad_norm": 0.11335550248622894, + "learning_rate": 1.6527375603481077e-05, + "loss": 0.009178735315799713, + "step": 5580, + "token_acc": 0.9958833647856511 + }, + { + "epoch": 0.2959570624011436, + "grad_norm": 0.05637415498495102, + "learning_rate": 1.651437534335586e-05, + "loss": 0.00958641916513443, + "step": 5590, + "token_acc": 0.9956717538153615 + }, + { + "epoch": 0.2964865025843299, + "grad_norm": 0.08341427892446518, + "learning_rate": 1.65013559285874e-05, + "loss": 0.009712567180395126, + "step": 5600, + "token_acc": 0.9957118782640615 + }, + { + "epoch": 0.2970159427675162, + "grad_norm": 0.09955371916294098, + "learning_rate": 1.648831739745754e-05, + "loss": 0.009708032011985779, + "step": 5610, + "token_acc": 0.9958019312785331 + }, + { + "epoch": 0.2975453829507025, + "grad_norm": 0.06681809574365616, + "learning_rate": 1.647525978830432e-05, + "loss": 0.009978721290826798, + "step": 5620, + "token_acc": 0.9956960547168238 + }, + { + "epoch": 0.2980748231338888, + "grad_norm": 0.10781566798686981, + "learning_rate": 1.646218313952188e-05, + "loss": 0.009703484177589417, + "step": 5630, + "token_acc": 0.9958117105241202 + }, + { + "epoch": 0.2986042633170751, + "grad_norm": 0.08826080709695816, + "learning_rate": 1.6449087489560347e-05, + "loss": 0.010396993905305862, + "step": 5640, + "token_acc": 0.9953953953953953 + }, + { + "epoch": 0.2991337035002614, + "grad_norm": 0.07920516282320023, + "learning_rate": 1.6435972876925716e-05, + "loss": 0.009586150199174881, + "step": 5650, + "token_acc": 0.9957675932882545 + }, + { + "epoch": 0.2996631436834477, + "grad_norm": 0.09159889817237854, + "learning_rate": 1.6422839340179736e-05, + "loss": 0.009946726262569427, + "step": 5660, + "token_acc": 0.9956647157887716 + }, + { + "epoch": 0.300192583866634, + "grad_norm": 0.09318218380212784, + "learning_rate": 1.64096869179398e-05, + "loss": 0.009697252511978149, + "step": 5670, + "token_acc": 0.9959112149532711 + }, + { + "epoch": 0.30072202404982035, + "grad_norm": 0.08872237801551819, + "learning_rate": 1.6396515648878845e-05, + "loss": 0.009905125200748443, + "step": 5680, + "token_acc": 0.9956585264036736 + }, + { + "epoch": 0.30125146423300664, + "grad_norm": 0.08196310698986053, + "learning_rate": 1.6383325571725204e-05, + "loss": 0.009628520160913468, + "step": 5690, + "token_acc": 0.995655436957972 + }, + { + "epoch": 0.30178090441619293, + "grad_norm": 0.11664915084838867, + "learning_rate": 1.6370116725262523e-05, + "loss": 0.009039324522018433, + "step": 5700, + "token_acc": 0.9960218838089837 + }, + { + "epoch": 0.3023103445993792, + "grad_norm": 0.0882045105099678, + "learning_rate": 1.635688914832964e-05, + "loss": 0.009757035970687866, + "step": 5710, + "token_acc": 0.9957274585888931 + }, + { + "epoch": 0.3028397847825655, + "grad_norm": 0.10893391072750092, + "learning_rate": 1.634364287982046e-05, + "loss": 0.009365728497505188, + "step": 5720, + "token_acc": 0.9959856451343683 + }, + { + "epoch": 0.30336922496575186, + "grad_norm": 0.11440259218215942, + "learning_rate": 1.6330377958683856e-05, + "loss": 0.009367022663354874, + "step": 5730, + "token_acc": 0.996018795957033 + }, + { + "epoch": 0.30389866514893815, + "grad_norm": 0.09525571763515472, + "learning_rate": 1.6317094423923534e-05, + "loss": 0.009122355282306671, + "step": 5740, + "token_acc": 0.9960885701180101 + }, + { + "epoch": 0.30442810533212444, + "grad_norm": 0.08001086115837097, + "learning_rate": 1.6303792314597943e-05, + "loss": 0.00919048860669136, + "step": 5750, + "token_acc": 0.9960473978702646 + }, + { + "epoch": 0.30495754551531074, + "grad_norm": 0.09881721436977386, + "learning_rate": 1.6290471669820146e-05, + "loss": 0.008948402106761932, + "step": 5760, + "token_acc": 0.9960394883852786 + }, + { + "epoch": 0.305486985698497, + "grad_norm": 0.07967948168516159, + "learning_rate": 1.6277132528757703e-05, + "loss": 0.009203647822141647, + "step": 5770, + "token_acc": 0.9961981307476175 + }, + { + "epoch": 0.3060164258816834, + "grad_norm": 0.07709358632564545, + "learning_rate": 1.626377493063256e-05, + "loss": 0.00967738702893257, + "step": 5780, + "token_acc": 0.9958345159189943 + }, + { + "epoch": 0.30654586606486967, + "grad_norm": 0.11473223567008972, + "learning_rate": 1.6250398914720936e-05, + "loss": 0.009214276075363159, + "step": 5790, + "token_acc": 0.9958866964248467 + }, + { + "epoch": 0.30707530624805596, + "grad_norm": 0.13097169995307922, + "learning_rate": 1.623700452035321e-05, + "loss": 0.009174647182226181, + "step": 5800, + "token_acc": 0.9960029706522918 + }, + { + "epoch": 0.30760474643124225, + "grad_norm": 0.0574369803071022, + "learning_rate": 1.6223591786913778e-05, + "loss": 0.010016897320747375, + "step": 5810, + "token_acc": 0.9957468101075807 + }, + { + "epoch": 0.3081341866144286, + "grad_norm": 0.08412422984838486, + "learning_rate": 1.6210160753840997e-05, + "loss": 0.009437409788370132, + "step": 5820, + "token_acc": 0.995954962468724 + }, + { + "epoch": 0.3086636267976149, + "grad_norm": 0.07035002112388611, + "learning_rate": 1.6196711460626993e-05, + "loss": 0.009588725864887238, + "step": 5830, + "token_acc": 0.9957728159549101 + }, + { + "epoch": 0.3091930669808012, + "grad_norm": 0.09316197037696838, + "learning_rate": 1.618324394681761e-05, + "loss": 0.009136982262134552, + "step": 5840, + "token_acc": 0.995995895793188 + }, + { + "epoch": 0.30972250716398747, + "grad_norm": 0.07297364622354507, + "learning_rate": 1.6169758252012262e-05, + "loss": 0.009341055154800415, + "step": 5850, + "token_acc": 0.9958718695010383 + }, + { + "epoch": 0.31025194734717376, + "grad_norm": 0.06908360868692398, + "learning_rate": 1.615625441586381e-05, + "loss": 0.009629489481449127, + "step": 5860, + "token_acc": 0.9957351627899212 + }, + { + "epoch": 0.3107813875303601, + "grad_norm": 0.08213352411985397, + "learning_rate": 1.6142732478078468e-05, + "loss": 0.009458699822425842, + "step": 5870, + "token_acc": 0.9959706684686037 + }, + { + "epoch": 0.3113108277135464, + "grad_norm": 0.10233397036790848, + "learning_rate": 1.612919247841567e-05, + "loss": 0.009136594831943512, + "step": 5880, + "token_acc": 0.9961392215069544 + }, + { + "epoch": 0.3118402678967327, + "grad_norm": 0.08014033734798431, + "learning_rate": 1.611563445668797e-05, + "loss": 0.009805220365524291, + "step": 5890, + "token_acc": 0.9956970955394892 + }, + { + "epoch": 0.312369708079919, + "grad_norm": 0.06626968830823898, + "learning_rate": 1.6102058452760894e-05, + "loss": 0.009142603725194931, + "step": 5900, + "token_acc": 0.9960734293717487 + }, + { + "epoch": 0.3128991482631053, + "grad_norm": 0.15002653002738953, + "learning_rate": 1.608846450655286e-05, + "loss": 0.009353496134281158, + "step": 5910, + "token_acc": 0.9959207507820647 + }, + { + "epoch": 0.3134285884462916, + "grad_norm": 0.08285783976316452, + "learning_rate": 1.6074852658035034e-05, + "loss": 0.0094138503074646, + "step": 5920, + "token_acc": 0.9958853232066102 + }, + { + "epoch": 0.3139580286294779, + "grad_norm": 0.09231264889240265, + "learning_rate": 1.6061222947231224e-05, + "loss": 0.03251364827156067, + "step": 5930, + "token_acc": 0.9946847124988527 + }, + { + "epoch": 0.3144874688126642, + "grad_norm": 0.11155049502849579, + "learning_rate": 1.6047575414217756e-05, + "loss": 0.00907934010028839, + "step": 5940, + "token_acc": 0.996103722738574 + }, + { + "epoch": 0.3150169089958505, + "grad_norm": 0.08581361919641495, + "learning_rate": 1.6033910099123366e-05, + "loss": 0.0087142214179039, + "step": 5950, + "token_acc": 0.996376357822141 + }, + { + "epoch": 0.3155463491790368, + "grad_norm": 0.07173480093479156, + "learning_rate": 1.602022704212907e-05, + "loss": 0.009482725709676742, + "step": 5960, + "token_acc": 0.9959811899679829 + }, + { + "epoch": 0.31607578936222314, + "grad_norm": 0.15052035450935364, + "learning_rate": 1.600652628346805e-05, + "loss": 0.009292979538440705, + "step": 5970, + "token_acc": 0.9959850419859435 + }, + { + "epoch": 0.31660522954540943, + "grad_norm": 0.07231432944536209, + "learning_rate": 1.5992807863425552e-05, + "loss": 0.0085860975086689, + "step": 5980, + "token_acc": 0.9962354551676934 + }, + { + "epoch": 0.3171346697285957, + "grad_norm": 0.08485893905162811, + "learning_rate": 1.5979071822338733e-05, + "loss": 0.009241646528244019, + "step": 5990, + "token_acc": 0.9959022208127123 + }, + { + "epoch": 0.317664109911782, + "grad_norm": 0.14029960334300995, + "learning_rate": 1.596531820059658e-05, + "loss": 0.008970920741558076, + "step": 6000, + "token_acc": 0.9962189818794915 + }, + { + "epoch": 0.317664109911782, + "eval_loss": 0.011477283202111721, + "eval_runtime": 710.3112, + "eval_samples_per_second": 16.863, + "eval_steps_per_second": 1.054, + "eval_token_acc": 0.995017686943771, + "step": 6000 + }, + { + "epoch": 0.31819355009496836, + "grad_norm": 0.0861055850982666, + "learning_rate": 1.5951547038639752e-05, + "loss": 0.009113739430904388, + "step": 6010, + "token_acc": 0.9960300912411802 + }, + { + "epoch": 0.31872299027815465, + "grad_norm": 0.08600447326898575, + "learning_rate": 1.5937758376960507e-05, + "loss": 0.009130691736936569, + "step": 6020, + "token_acc": 0.9960035042342831 + }, + { + "epoch": 0.31925243046134094, + "grad_norm": 0.08892437070608139, + "learning_rate": 1.5923952256102547e-05, + "loss": 0.008988317847251893, + "step": 6030, + "token_acc": 0.9960308526162185 + }, + { + "epoch": 0.31978187064452723, + "grad_norm": 0.06886985898017883, + "learning_rate": 1.5910128716660908e-05, + "loss": 0.00877256840467453, + "step": 6040, + "token_acc": 0.9962185715716718 + }, + { + "epoch": 0.3203113108277135, + "grad_norm": 0.07213792204856873, + "learning_rate": 1.589628779928184e-05, + "loss": 0.00880722850561142, + "step": 6050, + "token_acc": 0.9960985369513568 + }, + { + "epoch": 0.32084075101089987, + "grad_norm": 0.06557095795869827, + "learning_rate": 1.588242954466271e-05, + "loss": 0.009177390486001968, + "step": 6060, + "token_acc": 0.9960527743238394 + }, + { + "epoch": 0.32137019119408616, + "grad_norm": 0.07065486162900925, + "learning_rate": 1.5868553993551833e-05, + "loss": 0.009276406466960907, + "step": 6070, + "token_acc": 0.9960436698717948 + }, + { + "epoch": 0.32189963137727245, + "grad_norm": 0.09328465163707733, + "learning_rate": 1.5854661186748416e-05, + "loss": 0.009605391323566437, + "step": 6080, + "token_acc": 0.9958392048628772 + }, + { + "epoch": 0.32242907156045875, + "grad_norm": 0.07949452102184296, + "learning_rate": 1.584075116510237e-05, + "loss": 0.009189733862876892, + "step": 6090, + "token_acc": 0.9960864812543286 + }, + { + "epoch": 0.32295851174364504, + "grad_norm": 0.07642360776662827, + "learning_rate": 1.5826823969514245e-05, + "loss": 0.009498965740203858, + "step": 6100, + "token_acc": 0.9959803185722625 + }, + { + "epoch": 0.3234879519268314, + "grad_norm": 0.09671497344970703, + "learning_rate": 1.5812879640935088e-05, + "loss": 0.008847469091415405, + "step": 6110, + "token_acc": 0.9962866226624498 + }, + { + "epoch": 0.3240173921100177, + "grad_norm": 0.07606108486652374, + "learning_rate": 1.5798918220366316e-05, + "loss": 0.009391656517982483, + "step": 6120, + "token_acc": 0.995936316149597 + }, + { + "epoch": 0.32454683229320397, + "grad_norm": 0.10382432490587234, + "learning_rate": 1.5784939748859605e-05, + "loss": 0.009515581279993057, + "step": 6130, + "token_acc": 0.9959430355443512 + }, + { + "epoch": 0.32507627247639026, + "grad_norm": 0.06706821918487549, + "learning_rate": 1.5770944267516764e-05, + "loss": 0.00860891044139862, + "step": 6140, + "token_acc": 0.9963469557964971 + }, + { + "epoch": 0.3256057126595766, + "grad_norm": 0.08219406753778458, + "learning_rate": 1.575693181748963e-05, + "loss": 0.009296966344118118, + "step": 6150, + "token_acc": 0.9961355479509223 + }, + { + "epoch": 0.3261351528427629, + "grad_norm": 0.14151635766029358, + "learning_rate": 1.574290243997992e-05, + "loss": 0.009372777491807937, + "step": 6160, + "token_acc": 0.9958213436757163 + }, + { + "epoch": 0.3266645930259492, + "grad_norm": 0.07452268153429031, + "learning_rate": 1.5728856176239127e-05, + "loss": 0.008828724920749664, + "step": 6170, + "token_acc": 0.996195371034759 + }, + { + "epoch": 0.3271940332091355, + "grad_norm": 0.08804163336753845, + "learning_rate": 1.5714793067568403e-05, + "loss": 0.00887804850935936, + "step": 6180, + "token_acc": 0.9961437979416895 + }, + { + "epoch": 0.3277234733923218, + "grad_norm": 0.059914544224739075, + "learning_rate": 1.5700713155318418e-05, + "loss": 0.009255196154117584, + "step": 6190, + "token_acc": 0.995953038558781 + }, + { + "epoch": 0.3282529135755081, + "grad_norm": 0.07684949040412903, + "learning_rate": 1.5686616480889258e-05, + "loss": 0.008642568439245223, + "step": 6200, + "token_acc": 0.9962766312696186 + }, + { + "epoch": 0.3287823537586944, + "grad_norm": 0.08286368101835251, + "learning_rate": 1.5672503085730303e-05, + "loss": 0.008778942376375198, + "step": 6210, + "token_acc": 0.9962216309542675 + }, + { + "epoch": 0.3293117939418807, + "grad_norm": 0.1016073077917099, + "learning_rate": 1.5658373011340082e-05, + "loss": 0.009228568524122238, + "step": 6220, + "token_acc": 0.9963564508329303 + }, + { + "epoch": 0.329841234125067, + "grad_norm": 0.10347961634397507, + "learning_rate": 1.5644226299266183e-05, + "loss": 0.008666118979454041, + "step": 6230, + "token_acc": 0.9962085466373817 + }, + { + "epoch": 0.3303706743082533, + "grad_norm": 0.11229892075061798, + "learning_rate": 1.56300629911051e-05, + "loss": 0.008447064459323883, + "step": 6240, + "token_acc": 0.9964209437362344 + }, + { + "epoch": 0.33090011449143963, + "grad_norm": 0.07292419672012329, + "learning_rate": 1.5615883128502133e-05, + "loss": 0.008586619794368745, + "step": 6250, + "token_acc": 0.9962774703491332 + }, + { + "epoch": 0.3314295546746259, + "grad_norm": 0.07681415975093842, + "learning_rate": 1.5601686753151266e-05, + "loss": 0.008841368556022643, + "step": 6260, + "token_acc": 0.9962289022935282 + }, + { + "epoch": 0.3319589948578122, + "grad_norm": 0.06035500392317772, + "learning_rate": 1.5587473906795024e-05, + "loss": 0.009253267943859101, + "step": 6270, + "token_acc": 0.9960126128230368 + }, + { + "epoch": 0.3324884350409985, + "grad_norm": 0.07261207699775696, + "learning_rate": 1.5573244631224364e-05, + "loss": 0.00843152478337288, + "step": 6280, + "token_acc": 0.9965221307578752 + }, + { + "epoch": 0.3330178752241848, + "grad_norm": 0.0901603028178215, + "learning_rate": 1.555899896827856e-05, + "loss": 0.009364788979291916, + "step": 6290, + "token_acc": 0.9959241168919932 + }, + { + "epoch": 0.33354731540737115, + "grad_norm": 0.0729965791106224, + "learning_rate": 1.5544736959845063e-05, + "loss": 0.008965513110160828, + "step": 6300, + "token_acc": 0.9960481887531786 + }, + { + "epoch": 0.33407675559055744, + "grad_norm": 0.11340804398059845, + "learning_rate": 1.5530458647859385e-05, + "loss": 0.009127754718065262, + "step": 6310, + "token_acc": 0.995766815843965 + }, + { + "epoch": 0.33460619577374373, + "grad_norm": 0.12861299514770508, + "learning_rate": 1.551616407430498e-05, + "loss": 0.009375978261232376, + "step": 6320, + "token_acc": 0.9959953947039095 + }, + { + "epoch": 0.33513563595693, + "grad_norm": 0.08486894518136978, + "learning_rate": 1.550185328121312e-05, + "loss": 0.00914788842201233, + "step": 6330, + "token_acc": 0.995918469551282 + }, + { + "epoch": 0.33566507614011637, + "grad_norm": 0.13351406157016754, + "learning_rate": 1.5487526310662762e-05, + "loss": 0.009574226289987563, + "step": 6340, + "token_acc": 0.9959084836339346 + }, + { + "epoch": 0.33619451632330266, + "grad_norm": 0.05893190577626228, + "learning_rate": 1.547318320478043e-05, + "loss": 0.008804354071617126, + "step": 6350, + "token_acc": 0.9960209214444806 + }, + { + "epoch": 0.33672395650648895, + "grad_norm": 0.11144118010997772, + "learning_rate": 1.5458824005740103e-05, + "loss": 0.008796499669551849, + "step": 6360, + "token_acc": 0.9961771849724974 + }, + { + "epoch": 0.33725339668967524, + "grad_norm": 0.06568700075149536, + "learning_rate": 1.5444448755763062e-05, + "loss": 0.008684404194355011, + "step": 6370, + "token_acc": 0.996105152541242 + }, + { + "epoch": 0.33778283687286154, + "grad_norm": 0.06569815427064896, + "learning_rate": 1.54300574971178e-05, + "loss": 0.008211921155452728, + "step": 6380, + "token_acc": 0.996572915190074 + }, + { + "epoch": 0.3383122770560479, + "grad_norm": 0.07777664810419083, + "learning_rate": 1.5415650272119873e-05, + "loss": 0.009406762570142746, + "step": 6390, + "token_acc": 0.9958904680531822 + }, + { + "epoch": 0.3388417172392342, + "grad_norm": 0.08153358846902847, + "learning_rate": 1.540122712313178e-05, + "loss": 0.00931502878665924, + "step": 6400, + "token_acc": 0.9960531695635123 + }, + { + "epoch": 0.33937115742242047, + "grad_norm": 0.08143303543329239, + "learning_rate": 1.538678809256286e-05, + "loss": 0.008841399103403091, + "step": 6410, + "token_acc": 0.9962203384173814 + }, + { + "epoch": 0.33990059760560676, + "grad_norm": 0.07863390445709229, + "learning_rate": 1.5372333222869126e-05, + "loss": 0.008432415127754212, + "step": 6420, + "token_acc": 0.9962778431519825 + }, + { + "epoch": 0.34043003778879305, + "grad_norm": 0.06499326229095459, + "learning_rate": 1.535786255655318e-05, + "loss": 0.008718221634626388, + "step": 6430, + "token_acc": 0.9961719375500401 + }, + { + "epoch": 0.3409594779719794, + "grad_norm": 0.07058816403150558, + "learning_rate": 1.5343376136164067e-05, + "loss": 0.008544119447469712, + "step": 6440, + "token_acc": 0.9963618157543391 + }, + { + "epoch": 0.3414889181551657, + "grad_norm": 0.11667972058057785, + "learning_rate": 1.5328874004297154e-05, + "loss": 0.008464707434177399, + "step": 6450, + "token_acc": 0.9962048544499125 + }, + { + "epoch": 0.342018358338352, + "grad_norm": 0.10080605000257492, + "learning_rate": 1.5314356203594008e-05, + "loss": 0.008374043554067612, + "step": 6460, + "token_acc": 0.9963040830288165 + }, + { + "epoch": 0.34254779852153827, + "grad_norm": 0.08833302557468414, + "learning_rate": 1.529982277674227e-05, + "loss": 0.009146318584680558, + "step": 6470, + "token_acc": 0.9960195932808732 + }, + { + "epoch": 0.3430772387047246, + "grad_norm": 0.06996115297079086, + "learning_rate": 1.5285273766475513e-05, + "loss": 0.009018532186746597, + "step": 6480, + "token_acc": 0.9959539163586916 + }, + { + "epoch": 0.3436066788879109, + "grad_norm": 0.07584096491336823, + "learning_rate": 1.5270709215573152e-05, + "loss": 0.008696052432060241, + "step": 6490, + "token_acc": 0.9961458246433637 + }, + { + "epoch": 0.3441361190710972, + "grad_norm": 0.09288084506988525, + "learning_rate": 1.5256129166860284e-05, + "loss": 0.008919874578714371, + "step": 6500, + "token_acc": 0.996145181476846 + }, + { + "epoch": 0.3446655592542835, + "grad_norm": 0.07339269667863846, + "learning_rate": 1.5241533663207575e-05, + "loss": 0.008549313247203826, + "step": 6510, + "token_acc": 0.9964118525688632 + }, + { + "epoch": 0.3451949994374698, + "grad_norm": 0.08147459477186203, + "learning_rate": 1.5226922747531143e-05, + "loss": 0.009033076465129852, + "step": 6520, + "token_acc": 0.9960851745811805 + }, + { + "epoch": 0.34572443962065613, + "grad_norm": 0.10153432190418243, + "learning_rate": 1.5212296462792418e-05, + "loss": 0.008526679873466492, + "step": 6530, + "token_acc": 0.9964823659839789 + }, + { + "epoch": 0.3462538798038424, + "grad_norm": 0.07803604751825333, + "learning_rate": 1.5197654851998011e-05, + "loss": 0.00866355448961258, + "step": 6540, + "token_acc": 0.9965102979604444 + }, + { + "epoch": 0.3467833199870287, + "grad_norm": 0.11945109069347382, + "learning_rate": 1.5182997958199617e-05, + "loss": 0.008174373209476471, + "step": 6550, + "token_acc": 0.9965364135605669 + }, + { + "epoch": 0.347312760170215, + "grad_norm": 0.06914705038070679, + "learning_rate": 1.5168325824493853e-05, + "loss": 0.008564507961273194, + "step": 6560, + "token_acc": 0.9964615063634467 + }, + { + "epoch": 0.3478422003534013, + "grad_norm": 0.07243646681308746, + "learning_rate": 1.5153638494022152e-05, + "loss": 0.009045325964689255, + "step": 6570, + "token_acc": 0.9960967472894079 + }, + { + "epoch": 0.34837164053658765, + "grad_norm": 0.10352995991706848, + "learning_rate": 1.5138936009970633e-05, + "loss": 0.008151848614215852, + "step": 6580, + "token_acc": 0.9965321195751846 + }, + { + "epoch": 0.34890108071977394, + "grad_norm": 0.07816743850708008, + "learning_rate": 1.5124218415569969e-05, + "loss": 0.008650150895118714, + "step": 6590, + "token_acc": 0.9962153104862576 + }, + { + "epoch": 0.34943052090296023, + "grad_norm": 0.13017474114894867, + "learning_rate": 1.5109485754095261e-05, + "loss": 0.008454060554504395, + "step": 6600, + "token_acc": 0.996369675524102 + }, + { + "epoch": 0.3499599610861465, + "grad_norm": 0.07812652736902237, + "learning_rate": 1.509473806886592e-05, + "loss": 0.00891561284661293, + "step": 6610, + "token_acc": 0.9961234170619669 + }, + { + "epoch": 0.3504894012693328, + "grad_norm": 0.07754290848970413, + "learning_rate": 1.5079975403245523e-05, + "loss": 0.008398633450269699, + "step": 6620, + "token_acc": 0.9963806791647208 + }, + { + "epoch": 0.35101884145251916, + "grad_norm": 0.07319533824920654, + "learning_rate": 1.506519780064171e-05, + "loss": 0.009117646515369416, + "step": 6630, + "token_acc": 0.9959880893796969 + }, + { + "epoch": 0.35154828163570545, + "grad_norm": 0.07587375491857529, + "learning_rate": 1.505040530450602e-05, + "loss": 0.007751373946666718, + "step": 6640, + "token_acc": 0.9967133526305691 + }, + { + "epoch": 0.35207772181889174, + "grad_norm": 0.07332997769117355, + "learning_rate": 1.5035597958333799e-05, + "loss": 0.008114126324653626, + "step": 6650, + "token_acc": 0.9966032381906192 + }, + { + "epoch": 0.35260716200207803, + "grad_norm": 0.11108594387769699, + "learning_rate": 1.5020775805664055e-05, + "loss": 0.008764214813709259, + "step": 6660, + "token_acc": 0.9961375144947485 + }, + { + "epoch": 0.3531366021852644, + "grad_norm": 0.1248961016535759, + "learning_rate": 1.5005938890079328e-05, + "loss": 0.008434607088565827, + "step": 6670, + "token_acc": 0.9963794412326584 + }, + { + "epoch": 0.3536660423684507, + "grad_norm": 0.14208568632602692, + "learning_rate": 1.4991087255205574e-05, + "loss": 0.008077596127986909, + "step": 6680, + "token_acc": 0.9966114426407378 + }, + { + "epoch": 0.35419548255163696, + "grad_norm": 0.15353021025657654, + "learning_rate": 1.4976220944712025e-05, + "loss": 0.009166976809501648, + "step": 6690, + "token_acc": 0.9960055705196094 + }, + { + "epoch": 0.35472492273482326, + "grad_norm": 0.07467907667160034, + "learning_rate": 1.4961340002311063e-05, + "loss": 0.009131768345832824, + "step": 6700, + "token_acc": 0.9960319775923441 + }, + { + "epoch": 0.35525436291800955, + "grad_norm": 0.07088314741849899, + "learning_rate": 1.4946444471758092e-05, + "loss": 0.008214731514453889, + "step": 6710, + "token_acc": 0.9965887102155201 + }, + { + "epoch": 0.3557838031011959, + "grad_norm": 0.07474954426288605, + "learning_rate": 1.4931534396851416e-05, + "loss": 0.008914574235677718, + "step": 6720, + "token_acc": 0.9961196321578183 + }, + { + "epoch": 0.3563132432843822, + "grad_norm": 0.09953784197568893, + "learning_rate": 1.4916609821432107e-05, + "loss": 0.008826042711734771, + "step": 6730, + "token_acc": 0.9962184137372591 + }, + { + "epoch": 0.3568426834675685, + "grad_norm": 0.1087934747338295, + "learning_rate": 1.4901670789383859e-05, + "loss": 0.008589546382427215, + "step": 6740, + "token_acc": 0.9962563991395554 + }, + { + "epoch": 0.35737212365075477, + "grad_norm": 0.07643584907054901, + "learning_rate": 1.488671734463289e-05, + "loss": 0.008497269451618194, + "step": 6750, + "token_acc": 0.9963855522258487 + }, + { + "epoch": 0.35790156383394106, + "grad_norm": 0.08400347083806992, + "learning_rate": 1.4871749531147787e-05, + "loss": 0.008091196417808533, + "step": 6760, + "token_acc": 0.9964781973711663 + }, + { + "epoch": 0.3584310040171274, + "grad_norm": 0.10275434702634811, + "learning_rate": 1.4856767392939394e-05, + "loss": 0.008578431606292725, + "step": 6770, + "token_acc": 0.9962041579070311 + }, + { + "epoch": 0.3589604442003137, + "grad_norm": 0.0981011614203453, + "learning_rate": 1.4841770974060663e-05, + "loss": 0.00843815952539444, + "step": 6780, + "token_acc": 0.99636387897388 + }, + { + "epoch": 0.3594898843835, + "grad_norm": 0.07873008400201797, + "learning_rate": 1.4826760318606553e-05, + "loss": 0.008255517482757569, + "step": 6790, + "token_acc": 0.9964124512969406 + }, + { + "epoch": 0.3600193245666863, + "grad_norm": 0.08992158621549606, + "learning_rate": 1.4811735470713864e-05, + "loss": 0.007780134677886963, + "step": 6800, + "token_acc": 0.9965281544971248 + }, + { + "epoch": 0.36054876474987263, + "grad_norm": 0.10192002356052399, + "learning_rate": 1.4796696474561145e-05, + "loss": 0.008712895214557648, + "step": 6810, + "token_acc": 0.9963452793191205 + }, + { + "epoch": 0.3610782049330589, + "grad_norm": 0.11269593983888626, + "learning_rate": 1.4781643374368536e-05, + "loss": 0.008835648745298385, + "step": 6820, + "token_acc": 0.9960711038446459 + }, + { + "epoch": 0.3616076451162452, + "grad_norm": 0.07689455151557922, + "learning_rate": 1.476657621439765e-05, + "loss": 0.008530103415250779, + "step": 6830, + "token_acc": 0.9963126413000859 + }, + { + "epoch": 0.3621370852994315, + "grad_norm": 0.070935919880867, + "learning_rate": 1.4751495038951444e-05, + "loss": 0.008115139603614808, + "step": 6840, + "token_acc": 0.9965283283400236 + }, + { + "epoch": 0.3626665254826178, + "grad_norm": 0.10170845687389374, + "learning_rate": 1.473639989237408e-05, + "loss": 0.00894813984632492, + "step": 6850, + "token_acc": 0.9961688382147955 + }, + { + "epoch": 0.36319596566580414, + "grad_norm": 0.13983041048049927, + "learning_rate": 1.4721290819050804e-05, + "loss": 0.008756274729967118, + "step": 6860, + "token_acc": 0.9962511480337313 + }, + { + "epoch": 0.36372540584899044, + "grad_norm": 0.07963170111179352, + "learning_rate": 1.4706167863407812e-05, + "loss": 0.008202113211154938, + "step": 6870, + "token_acc": 0.9965096568942626 + }, + { + "epoch": 0.3642548460321767, + "grad_norm": 0.07169204950332642, + "learning_rate": 1.4691031069912118e-05, + "loss": 0.0080909863114357, + "step": 6880, + "token_acc": 0.9964880083420229 + }, + { + "epoch": 0.364784286215363, + "grad_norm": 0.09158442914485931, + "learning_rate": 1.4675880483071428e-05, + "loss": 0.008538705110549927, + "step": 6890, + "token_acc": 0.9962029541850955 + }, + { + "epoch": 0.3653137263985493, + "grad_norm": 0.09928670525550842, + "learning_rate": 1.4660716147433995e-05, + "loss": 0.008499117940664292, + "step": 6900, + "token_acc": 0.9964229133661302 + }, + { + "epoch": 0.36584316658173566, + "grad_norm": 0.07625437527894974, + "learning_rate": 1.4645538107588507e-05, + "loss": 0.008902019262313843, + "step": 6910, + "token_acc": 0.9963225483655771 + }, + { + "epoch": 0.36637260676492195, + "grad_norm": 0.08107022941112518, + "learning_rate": 1.4630346408163948e-05, + "loss": 0.00802530199289322, + "step": 6920, + "token_acc": 0.9965396769809307 + }, + { + "epoch": 0.36690204694810824, + "grad_norm": 0.07106967270374298, + "learning_rate": 1.4615141093829462e-05, + "loss": 0.008048564195632935, + "step": 6930, + "token_acc": 0.9965632585647434 + }, + { + "epoch": 0.36743148713129453, + "grad_norm": 0.1261410415172577, + "learning_rate": 1.4599922209294229e-05, + "loss": 0.008151649683713912, + "step": 6940, + "token_acc": 0.9965276328639516 + }, + { + "epoch": 0.3679609273144808, + "grad_norm": 0.08320154994726181, + "learning_rate": 1.458468979930733e-05, + "loss": 0.008468060195446015, + "step": 6950, + "token_acc": 0.9963103635377103 + }, + { + "epoch": 0.36849036749766717, + "grad_norm": 0.08258127421140671, + "learning_rate": 1.4569443908657612e-05, + "loss": 0.00821821540594101, + "step": 6960, + "token_acc": 0.9965705130879567 + }, + { + "epoch": 0.36901980768085346, + "grad_norm": 0.08468693494796753, + "learning_rate": 1.4554184582173564e-05, + "loss": 0.008022025972604752, + "step": 6970, + "token_acc": 0.9965547445255475 + }, + { + "epoch": 0.36954924786403975, + "grad_norm": 0.08238520473241806, + "learning_rate": 1.4538911864723182e-05, + "loss": 0.008397844433784486, + "step": 6980, + "token_acc": 0.9963444557578994 + }, + { + "epoch": 0.37007868804722605, + "grad_norm": 0.08201859146356583, + "learning_rate": 1.4523625801213833e-05, + "loss": 0.008848610520362853, + "step": 6990, + "token_acc": 0.9962962345050802 + }, + { + "epoch": 0.3706081282304124, + "grad_norm": 0.06624449044466019, + "learning_rate": 1.4508326436592129e-05, + "loss": 0.008092763274908066, + "step": 7000, + "token_acc": 0.9966216216216216 + }, + { + "epoch": 0.3711375684135987, + "grad_norm": 0.1265631765127182, + "learning_rate": 1.4493013815843787e-05, + "loss": 0.008300593495368958, + "step": 7010, + "token_acc": 0.9962472166857086 + }, + { + "epoch": 0.371667008596785, + "grad_norm": 0.10278423130512238, + "learning_rate": 1.4477687983993509e-05, + "loss": 0.008206693828105927, + "step": 7020, + "token_acc": 0.9965638318279247 + }, + { + "epoch": 0.37219644877997127, + "grad_norm": 0.06805432587862015, + "learning_rate": 1.4462348986104839e-05, + "loss": 0.008539116382598877, + "step": 7030, + "token_acc": 0.9965042257986467 + }, + { + "epoch": 0.37272588896315756, + "grad_norm": 0.08357007056474686, + "learning_rate": 1.4446996867280036e-05, + "loss": 0.008512961119413376, + "step": 7040, + "token_acc": 0.9965128016885382 + }, + { + "epoch": 0.3732553291463439, + "grad_norm": 0.07158810645341873, + "learning_rate": 1.4431631672659937e-05, + "loss": 0.008031125366687774, + "step": 7050, + "token_acc": 0.9964886527602859 + }, + { + "epoch": 0.3737847693295302, + "grad_norm": 0.0848868191242218, + "learning_rate": 1.4416253447423822e-05, + "loss": 0.008569315075874329, + "step": 7060, + "token_acc": 0.9961365475922264 + }, + { + "epoch": 0.3743142095127165, + "grad_norm": 0.10126155614852905, + "learning_rate": 1.4400862236789298e-05, + "loss": 0.008170506358146668, + "step": 7070, + "token_acc": 0.9965553479682392 + }, + { + "epoch": 0.3748436496959028, + "grad_norm": 0.08059631288051605, + "learning_rate": 1.4385458086012138e-05, + "loss": 0.008338314294815064, + "step": 7080, + "token_acc": 0.9963692513145814 + }, + { + "epoch": 0.3753730898790891, + "grad_norm": 0.0937328189611435, + "learning_rate": 1.4370041040386182e-05, + "loss": 0.008450078964233398, + "step": 7090, + "token_acc": 0.9964055776095007 + }, + { + "epoch": 0.3759025300622754, + "grad_norm": 0.06613848358392715, + "learning_rate": 1.4354611145243169e-05, + "loss": 0.008243384957313537, + "step": 7100, + "token_acc": 0.9964874809147568 + }, + { + "epoch": 0.3764319702454617, + "grad_norm": 0.0815696269273758, + "learning_rate": 1.4339168445952636e-05, + "loss": 0.007915447652339935, + "step": 7110, + "token_acc": 0.9966137601441237 + }, + { + "epoch": 0.376961410428648, + "grad_norm": 0.10713789612054825, + "learning_rate": 1.432371298792175e-05, + "loss": 0.008450544625520705, + "step": 7120, + "token_acc": 0.9963109794266161 + }, + { + "epoch": 0.3774908506118343, + "grad_norm": 0.07149320840835571, + "learning_rate": 1.4308244816595206e-05, + "loss": 0.007767795026302338, + "step": 7130, + "token_acc": 0.9965394461446094 + }, + { + "epoch": 0.37802029079502064, + "grad_norm": 0.08364174515008926, + "learning_rate": 1.4292763977455084e-05, + "loss": 0.008118488639593125, + "step": 7140, + "token_acc": 0.9964550246897104 + }, + { + "epoch": 0.37854973097820693, + "grad_norm": 0.09011761844158173, + "learning_rate": 1.42772705160207e-05, + "loss": 0.008179964125156402, + "step": 7150, + "token_acc": 0.9964158004851172 + }, + { + "epoch": 0.3790791711613932, + "grad_norm": 0.0949261337518692, + "learning_rate": 1.4261764477848497e-05, + "loss": 0.0076960533857345585, + "step": 7160, + "token_acc": 0.9966965037998549 + }, + { + "epoch": 0.3796086113445795, + "grad_norm": 0.10810185968875885, + "learning_rate": 1.4246245908531883e-05, + "loss": 0.008588281273841859, + "step": 7170, + "token_acc": 0.9963795620437956 + }, + { + "epoch": 0.3801380515277658, + "grad_norm": 0.08978131413459778, + "learning_rate": 1.423071485370113e-05, + "loss": 0.008015083521604538, + "step": 7180, + "token_acc": 0.9966467025908811 + }, + { + "epoch": 0.38066749171095215, + "grad_norm": 0.0956328883767128, + "learning_rate": 1.4215171359023207e-05, + "loss": 0.007883992046117783, + "step": 7190, + "token_acc": 0.9965448460620425 + }, + { + "epoch": 0.38119693189413845, + "grad_norm": 0.09252512454986572, + "learning_rate": 1.4199615470201667e-05, + "loss": 0.007978732883930206, + "step": 7200, + "token_acc": 0.9967143119235452 + }, + { + "epoch": 0.38172637207732474, + "grad_norm": 0.07298839092254639, + "learning_rate": 1.4184047232976506e-05, + "loss": 0.008083684742450714, + "step": 7210, + "token_acc": 0.9965619967122007 + }, + { + "epoch": 0.38225581226051103, + "grad_norm": 0.07473775744438171, + "learning_rate": 1.416846669312403e-05, + "loss": 0.00801447406411171, + "step": 7220, + "token_acc": 0.9966136754132817 + }, + { + "epoch": 0.3827852524436973, + "grad_norm": 0.06686414778232574, + "learning_rate": 1.4152873896456716e-05, + "loss": 0.008310705423355103, + "step": 7230, + "token_acc": 0.9964955318030488 + }, + { + "epoch": 0.38331469262688367, + "grad_norm": 0.09262493997812271, + "learning_rate": 1.4137268888823083e-05, + "loss": 0.008094839751720428, + "step": 7240, + "token_acc": 0.9966215652580123 + }, + { + "epoch": 0.38384413281006996, + "grad_norm": 0.08500389754772186, + "learning_rate": 1.4121651716107554e-05, + "loss": 0.008263372629880906, + "step": 7250, + "token_acc": 0.9965145754119138 + }, + { + "epoch": 0.38437357299325625, + "grad_norm": 0.07279597967863083, + "learning_rate": 1.4106022424230323e-05, + "loss": 0.008444715291261673, + "step": 7260, + "token_acc": 0.9962539629567829 + }, + { + "epoch": 0.38490301317644254, + "grad_norm": 0.07070887833833694, + "learning_rate": 1.4090381059147218e-05, + "loss": 0.0077133625745773315, + "step": 7270, + "token_acc": 0.9967804357254862 + }, + { + "epoch": 0.38543245335962883, + "grad_norm": 0.06875324249267578, + "learning_rate": 1.407472766684956e-05, + "loss": 0.007539429515600204, + "step": 7280, + "token_acc": 0.9968985268002301 + }, + { + "epoch": 0.3859618935428152, + "grad_norm": 0.08456079661846161, + "learning_rate": 1.4059062293364049e-05, + "loss": 0.0075863063335418705, + "step": 7290, + "token_acc": 0.9968069762982601 + }, + { + "epoch": 0.3864913337260015, + "grad_norm": 0.11147870123386383, + "learning_rate": 1.4043384984752601e-05, + "loss": 0.008118689060211182, + "step": 7300, + "token_acc": 0.9964868946986323 + }, + { + "epoch": 0.38702077390918777, + "grad_norm": 0.05997251719236374, + "learning_rate": 1.4027695787112231e-05, + "loss": 0.00764729231595993, + "step": 7310, + "token_acc": 0.9967971174056651 + }, + { + "epoch": 0.38755021409237406, + "grad_norm": 0.08413419872522354, + "learning_rate": 1.4011994746574915e-05, + "loss": 0.007385222613811493, + "step": 7320, + "token_acc": 0.9968388743389994 + }, + { + "epoch": 0.3880796542755604, + "grad_norm": 0.13552887737751007, + "learning_rate": 1.3996281909307445e-05, + "loss": 0.007742879539728164, + "step": 7330, + "token_acc": 0.9966222696679816 + }, + { + "epoch": 0.3886090944587467, + "grad_norm": 0.07952715456485748, + "learning_rate": 1.39805573215113e-05, + "loss": 0.00809568539261818, + "step": 7340, + "token_acc": 0.9965871731112632 + }, + { + "epoch": 0.389138534641933, + "grad_norm": 0.09367859363555908, + "learning_rate": 1.3964821029422512e-05, + "loss": 0.007555045187473297, + "step": 7350, + "token_acc": 0.9968778174775437 + }, + { + "epoch": 0.3896679748251193, + "grad_norm": 0.06426790356636047, + "learning_rate": 1.3949073079311533e-05, + "loss": 0.008002289384603501, + "step": 7360, + "token_acc": 0.9966116406002237 + }, + { + "epoch": 0.39019741500830557, + "grad_norm": 0.08233428001403809, + "learning_rate": 1.3933313517483086e-05, + "loss": 0.007886958122253419, + "step": 7370, + "token_acc": 0.9965963978543959 + }, + { + "epoch": 0.3907268551914919, + "grad_norm": 0.08105305582284927, + "learning_rate": 1.3917542390276041e-05, + "loss": 0.007922616600990296, + "step": 7380, + "token_acc": 0.9965973345787533 + }, + { + "epoch": 0.3912562953746782, + "grad_norm": 0.07301913946866989, + "learning_rate": 1.3901759744063274e-05, + "loss": 0.007748831808567047, + "step": 7390, + "token_acc": 0.996660878856685 + }, + { + "epoch": 0.3917857355578645, + "grad_norm": 0.06495894491672516, + "learning_rate": 1.3885965625251529e-05, + "loss": 0.007269527018070221, + "step": 7400, + "token_acc": 0.9969650986342944 + }, + { + "epoch": 0.3923151757410508, + "grad_norm": 0.06727399677038193, + "learning_rate": 1.3870160080281285e-05, + "loss": 0.007925257086753845, + "step": 7410, + "token_acc": 0.9966041734458044 + }, + { + "epoch": 0.3928446159242371, + "grad_norm": 0.17205610871315002, + "learning_rate": 1.385434315562662e-05, + "loss": 0.008063279092311859, + "step": 7420, + "token_acc": 0.9965369338095397 + }, + { + "epoch": 0.39337405610742343, + "grad_norm": 0.0811929702758789, + "learning_rate": 1.3838514897795067e-05, + "loss": 0.008019857108592987, + "step": 7430, + "token_acc": 0.9967133526305691 + }, + { + "epoch": 0.3939034962906097, + "grad_norm": 0.06627713888883591, + "learning_rate": 1.382267535332749e-05, + "loss": 0.007460875809192658, + "step": 7440, + "token_acc": 0.9968557130942453 + }, + { + "epoch": 0.394432936473796, + "grad_norm": 0.08638838678598404, + "learning_rate": 1.3806824568797932e-05, + "loss": 0.008147361129522324, + "step": 7450, + "token_acc": 0.9964785794155443 + }, + { + "epoch": 0.3949623766569823, + "grad_norm": 0.06839636713266373, + "learning_rate": 1.3790962590813492e-05, + "loss": 0.008132056891918182, + "step": 7460, + "token_acc": 0.9964954733197047 + }, + { + "epoch": 0.39549181684016865, + "grad_norm": 0.08983190357685089, + "learning_rate": 1.377508946601418e-05, + "loss": 0.00815693810582161, + "step": 7470, + "token_acc": 0.996411223689262 + }, + { + "epoch": 0.39602125702335494, + "grad_norm": 0.06986068934202194, + "learning_rate": 1.3759205241072782e-05, + "loss": 0.007769600301980972, + "step": 7480, + "token_acc": 0.996563086548488 + }, + { + "epoch": 0.39655069720654124, + "grad_norm": 0.0918274000287056, + "learning_rate": 1.374330996269472e-05, + "loss": 0.007907094806432724, + "step": 7490, + "token_acc": 0.996542479893769 + }, + { + "epoch": 0.3970801373897275, + "grad_norm": 0.08254879713058472, + "learning_rate": 1.3727403677617921e-05, + "loss": 0.007771669328212738, + "step": 7500, + "token_acc": 0.9965883402984577 + }, + { + "epoch": 0.3976095775729138, + "grad_norm": 0.09408901631832123, + "learning_rate": 1.3711486432612671e-05, + "loss": 0.008191441744565963, + "step": 7510, + "token_acc": 0.9964803709789073 + }, + { + "epoch": 0.39813901775610017, + "grad_norm": 0.09008632600307465, + "learning_rate": 1.3695558274481486e-05, + "loss": 0.007587350159883499, + "step": 7520, + "token_acc": 0.996837502711904 + }, + { + "epoch": 0.39866845793928646, + "grad_norm": 0.08176735043525696, + "learning_rate": 1.367961925005897e-05, + "loss": 0.008373795449733734, + "step": 7530, + "token_acc": 0.9964034480168896 + }, + { + "epoch": 0.39919789812247275, + "grad_norm": 0.0642644464969635, + "learning_rate": 1.3663669406211681e-05, + "loss": 0.0078595831990242, + "step": 7540, + "token_acc": 0.9966452754295633 + }, + { + "epoch": 0.39972733830565904, + "grad_norm": 0.08710912615060806, + "learning_rate": 1.3647708789837981e-05, + "loss": 0.007826000452041626, + "step": 7550, + "token_acc": 0.9966428660904422 + }, + { + "epoch": 0.40025677848884533, + "grad_norm": 0.06794148683547974, + "learning_rate": 1.3631737447867917e-05, + "loss": 0.0074563220143318175, + "step": 7560, + "token_acc": 0.9967275246896575 + }, + { + "epoch": 0.4007862186720317, + "grad_norm": 0.07518324255943298, + "learning_rate": 1.3615755427263063e-05, + "loss": 0.00782143771648407, + "step": 7570, + "token_acc": 0.9968456627863312 + }, + { + "epoch": 0.40131565885521797, + "grad_norm": 0.07756779342889786, + "learning_rate": 1.3599762775016401e-05, + "loss": 0.00792180746793747, + "step": 7580, + "token_acc": 0.9965213764337852 + }, + { + "epoch": 0.40184509903840426, + "grad_norm": 0.06189802289009094, + "learning_rate": 1.3583759538152174e-05, + "loss": 0.007234364748001099, + "step": 7590, + "token_acc": 0.9970521674502502 + }, + { + "epoch": 0.40237453922159055, + "grad_norm": 0.0644647628068924, + "learning_rate": 1.3567745763725744e-05, + "loss": 0.007789485156536102, + "step": 7600, + "token_acc": 0.9968148086383724 + }, + { + "epoch": 0.40290397940477685, + "grad_norm": 0.06582963466644287, + "learning_rate": 1.355172149882345e-05, + "loss": 0.008021900057792663, + "step": 7610, + "token_acc": 0.9964985410587744 + }, + { + "epoch": 0.4034334195879632, + "grad_norm": 0.06564021110534668, + "learning_rate": 1.3535686790562492e-05, + "loss": 0.007582755386829376, + "step": 7620, + "token_acc": 0.9965886533104242 + }, + { + "epoch": 0.4039628597711495, + "grad_norm": 0.07986405491828918, + "learning_rate": 1.3519641686090767e-05, + "loss": 0.008347317576408386, + "step": 7630, + "token_acc": 0.996397418149674 + }, + { + "epoch": 0.4044922999543358, + "grad_norm": 0.08133228868246078, + "learning_rate": 1.3503586232586745e-05, + "loss": 0.007617518305778503, + "step": 7640, + "token_acc": 0.9966133081972957 + }, + { + "epoch": 0.40502174013752207, + "grad_norm": 0.06672562658786774, + "learning_rate": 1.3487520477259324e-05, + "loss": 0.007324489951133728, + "step": 7650, + "token_acc": 0.9970115863634846 + }, + { + "epoch": 0.4055511803207084, + "grad_norm": 0.09303580969572067, + "learning_rate": 1.3471444467347694e-05, + "loss": 0.0074172839522361755, + "step": 7660, + "token_acc": 0.9967712061672465 + }, + { + "epoch": 0.4060806205038947, + "grad_norm": 0.1255025714635849, + "learning_rate": 1.3455358250121193e-05, + "loss": 0.007051350176334381, + "step": 7670, + "token_acc": 0.9967706673119769 + }, + { + "epoch": 0.406610060687081, + "grad_norm": 0.07746310532093048, + "learning_rate": 1.3439261872879179e-05, + "loss": 0.00781775414943695, + "step": 7680, + "token_acc": 0.9965317726310173 + }, + { + "epoch": 0.4071395008702673, + "grad_norm": 0.07184544205665588, + "learning_rate": 1.3423155382950882e-05, + "loss": 0.007733357697725296, + "step": 7690, + "token_acc": 0.9966031247913467 + }, + { + "epoch": 0.4076689410534536, + "grad_norm": 0.09585249423980713, + "learning_rate": 1.340703882769527e-05, + "loss": 0.007455173134803772, + "step": 7700, + "token_acc": 0.9967613769156288 + }, + { + "epoch": 0.40819838123663993, + "grad_norm": 0.06763704121112823, + "learning_rate": 1.3390912254500892e-05, + "loss": 0.007343024015426636, + "step": 7710, + "token_acc": 0.9967519997328124 + }, + { + "epoch": 0.4087278214198262, + "grad_norm": 0.08368167281150818, + "learning_rate": 1.3374775710785773e-05, + "loss": 0.007527545094490051, + "step": 7720, + "token_acc": 0.9967286443902927 + }, + { + "epoch": 0.4092572616030125, + "grad_norm": 0.0867418721318245, + "learning_rate": 1.3358629243997245e-05, + "loss": 0.008240339905023574, + "step": 7730, + "token_acc": 0.9964122716994985 + }, + { + "epoch": 0.4097867017861988, + "grad_norm": 0.09582404792308807, + "learning_rate": 1.3342472901611813e-05, + "loss": 0.007841107994318008, + "step": 7740, + "token_acc": 0.9967479674796748 + }, + { + "epoch": 0.4103161419693851, + "grad_norm": 0.07575418055057526, + "learning_rate": 1.3326306731135035e-05, + "loss": 0.007536052167415619, + "step": 7750, + "token_acc": 0.9968868672536828 + }, + { + "epoch": 0.41084558215257144, + "grad_norm": 0.0782008096575737, + "learning_rate": 1.3310130780101353e-05, + "loss": 0.0075279176235198975, + "step": 7760, + "token_acc": 0.9968444251511003 + }, + { + "epoch": 0.41137502233575773, + "grad_norm": 0.07469892501831055, + "learning_rate": 1.329394509607397e-05, + "loss": 0.007520338892936707, + "step": 7770, + "token_acc": 0.9969688366344901 + }, + { + "epoch": 0.411904462518944, + "grad_norm": 0.08782609552145004, + "learning_rate": 1.327774972664471e-05, + "loss": 0.007304920256137848, + "step": 7780, + "token_acc": 0.9968532198155335 + }, + { + "epoch": 0.4124339027021303, + "grad_norm": 0.06542053073644638, + "learning_rate": 1.3261544719433878e-05, + "loss": 0.007968396693468095, + "step": 7790, + "token_acc": 0.9965279516584039 + }, + { + "epoch": 0.41296334288531666, + "grad_norm": 0.07319057732820511, + "learning_rate": 1.3245330122090117e-05, + "loss": 0.0075017563998699185, + "step": 7800, + "token_acc": 0.9969399581432967 + }, + { + "epoch": 0.41349278306850296, + "grad_norm": 0.09119752049446106, + "learning_rate": 1.3229105982290262e-05, + "loss": 0.007177518308162689, + "step": 7810, + "token_acc": 0.9968862175473746 + }, + { + "epoch": 0.41402222325168925, + "grad_norm": 0.08031339198350906, + "learning_rate": 1.3212872347739212e-05, + "loss": 0.007702425122261047, + "step": 7820, + "token_acc": 0.9967446558099546 + }, + { + "epoch": 0.41455166343487554, + "grad_norm": 0.08865433931350708, + "learning_rate": 1.3196629266169783e-05, + "loss": 0.0076177999377250675, + "step": 7830, + "token_acc": 0.9966637752719023 + }, + { + "epoch": 0.41508110361806183, + "grad_norm": 0.07217568159103394, + "learning_rate": 1.3180376785342568e-05, + "loss": 0.0075017943978309635, + "step": 7840, + "token_acc": 0.9966446314102564 + }, + { + "epoch": 0.4156105438012482, + "grad_norm": 0.09430079162120819, + "learning_rate": 1.3164114953045798e-05, + "loss": 0.007469096779823303, + "step": 7850, + "token_acc": 0.9967972776860331 + }, + { + "epoch": 0.41613998398443447, + "grad_norm": 0.0758441910147667, + "learning_rate": 1.3147843817095205e-05, + "loss": 0.007645830512046814, + "step": 7860, + "token_acc": 0.996662578325115 + }, + { + "epoch": 0.41666942416762076, + "grad_norm": 0.08946306258440018, + "learning_rate": 1.3131563425333866e-05, + "loss": 0.007801616936922074, + "step": 7870, + "token_acc": 0.9966953183676875 + }, + { + "epoch": 0.41719886435080705, + "grad_norm": 0.07247814536094666, + "learning_rate": 1.3115273825632083e-05, + "loss": 0.007302309572696686, + "step": 7880, + "token_acc": 0.9969370467121241 + }, + { + "epoch": 0.41772830453399334, + "grad_norm": 0.08185109496116638, + "learning_rate": 1.3098975065887228e-05, + "loss": 0.007757198065519333, + "step": 7890, + "token_acc": 0.9967066866766716 + }, + { + "epoch": 0.4182577447171797, + "grad_norm": 0.08948971331119537, + "learning_rate": 1.308266719402361e-05, + "loss": 0.008090752363204955, + "step": 7900, + "token_acc": 0.9967158456280737 + }, + { + "epoch": 0.418787184900366, + "grad_norm": 0.0731845274567604, + "learning_rate": 1.3066350257992332e-05, + "loss": 0.00762658566236496, + "step": 7910, + "token_acc": 0.9967633220995028 + }, + { + "epoch": 0.4193166250835523, + "grad_norm": 0.09836041927337646, + "learning_rate": 1.3050024305771141e-05, + "loss": 0.006870850920677185, + "step": 7920, + "token_acc": 0.9972211327430675 + }, + { + "epoch": 0.41984606526673857, + "grad_norm": 0.08738066256046295, + "learning_rate": 1.3033689385364306e-05, + "loss": 0.007510527223348618, + "step": 7930, + "token_acc": 0.9969322851974425 + }, + { + "epoch": 0.4203755054499249, + "grad_norm": 0.08786273747682571, + "learning_rate": 1.3017345544802455e-05, + "loss": 0.006907186657190323, + "step": 7940, + "token_acc": 0.997229200223667 + }, + { + "epoch": 0.4209049456331112, + "grad_norm": 0.07225662469863892, + "learning_rate": 1.3000992832142453e-05, + "loss": 0.007495681941509247, + "step": 7950, + "token_acc": 0.9969217164975974 + }, + { + "epoch": 0.4214343858162975, + "grad_norm": 0.08197043091058731, + "learning_rate": 1.298463129546725e-05, + "loss": 0.007363441586494446, + "step": 7960, + "token_acc": 0.9968947728676606 + }, + { + "epoch": 0.4219638259994838, + "grad_norm": 0.08194158971309662, + "learning_rate": 1.296826098288574e-05, + "loss": 0.007524023205041886, + "step": 7970, + "token_acc": 0.9968457680721635 + }, + { + "epoch": 0.4224932661826701, + "grad_norm": 0.09474582970142365, + "learning_rate": 1.295188194253262e-05, + "loss": 0.007502316683530808, + "step": 7980, + "token_acc": 0.9967624848762986 + }, + { + "epoch": 0.4230227063658564, + "grad_norm": 0.08394216746091843, + "learning_rate": 1.2935494222568255e-05, + "loss": 0.007041873037815094, + "step": 7990, + "token_acc": 0.9969458259621483 + }, + { + "epoch": 0.4235521465490427, + "grad_norm": 0.0847509428858757, + "learning_rate": 1.2919097871178526e-05, + "loss": 0.0073317036032676695, + "step": 8000, + "token_acc": 0.9968964567881665 + }, + { + "epoch": 0.424081586732229, + "grad_norm": 0.07296353578567505, + "learning_rate": 1.2902692936574697e-05, + "loss": 0.006668738275766373, + "step": 8010, + "token_acc": 0.9971054387721054 + }, + { + "epoch": 0.4246110269154153, + "grad_norm": 0.10959818959236145, + "learning_rate": 1.2886279466993265e-05, + "loss": 0.007273763418197632, + "step": 8020, + "token_acc": 0.9968791722296395 + }, + { + "epoch": 0.4251404670986016, + "grad_norm": 0.05003484711050987, + "learning_rate": 1.2869857510695833e-05, + "loss": 0.006755241751670837, + "step": 8030, + "token_acc": 0.9971214497880719 + }, + { + "epoch": 0.42566990728178794, + "grad_norm": 0.10306958854198456, + "learning_rate": 1.2853427115968946e-05, + "loss": 0.0070476517081260685, + "step": 8040, + "token_acc": 0.9969209202192906 + }, + { + "epoch": 0.42619934746497423, + "grad_norm": 0.08021682500839233, + "learning_rate": 1.283698833112397e-05, + "loss": 0.0077625267207622525, + "step": 8050, + "token_acc": 0.9966364250957742 + }, + { + "epoch": 0.4267287876481605, + "grad_norm": 0.07751452177762985, + "learning_rate": 1.2820541204496934e-05, + "loss": 0.007402804493904114, + "step": 8060, + "token_acc": 0.996746529631607 + }, + { + "epoch": 0.4272582278313468, + "grad_norm": 0.07828596979379654, + "learning_rate": 1.28040857844484e-05, + "loss": 0.007048720121383667, + "step": 8070, + "token_acc": 0.9970560517734578 + }, + { + "epoch": 0.4277876680145331, + "grad_norm": 0.06443706899881363, + "learning_rate": 1.2787622119363318e-05, + "loss": 0.007646536827087403, + "step": 8080, + "token_acc": 0.9968956280094466 + }, + { + "epoch": 0.42831710819771945, + "grad_norm": 0.06105376034975052, + "learning_rate": 1.2771150257650872e-05, + "loss": 0.007174349576234818, + "step": 8090, + "token_acc": 0.9969384775808133 + }, + { + "epoch": 0.42884654838090575, + "grad_norm": 0.06430378556251526, + "learning_rate": 1.2754670247744353e-05, + "loss": 0.007035088539123535, + "step": 8100, + "token_acc": 0.9971055836377892 + }, + { + "epoch": 0.42937598856409204, + "grad_norm": 0.09101198613643646, + "learning_rate": 1.2738182138101016e-05, + "loss": 0.007310198247432708, + "step": 8110, + "token_acc": 0.997004214127759 + }, + { + "epoch": 0.42990542874727833, + "grad_norm": 0.07724619656801224, + "learning_rate": 1.2721685977201922e-05, + "loss": 0.007639773935079574, + "step": 8120, + "token_acc": 0.9965376561183371 + }, + { + "epoch": 0.4304348689304647, + "grad_norm": 0.11012060195207596, + "learning_rate": 1.2705181813551813e-05, + "loss": 0.0071842528879642485, + "step": 8130, + "token_acc": 0.9969633009919328 + }, + { + "epoch": 0.43096430911365097, + "grad_norm": 0.061232708394527435, + "learning_rate": 1.2688669695678955e-05, + "loss": 0.006945900619029999, + "step": 8140, + "token_acc": 0.9969953427813111 + }, + { + "epoch": 0.43149374929683726, + "grad_norm": 0.0658828541636467, + "learning_rate": 1.267214967213501e-05, + "loss": 0.006726765632629394, + "step": 8150, + "token_acc": 0.9969313642920874 + }, + { + "epoch": 0.43202318948002355, + "grad_norm": 0.08629443496465683, + "learning_rate": 1.265562179149488e-05, + "loss": 0.007222894579172134, + "step": 8160, + "token_acc": 0.9969039472586164 + }, + { + "epoch": 0.43255262966320984, + "grad_norm": 0.06964782625436783, + "learning_rate": 1.263908610235657e-05, + "loss": 0.006823251396417618, + "step": 8170, + "token_acc": 0.9970120353214149 + }, + { + "epoch": 0.4330820698463962, + "grad_norm": 0.11429233849048615, + "learning_rate": 1.2622542653341054e-05, + "loss": 0.007249373197555542, + "step": 8180, + "token_acc": 0.9969057031810372 + }, + { + "epoch": 0.4336115100295825, + "grad_norm": 0.10671359300613403, + "learning_rate": 1.2605991493092109e-05, + "loss": 0.006735020130872726, + "step": 8190, + "token_acc": 0.9971627891451651 + }, + { + "epoch": 0.4341409502127688, + "grad_norm": 0.09352430701255798, + "learning_rate": 1.2589432670276191e-05, + "loss": 0.0076766788959503176, + "step": 8200, + "token_acc": 0.9967223779023219 + }, + { + "epoch": 0.43467039039595506, + "grad_norm": 0.08164774626493454, + "learning_rate": 1.2572866233582295e-05, + "loss": 0.006597428023815155, + "step": 8210, + "token_acc": 0.9971966827412438 + }, + { + "epoch": 0.43519983057914136, + "grad_norm": 0.08605318516492844, + "learning_rate": 1.2556292231721792e-05, + "loss": 0.006686335802078247, + "step": 8220, + "token_acc": 0.9970440388116034 + }, + { + "epoch": 0.4357292707623277, + "grad_norm": 0.0780930444598198, + "learning_rate": 1.2539710713428303e-05, + "loss": 0.0073903828859329225, + "step": 8230, + "token_acc": 0.9969459024190789 + }, + { + "epoch": 0.436258710945514, + "grad_norm": 0.07365619391202927, + "learning_rate": 1.2523121727457548e-05, + "loss": 0.007713119685649872, + "step": 8240, + "token_acc": 0.9965202152960321 + }, + { + "epoch": 0.4367881511287003, + "grad_norm": 0.07649888843297958, + "learning_rate": 1.2506525322587207e-05, + "loss": 0.007298261672258377, + "step": 8250, + "token_acc": 0.9969217935349323 + }, + { + "epoch": 0.4373175913118866, + "grad_norm": 0.07456541806459427, + "learning_rate": 1.2489921547616773e-05, + "loss": 0.007129441946744919, + "step": 8260, + "token_acc": 0.9968946951041362 + }, + { + "epoch": 0.4378470314950729, + "grad_norm": 0.07608219236135483, + "learning_rate": 1.2473310451367409e-05, + "loss": 0.006800711154937744, + "step": 8270, + "token_acc": 0.9972210167907334 + }, + { + "epoch": 0.4383764716782592, + "grad_norm": 0.09409811347723007, + "learning_rate": 1.2456692082681807e-05, + "loss": 0.007800394296646118, + "step": 8280, + "token_acc": 0.9968075352171376 + }, + { + "epoch": 0.4389059118614455, + "grad_norm": 0.0708824098110199, + "learning_rate": 1.2440066490424045e-05, + "loss": 0.006974401324987412, + "step": 8290, + "token_acc": 0.9969633516588943 + }, + { + "epoch": 0.4394353520446318, + "grad_norm": 0.09019751101732254, + "learning_rate": 1.242343372347943e-05, + "loss": 0.007305324077606201, + "step": 8300, + "token_acc": 0.9969604863221885 + }, + { + "epoch": 0.4399647922278181, + "grad_norm": 0.09046681970357895, + "learning_rate": 1.2406793830754377e-05, + "loss": 0.006630456447601319, + "step": 8310, + "token_acc": 0.9972630630329934 + }, + { + "epoch": 0.44049423241100444, + "grad_norm": 0.13311974704265594, + "learning_rate": 1.2390146861176249e-05, + "loss": 0.007000093907117843, + "step": 8320, + "token_acc": 0.9969693933777488 + }, + { + "epoch": 0.44102367259419073, + "grad_norm": 0.10188629478216171, + "learning_rate": 1.2373492863693217e-05, + "loss": 0.006829223036766053, + "step": 8330, + "token_acc": 0.9970721453417082 + }, + { + "epoch": 0.441553112777377, + "grad_norm": 0.08324191719293594, + "learning_rate": 1.2356831887274124e-05, + "loss": 0.007169118523597718, + "step": 8340, + "token_acc": 0.9970722674496196 + }, + { + "epoch": 0.4420825529605633, + "grad_norm": 0.0850115641951561, + "learning_rate": 1.2340163980908315e-05, + "loss": 0.00763884037733078, + "step": 8350, + "token_acc": 0.9967055879899916 + }, + { + "epoch": 0.4426119931437496, + "grad_norm": 0.08366356790065765, + "learning_rate": 1.2323489193605531e-05, + "loss": 0.0069373130798339845, + "step": 8360, + "token_acc": 0.996947201601468 + }, + { + "epoch": 0.44314143332693595, + "grad_norm": 0.08251821249723434, + "learning_rate": 1.2306807574395738e-05, + "loss": 0.007184948027133942, + "step": 8370, + "token_acc": 0.9970047390201575 + }, + { + "epoch": 0.44367087351012224, + "grad_norm": 0.10406091064214706, + "learning_rate": 1.2290119172328988e-05, + "loss": 0.007324790954589844, + "step": 8380, + "token_acc": 0.9969735119767219 + }, + { + "epoch": 0.44420031369330853, + "grad_norm": 0.08435743302106857, + "learning_rate": 1.2273424036475286e-05, + "loss": 0.007114642858505249, + "step": 8390, + "token_acc": 0.9968053783082684 + }, + { + "epoch": 0.4447297538764948, + "grad_norm": 0.09049350023269653, + "learning_rate": 1.2256722215924418e-05, + "loss": 0.006696093082427979, + "step": 8400, + "token_acc": 0.9973641186480264 + }, + { + "epoch": 0.4452591940596811, + "grad_norm": 0.08535543829202652, + "learning_rate": 1.2240013759785849e-05, + "loss": 0.00725688487291336, + "step": 8410, + "token_acc": 0.9969633769917411 + }, + { + "epoch": 0.44578863424286747, + "grad_norm": 0.09684964269399643, + "learning_rate": 1.222329871718854e-05, + "loss": 0.0070689678192138675, + "step": 8420, + "token_acc": 0.9970549219512602 + }, + { + "epoch": 0.44631807442605376, + "grad_norm": 0.06480246782302856, + "learning_rate": 1.2206577137280817e-05, + "loss": 0.006816080212593079, + "step": 8430, + "token_acc": 0.9972260053641704 + }, + { + "epoch": 0.44684751460924005, + "grad_norm": 0.06663979589939117, + "learning_rate": 1.2189849069230241e-05, + "loss": 0.007254540920257568, + "step": 8440, + "token_acc": 0.9968616714937943 + }, + { + "epoch": 0.44737695479242634, + "grad_norm": 0.0791064128279686, + "learning_rate": 1.217311456222344e-05, + "loss": 0.007144643366336823, + "step": 8450, + "token_acc": 0.9968695216629101 + }, + { + "epoch": 0.4479063949756127, + "grad_norm": 0.08421915769577026, + "learning_rate": 1.2156373665465977e-05, + "loss": 0.006950680911540985, + "step": 8460, + "token_acc": 0.9969700510834363 + }, + { + "epoch": 0.448435835158799, + "grad_norm": 0.0712527334690094, + "learning_rate": 1.2139626428182205e-05, + "loss": 0.00698014497756958, + "step": 8470, + "token_acc": 0.9969459533882394 + }, + { + "epoch": 0.44896527534198527, + "grad_norm": 0.10700621455907822, + "learning_rate": 1.2122872899615117e-05, + "loss": 0.007061784714460373, + "step": 8480, + "token_acc": 0.9970428782651553 + }, + { + "epoch": 0.44949471552517156, + "grad_norm": 0.0860348492860794, + "learning_rate": 1.2106113129026211e-05, + "loss": 0.007023942470550537, + "step": 8490, + "token_acc": 0.9969471252095723 + }, + { + "epoch": 0.45002415570835785, + "grad_norm": 0.0944226011633873, + "learning_rate": 1.2089347165695335e-05, + "loss": 0.007027996331453323, + "step": 8500, + "token_acc": 0.997055241336737 + }, + { + "epoch": 0.4505535958915442, + "grad_norm": 0.08265798538923264, + "learning_rate": 1.2072575058920542e-05, + "loss": 0.006468549370765686, + "step": 8510, + "token_acc": 0.9971304159228549 + }, + { + "epoch": 0.4510830360747305, + "grad_norm": 0.08250034600496292, + "learning_rate": 1.2055796858017955e-05, + "loss": 0.006801193207502365, + "step": 8520, + "token_acc": 0.9969285726208957 + }, + { + "epoch": 0.4516124762579168, + "grad_norm": 0.09428032487630844, + "learning_rate": 1.2039012612321612e-05, + "loss": 0.007062745839357376, + "step": 8530, + "token_acc": 0.997062774843334 + }, + { + "epoch": 0.4521419164411031, + "grad_norm": 0.0884341150522232, + "learning_rate": 1.2022222371183332e-05, + "loss": 0.006854654103517532, + "step": 8540, + "token_acc": 0.9968808900305239 + }, + { + "epoch": 0.45267135662428937, + "grad_norm": 0.0797189325094223, + "learning_rate": 1.2005426183972552e-05, + "loss": 0.007164597511291504, + "step": 8550, + "token_acc": 0.9968306129427759 + }, + { + "epoch": 0.4532007968074757, + "grad_norm": 0.08434195071458817, + "learning_rate": 1.1988624100076202e-05, + "loss": 0.006730017066001892, + "step": 8560, + "token_acc": 0.9971064525274762 + }, + { + "epoch": 0.453730236990662, + "grad_norm": 0.08552851527929306, + "learning_rate": 1.1971816168898544e-05, + "loss": 0.00668661817908287, + "step": 8570, + "token_acc": 0.9972564814583177 + }, + { + "epoch": 0.4542596771738483, + "grad_norm": 0.09458556026220322, + "learning_rate": 1.1955002439861034e-05, + "loss": 0.006946419924497604, + "step": 8580, + "token_acc": 0.9971474088964143 + }, + { + "epoch": 0.4547891173570346, + "grad_norm": 0.08387396484613419, + "learning_rate": 1.193818296240218e-05, + "loss": 0.006616406142711639, + "step": 8590, + "token_acc": 0.9972661654636683 + }, + { + "epoch": 0.45531855754022094, + "grad_norm": 0.09891533851623535, + "learning_rate": 1.1921357785977393e-05, + "loss": 0.00710974931716919, + "step": 8600, + "token_acc": 0.9968519848360861 + }, + { + "epoch": 0.45584799772340723, + "grad_norm": 0.06725495308637619, + "learning_rate": 1.1904526960058834e-05, + "loss": 0.006500191241502762, + "step": 8610, + "token_acc": 0.997363370880267 + }, + { + "epoch": 0.4563774379065935, + "grad_norm": 0.09898929297924042, + "learning_rate": 1.1887690534135277e-05, + "loss": 0.007061667740345001, + "step": 8620, + "token_acc": 0.9970701657735264 + }, + { + "epoch": 0.4569068780897798, + "grad_norm": 0.07672589272260666, + "learning_rate": 1.1870848557711969e-05, + "loss": 0.0071324102580547334, + "step": 8630, + "token_acc": 0.9970624561872017 + }, + { + "epoch": 0.4574363182729661, + "grad_norm": 0.06281300634145737, + "learning_rate": 1.1854001080310473e-05, + "loss": 0.006548981368541718, + "step": 8640, + "token_acc": 0.9973396714202318 + }, + { + "epoch": 0.45796575845615245, + "grad_norm": 0.07197356969118118, + "learning_rate": 1.1837148151468526e-05, + "loss": 0.006644316762685776, + "step": 8650, + "token_acc": 0.997188011081072 + }, + { + "epoch": 0.45849519863933874, + "grad_norm": 0.07677491754293442, + "learning_rate": 1.1820289820739897e-05, + "loss": 0.00696353018283844, + "step": 8660, + "token_acc": 0.996953204561011 + }, + { + "epoch": 0.45902463882252503, + "grad_norm": 0.10775228589773178, + "learning_rate": 1.1803426137694232e-05, + "loss": 0.006690756976604461, + "step": 8670, + "token_acc": 0.9970042391268067 + }, + { + "epoch": 0.4595540790057113, + "grad_norm": 0.08301912248134613, + "learning_rate": 1.1786557151916922e-05, + "loss": 0.006932760775089264, + "step": 8680, + "token_acc": 0.9972217818973644 + }, + { + "epoch": 0.4600835191888976, + "grad_norm": 0.07021474838256836, + "learning_rate": 1.1769682913008951e-05, + "loss": 0.006398163735866547, + "step": 8690, + "token_acc": 0.9972632913927176 + }, + { + "epoch": 0.46061295937208396, + "grad_norm": 0.09164784103631973, + "learning_rate": 1.1752803470586742e-05, + "loss": 0.00654696524143219, + "step": 8700, + "token_acc": 0.9971455041690663 + }, + { + "epoch": 0.46114239955527025, + "grad_norm": 0.08428556472063065, + "learning_rate": 1.1735918874282022e-05, + "loss": 0.006864959001541137, + "step": 8710, + "token_acc": 0.997087589291675 + }, + { + "epoch": 0.46167183973845655, + "grad_norm": 0.059441205114126205, + "learning_rate": 1.1719029173741675e-05, + "loss": 0.00677691251039505, + "step": 8720, + "token_acc": 0.9969222556779795 + }, + { + "epoch": 0.46220127992164284, + "grad_norm": 0.08180636167526245, + "learning_rate": 1.1702134418627586e-05, + "loss": 0.006769742071628571, + "step": 8730, + "token_acc": 0.9970880995886627 + }, + { + "epoch": 0.46273072010482913, + "grad_norm": 0.0650036558508873, + "learning_rate": 1.1685234658616508e-05, + "loss": 0.006812536716461181, + "step": 8740, + "token_acc": 0.9970022879473605 + }, + { + "epoch": 0.4632601602880155, + "grad_norm": 0.06636828184127808, + "learning_rate": 1.1668329943399912e-05, + "loss": 0.00685785710811615, + "step": 8750, + "token_acc": 0.9971378504672898 + }, + { + "epoch": 0.46378960047120177, + "grad_norm": 0.06768668442964554, + "learning_rate": 1.1651420322683833e-05, + "loss": 0.007222312688827515, + "step": 8760, + "token_acc": 0.9968721849664698 + }, + { + "epoch": 0.46431904065438806, + "grad_norm": 0.06543266028165817, + "learning_rate": 1.1634505846188737e-05, + "loss": 0.006589705497026444, + "step": 8770, + "token_acc": 0.9972818156804216 + }, + { + "epoch": 0.46484848083757435, + "grad_norm": 0.07365161925554276, + "learning_rate": 1.1617586563649357e-05, + "loss": 0.007308214157819748, + "step": 8780, + "token_acc": 0.996862588031107 + }, + { + "epoch": 0.4653779210207607, + "grad_norm": 0.06949863582849503, + "learning_rate": 1.160066252481457e-05, + "loss": 0.006833705306053162, + "step": 8790, + "token_acc": 0.9970394215613507 + }, + { + "epoch": 0.465907361203947, + "grad_norm": 0.06287001073360443, + "learning_rate": 1.158373377944723e-05, + "loss": 0.006444665789604187, + "step": 8800, + "token_acc": 0.9973618300217064 + }, + { + "epoch": 0.4664368013871333, + "grad_norm": 0.09748999029397964, + "learning_rate": 1.1566800377324034e-05, + "loss": 0.006535249948501587, + "step": 8810, + "token_acc": 0.9972309334756789 + }, + { + "epoch": 0.4669662415703196, + "grad_norm": 0.08026882261037827, + "learning_rate": 1.1549862368235372e-05, + "loss": 0.007122346758842468, + "step": 8820, + "token_acc": 0.9971970335271496 + }, + { + "epoch": 0.46749568175350587, + "grad_norm": 0.10175097733736038, + "learning_rate": 1.1532919801985173e-05, + "loss": 0.00696885734796524, + "step": 8830, + "token_acc": 0.9969380688976214 + }, + { + "epoch": 0.4680251219366922, + "grad_norm": 0.08993886411190033, + "learning_rate": 1.1515972728390777e-05, + "loss": 0.006513018906116485, + "step": 8840, + "token_acc": 0.9973452657238743 + }, + { + "epoch": 0.4685545621198785, + "grad_norm": 0.05891492962837219, + "learning_rate": 1.1499021197282764e-05, + "loss": 0.006062068790197372, + "step": 8850, + "token_acc": 0.9973967893735607 + }, + { + "epoch": 0.4690840023030648, + "grad_norm": 0.07915236800909042, + "learning_rate": 1.1482065258504834e-05, + "loss": 0.006537307798862457, + "step": 8860, + "token_acc": 0.9972968688731113 + }, + { + "epoch": 0.4696134424862511, + "grad_norm": 0.09380991756916046, + "learning_rate": 1.1465104961913638e-05, + "loss": 0.006617280095815659, + "step": 8870, + "token_acc": 0.9971458138603285 + }, + { + "epoch": 0.4701428826694374, + "grad_norm": 0.09571976214647293, + "learning_rate": 1.1448140357378646e-05, + "loss": 0.006998862326145172, + "step": 8880, + "token_acc": 0.9970116611991753 + }, + { + "epoch": 0.4706723228526237, + "grad_norm": 0.06741785258054733, + "learning_rate": 1.143117149478199e-05, + "loss": 0.006386888027191162, + "step": 8890, + "token_acc": 0.997356791461686 + }, + { + "epoch": 0.47120176303581, + "grad_norm": 0.0710383802652359, + "learning_rate": 1.1414198424018325e-05, + "loss": 0.00645061582326889, + "step": 8900, + "token_acc": 0.9971465520837679 + }, + { + "epoch": 0.4717312032189963, + "grad_norm": 0.09290769696235657, + "learning_rate": 1.1397221194994681e-05, + "loss": 0.006536872684955597, + "step": 8910, + "token_acc": 0.997336784104191 + }, + { + "epoch": 0.4722606434021826, + "grad_norm": 0.08055935800075531, + "learning_rate": 1.1380239857630314e-05, + "loss": 0.006734628975391388, + "step": 8920, + "token_acc": 0.99707990221844 + }, + { + "epoch": 0.47279008358536895, + "grad_norm": 0.07397284358739853, + "learning_rate": 1.1363254461856559e-05, + "loss": 0.006785006076097488, + "step": 8930, + "token_acc": 0.9971120459401709 + }, + { + "epoch": 0.47331952376855524, + "grad_norm": 0.0891742929816246, + "learning_rate": 1.1346265057616681e-05, + "loss": 0.006706812977790832, + "step": 8940, + "token_acc": 0.9973735554555005 + }, + { + "epoch": 0.47384896395174153, + "grad_norm": 0.09550976008176804, + "learning_rate": 1.1329271694865737e-05, + "loss": 0.00651034414768219, + "step": 8950, + "token_acc": 0.9970796585703677 + }, + { + "epoch": 0.4743784041349278, + "grad_norm": 0.09056397527456284, + "learning_rate": 1.1312274423570422e-05, + "loss": 0.006887002289295197, + "step": 8960, + "token_acc": 0.9969552124260701 + }, + { + "epoch": 0.4749078443181141, + "grad_norm": 0.0672697201371193, + "learning_rate": 1.1295273293708921e-05, + "loss": 0.006419669091701508, + "step": 8970, + "token_acc": 0.99713024835031 + }, + { + "epoch": 0.47543728450130046, + "grad_norm": 0.06103379651904106, + "learning_rate": 1.1278268355270769e-05, + "loss": 0.006085826456546784, + "step": 8980, + "token_acc": 0.9972978608064718 + }, + { + "epoch": 0.47596672468448675, + "grad_norm": 0.07626320421695709, + "learning_rate": 1.1261259658256696e-05, + "loss": 0.006189360097050667, + "step": 8990, + "token_acc": 0.9972561840108085 + }, + { + "epoch": 0.47649616486767304, + "grad_norm": 0.0696171224117279, + "learning_rate": 1.1244247252678482e-05, + "loss": 0.006449063867330551, + "step": 9000, + "token_acc": 0.9973563726430436 + }, + { + "epoch": 0.47649616486767304, + "eval_loss": 0.008982530795037746, + "eval_runtime": 710.4375, + "eval_samples_per_second": 16.86, + "eval_steps_per_second": 1.054, + "eval_token_acc": 0.9964215825310436, + "step": 9000 + }, + { + "epoch": 0.47702560505085934, + "grad_norm": 0.06868080794811249, + "learning_rate": 1.1227231188558815e-05, + "loss": 0.0064639545977115635, + "step": 9010, + "token_acc": 0.9972720219573034 + }, + { + "epoch": 0.4775550452340456, + "grad_norm": 0.07376545667648315, + "learning_rate": 1.1210211515931144e-05, + "loss": 0.006520270556211472, + "step": 9020, + "token_acc": 0.9973395604853843 + }, + { + "epoch": 0.478084485417232, + "grad_norm": 0.07147303223609924, + "learning_rate": 1.1193188284839518e-05, + "loss": 0.006291882693767547, + "step": 9030, + "token_acc": 0.9973486301255649 + }, + { + "epoch": 0.47861392560041827, + "grad_norm": 0.06953991204500198, + "learning_rate": 1.117616154533846e-05, + "loss": 0.006216222792863846, + "step": 9040, + "token_acc": 0.9974230673004754 + }, + { + "epoch": 0.47914336578360456, + "grad_norm": 0.08075948059558868, + "learning_rate": 1.11591313474928e-05, + "loss": 0.0064853310585021974, + "step": 9050, + "token_acc": 0.9973301963140023 + }, + { + "epoch": 0.47967280596679085, + "grad_norm": 0.07661456614732742, + "learning_rate": 1.1142097741377542e-05, + "loss": 0.0065724849700927734, + "step": 9060, + "token_acc": 0.9971641616761473 + }, + { + "epoch": 0.48020224614997714, + "grad_norm": 0.10013527423143387, + "learning_rate": 1.1125060777077713e-05, + "loss": 0.006989073753356933, + "step": 9070, + "token_acc": 0.997069994574064 + }, + { + "epoch": 0.4807316863331635, + "grad_norm": 0.06834859400987625, + "learning_rate": 1.1108020504688206e-05, + "loss": 0.005808959156274796, + "step": 9080, + "token_acc": 0.9975791977962353 + }, + { + "epoch": 0.4812611265163498, + "grad_norm": 0.13167577981948853, + "learning_rate": 1.1090976974313654e-05, + "loss": 0.006052143126726151, + "step": 9090, + "token_acc": 0.9974130449295681 + }, + { + "epoch": 0.48179056669953607, + "grad_norm": 0.07746642827987671, + "learning_rate": 1.1073930236068258e-05, + "loss": 0.006605914980173111, + "step": 9100, + "token_acc": 0.9971136600625652 + }, + { + "epoch": 0.48232000688272236, + "grad_norm": 0.08279495686292648, + "learning_rate": 1.1056880340075657e-05, + "loss": 0.006430205702781677, + "step": 9110, + "token_acc": 0.9971806549555411 + }, + { + "epoch": 0.4828494470659087, + "grad_norm": 0.07860155403614044, + "learning_rate": 1.1039827336468774e-05, + "loss": 0.00636928528547287, + "step": 9120, + "token_acc": 0.9971876356110425 + }, + { + "epoch": 0.483378887249095, + "grad_norm": 0.07272429019212723, + "learning_rate": 1.1022771275389665e-05, + "loss": 0.006800740957260132, + "step": 9130, + "token_acc": 0.9971380176391565 + }, + { + "epoch": 0.4839083274322813, + "grad_norm": 0.0793454721570015, + "learning_rate": 1.1005712206989385e-05, + "loss": 0.0063656061887741085, + "step": 9140, + "token_acc": 0.9974236259338314 + }, + { + "epoch": 0.4844377676154676, + "grad_norm": 0.08458255976438522, + "learning_rate": 1.0988650181427827e-05, + "loss": 0.0064185261726379395, + "step": 9150, + "token_acc": 0.9972132994034458 + }, + { + "epoch": 0.4849672077986539, + "grad_norm": 0.08457297086715698, + "learning_rate": 1.0971585248873575e-05, + "loss": 0.006569202244281769, + "step": 9160, + "token_acc": 0.9973154810041102 + }, + { + "epoch": 0.4854966479818402, + "grad_norm": 0.07983355224132538, + "learning_rate": 1.0954517459503766e-05, + "loss": 0.007108770310878754, + "step": 9170, + "token_acc": 0.9970314780070878 + }, + { + "epoch": 0.4860260881650265, + "grad_norm": 0.06524936854839325, + "learning_rate": 1.0937446863503937e-05, + "loss": 0.006485345959663391, + "step": 9180, + "token_acc": 0.9971224821719004 + }, + { + "epoch": 0.4865555283482128, + "grad_norm": 0.09499799460172653, + "learning_rate": 1.0920373511067878e-05, + "loss": 0.006753510236740113, + "step": 9190, + "token_acc": 0.9970382359566498 + }, + { + "epoch": 0.4870849685313991, + "grad_norm": 0.08762476593255997, + "learning_rate": 1.0903297452397481e-05, + "loss": 0.005918784439563752, + "step": 9200, + "token_acc": 0.9974793422919622 + }, + { + "epoch": 0.4876144087145854, + "grad_norm": 0.07055126875638962, + "learning_rate": 1.0886218737702592e-05, + "loss": 0.006378862261772156, + "step": 9210, + "token_acc": 0.9972630173564753 + }, + { + "epoch": 0.48814384889777174, + "grad_norm": 0.08438345789909363, + "learning_rate": 1.0869137417200878e-05, + "loss": 0.006247039884328842, + "step": 9220, + "token_acc": 0.9972217355392586 + }, + { + "epoch": 0.48867328908095803, + "grad_norm": 0.08413192629814148, + "learning_rate": 1.0852053541117653e-05, + "loss": 0.006498625129461288, + "step": 9230, + "token_acc": 0.9972468797971034 + }, + { + "epoch": 0.4892027292641443, + "grad_norm": 0.07180822640657425, + "learning_rate": 1.0834967159685762e-05, + "loss": 0.006661414355039597, + "step": 9240, + "token_acc": 0.9972636338608625 + }, + { + "epoch": 0.4897321694473306, + "grad_norm": 0.10653575509786606, + "learning_rate": 1.0817878323145404e-05, + "loss": 0.006329770386219025, + "step": 9250, + "token_acc": 0.9974457215836526 + }, + { + "epoch": 0.49026160963051696, + "grad_norm": 0.08672437816858292, + "learning_rate": 1.0800787081744002e-05, + "loss": 0.006199394911527633, + "step": 9260, + "token_acc": 0.9974384006274667 + }, + { + "epoch": 0.49079104981370325, + "grad_norm": 0.07458885759115219, + "learning_rate": 1.0783693485736047e-05, + "loss": 0.006351241469383239, + "step": 9270, + "token_acc": 0.9971567932630174 + }, + { + "epoch": 0.49132048999688954, + "grad_norm": 0.09359791874885559, + "learning_rate": 1.0766597585382957e-05, + "loss": 0.006542753428220749, + "step": 9280, + "token_acc": 0.9973534150963874 + }, + { + "epoch": 0.49184993018007583, + "grad_norm": 0.08037353307008743, + "learning_rate": 1.0749499430952923e-05, + "loss": 0.006577141582965851, + "step": 9290, + "token_acc": 0.9972976354310021 + }, + { + "epoch": 0.4923793703632621, + "grad_norm": 0.09727724641561508, + "learning_rate": 1.0732399072720767e-05, + "loss": 0.006605660915374756, + "step": 9300, + "token_acc": 0.9972877791501151 + }, + { + "epoch": 0.4929088105464485, + "grad_norm": 0.0941292867064476, + "learning_rate": 1.0715296560967791e-05, + "loss": 0.007075536996126175, + "step": 9310, + "token_acc": 0.9969545010805263 + }, + { + "epoch": 0.49343825072963476, + "grad_norm": 0.06931695342063904, + "learning_rate": 1.0698191945981624e-05, + "loss": 0.006279227137565613, + "step": 9320, + "token_acc": 0.997245409015025 + }, + { + "epoch": 0.49396769091282106, + "grad_norm": 0.11702948063611984, + "learning_rate": 1.0681085278056082e-05, + "loss": 0.006294234097003937, + "step": 9330, + "token_acc": 0.9973388946819604 + }, + { + "epoch": 0.49449713109600735, + "grad_norm": 0.08520205318927765, + "learning_rate": 1.066397660749102e-05, + "loss": 0.006253381073474884, + "step": 9340, + "token_acc": 0.9974563408002936 + }, + { + "epoch": 0.49502657127919364, + "grad_norm": 0.08152945339679718, + "learning_rate": 1.064686598459218e-05, + "loss": 0.006217408180236817, + "step": 9350, + "token_acc": 0.9973133531355338 + }, + { + "epoch": 0.49555601146238, + "grad_norm": 0.10385942459106445, + "learning_rate": 1.0629753459671039e-05, + "loss": 0.006504766643047333, + "step": 9360, + "token_acc": 0.9972717033081641 + }, + { + "epoch": 0.4960854516455663, + "grad_norm": 0.07816924154758453, + "learning_rate": 1.0612639083044678e-05, + "loss": 0.006310644000768662, + "step": 9370, + "token_acc": 0.9972224539160898 + }, + { + "epoch": 0.49661489182875257, + "grad_norm": 0.10103234648704529, + "learning_rate": 1.059552290503561e-05, + "loss": 0.006250058114528656, + "step": 9380, + "token_acc": 0.9974631150277055 + }, + { + "epoch": 0.49714433201193886, + "grad_norm": 0.0808364525437355, + "learning_rate": 1.0578404975971656e-05, + "loss": 0.005817214027047157, + "step": 9390, + "token_acc": 0.9975887128398956 + }, + { + "epoch": 0.49767377219512515, + "grad_norm": 0.08603619039058685, + "learning_rate": 1.056128534618578e-05, + "loss": 0.0058484189212322235, + "step": 9400, + "token_acc": 0.9975145333077559 + }, + { + "epoch": 0.4982032123783115, + "grad_norm": 0.0809829905629158, + "learning_rate": 1.0544164066015943e-05, + "loss": 0.006169483065605164, + "step": 9410, + "token_acc": 0.9973388280832889 + }, + { + "epoch": 0.4987326525614978, + "grad_norm": 0.10229097306728363, + "learning_rate": 1.052704118580497e-05, + "loss": 0.00620715394616127, + "step": 9420, + "token_acc": 0.9973800584063413 + }, + { + "epoch": 0.4992620927446841, + "grad_norm": 0.06899844110012054, + "learning_rate": 1.0509916755900377e-05, + "loss": 0.006371918320655823, + "step": 9430, + "token_acc": 0.997337939782362 + }, + { + "epoch": 0.4997915329278704, + "grad_norm": 0.08855949342250824, + "learning_rate": 1.0492790826654246e-05, + "loss": 0.006330734491348267, + "step": 9440, + "token_acc": 0.9974059121847058 + }, + { + "epoch": 0.5003209731110567, + "grad_norm": 0.08690545707941055, + "learning_rate": 1.0475663448423068e-05, + "loss": 0.006379410624504089, + "step": 9450, + "token_acc": 0.997399111363049 + }, + { + "epoch": 0.500850413294243, + "grad_norm": 0.07448697835206985, + "learning_rate": 1.0458534671567593e-05, + "loss": 0.006173017621040344, + "step": 9460, + "token_acc": 0.9974715655431962 + }, + { + "epoch": 0.5013798534774293, + "grad_norm": 0.07541800290346146, + "learning_rate": 1.0441404546452673e-05, + "loss": 0.006297103315591812, + "step": 9470, + "token_acc": 0.9974487030907377 + }, + { + "epoch": 0.5019092936606157, + "grad_norm": 0.09205257147550583, + "learning_rate": 1.0424273123447141e-05, + "loss": 0.006992320716381073, + "step": 9480, + "token_acc": 0.9970886582802229 + }, + { + "epoch": 0.5024387338438019, + "grad_norm": 0.07118821144104004, + "learning_rate": 1.0407140452923638e-05, + "loss": 0.006107927858829498, + "step": 9490, + "token_acc": 0.9974313426238668 + }, + { + "epoch": 0.5029681740269882, + "grad_norm": 0.07505251467227936, + "learning_rate": 1.0390006585258474e-05, + "loss": 0.005801636725664139, + "step": 9500, + "token_acc": 0.9974555981012923 + }, + { + "epoch": 0.5034976142101745, + "grad_norm": 0.06989327073097229, + "learning_rate": 1.0372871570831477e-05, + "loss": 0.0060681581497192385, + "step": 9510, + "token_acc": 0.9973632608786349 + }, + { + "epoch": 0.5040270543933608, + "grad_norm": 0.07721663266420364, + "learning_rate": 1.035573546002585e-05, + "loss": 0.005885158479213714, + "step": 9520, + "token_acc": 0.9976148181507335 + }, + { + "epoch": 0.5045564945765472, + "grad_norm": 0.07110454142093658, + "learning_rate": 1.0338598303228023e-05, + "loss": 0.0060627788305282595, + "step": 9530, + "token_acc": 0.9976301537896678 + }, + { + "epoch": 0.5050859347597334, + "grad_norm": 0.08759637176990509, + "learning_rate": 1.032146015082749e-05, + "loss": 0.006326910853385925, + "step": 9540, + "token_acc": 0.9972403328275333 + }, + { + "epoch": 0.5056153749429197, + "grad_norm": 0.07057558000087738, + "learning_rate": 1.030432105321668e-05, + "loss": 0.006206736341118813, + "step": 9550, + "token_acc": 0.9974224439235575 + }, + { + "epoch": 0.506144815126106, + "grad_norm": 0.06700477749109268, + "learning_rate": 1.0287181060790806e-05, + "loss": 0.0060766667127609255, + "step": 9560, + "token_acc": 0.9974956173303281 + }, + { + "epoch": 0.5066742553092923, + "grad_norm": 0.074788898229599, + "learning_rate": 1.0270040223947706e-05, + "loss": 0.0061182532459497455, + "step": 9570, + "token_acc": 0.9972717260712856 + }, + { + "epoch": 0.5072036954924787, + "grad_norm": 0.10445187240839005, + "learning_rate": 1.0252898593087698e-05, + "loss": 0.005924729257822036, + "step": 9580, + "token_acc": 0.9977628075096208 + }, + { + "epoch": 0.5077331356756649, + "grad_norm": 0.07183568924665451, + "learning_rate": 1.0235756218613441e-05, + "loss": 0.006352004408836365, + "step": 9590, + "token_acc": 0.9975150514501092 + }, + { + "epoch": 0.5082625758588513, + "grad_norm": 0.07877341657876968, + "learning_rate": 1.0218613150929782e-05, + "loss": 0.005877148360013962, + "step": 9600, + "token_acc": 0.9975464628172281 + }, + { + "epoch": 0.5087920160420375, + "grad_norm": 0.08823011070489883, + "learning_rate": 1.0201469440443596e-05, + "loss": 0.005926524847745895, + "step": 9610, + "token_acc": 0.997515030728563 + }, + { + "epoch": 0.5093214562252238, + "grad_norm": 0.10528596490621567, + "learning_rate": 1.018432513756366e-05, + "loss": 0.005964019149541855, + "step": 9620, + "token_acc": 0.9975796624881069 + }, + { + "epoch": 0.5098508964084102, + "grad_norm": 0.08236780762672424, + "learning_rate": 1.0167180292700482e-05, + "loss": 0.006844854354858399, + "step": 9630, + "token_acc": 0.9971474088964143 + }, + { + "epoch": 0.5103803365915964, + "grad_norm": 0.09230993688106537, + "learning_rate": 1.0150034956266173e-05, + "loss": 0.0061198890209198, + "step": 9640, + "token_acc": 0.9973313985973163 + }, + { + "epoch": 0.5109097767747828, + "grad_norm": 0.0912308320403099, + "learning_rate": 1.0132889178674283e-05, + "loss": 0.006087886542081833, + "step": 9650, + "token_acc": 0.9973391610433158 + }, + { + "epoch": 0.5114392169579691, + "grad_norm": 0.07799002528190613, + "learning_rate": 1.0115743010339666e-05, + "loss": 0.005520019680261612, + "step": 9660, + "token_acc": 0.9976295010266598 + }, + { + "epoch": 0.5119686571411554, + "grad_norm": 0.0757814273238182, + "learning_rate": 1.0098596501678315e-05, + "loss": 0.005552807077765465, + "step": 9670, + "token_acc": 0.9976989837178087 + }, + { + "epoch": 0.5124980973243417, + "grad_norm": 0.06419990211725235, + "learning_rate": 1.0081449703107234e-05, + "loss": 0.006049920618534088, + "step": 9680, + "token_acc": 0.9975984389853405 + }, + { + "epoch": 0.5130275375075279, + "grad_norm": 0.06704181432723999, + "learning_rate": 1.0064302665044268e-05, + "loss": 0.005979394540190697, + "step": 9690, + "token_acc": 0.9973137565696171 + }, + { + "epoch": 0.5135569776907143, + "grad_norm": 0.0717286467552185, + "learning_rate": 1.0047155437907976e-05, + "loss": 0.006159737706184387, + "step": 9700, + "token_acc": 0.9972968237679273 + }, + { + "epoch": 0.5140864178739006, + "grad_norm": 0.0916777104139328, + "learning_rate": 1.0030008072117472e-05, + "loss": 0.0060527905821800235, + "step": 9710, + "token_acc": 0.9974986659551761 + }, + { + "epoch": 0.5146158580570869, + "grad_norm": 0.08642859011888504, + "learning_rate": 1.0012860618092275e-05, + "loss": 0.005933391302824021, + "step": 9720, + "token_acc": 0.9975449671827246 + }, + { + "epoch": 0.5151452982402732, + "grad_norm": 0.11616049706935883, + "learning_rate": 9.995713126252164e-06, + "loss": 0.006013680249452591, + "step": 9730, + "token_acc": 0.9974654416301212 + }, + { + "epoch": 0.5156747384234595, + "grad_norm": 0.06490373611450195, + "learning_rate": 9.978565647017024e-06, + "loss": 0.006184115633368492, + "step": 9740, + "token_acc": 0.9975220264317181 + }, + { + "epoch": 0.5162041786066458, + "grad_norm": 0.08089577406644821, + "learning_rate": 9.961418230806714e-06, + "loss": 0.00578242614865303, + "step": 9750, + "token_acc": 0.9975644543793947 + }, + { + "epoch": 0.5167336187898321, + "grad_norm": 0.09759592264890671, + "learning_rate": 9.944270928040898e-06, + "loss": 0.006061964109539986, + "step": 9760, + "token_acc": 0.9973715611258063 + }, + { + "epoch": 0.5172630589730184, + "grad_norm": 0.05860958620905876, + "learning_rate": 9.927123789138913e-06, + "loss": 0.005434776842594147, + "step": 9770, + "token_acc": 0.9977225707421249 + }, + { + "epoch": 0.5177924991562047, + "grad_norm": 0.09341993927955627, + "learning_rate": 9.909976864519611e-06, + "loss": 0.005640207976102829, + "step": 9780, + "token_acc": 0.9977037791620046 + }, + { + "epoch": 0.518321939339391, + "grad_norm": 0.09343712776899338, + "learning_rate": 9.892830204601212e-06, + "loss": 0.0058887191116809845, + "step": 9790, + "token_acc": 0.9975718231352686 + }, + { + "epoch": 0.5188513795225773, + "grad_norm": 0.08606129884719849, + "learning_rate": 9.87568385980116e-06, + "loss": 0.0055746838450431825, + "step": 9800, + "token_acc": 0.9976213725932046 + }, + { + "epoch": 0.5193808197057637, + "grad_norm": 0.11738255620002747, + "learning_rate": 9.858537880535974e-06, + "loss": 0.0055122010409832, + "step": 9810, + "token_acc": 0.9978956685845031 + }, + { + "epoch": 0.5199102598889499, + "grad_norm": 0.09453237801790237, + "learning_rate": 9.841392317221096e-06, + "loss": 0.005794931948184967, + "step": 9820, + "token_acc": 0.9976141416332285 + }, + { + "epoch": 0.5204397000721362, + "grad_norm": 0.08104211837053299, + "learning_rate": 9.824247220270749e-06, + "loss": 0.006003174185752869, + "step": 9830, + "token_acc": 0.9974561922951818 + }, + { + "epoch": 0.5209691402553225, + "grad_norm": 0.08040585368871689, + "learning_rate": 9.807102640097774e-06, + "loss": 0.006172920763492584, + "step": 9840, + "token_acc": 0.9974471706140973 + }, + { + "epoch": 0.5214985804385088, + "grad_norm": 0.06752321869134903, + "learning_rate": 9.789958627113503e-06, + "loss": 0.005765810981392861, + "step": 9850, + "token_acc": 0.9975799856468114 + }, + { + "epoch": 0.5220280206216952, + "grad_norm": 0.08147775381803513, + "learning_rate": 9.772815231727598e-06, + "loss": 0.006705331057310105, + "step": 9860, + "token_acc": 0.9971882691585666 + }, + { + "epoch": 0.5225574608048814, + "grad_norm": 0.10563169419765472, + "learning_rate": 9.755672504347907e-06, + "loss": 0.006237004697322845, + "step": 9870, + "token_acc": 0.9973712101613159 + }, + { + "epoch": 0.5230869009880678, + "grad_norm": 0.11307316273450851, + "learning_rate": 9.738530495380303e-06, + "loss": 0.006556305289268494, + "step": 9880, + "token_acc": 0.9973062232081797 + }, + { + "epoch": 0.523616341171254, + "grad_norm": 0.07456500828266144, + "learning_rate": 9.721389255228559e-06, + "loss": 0.005898710340261459, + "step": 9890, + "token_acc": 0.9976225234619395 + }, + { + "epoch": 0.5241457813544403, + "grad_norm": 0.09724334627389908, + "learning_rate": 9.704248834294184e-06, + "loss": 0.005645465850830078, + "step": 9900, + "token_acc": 0.9975654087945438 + }, + { + "epoch": 0.5246752215376267, + "grad_norm": 0.08709348738193512, + "learning_rate": 9.687109282976273e-06, + "loss": 0.0061257675290107725, + "step": 9910, + "token_acc": 0.9973894912427023 + }, + { + "epoch": 0.5252046617208129, + "grad_norm": 0.06941056251525879, + "learning_rate": 9.669970651671373e-06, + "loss": 0.005428103357553482, + "step": 9920, + "token_acc": 0.9977058097240298 + }, + { + "epoch": 0.5257341019039993, + "grad_norm": 0.08480376750230789, + "learning_rate": 9.652832990773315e-06, + "loss": 0.006123346090316772, + "step": 9930, + "token_acc": 0.9974898886711421 + }, + { + "epoch": 0.5262635420871855, + "grad_norm": 0.09106945991516113, + "learning_rate": 9.635696350673088e-06, + "loss": 0.005818301439285278, + "step": 9940, + "token_acc": 0.9976885654920351 + }, + { + "epoch": 0.5267929822703719, + "grad_norm": 0.06666842103004456, + "learning_rate": 9.618560781758673e-06, + "loss": 0.005581918358802796, + "step": 9950, + "token_acc": 0.9975953911664023 + }, + { + "epoch": 0.5273224224535582, + "grad_norm": 0.08956436067819595, + "learning_rate": 9.601426334414898e-06, + "loss": 0.005594116449356079, + "step": 9960, + "token_acc": 0.9976652268899154 + }, + { + "epoch": 0.5278518626367444, + "grad_norm": 0.08601323515176773, + "learning_rate": 9.5842930590233e-06, + "loss": 0.005335238575935364, + "step": 9970, + "token_acc": 0.9978230225788425 + }, + { + "epoch": 0.5283813028199308, + "grad_norm": 0.07210409641265869, + "learning_rate": 9.56716100596197e-06, + "loss": 0.005897215753793717, + "step": 9980, + "token_acc": 0.9974215835982677 + }, + { + "epoch": 0.5289107430031171, + "grad_norm": 0.0716291144490242, + "learning_rate": 9.5500302256054e-06, + "loss": 0.006033800542354584, + "step": 9990, + "token_acc": 0.9976139226777461 + }, + { + "epoch": 0.5294401831863034, + "grad_norm": 0.07986830919981003, + "learning_rate": 9.532900768324345e-06, + "loss": 0.005965420603752136, + "step": 10000, + "token_acc": 0.997496829317135 + }, + { + "epoch": 0.5299696233694897, + "grad_norm": 0.08717332780361176, + "learning_rate": 9.515772684485662e-06, + "loss": 0.005752303823828697, + "step": 10010, + "token_acc": 0.9976394044292447 + }, + { + "epoch": 0.530499063552676, + "grad_norm": 0.09234462678432465, + "learning_rate": 9.498646024452178e-06, + "loss": 0.005749382823705673, + "step": 10020, + "token_acc": 0.9974636441759763 + }, + { + "epoch": 0.5310285037358623, + "grad_norm": 0.0769585520029068, + "learning_rate": 9.481520838582527e-06, + "loss": 0.006350738555192947, + "step": 10030, + "token_acc": 0.997323059217933 + }, + { + "epoch": 0.5315579439190486, + "grad_norm": 0.07705565541982651, + "learning_rate": 9.464397177231014e-06, + "loss": 0.0055728830397129055, + "step": 10040, + "token_acc": 0.997539636867082 + }, + { + "epoch": 0.5320873841022349, + "grad_norm": 0.09076856076717377, + "learning_rate": 9.447275090747461e-06, + "loss": 0.005577860027551651, + "step": 10050, + "token_acc": 0.9977807627168136 + }, + { + "epoch": 0.5326168242854212, + "grad_norm": 0.0796055793762207, + "learning_rate": 9.430154629477052e-06, + "loss": 0.005743794143199921, + "step": 10060, + "token_acc": 0.9976966986013286 + }, + { + "epoch": 0.5331462644686075, + "grad_norm": 0.09636925160884857, + "learning_rate": 9.413035843760199e-06, + "loss": 0.006041829288005829, + "step": 10070, + "token_acc": 0.9973300181058147 + }, + { + "epoch": 0.5336757046517938, + "grad_norm": 0.07640992850065231, + "learning_rate": 9.395918783932384e-06, + "loss": 0.005545875430107117, + "step": 10080, + "token_acc": 0.9976796788274866 + }, + { + "epoch": 0.5342051448349802, + "grad_norm": 0.09229988604784012, + "learning_rate": 9.378803500324021e-06, + "loss": 0.0058249209076166155, + "step": 10090, + "token_acc": 0.9974642577114926 + }, + { + "epoch": 0.5347345850181664, + "grad_norm": 0.0810672715306282, + "learning_rate": 9.361690043260293e-06, + "loss": 0.0056256964802742, + "step": 10100, + "token_acc": 0.9976643894463769 + }, + { + "epoch": 0.5352640252013527, + "grad_norm": 0.09449639171361923, + "learning_rate": 9.344578463061013e-06, + "loss": 0.005812922120094299, + "step": 10110, + "token_acc": 0.9976715849913206 + }, + { + "epoch": 0.535793465384539, + "grad_norm": 0.07625672966241837, + "learning_rate": 9.327468810040482e-06, + "loss": 0.005629872158169747, + "step": 10120, + "token_acc": 0.9977306479333878 + }, + { + "epoch": 0.5363229055677253, + "grad_norm": 0.08401705324649811, + "learning_rate": 9.310361134507326e-06, + "loss": 0.005926584079861641, + "step": 10130, + "token_acc": 0.9975380763613603 + }, + { + "epoch": 0.5368523457509117, + "grad_norm": 0.08435894548892975, + "learning_rate": 9.293255486764364e-06, + "loss": 0.005681088566780091, + "step": 10140, + "token_acc": 0.9976152755774201 + }, + { + "epoch": 0.5373817859340979, + "grad_norm": 0.09608595818281174, + "learning_rate": 9.276151917108451e-06, + "loss": 0.005818182229995727, + "step": 10150, + "token_acc": 0.9975734227247711 + }, + { + "epoch": 0.5379112261172843, + "grad_norm": 0.08716209977865219, + "learning_rate": 9.259050475830324e-06, + "loss": 0.005730519816279411, + "step": 10160, + "token_acc": 0.9977151624819673 + }, + { + "epoch": 0.5384406663004705, + "grad_norm": 0.0732644647359848, + "learning_rate": 9.241951213214471e-06, + "loss": 0.005736709386110306, + "step": 10170, + "token_acc": 0.997555257031765 + }, + { + "epoch": 0.5389701064836568, + "grad_norm": 0.09009082615375519, + "learning_rate": 9.224854179538968e-06, + "loss": 0.00554574690759182, + "step": 10180, + "token_acc": 0.9977137707763166 + }, + { + "epoch": 0.5394995466668432, + "grad_norm": 0.0633961483836174, + "learning_rate": 9.207759425075341e-06, + "loss": 0.0058026410639286045, + "step": 10190, + "token_acc": 0.9976723425911249 + }, + { + "epoch": 0.5400289868500294, + "grad_norm": 0.08042139559984207, + "learning_rate": 9.190667000088413e-06, + "loss": 0.00541747659444809, + "step": 10200, + "token_acc": 0.9978231133649162 + }, + { + "epoch": 0.5405584270332158, + "grad_norm": 0.14404639601707458, + "learning_rate": 9.17357695483616e-06, + "loss": 0.005680832639336586, + "step": 10210, + "token_acc": 0.9977226277372263 + }, + { + "epoch": 0.541087867216402, + "grad_norm": 0.11139760911464691, + "learning_rate": 9.156489339569555e-06, + "loss": 0.005819535255432129, + "step": 10220, + "token_acc": 0.9976050402636959 + }, + { + "epoch": 0.5416173073995884, + "grad_norm": 0.08715484291315079, + "learning_rate": 9.139404204532428e-06, + "loss": 0.005735942348837853, + "step": 10230, + "token_acc": 0.9975622380658196 + }, + { + "epoch": 0.5421467475827747, + "grad_norm": 0.06409761309623718, + "learning_rate": 9.12232159996132e-06, + "loss": 0.005963046103715897, + "step": 10240, + "token_acc": 0.9973363171035162 + }, + { + "epoch": 0.5426761877659609, + "grad_norm": 0.08512922376394272, + "learning_rate": 9.105241576085328e-06, + "loss": 0.005926216393709183, + "step": 10250, + "token_acc": 0.9975474247960359 + }, + { + "epoch": 0.5432056279491473, + "grad_norm": 0.07714362442493439, + "learning_rate": 9.08816418312596e-06, + "loss": 0.005915514007210731, + "step": 10260, + "token_acc": 0.9975716205052031 + }, + { + "epoch": 0.5437350681323335, + "grad_norm": 0.08463723212480545, + "learning_rate": 9.071089471296996e-06, + "loss": 0.005881936848163604, + "step": 10270, + "token_acc": 0.9975884311451197 + }, + { + "epoch": 0.5442645083155199, + "grad_norm": 0.07101252675056458, + "learning_rate": 9.054017490804319e-06, + "loss": 0.005859004706144333, + "step": 10280, + "token_acc": 0.9974826410596247 + }, + { + "epoch": 0.5447939484987062, + "grad_norm": 0.08222519606351852, + "learning_rate": 9.03694829184579e-06, + "loss": 0.005490032956004143, + "step": 10290, + "token_acc": 0.9977225517431239 + }, + { + "epoch": 0.5453233886818925, + "grad_norm": 0.08493092656135559, + "learning_rate": 9.019881924611088e-06, + "loss": 0.00568794384598732, + "step": 10300, + "token_acc": 0.9975629908445238 + }, + { + "epoch": 0.5458528288650788, + "grad_norm": 0.08972533792257309, + "learning_rate": 9.002818439281572e-06, + "loss": 0.005533015355467796, + "step": 10310, + "token_acc": 0.9977470711925503 + }, + { + "epoch": 0.5463822690482651, + "grad_norm": 0.09296856820583344, + "learning_rate": 8.985757886030119e-06, + "loss": 0.00558307021856308, + "step": 10320, + "token_acc": 0.9977063072380603 + }, + { + "epoch": 0.5469117092314514, + "grad_norm": 0.07048169523477554, + "learning_rate": 8.968700315020987e-06, + "loss": 0.005209856852889061, + "step": 10330, + "token_acc": 0.99786604426291 + }, + { + "epoch": 0.5474411494146377, + "grad_norm": 0.08755260705947876, + "learning_rate": 8.951645776409668e-06, + "loss": 0.005423923954367638, + "step": 10340, + "token_acc": 0.9975798442754972 + }, + { + "epoch": 0.547970589597824, + "grad_norm": 0.10201948136091232, + "learning_rate": 8.934594320342734e-06, + "loss": 0.005968772619962692, + "step": 10350, + "token_acc": 0.9974406429238362 + }, + { + "epoch": 0.5485000297810103, + "grad_norm": 0.06459422409534454, + "learning_rate": 8.917545996957696e-06, + "loss": 0.005491361021995544, + "step": 10360, + "token_acc": 0.9978060296644811 + }, + { + "epoch": 0.5490294699641967, + "grad_norm": 0.06837255507707596, + "learning_rate": 8.900500856382855e-06, + "loss": 0.006030368432402611, + "step": 10370, + "token_acc": 0.9975208267249871 + }, + { + "epoch": 0.5495589101473829, + "grad_norm": 0.08479826897382736, + "learning_rate": 8.883458948737144e-06, + "loss": 0.005318785086274147, + "step": 10380, + "token_acc": 0.9979291571335527 + }, + { + "epoch": 0.5500883503305692, + "grad_norm": 0.09189905971288681, + "learning_rate": 8.866420324130003e-06, + "loss": 0.005710026249289513, + "step": 10390, + "token_acc": 0.9978213143901767 + }, + { + "epoch": 0.5506177905137555, + "grad_norm": 0.0816366970539093, + "learning_rate": 8.84938503266121e-06, + "loss": 0.00552150011062622, + "step": 10400, + "token_acc": 0.9976040005676933 + }, + { + "epoch": 0.5511472306969418, + "grad_norm": 0.0827329158782959, + "learning_rate": 8.832353124420746e-06, + "loss": 0.0058914490044116976, + "step": 10410, + "token_acc": 0.9976724785183949 + }, + { + "epoch": 0.5516766708801282, + "grad_norm": 0.07583095133304596, + "learning_rate": 8.815324649488647e-06, + "loss": 0.005840864032506943, + "step": 10420, + "token_acc": 0.9975393906029744 + }, + { + "epoch": 0.5522061110633144, + "grad_norm": 0.079261913895607, + "learning_rate": 8.798299657934841e-06, + "loss": 0.005099338293075561, + "step": 10430, + "token_acc": 0.9980569894176806 + }, + { + "epoch": 0.5527355512465008, + "grad_norm": 0.07089421153068542, + "learning_rate": 8.78127819981903e-06, + "loss": 0.005278304219245911, + "step": 10440, + "token_acc": 0.997779410291515 + }, + { + "epoch": 0.553264991429687, + "grad_norm": 0.09767837077379227, + "learning_rate": 8.764260325190513e-06, + "loss": 0.0057599864900112156, + "step": 10450, + "token_acc": 0.9976306480678102 + }, + { + "epoch": 0.5537944316128733, + "grad_norm": 0.09542081505060196, + "learning_rate": 8.747246084088063e-06, + "loss": 0.005584334954619408, + "step": 10460, + "token_acc": 0.9977300418940798 + }, + { + "epoch": 0.5543238717960597, + "grad_norm": 0.07572547346353531, + "learning_rate": 8.73023552653976e-06, + "loss": 0.005530974268913269, + "step": 10470, + "token_acc": 0.9977055417424533 + }, + { + "epoch": 0.5548533119792459, + "grad_norm": 0.0940098837018013, + "learning_rate": 8.71322870256286e-06, + "loss": 0.005657859519124031, + "step": 10480, + "token_acc": 0.9976153551119783 + }, + { + "epoch": 0.5553827521624323, + "grad_norm": 0.08741765469312668, + "learning_rate": 8.69622566216364e-06, + "loss": 0.0053317703306674956, + "step": 10490, + "token_acc": 0.9977565011425998 + }, + { + "epoch": 0.5559121923456185, + "grad_norm": 0.09390822798013687, + "learning_rate": 8.679226455337246e-06, + "loss": 0.005764786899089813, + "step": 10500, + "token_acc": 0.9975888737787938 + }, + { + "epoch": 0.5564416325288049, + "grad_norm": 0.07534737139940262, + "learning_rate": 8.662231132067557e-06, + "loss": 0.0054897740483284, + "step": 10510, + "token_acc": 0.9976895102095289 + }, + { + "epoch": 0.5569710727119912, + "grad_norm": 0.07866326719522476, + "learning_rate": 8.645239742327034e-06, + "loss": 0.00528850331902504, + "step": 10520, + "token_acc": 0.9977730328456803 + }, + { + "epoch": 0.5575005128951774, + "grad_norm": 0.058244187384843826, + "learning_rate": 8.628252336076568e-06, + "loss": 0.005400420725345611, + "step": 10530, + "token_acc": 0.9976907429638522 + }, + { + "epoch": 0.5580299530783638, + "grad_norm": 0.08737803995609283, + "learning_rate": 8.611268963265342e-06, + "loss": 0.005433228611946106, + "step": 10540, + "token_acc": 0.9978140070419009 + }, + { + "epoch": 0.55855939326155, + "grad_norm": 0.09868389368057251, + "learning_rate": 8.594289673830671e-06, + "loss": 0.005635057389736175, + "step": 10550, + "token_acc": 0.9976789814235024 + }, + { + "epoch": 0.5590888334447364, + "grad_norm": 0.07014649361371994, + "learning_rate": 8.57731451769787e-06, + "loss": 0.005669058859348297, + "step": 10560, + "token_acc": 0.9976317742513822 + }, + { + "epoch": 0.5596182736279227, + "grad_norm": 0.09194966405630112, + "learning_rate": 8.560343544780102e-06, + "loss": 0.0055405043065547945, + "step": 10570, + "token_acc": 0.9976646037332978 + }, + { + "epoch": 0.560147713811109, + "grad_norm": 0.06665540486574173, + "learning_rate": 8.543376804978223e-06, + "loss": 0.0054998710751533505, + "step": 10580, + "token_acc": 0.9977068045363575 + }, + { + "epoch": 0.5606771539942953, + "grad_norm": 0.07561322301626205, + "learning_rate": 8.526414348180652e-06, + "loss": 0.0055532239377498625, + "step": 10590, + "token_acc": 0.9978732986389112 + }, + { + "epoch": 0.5612065941774815, + "grad_norm": 0.07968828082084656, + "learning_rate": 8.509456224263203e-06, + "loss": 0.0055864803493022915, + "step": 10600, + "token_acc": 0.9976723231716474 + }, + { + "epoch": 0.5617360343606679, + "grad_norm": 0.07937763631343842, + "learning_rate": 8.492502483088956e-06, + "loss": 0.0052876144647598265, + "step": 10610, + "token_acc": 0.9978570296930633 + }, + { + "epoch": 0.5622654745438542, + "grad_norm": 0.06759688258171082, + "learning_rate": 8.475553174508105e-06, + "loss": 0.005193571001291275, + "step": 10620, + "token_acc": 0.9978633013663186 + }, + { + "epoch": 0.5627949147270405, + "grad_norm": 0.08593278378248215, + "learning_rate": 8.458608348357808e-06, + "loss": 0.005326055362820626, + "step": 10630, + "token_acc": 0.9977219055884243 + }, + { + "epoch": 0.5633243549102268, + "grad_norm": 0.08525919914245605, + "learning_rate": 8.441668054462045e-06, + "loss": 0.005171322077512741, + "step": 10640, + "token_acc": 0.9979057504255248 + }, + { + "epoch": 0.5638537950934132, + "grad_norm": 0.09290236979722977, + "learning_rate": 8.424732342631466e-06, + "loss": 0.005482325330376625, + "step": 10650, + "token_acc": 0.9977303828310137 + }, + { + "epoch": 0.5643832352765994, + "grad_norm": 0.07777766138315201, + "learning_rate": 8.407801262663252e-06, + "loss": 0.004773225262761116, + "step": 10660, + "token_acc": 0.9979968283114932 + }, + { + "epoch": 0.5649126754597857, + "grad_norm": 0.09356539696455002, + "learning_rate": 8.390874864340963e-06, + "loss": 0.005722983926534653, + "step": 10670, + "token_acc": 0.9977797253870874 + }, + { + "epoch": 0.565442115642972, + "grad_norm": 0.09650908410549164, + "learning_rate": 8.373953197434396e-06, + "loss": 0.005484995990991592, + "step": 10680, + "token_acc": 0.997781058918725 + }, + { + "epoch": 0.5659715558261583, + "grad_norm": 0.0700053796172142, + "learning_rate": 8.357036311699427e-06, + "loss": 0.005631019175052643, + "step": 10690, + "token_acc": 0.9976724979352805 + }, + { + "epoch": 0.5665009960093447, + "grad_norm": 0.08488014340400696, + "learning_rate": 8.340124256877887e-06, + "loss": 0.005525138601660729, + "step": 10700, + "token_acc": 0.997639522566331 + }, + { + "epoch": 0.5670304361925309, + "grad_norm": 0.08865460008382797, + "learning_rate": 8.323217082697392e-06, + "loss": 0.005481350421905518, + "step": 10710, + "token_acc": 0.9978473630196991 + }, + { + "epoch": 0.5675598763757173, + "grad_norm": 0.08897771686315536, + "learning_rate": 8.306314838871208e-06, + "loss": 0.00523499920964241, + "step": 10720, + "token_acc": 0.9979386089365893 + }, + { + "epoch": 0.5680893165589035, + "grad_norm": 0.07740037888288498, + "learning_rate": 8.289417575098107e-06, + "loss": 0.005479452013969421, + "step": 10730, + "token_acc": 0.9976641166606879 + }, + { + "epoch": 0.5686187567420898, + "grad_norm": 0.0868200734257698, + "learning_rate": 8.272525341062218e-06, + "loss": 0.004969134554266929, + "step": 10740, + "token_acc": 0.9978044545363476 + }, + { + "epoch": 0.5691481969252762, + "grad_norm": 0.07562980055809021, + "learning_rate": 8.255638186432883e-06, + "loss": 0.005410711839795112, + "step": 10750, + "token_acc": 0.997804985936887 + }, + { + "epoch": 0.5696776371084624, + "grad_norm": 0.0700678825378418, + "learning_rate": 8.238756160864504e-06, + "loss": 0.005320299789309502, + "step": 10760, + "token_acc": 0.9978061943728469 + }, + { + "epoch": 0.5702070772916488, + "grad_norm": 0.09050122648477554, + "learning_rate": 8.221879313996399e-06, + "loss": 0.005297638475894928, + "step": 10770, + "token_acc": 0.9977884230204299 + }, + { + "epoch": 0.570736517474835, + "grad_norm": 0.09789244085550308, + "learning_rate": 8.205007695452669e-06, + "loss": 0.005724583566188812, + "step": 10780, + "token_acc": 0.9975051524026467 + }, + { + "epoch": 0.5712659576580214, + "grad_norm": 0.10143784433603287, + "learning_rate": 8.18814135484203e-06, + "loss": 0.005482998117804528, + "step": 10790, + "token_acc": 0.9977469959946595 + }, + { + "epoch": 0.5717953978412077, + "grad_norm": 0.0662374496459961, + "learning_rate": 8.171280341757692e-06, + "loss": 0.0052685577422380446, + "step": 10800, + "token_acc": 0.9977553404539385 + }, + { + "epoch": 0.5723248380243939, + "grad_norm": 0.0728992149233818, + "learning_rate": 8.154424705777184e-06, + "loss": 0.005137554556131363, + "step": 10810, + "token_acc": 0.997881497297658 + }, + { + "epoch": 0.5728542782075803, + "grad_norm": 0.0719781368970871, + "learning_rate": 8.13757449646224e-06, + "loss": 0.004936916008591652, + "step": 10820, + "token_acc": 0.9978961604928995 + }, + { + "epoch": 0.5733837183907665, + "grad_norm": 0.10614340752363205, + "learning_rate": 8.120729763358627e-06, + "loss": 0.005483512207865715, + "step": 10830, + "token_acc": 0.9977154101422449 + }, + { + "epoch": 0.5739131585739529, + "grad_norm": 0.07656998932361603, + "learning_rate": 8.103890555996016e-06, + "loss": 0.004678269848227501, + "step": 10840, + "token_acc": 0.9979822235562893 + }, + { + "epoch": 0.5744425987571392, + "grad_norm": 0.09332893788814545, + "learning_rate": 8.087056923887832e-06, + "loss": 0.005308455973863602, + "step": 10850, + "token_acc": 0.9977562953015656 + }, + { + "epoch": 0.5749720389403254, + "grad_norm": 0.07631412148475647, + "learning_rate": 8.070228916531094e-06, + "loss": 0.005129978805780411, + "step": 10860, + "token_acc": 0.9978547937430093 + }, + { + "epoch": 0.5755014791235118, + "grad_norm": 0.07666252553462982, + "learning_rate": 8.053406583406295e-06, + "loss": 0.00522054098546505, + "step": 10870, + "token_acc": 0.9979636460750113 + }, + { + "epoch": 0.576030919306698, + "grad_norm": 0.09561937302350998, + "learning_rate": 8.036589973977242e-06, + "loss": 0.005222373455762863, + "step": 10880, + "token_acc": 0.9977729028168192 + }, + { + "epoch": 0.5765603594898844, + "grad_norm": 0.0971507877111435, + "learning_rate": 8.019779137690906e-06, + "loss": 0.005486468598246574, + "step": 10890, + "token_acc": 0.9977311590274013 + }, + { + "epoch": 0.5770897996730707, + "grad_norm": 0.1078389436006546, + "learning_rate": 8.002974123977292e-06, + "loss": 0.005656520277261734, + "step": 10900, + "token_acc": 0.997672245490497 + }, + { + "epoch": 0.577619239856257, + "grad_norm": 0.10358531773090363, + "learning_rate": 7.986174982249272e-06, + "loss": 0.004924916848540306, + "step": 10910, + "token_acc": 0.9980050083472454 + }, + { + "epoch": 0.5781486800394433, + "grad_norm": 0.06699983030557632, + "learning_rate": 7.969381761902465e-06, + "loss": 0.005164104700088501, + "step": 10920, + "token_acc": 0.9979489915875306 + }, + { + "epoch": 0.5786781202226295, + "grad_norm": 0.07490614056587219, + "learning_rate": 7.95259451231507e-06, + "loss": 0.005282736569643021, + "step": 10930, + "token_acc": 0.9978486195308657 + }, + { + "epoch": 0.5792075604058159, + "grad_norm": 0.06401417404413223, + "learning_rate": 7.935813282847737e-06, + "loss": 0.005152691155672073, + "step": 10940, + "token_acc": 0.9978313092219405 + }, + { + "epoch": 0.5797370005890022, + "grad_norm": 0.09496421366930008, + "learning_rate": 7.919038122843412e-06, + "loss": 0.005290433019399643, + "step": 10950, + "token_acc": 0.9977471651828551 + }, + { + "epoch": 0.5802664407721885, + "grad_norm": 0.07934817671775818, + "learning_rate": 7.902269081627185e-06, + "loss": 0.004965965449810028, + "step": 10960, + "token_acc": 0.9979629150351899 + }, + { + "epoch": 0.5807958809553748, + "grad_norm": 0.0899849683046341, + "learning_rate": 7.885506208506175e-06, + "loss": 0.004542484134435654, + "step": 10970, + "token_acc": 0.9980069548104106 + }, + { + "epoch": 0.5813253211385612, + "grad_norm": 0.08526492863893509, + "learning_rate": 7.868749552769341e-06, + "loss": 0.005321484059095383, + "step": 10980, + "token_acc": 0.9977981284091477 + }, + { + "epoch": 0.5818547613217474, + "grad_norm": 0.08841445297002792, + "learning_rate": 7.851999163687378e-06, + "loss": 0.00540248267352581, + "step": 10990, + "token_acc": 0.9976556959913235 + }, + { + "epoch": 0.5823842015049338, + "grad_norm": 0.09864561259746552, + "learning_rate": 7.835255090512546e-06, + "loss": 0.004821636155247689, + "step": 11000, + "token_acc": 0.9981484416050176 + }, + { + "epoch": 0.58291364168812, + "grad_norm": 0.10166633129119873, + "learning_rate": 7.818517382478542e-06, + "loss": 0.004606721550226211, + "step": 11010, + "token_acc": 0.9980244237902721 + }, + { + "epoch": 0.5834430818713063, + "grad_norm": 0.08097875118255615, + "learning_rate": 7.801786088800338e-06, + "loss": 0.004694987833499908, + "step": 11020, + "token_acc": 0.9980558545478819 + }, + { + "epoch": 0.5839725220544927, + "grad_norm": 0.07913080602884293, + "learning_rate": 7.78506125867405e-06, + "loss": 0.00548882782459259, + "step": 11030, + "token_acc": 0.9977813365362159 + }, + { + "epoch": 0.5845019622376789, + "grad_norm": 0.07013921439647675, + "learning_rate": 7.768342941276787e-06, + "loss": 0.005353599041700363, + "step": 11040, + "token_acc": 0.9976487852056896 + }, + { + "epoch": 0.5850314024208653, + "grad_norm": 0.07369482517242432, + "learning_rate": 7.751631185766511e-06, + "loss": 0.004796169698238373, + "step": 11050, + "token_acc": 0.9980895014349597 + }, + { + "epoch": 0.5855608426040515, + "grad_norm": 0.07227800041437149, + "learning_rate": 7.734926041281888e-06, + "loss": 0.0049373660236597065, + "step": 11060, + "token_acc": 0.9978802243290159 + }, + { + "epoch": 0.5860902827872378, + "grad_norm": 0.08486678451299667, + "learning_rate": 7.718227556942148e-06, + "loss": 0.004980228468775749, + "step": 11070, + "token_acc": 0.9980306583108722 + }, + { + "epoch": 0.5866197229704242, + "grad_norm": 0.09606419503688812, + "learning_rate": 7.701535781846928e-06, + "loss": 0.004831057786941528, + "step": 11080, + "token_acc": 0.9980223466484199 + }, + { + "epoch": 0.5871491631536104, + "grad_norm": 0.08035236597061157, + "learning_rate": 7.68485076507615e-06, + "loss": 0.004803618416190147, + "step": 11090, + "token_acc": 0.9980386756470284 + }, + { + "epoch": 0.5876786033367968, + "grad_norm": 0.08023076504468918, + "learning_rate": 7.668172555689857e-06, + "loss": 0.005093364417552948, + "step": 11100, + "token_acc": 0.9978641214113485 + }, + { + "epoch": 0.588208043519983, + "grad_norm": 0.0665818303823471, + "learning_rate": 7.651501202728078e-06, + "loss": 0.004961021989583969, + "step": 11110, + "token_acc": 0.9980561302809852 + }, + { + "epoch": 0.5887374837031694, + "grad_norm": 0.07671774923801422, + "learning_rate": 7.634836755210681e-06, + "loss": 0.005055252835154534, + "step": 11120, + "token_acc": 0.9978636758128046 + }, + { + "epoch": 0.5892669238863557, + "grad_norm": 0.1001303568482399, + "learning_rate": 7.618179262137227e-06, + "loss": 0.0048271968960762026, + "step": 11130, + "token_acc": 0.9981224809953354 + }, + { + "epoch": 0.589796364069542, + "grad_norm": 0.09907828271389008, + "learning_rate": 7.60152877248683e-06, + "loss": 0.005075004696846008, + "step": 11140, + "token_acc": 0.997739066593249 + }, + { + "epoch": 0.5903258042527283, + "grad_norm": 0.07219123840332031, + "learning_rate": 7.584885335218015e-06, + "loss": 0.00529729463160038, + "step": 11150, + "token_acc": 0.9976729056150535 + }, + { + "epoch": 0.5908552444359145, + "grad_norm": 0.0762040913105011, + "learning_rate": 7.568248999268568e-06, + "loss": 0.0047658286988735196, + "step": 11160, + "token_acc": 0.9980389702507615 + }, + { + "epoch": 0.5913846846191009, + "grad_norm": 0.07680967450141907, + "learning_rate": 7.551619813555393e-06, + "loss": 0.005047033727169037, + "step": 11170, + "token_acc": 0.9978637827734108 + }, + { + "epoch": 0.5919141248022872, + "grad_norm": 0.10094088315963745, + "learning_rate": 7.534997826974369e-06, + "loss": 0.004924702644348145, + "step": 11180, + "token_acc": 0.9979309194059737 + }, + { + "epoch": 0.5924435649854735, + "grad_norm": 0.09660059958696365, + "learning_rate": 7.51838308840021e-06, + "loss": 0.0048106517642736435, + "step": 11190, + "token_acc": 0.9979398984128043 + }, + { + "epoch": 0.5929730051686598, + "grad_norm": 0.0771661177277565, + "learning_rate": 7.501775646686316e-06, + "loss": 0.005234928429126739, + "step": 11200, + "token_acc": 0.9979971793609334 + }, + { + "epoch": 0.593502445351846, + "grad_norm": 0.06382273137569427, + "learning_rate": 7.485175550664634e-06, + "loss": 0.004693141579627991, + "step": 11210, + "token_acc": 0.9980566653043863 + }, + { + "epoch": 0.5940318855350324, + "grad_norm": 0.0865858867764473, + "learning_rate": 7.468582849145512e-06, + "loss": 0.005297693610191345, + "step": 11220, + "token_acc": 0.9977970995143606 + }, + { + "epoch": 0.5945613257182187, + "grad_norm": 0.0804460346698761, + "learning_rate": 7.451997590917548e-06, + "loss": 0.004745312407612801, + "step": 11230, + "token_acc": 0.9980219009781992 + }, + { + "epoch": 0.595090765901405, + "grad_norm": 0.08067922294139862, + "learning_rate": 7.435419824747463e-06, + "loss": 0.004882035776972771, + "step": 11240, + "token_acc": 0.9980396736682293 + }, + { + "epoch": 0.5956202060845913, + "grad_norm": 0.08097436279058456, + "learning_rate": 7.418849599379951e-06, + "loss": 0.004671842604875564, + "step": 11250, + "token_acc": 0.9980901865612516 + }, + { + "epoch": 0.5961496462677776, + "grad_norm": 0.08745463192462921, + "learning_rate": 7.402286963537518e-06, + "loss": 0.005312366411089897, + "step": 11260, + "token_acc": 0.9978460690760638 + }, + { + "epoch": 0.5966790864509639, + "grad_norm": 0.1108814999461174, + "learning_rate": 7.385731965920368e-06, + "loss": 0.004885765910148621, + "step": 11270, + "token_acc": 0.9979888174914462 + }, + { + "epoch": 0.5972085266341502, + "grad_norm": 0.07561526447534561, + "learning_rate": 7.369184655206243e-06, + "loss": 0.005072978138923645, + "step": 11280, + "token_acc": 0.997872340425532 + }, + { + "epoch": 0.5977379668173365, + "grad_norm": 0.07288990169763565, + "learning_rate": 7.352645080050281e-06, + "loss": 0.0050184354186058044, + "step": 11290, + "token_acc": 0.9979477082742396 + }, + { + "epoch": 0.5982674070005228, + "grad_norm": 0.0877259150147438, + "learning_rate": 7.336113289084872e-06, + "loss": 0.005053167417645455, + "step": 11300, + "token_acc": 0.9978811438486436 + }, + { + "epoch": 0.5987968471837092, + "grad_norm": 0.09617678821086884, + "learning_rate": 7.319589330919523e-06, + "loss": 0.0047162756323814396, + "step": 11310, + "token_acc": 0.9980055244469295 + }, + { + "epoch": 0.5993262873668954, + "grad_norm": 0.10047056525945663, + "learning_rate": 7.303073254140708e-06, + "loss": 0.004900079220533371, + "step": 11320, + "token_acc": 0.9980644407735563 + }, + { + "epoch": 0.5998557275500818, + "grad_norm": 0.09541192650794983, + "learning_rate": 7.286565107311725e-06, + "loss": 0.004744534194469452, + "step": 11330, + "token_acc": 0.9978975822195525 + }, + { + "epoch": 0.600385167733268, + "grad_norm": 0.07697328180074692, + "learning_rate": 7.270064938972556e-06, + "loss": 0.004664310067892074, + "step": 11340, + "token_acc": 0.9980476897662234 + }, + { + "epoch": 0.6009146079164543, + "grad_norm": 0.08482483774423599, + "learning_rate": 7.253572797639724e-06, + "loss": 0.00455038771033287, + "step": 11350, + "token_acc": 0.9981966186599874 + }, + { + "epoch": 0.6014440480996407, + "grad_norm": 0.10197839140892029, + "learning_rate": 7.237088731806148e-06, + "loss": 0.0053408615291118625, + "step": 11360, + "token_acc": 0.9977717132770837 + }, + { + "epoch": 0.6019734882828269, + "grad_norm": 0.068539559841156, + "learning_rate": 7.220612789941005e-06, + "loss": 0.005031901597976685, + "step": 11370, + "token_acc": 0.9979475712926964 + }, + { + "epoch": 0.6025029284660133, + "grad_norm": 0.07528601586818695, + "learning_rate": 7.204145020489581e-06, + "loss": 0.004945235699415207, + "step": 11380, + "token_acc": 0.9981058709655886 + }, + { + "epoch": 0.6030323686491995, + "grad_norm": 0.09659932553768158, + "learning_rate": 7.187685471873138e-06, + "loss": 0.00477801039814949, + "step": 11390, + "token_acc": 0.9979889853137517 + }, + { + "epoch": 0.6035618088323859, + "grad_norm": 0.09083124995231628, + "learning_rate": 7.171234192488756e-06, + "loss": 0.004640892893075943, + "step": 11400, + "token_acc": 0.9981812265874638 + }, + { + "epoch": 0.6040912490155722, + "grad_norm": 0.0744200348854065, + "learning_rate": 7.154791230709208e-06, + "loss": 0.004779967293143272, + "step": 11410, + "token_acc": 0.9979985155657113 + }, + { + "epoch": 0.6046206891987584, + "grad_norm": 0.0739501565694809, + "learning_rate": 7.13835663488281e-06, + "loss": 0.004353340715169907, + "step": 11420, + "token_acc": 0.9981734321959682 + }, + { + "epoch": 0.6051501293819448, + "grad_norm": 0.07321074604988098, + "learning_rate": 7.121930453333277e-06, + "loss": 0.00450860857963562, + "step": 11430, + "token_acc": 0.9982318303891641 + }, + { + "epoch": 0.605679569565131, + "grad_norm": 0.10983549803495407, + "learning_rate": 7.105512734359586e-06, + "loss": 0.005035143345594406, + "step": 11440, + "token_acc": 0.9978652790980804 + }, + { + "epoch": 0.6062090097483174, + "grad_norm": 0.08457458764314651, + "learning_rate": 7.08910352623582e-06, + "loss": 0.0048677843064069744, + "step": 11450, + "token_acc": 0.998006555845629 + }, + { + "epoch": 0.6067384499315037, + "grad_norm": 0.07446708530187607, + "learning_rate": 7.07270287721105e-06, + "loss": 0.0046407092362642285, + "step": 11460, + "token_acc": 0.9982052074028934 + }, + { + "epoch": 0.60726789011469, + "grad_norm": 0.0804063156247139, + "learning_rate": 7.056310835509178e-06, + "loss": 0.004639324173331261, + "step": 11470, + "token_acc": 0.9980461574429712 + }, + { + "epoch": 0.6077973302978763, + "grad_norm": 0.09055532515048981, + "learning_rate": 7.039927449328792e-06, + "loss": 0.004808081313967705, + "step": 11480, + "token_acc": 0.9980221485975614 + }, + { + "epoch": 0.6083267704810625, + "grad_norm": 0.07311367988586426, + "learning_rate": 7.023552766843034e-06, + "loss": 0.004499221593141556, + "step": 11490, + "token_acc": 0.9981564744450654 + }, + { + "epoch": 0.6088562106642489, + "grad_norm": 0.08466446399688721, + "learning_rate": 7.007186836199448e-06, + "loss": 0.004766301438212395, + "step": 11500, + "token_acc": 0.9979574482275653 + }, + { + "epoch": 0.6093856508474352, + "grad_norm": 0.08972243964672089, + "learning_rate": 6.990829705519853e-06, + "loss": 0.004839757829904557, + "step": 11510, + "token_acc": 0.9980470380076449 + }, + { + "epoch": 0.6099150910306215, + "grad_norm": 0.09134890139102936, + "learning_rate": 6.974481422900186e-06, + "loss": 0.00556851476430893, + "step": 11520, + "token_acc": 0.9977480295258351 + }, + { + "epoch": 0.6104445312138078, + "grad_norm": 0.06738061457872391, + "learning_rate": 6.958142036410374e-06, + "loss": 0.004838953167200089, + "step": 11530, + "token_acc": 0.9979642746894267 + }, + { + "epoch": 0.610973971396994, + "grad_norm": 0.07727788388729095, + "learning_rate": 6.941811594094176e-06, + "loss": 0.004819876700639725, + "step": 11540, + "token_acc": 0.9979653611067102 + }, + { + "epoch": 0.6115034115801804, + "grad_norm": 0.08591711521148682, + "learning_rate": 6.925490143969063e-06, + "loss": 0.004954057931900025, + "step": 11550, + "token_acc": 0.9979566475675766 + }, + { + "epoch": 0.6120328517633667, + "grad_norm": 0.0821395143866539, + "learning_rate": 6.909177734026058e-06, + "loss": 0.004397475719451904, + "step": 11560, + "token_acc": 0.9981239994663821 + }, + { + "epoch": 0.612562291946553, + "grad_norm": 0.09497993439435959, + "learning_rate": 6.892874412229606e-06, + "loss": 0.004662252217531204, + "step": 11570, + "token_acc": 0.9982063154382013 + }, + { + "epoch": 0.6130917321297393, + "grad_norm": 0.08247041702270508, + "learning_rate": 6.8765802265174274e-06, + "loss": 0.00491764098405838, + "step": 11580, + "token_acc": 0.9980991129193624 + }, + { + "epoch": 0.6136211723129256, + "grad_norm": 0.09365751594305038, + "learning_rate": 6.8602952248003816e-06, + "loss": 0.004528368264436722, + "step": 11590, + "token_acc": 0.9980636168632262 + }, + { + "epoch": 0.6141506124961119, + "grad_norm": 0.06582755595445633, + "learning_rate": 6.844019454962325e-06, + "loss": 0.00482339933514595, + "step": 11600, + "token_acc": 0.9979151029939121 + }, + { + "epoch": 0.6146800526792983, + "grad_norm": 0.09927351772785187, + "learning_rate": 6.827752964859965e-06, + "loss": 0.005013743415474892, + "step": 11610, + "token_acc": 0.9979643596075553 + }, + { + "epoch": 0.6152094928624845, + "grad_norm": 0.08178234845399857, + "learning_rate": 6.811495802322723e-06, + "loss": 0.004760830104351044, + "step": 11620, + "token_acc": 0.9981395271228579 + }, + { + "epoch": 0.6157389330456708, + "grad_norm": 0.08250564336776733, + "learning_rate": 6.7952480151525955e-06, + "loss": 0.005050058290362358, + "step": 11630, + "token_acc": 0.9980393792758218 + }, + { + "epoch": 0.6162683732288572, + "grad_norm": 0.07570900022983551, + "learning_rate": 6.779009651124015e-06, + "loss": 0.004355224221944809, + "step": 11640, + "token_acc": 0.9982311960285345 + }, + { + "epoch": 0.6167978134120434, + "grad_norm": 0.08450239896774292, + "learning_rate": 6.762780757983701e-06, + "loss": 0.004363410174846649, + "step": 11650, + "token_acc": 0.9982794046405961 + }, + { + "epoch": 0.6173272535952298, + "grad_norm": 0.07970098406076431, + "learning_rate": 6.746561383450532e-06, + "loss": 0.004891921579837799, + "step": 11660, + "token_acc": 0.997989337649444 + }, + { + "epoch": 0.617856693778416, + "grad_norm": 0.09942697733640671, + "learning_rate": 6.730351575215389e-06, + "loss": 0.004778019338846207, + "step": 11670, + "token_acc": 0.9980645860967207 + }, + { + "epoch": 0.6183861339616024, + "grad_norm": 0.07975873351097107, + "learning_rate": 6.714151380941032e-06, + "loss": 0.005003124102950096, + "step": 11680, + "token_acc": 0.9979040191059931 + }, + { + "epoch": 0.6189155741447887, + "grad_norm": 0.0885813757777214, + "learning_rate": 6.6979608482619495e-06, + "loss": 0.004332273826003074, + "step": 11690, + "token_acc": 0.9979986490881345 + }, + { + "epoch": 0.6194450143279749, + "grad_norm": 0.08120742440223694, + "learning_rate": 6.6817800247842234e-06, + "loss": 0.004646043106913566, + "step": 11700, + "token_acc": 0.9980324643384161 + }, + { + "epoch": 0.6199744545111613, + "grad_norm": 0.09990958869457245, + "learning_rate": 6.665608958085386e-06, + "loss": 0.004864935949444771, + "step": 11710, + "token_acc": 0.9979055232436853 + }, + { + "epoch": 0.6205038946943475, + "grad_norm": 0.09443415701389313, + "learning_rate": 6.6494476957142785e-06, + "loss": 0.0047955639660358425, + "step": 11720, + "token_acc": 0.998039559192799 + }, + { + "epoch": 0.6210333348775339, + "grad_norm": 0.07422012090682983, + "learning_rate": 6.633296285190917e-06, + "loss": 0.00441572442650795, + "step": 11730, + "token_acc": 0.9982308417688244 + }, + { + "epoch": 0.6215627750607202, + "grad_norm": 0.07735058665275574, + "learning_rate": 6.617154774006348e-06, + "loss": 0.004788857325911522, + "step": 11740, + "token_acc": 0.9980974156562664 + }, + { + "epoch": 0.6220922152439065, + "grad_norm": 0.07339327037334442, + "learning_rate": 6.601023209622512e-06, + "loss": 0.00441078320145607, + "step": 11750, + "token_acc": 0.9981648774221533 + }, + { + "epoch": 0.6226216554270928, + "grad_norm": 0.09270325303077698, + "learning_rate": 6.584901639472103e-06, + "loss": 0.004966029524803161, + "step": 11760, + "token_acc": 0.9979976639412648 + }, + { + "epoch": 0.623151095610279, + "grad_norm": 0.10555166751146317, + "learning_rate": 6.568790110958424e-06, + "loss": 0.0045115090906620026, + "step": 11770, + "token_acc": 0.9981311997864228 + }, + { + "epoch": 0.6236805357934654, + "grad_norm": 0.05777881294488907, + "learning_rate": 6.552688671455254e-06, + "loss": 0.004300295561552048, + "step": 11780, + "token_acc": 0.9981566897149958 + }, + { + "epoch": 0.6242099759766517, + "grad_norm": 0.0750829428434372, + "learning_rate": 6.536597368306707e-06, + "loss": 0.00451211966574192, + "step": 11790, + "token_acc": 0.9982317271521632 + }, + { + "epoch": 0.624739416159838, + "grad_norm": 0.07661847025156021, + "learning_rate": 6.520516248827099e-06, + "loss": 0.00444275289773941, + "step": 11800, + "token_acc": 0.9981551817688551 + }, + { + "epoch": 0.6252688563430243, + "grad_norm": 0.08125019073486328, + "learning_rate": 6.504445360300785e-06, + "loss": 0.00469810776412487, + "step": 11810, + "token_acc": 0.998147483665312 + }, + { + "epoch": 0.6257982965262106, + "grad_norm": 0.07840893417596817, + "learning_rate": 6.488384749982054e-06, + "loss": 0.004289882630109787, + "step": 11820, + "token_acc": 0.9982726104444481 + }, + { + "epoch": 0.6263277367093969, + "grad_norm": 0.07432328164577484, + "learning_rate": 6.472334465094969e-06, + "loss": 0.004120988398790359, + "step": 11830, + "token_acc": 0.9983311638463336 + }, + { + "epoch": 0.6268571768925832, + "grad_norm": 0.07049127668142319, + "learning_rate": 6.4562945528332255e-06, + "loss": 0.0046249967068433765, + "step": 11840, + "token_acc": 0.9981224183251971 + }, + { + "epoch": 0.6273866170757695, + "grad_norm": 0.09359797835350037, + "learning_rate": 6.4402650603600295e-06, + "loss": 0.004934298992156983, + "step": 11850, + "token_acc": 0.99803105263597 + }, + { + "epoch": 0.6279160572589558, + "grad_norm": 0.0778711810708046, + "learning_rate": 6.424246034807942e-06, + "loss": 0.004534978419542313, + "step": 11860, + "token_acc": 0.9981221205848968 + }, + { + "epoch": 0.6284454974421421, + "grad_norm": 0.07564844936132431, + "learning_rate": 6.40823752327875e-06, + "loss": 0.0046748299151659015, + "step": 11870, + "token_acc": 0.9980807743658211 + }, + { + "epoch": 0.6289749376253284, + "grad_norm": 0.07255757600069046, + "learning_rate": 6.392239572843328e-06, + "loss": 0.004446384310722351, + "step": 11880, + "token_acc": 0.9982485842722868 + }, + { + "epoch": 0.6295043778085148, + "grad_norm": 0.09227579832077026, + "learning_rate": 6.376252230541493e-06, + "loss": 0.004597494006156921, + "step": 11890, + "token_acc": 0.9981305135245662 + }, + { + "epoch": 0.630033817991701, + "grad_norm": 0.0705723688006401, + "learning_rate": 6.360275543381869e-06, + "loss": 0.0047139361500740055, + "step": 11900, + "token_acc": 0.9982631930527722 + }, + { + "epoch": 0.6305632581748873, + "grad_norm": 0.10290166735649109, + "learning_rate": 6.344309558341757e-06, + "loss": 0.004432447254657745, + "step": 11910, + "token_acc": 0.9981316362362479 + }, + { + "epoch": 0.6310926983580736, + "grad_norm": 0.07825620472431183, + "learning_rate": 6.328354322366982e-06, + "loss": 0.004492341354489326, + "step": 11920, + "token_acc": 0.9981315115571016 + }, + { + "epoch": 0.6316221385412599, + "grad_norm": 0.08164883404970169, + "learning_rate": 6.31240988237177e-06, + "loss": 0.004306265339255333, + "step": 11930, + "token_acc": 0.9982485112345494 + }, + { + "epoch": 0.6321515787244463, + "grad_norm": 0.07040442526340485, + "learning_rate": 6.296476285238595e-06, + "loss": 0.004504130035638809, + "step": 11940, + "token_acc": 0.9980964943771445 + }, + { + "epoch": 0.6326810189076325, + "grad_norm": 0.07695594429969788, + "learning_rate": 6.280553577818055e-06, + "loss": 0.004163908213376999, + "step": 11950, + "token_acc": 0.9983394664597258 + }, + { + "epoch": 0.6332104590908189, + "grad_norm": 0.08470465987920761, + "learning_rate": 6.264641806928726e-06, + "loss": 0.004248766601085663, + "step": 11960, + "token_acc": 0.9982052823132659 + }, + { + "epoch": 0.6337398992740052, + "grad_norm": 0.0920063704252243, + "learning_rate": 6.248741019357028e-06, + "loss": 0.004523378238081932, + "step": 11970, + "token_acc": 0.9981392321620787 + }, + { + "epoch": 0.6342693394571914, + "grad_norm": 0.09320084005594254, + "learning_rate": 6.232851261857087e-06, + "loss": 0.004471444711089134, + "step": 11980, + "token_acc": 0.9981728988336587 + }, + { + "epoch": 0.6347987796403778, + "grad_norm": 0.09099368751049042, + "learning_rate": 6.216972581150588e-06, + "loss": 0.0047561291605234144, + "step": 11990, + "token_acc": 0.9979133976563282 + }, + { + "epoch": 0.635328219823564, + "grad_norm": 0.08445628732442856, + "learning_rate": 6.20110502392666e-06, + "loss": 0.004448097571730614, + "step": 12000, + "token_acc": 0.9981892372266123 + }, + { + "epoch": 0.635328219823564, + "eval_loss": 0.00658675841987133, + "eval_runtime": 712.239, + "eval_samples_per_second": 16.817, + "eval_steps_per_second": 1.052, + "eval_token_acc": 0.9974125151850841, + "step": 12000 + }, + { + "epoch": 0.6358576600067504, + "grad_norm": 0.07349839806556702, + "learning_rate": 6.185248636841714e-06, + "loss": 0.004632272571325302, + "step": 12010, + "token_acc": 0.9979978977926823 + }, + { + "epoch": 0.6363871001899367, + "grad_norm": 0.07285565882921219, + "learning_rate": 6.169403466519323e-06, + "loss": 0.004561829566955567, + "step": 12020, + "token_acc": 0.9981731884118418 + }, + { + "epoch": 0.636916540373123, + "grad_norm": 0.06152743846178055, + "learning_rate": 6.1535695595500726e-06, + "loss": 0.004215127974748612, + "step": 12030, + "token_acc": 0.9982721057771767 + }, + { + "epoch": 0.6374459805563093, + "grad_norm": 0.08216620236635208, + "learning_rate": 6.137746962491433e-06, + "loss": 0.004393155872821808, + "step": 12040, + "token_acc": 0.998230738416344 + }, + { + "epoch": 0.6379754207394955, + "grad_norm": 0.10928614437580109, + "learning_rate": 6.121935721867624e-06, + "loss": 0.00498277060687542, + "step": 12050, + "token_acc": 0.9978553307964484 + }, + { + "epoch": 0.6385048609226819, + "grad_norm": 0.08736424893140793, + "learning_rate": 6.106135884169466e-06, + "loss": 0.004545886814594269, + "step": 12060, + "token_acc": 0.9980892304354719 + }, + { + "epoch": 0.6390343011058682, + "grad_norm": 0.05574479326605797, + "learning_rate": 6.090347495854257e-06, + "loss": 0.004200483858585358, + "step": 12070, + "token_acc": 0.998215164430062 + }, + { + "epoch": 0.6395637412890545, + "grad_norm": 0.09557504206895828, + "learning_rate": 6.074570603345625e-06, + "loss": 0.004940097033977508, + "step": 12080, + "token_acc": 0.9980487316755892 + }, + { + "epoch": 0.6400931814722408, + "grad_norm": 0.09401912987232208, + "learning_rate": 6.058805253033391e-06, + "loss": 0.00471298024058342, + "step": 12090, + "token_acc": 0.9980904413665436 + }, + { + "epoch": 0.640622621655427, + "grad_norm": 0.0804917961359024, + "learning_rate": 6.043051491273452e-06, + "loss": 0.0043868452310562136, + "step": 12100, + "token_acc": 0.9981813479715356 + }, + { + "epoch": 0.6411520618386134, + "grad_norm": 0.06310533732175827, + "learning_rate": 6.027309364387617e-06, + "loss": 0.004273098334670067, + "step": 12110, + "token_acc": 0.9983819445023645 + }, + { + "epoch": 0.6416815020217997, + "grad_norm": 0.06879176944494247, + "learning_rate": 6.011578918663492e-06, + "loss": 0.0037895090878009795, + "step": 12120, + "token_acc": 0.9984136393617821 + }, + { + "epoch": 0.642210942204986, + "grad_norm": 0.11281027644872665, + "learning_rate": 5.995860200354335e-06, + "loss": 0.004349033161997795, + "step": 12130, + "token_acc": 0.9982311812705458 + }, + { + "epoch": 0.6427403823881723, + "grad_norm": 0.07777886837720871, + "learning_rate": 5.980153255678921e-06, + "loss": 0.004372740536928177, + "step": 12140, + "token_acc": 0.9982045029396045 + }, + { + "epoch": 0.6432698225713586, + "grad_norm": 0.08442207425832748, + "learning_rate": 5.9644581308214015e-06, + "loss": 0.004363740980625153, + "step": 12150, + "token_acc": 0.998365181706717 + }, + { + "epoch": 0.6437992627545449, + "grad_norm": 0.08432236313819885, + "learning_rate": 5.948774871931178e-06, + "loss": 0.004592880979180336, + "step": 12160, + "token_acc": 0.9981146241761909 + }, + { + "epoch": 0.6443287029377313, + "grad_norm": 0.08524004369974136, + "learning_rate": 5.933103525122761e-06, + "loss": 0.004293068498373032, + "step": 12170, + "token_acc": 0.9982966500784877 + }, + { + "epoch": 0.6448581431209175, + "grad_norm": 0.07770753651857376, + "learning_rate": 5.917444136475636e-06, + "loss": 0.004217541217803955, + "step": 12180, + "token_acc": 0.9983322492953753 + }, + { + "epoch": 0.6453875833041038, + "grad_norm": 0.07285448908805847, + "learning_rate": 5.901796752034128e-06, + "loss": 0.004719100892543793, + "step": 12190, + "token_acc": 0.9979399155949223 + }, + { + "epoch": 0.6459170234872901, + "grad_norm": 0.08697082847356796, + "learning_rate": 5.886161417807257e-06, + "loss": 0.004762466251850128, + "step": 12200, + "token_acc": 0.9980488292975785 + }, + { + "epoch": 0.6464464636704764, + "grad_norm": 0.06786496192216873, + "learning_rate": 5.870538179768624e-06, + "loss": 0.004261656105518341, + "step": 12210, + "token_acc": 0.9983159228659324 + }, + { + "epoch": 0.6469759038536628, + "grad_norm": 0.07730599492788315, + "learning_rate": 5.85492708385625e-06, + "loss": 0.0043582689017057415, + "step": 12220, + "token_acc": 0.9981235300985772 + }, + { + "epoch": 0.647505344036849, + "grad_norm": 0.08336374908685684, + "learning_rate": 5.839328175972465e-06, + "loss": 0.003921012207865715, + "step": 12230, + "token_acc": 0.9984650677789364 + }, + { + "epoch": 0.6480347842200354, + "grad_norm": 0.10513246059417725, + "learning_rate": 5.823741501983753e-06, + "loss": 0.004284816235303879, + "step": 12240, + "token_acc": 0.9982570406384735 + }, + { + "epoch": 0.6485642244032216, + "grad_norm": 0.0765002965927124, + "learning_rate": 5.808167107720633e-06, + "loss": 0.004452315717935562, + "step": 12250, + "token_acc": 0.9981734169613665 + }, + { + "epoch": 0.6490936645864079, + "grad_norm": 0.06994980573654175, + "learning_rate": 5.792605038977509e-06, + "loss": 0.004187068343162537, + "step": 12260, + "token_acc": 0.9982558915815475 + }, + { + "epoch": 0.6496231047695943, + "grad_norm": 0.086262047290802, + "learning_rate": 5.777055341512556e-06, + "loss": 0.004642238467931747, + "step": 12270, + "token_acc": 0.9980718506894595 + }, + { + "epoch": 0.6501525449527805, + "grad_norm": 0.10271665453910828, + "learning_rate": 5.761518061047555e-06, + "loss": 0.0042902126908302305, + "step": 12280, + "token_acc": 0.998206584753468 + }, + { + "epoch": 0.6506819851359669, + "grad_norm": 0.07652673870325089, + "learning_rate": 5.745993243267791e-06, + "loss": 0.004588426649570465, + "step": 12290, + "token_acc": 0.9980894376772902 + }, + { + "epoch": 0.6512114253191532, + "grad_norm": 0.08382252603769302, + "learning_rate": 5.730480933821898e-06, + "loss": 0.004401729255914688, + "step": 12300, + "token_acc": 0.9981725329194412 + }, + { + "epoch": 0.6517408655023395, + "grad_norm": 0.07986811548471451, + "learning_rate": 5.714981178321735e-06, + "loss": 0.0044216692447662355, + "step": 12310, + "token_acc": 0.9982491391600871 + }, + { + "epoch": 0.6522703056855258, + "grad_norm": 0.08010311424732208, + "learning_rate": 5.6994940223422444e-06, + "loss": 0.003999342024326324, + "step": 12320, + "token_acc": 0.998349325974773 + }, + { + "epoch": 0.652799745868712, + "grad_norm": 0.07987107336521149, + "learning_rate": 5.684019511421321e-06, + "loss": 0.004407377168536186, + "step": 12330, + "token_acc": 0.9982316091522568 + }, + { + "epoch": 0.6533291860518984, + "grad_norm": 0.08694404363632202, + "learning_rate": 5.6685576910596804e-06, + "loss": 0.004183653742074966, + "step": 12340, + "token_acc": 0.9983638989290216 + }, + { + "epoch": 0.6538586262350847, + "grad_norm": 0.09605570882558823, + "learning_rate": 5.653108606720729e-06, + "loss": 0.004340231046080589, + "step": 12350, + "token_acc": 0.9981561975955482 + }, + { + "epoch": 0.654388066418271, + "grad_norm": 0.08879438787698746, + "learning_rate": 5.637672303830407e-06, + "loss": 0.004205046594142914, + "step": 12360, + "token_acc": 0.9982810844187813 + }, + { + "epoch": 0.6549175066014573, + "grad_norm": 0.07693406939506531, + "learning_rate": 5.62224882777709e-06, + "loss": 0.003913236036896706, + "step": 12370, + "token_acc": 0.9984308226495726 + }, + { + "epoch": 0.6554469467846435, + "grad_norm": 0.08469942212104797, + "learning_rate": 5.606838223911431e-06, + "loss": 0.004342402890324592, + "step": 12380, + "token_acc": 0.998123357938196 + }, + { + "epoch": 0.6559763869678299, + "grad_norm": 0.07652900367975235, + "learning_rate": 5.591440537546239e-06, + "loss": 0.004285915940999985, + "step": 12390, + "token_acc": 0.9981235770459265 + }, + { + "epoch": 0.6565058271510162, + "grad_norm": 0.11217132955789566, + "learning_rate": 5.576055813956329e-06, + "loss": 0.00395175963640213, + "step": 12400, + "token_acc": 0.9985149588693665 + }, + { + "epoch": 0.6570352673342025, + "grad_norm": 0.07602044939994812, + "learning_rate": 5.5606840983784125e-06, + "loss": 0.004246373474597931, + "step": 12410, + "token_acc": 0.9982835192854107 + }, + { + "epoch": 0.6575647075173888, + "grad_norm": 0.09871669858694077, + "learning_rate": 5.545325436010948e-06, + "loss": 0.004282528534531593, + "step": 12420, + "token_acc": 0.9982239935963713 + }, + { + "epoch": 0.6580941477005751, + "grad_norm": 0.06864234060049057, + "learning_rate": 5.529979872014015e-06, + "loss": 0.0045604448765516285, + "step": 12430, + "token_acc": 0.9982319336141112 + }, + { + "epoch": 0.6586235878837614, + "grad_norm": 0.09332428127527237, + "learning_rate": 5.514647451509175e-06, + "loss": 0.003665485978126526, + "step": 12440, + "token_acc": 0.9986560037398157 + }, + { + "epoch": 0.6591530280669478, + "grad_norm": 0.07839623093605042, + "learning_rate": 5.499328219579348e-06, + "loss": 0.004023130238056183, + "step": 12450, + "token_acc": 0.9983490644698663 + }, + { + "epoch": 0.659682468250134, + "grad_norm": 0.08892426639795303, + "learning_rate": 5.48402222126867e-06, + "loss": 0.004008069634437561, + "step": 12460, + "token_acc": 0.9983657950907151 + }, + { + "epoch": 0.6602119084333203, + "grad_norm": 0.0657774806022644, + "learning_rate": 5.468729501582368e-06, + "loss": 0.004086549580097199, + "step": 12470, + "token_acc": 0.9984308881340088 + }, + { + "epoch": 0.6607413486165066, + "grad_norm": 0.08652454614639282, + "learning_rate": 5.453450105486625e-06, + "loss": 0.00394832082092762, + "step": 12480, + "token_acc": 0.9983384959631296 + }, + { + "epoch": 0.6612707887996929, + "grad_norm": 0.0706365555524826, + "learning_rate": 5.438184077908451e-06, + "loss": 0.0041192807257175446, + "step": 12490, + "token_acc": 0.9983228896361255 + }, + { + "epoch": 0.6618002289828793, + "grad_norm": 0.08671177178621292, + "learning_rate": 5.422931463735534e-06, + "loss": 0.004038792848587036, + "step": 12500, + "token_acc": 0.9983813369822782 + }, + { + "epoch": 0.6623296691660655, + "grad_norm": 0.08677837252616882, + "learning_rate": 5.407692307816139e-06, + "loss": 0.004159288480877876, + "step": 12510, + "token_acc": 0.9982144645062243 + }, + { + "epoch": 0.6628591093492519, + "grad_norm": 0.06664512306451797, + "learning_rate": 5.392466654958949e-06, + "loss": 0.0037685029208660127, + "step": 12520, + "token_acc": 0.9984993246961132 + }, + { + "epoch": 0.6633885495324381, + "grad_norm": 0.07323741167783737, + "learning_rate": 5.3772545499329445e-06, + "loss": 0.004552188515663147, + "step": 12530, + "token_acc": 0.9982488325550367 + }, + { + "epoch": 0.6639179897156244, + "grad_norm": 0.06923746317625046, + "learning_rate": 5.362056037467274e-06, + "loss": 0.004518025740981102, + "step": 12540, + "token_acc": 0.9980393465600961 + }, + { + "epoch": 0.6644474298988108, + "grad_norm": 0.06403191387653351, + "learning_rate": 5.346871162251114e-06, + "loss": 0.0036100316792726517, + "step": 12550, + "token_acc": 0.9985408154756942 + }, + { + "epoch": 0.664976870081997, + "grad_norm": 0.059143468737602234, + "learning_rate": 5.3316999689335455e-06, + "loss": 0.0039113923907279965, + "step": 12560, + "token_acc": 0.9984648244560139 + }, + { + "epoch": 0.6655063102651834, + "grad_norm": 0.07519901543855667, + "learning_rate": 5.316542502123419e-06, + "loss": 0.004204827919602394, + "step": 12570, + "token_acc": 0.9982819445718623 + }, + { + "epoch": 0.6660357504483696, + "grad_norm": 0.0802607536315918, + "learning_rate": 5.301398806389224e-06, + "loss": 0.0038543429225683214, + "step": 12580, + "token_acc": 0.9984312547458716 + }, + { + "epoch": 0.666565190631556, + "grad_norm": 0.07866298407316208, + "learning_rate": 5.2862689262589615e-06, + "loss": 0.0038709498941898344, + "step": 12590, + "token_acc": 0.9982722644186629 + }, + { + "epoch": 0.6670946308147423, + "grad_norm": 0.09237617999315262, + "learning_rate": 5.27115290622e-06, + "loss": 0.0037260249257087706, + "step": 12600, + "token_acc": 0.9984736644563994 + }, + { + "epoch": 0.6676240709979285, + "grad_norm": 0.08292164653539658, + "learning_rate": 5.256050790718963e-06, + "loss": 0.0036970093846321107, + "step": 12610, + "token_acc": 0.9985892213169494 + }, + { + "epoch": 0.6681535111811149, + "grad_norm": 0.08283206075429916, + "learning_rate": 5.240962624161585e-06, + "loss": 0.004396694898605347, + "step": 12620, + "token_acc": 0.9983149394796333 + }, + { + "epoch": 0.6686829513643012, + "grad_norm": 0.11574340611696243, + "learning_rate": 5.2258884509125955e-06, + "loss": 0.004080820083618164, + "step": 12630, + "token_acc": 0.9983139127241161 + }, + { + "epoch": 0.6692123915474875, + "grad_norm": 0.08617594093084335, + "learning_rate": 5.21082831529556e-06, + "loss": 0.004264967888593674, + "step": 12640, + "token_acc": 0.9981560437543283 + }, + { + "epoch": 0.6697418317306738, + "grad_norm": 0.08009687811136246, + "learning_rate": 5.195782261592782e-06, + "loss": 0.004412015527486801, + "step": 12650, + "token_acc": 0.9981997449638699 + }, + { + "epoch": 0.67027127191386, + "grad_norm": 0.08463038504123688, + "learning_rate": 5.180750334045157e-06, + "loss": 0.004148414731025696, + "step": 12660, + "token_acc": 0.9982652788897785 + }, + { + "epoch": 0.6708007120970464, + "grad_norm": 0.06682711839675903, + "learning_rate": 5.165732576852045e-06, + "loss": 0.003902681916952133, + "step": 12670, + "token_acc": 0.9984321966759234 + }, + { + "epoch": 0.6713301522802327, + "grad_norm": 0.08391131460666656, + "learning_rate": 5.150729034171134e-06, + "loss": 0.004401078820228577, + "step": 12680, + "token_acc": 0.9982226450046311 + }, + { + "epoch": 0.671859592463419, + "grad_norm": 0.08375461399555206, + "learning_rate": 5.135739750118325e-06, + "loss": 0.004011020436882973, + "step": 12690, + "token_acc": 0.9982474003104606 + }, + { + "epoch": 0.6723890326466053, + "grad_norm": 0.08065548539161682, + "learning_rate": 5.120764768767584e-06, + "loss": 0.003962218761444092, + "step": 12700, + "token_acc": 0.9984229498347852 + }, + { + "epoch": 0.6729184728297916, + "grad_norm": 0.06963040679693222, + "learning_rate": 5.105804134150835e-06, + "loss": 0.004202116280794144, + "step": 12710, + "token_acc": 0.9984317781800285 + }, + { + "epoch": 0.6734479130129779, + "grad_norm": 0.08603809773921967, + "learning_rate": 5.090857890257798e-06, + "loss": 0.004096759855747223, + "step": 12720, + "token_acc": 0.9981984837238008 + }, + { + "epoch": 0.6739773531961643, + "grad_norm": 0.09896639734506607, + "learning_rate": 5.075926081035896e-06, + "loss": 0.004168228432536125, + "step": 12730, + "token_acc": 0.9983050847457627 + }, + { + "epoch": 0.6745067933793505, + "grad_norm": 0.08234656602144241, + "learning_rate": 5.0610087503900994e-06, + "loss": 0.003743991255760193, + "step": 12740, + "token_acc": 0.9984323927689948 + }, + { + "epoch": 0.6750362335625368, + "grad_norm": 0.09244005382061005, + "learning_rate": 5.046105942182815e-06, + "loss": 0.00409916490316391, + "step": 12750, + "token_acc": 0.9983977702302371 + }, + { + "epoch": 0.6755656737457231, + "grad_norm": 0.07756570726633072, + "learning_rate": 5.031217700233738e-06, + "loss": 0.003738586604595184, + "step": 12760, + "token_acc": 0.9986147388448925 + }, + { + "epoch": 0.6760951139289094, + "grad_norm": 0.07345060259103775, + "learning_rate": 5.0163440683197405e-06, + "loss": 0.003959799185395241, + "step": 12770, + "token_acc": 0.9984899048890372 + }, + { + "epoch": 0.6766245541120958, + "grad_norm": 0.08256395906209946, + "learning_rate": 5.0014850901747355e-06, + "loss": 0.0038456574082374574, + "step": 12780, + "token_acc": 0.9985142275235178 + }, + { + "epoch": 0.677153994295282, + "grad_norm": 0.09645641595125198, + "learning_rate": 4.9866408094895435e-06, + "loss": 0.004287015646696091, + "step": 12790, + "token_acc": 0.9982394219296269 + }, + { + "epoch": 0.6776834344784683, + "grad_norm": 0.08996686339378357, + "learning_rate": 4.971811269911778e-06, + "loss": 0.00408768393099308, + "step": 12800, + "token_acc": 0.9983309270865498 + }, + { + "epoch": 0.6782128746616546, + "grad_norm": 0.07943252474069595, + "learning_rate": 4.956996515045695e-06, + "loss": 0.0037150956690311433, + "step": 12810, + "token_acc": 0.9984739190259767 + }, + { + "epoch": 0.6787423148448409, + "grad_norm": 0.09274609386920929, + "learning_rate": 4.942196588452089e-06, + "loss": 0.004604731872677803, + "step": 12820, + "token_acc": 0.9981645099658766 + }, + { + "epoch": 0.6792717550280273, + "grad_norm": 0.0696905255317688, + "learning_rate": 4.927411533648151e-06, + "loss": 0.0034449130296707155, + "step": 12830, + "token_acc": 0.9986232790988736 + }, + { + "epoch": 0.6798011952112135, + "grad_norm": 0.10087598860263824, + "learning_rate": 4.912641394107344e-06, + "loss": 0.0038737528026103973, + "step": 12840, + "token_acc": 0.9984483966765658 + }, + { + "epoch": 0.6803306353943999, + "grad_norm": 0.08246959745883942, + "learning_rate": 4.897886213259272e-06, + "loss": 0.003932100906968117, + "step": 12850, + "token_acc": 0.9983558671340343 + }, + { + "epoch": 0.6808600755775861, + "grad_norm": 0.09181206673383713, + "learning_rate": 4.883146034489558e-06, + "loss": 0.0039091378450393675, + "step": 12860, + "token_acc": 0.9983397989404748 + }, + { + "epoch": 0.6813895157607724, + "grad_norm": 0.09051388502120972, + "learning_rate": 4.868420901139711e-06, + "loss": 0.004027166962623596, + "step": 12870, + "token_acc": 0.9984058757250761 + }, + { + "epoch": 0.6819189559439588, + "grad_norm": 0.08767787367105484, + "learning_rate": 4.8537108565070045e-06, + "loss": 0.003683003410696983, + "step": 12880, + "token_acc": 0.998506179742462 + }, + { + "epoch": 0.682448396127145, + "grad_norm": 0.0805836021900177, + "learning_rate": 4.839015943844342e-06, + "loss": 0.003956511616706848, + "step": 12890, + "token_acc": 0.9983575805577556 + }, + { + "epoch": 0.6829778363103314, + "grad_norm": 0.07881326973438263, + "learning_rate": 4.824336206360137e-06, + "loss": 0.003999922797083855, + "step": 12900, + "token_acc": 0.9982742234005303 + }, + { + "epoch": 0.6835072764935176, + "grad_norm": 0.1055709719657898, + "learning_rate": 4.8096716872181755e-06, + "loss": 0.004221456125378609, + "step": 12910, + "token_acc": 0.9983151081416978 + }, + { + "epoch": 0.684036716676704, + "grad_norm": 0.08504793047904968, + "learning_rate": 4.795022429537508e-06, + "loss": 0.0040035687386989595, + "step": 12920, + "token_acc": 0.9984551664245036 + }, + { + "epoch": 0.6845661568598903, + "grad_norm": 0.07294493168592453, + "learning_rate": 4.780388476392295e-06, + "loss": 0.003948199003934861, + "step": 12930, + "token_acc": 0.9984064343342484 + }, + { + "epoch": 0.6850955970430765, + "grad_norm": 0.060608625411987305, + "learning_rate": 4.765769870811709e-06, + "loss": 0.0039102703332901, + "step": 12940, + "token_acc": 0.998490345719171 + }, + { + "epoch": 0.6856250372262629, + "grad_norm": 0.07788658887147903, + "learning_rate": 4.751166655779789e-06, + "loss": 0.0036722108721733092, + "step": 12950, + "token_acc": 0.9984729512095395 + }, + { + "epoch": 0.6861544774094492, + "grad_norm": 0.1033112183213234, + "learning_rate": 4.7365788742353235e-06, + "loss": 0.0038107864558696745, + "step": 12960, + "token_acc": 0.9984411080637229 + }, + { + "epoch": 0.6866839175926355, + "grad_norm": 0.10828004777431488, + "learning_rate": 4.72200656907172e-06, + "loss": 0.003976759687066078, + "step": 12970, + "token_acc": 0.9984574978113144 + }, + { + "epoch": 0.6872133577758218, + "grad_norm": 0.06854714453220367, + "learning_rate": 4.707449783136876e-06, + "loss": 0.003819131106138229, + "step": 12980, + "token_acc": 0.9984566098810338 + }, + { + "epoch": 0.6877427979590081, + "grad_norm": 0.07833156734704971, + "learning_rate": 4.692908559233063e-06, + "loss": 0.003779006004333496, + "step": 12990, + "token_acc": 0.998506827717949 + }, + { + "epoch": 0.6882722381421944, + "grad_norm": 0.08585453778505325, + "learning_rate": 4.678382940116789e-06, + "loss": 0.003643868863582611, + "step": 13000, + "token_acc": 0.9984905723316071 + }, + { + "epoch": 0.6888016783253808, + "grad_norm": 0.05784493312239647, + "learning_rate": 4.663872968498682e-06, + "loss": 0.00374244824051857, + "step": 13010, + "token_acc": 0.9986408171774025 + }, + { + "epoch": 0.689331118508567, + "grad_norm": 0.07995244115591049, + "learning_rate": 4.649378687043361e-06, + "loss": 0.004127905517816543, + "step": 13020, + "token_acc": 0.9982653078244625 + }, + { + "epoch": 0.6898605586917533, + "grad_norm": 0.07328928261995316, + "learning_rate": 4.634900138369303e-06, + "loss": 0.0041158322244882585, + "step": 13030, + "token_acc": 0.9983820254705887 + }, + { + "epoch": 0.6903899988749396, + "grad_norm": 0.08040270209312439, + "learning_rate": 4.620437365048732e-06, + "loss": 0.0037848860025405883, + "step": 13040, + "token_acc": 0.9984136128714441 + }, + { + "epoch": 0.6909194390581259, + "grad_norm": 0.08328527957201004, + "learning_rate": 4.605990409607487e-06, + "loss": 0.004001417756080627, + "step": 13050, + "token_acc": 0.9983489818721545 + }, + { + "epoch": 0.6914488792413123, + "grad_norm": 0.0825948491692543, + "learning_rate": 4.5915593145248926e-06, + "loss": 0.004089251905679703, + "step": 13060, + "token_acc": 0.9982313731051916 + }, + { + "epoch": 0.6919783194244985, + "grad_norm": 0.08563896268606186, + "learning_rate": 4.577144122233641e-06, + "loss": 0.003675517439842224, + "step": 13070, + "token_acc": 0.9984820304094347 + }, + { + "epoch": 0.6925077596076848, + "grad_norm": 0.07986423373222351, + "learning_rate": 4.5627448751196635e-06, + "loss": 0.003576098382472992, + "step": 13080, + "token_acc": 0.9984977340822407 + }, + { + "epoch": 0.6930371997908711, + "grad_norm": 0.07291054725646973, + "learning_rate": 4.548361615522006e-06, + "loss": 0.0037225954234600066, + "step": 13090, + "token_acc": 0.998556239881161 + }, + { + "epoch": 0.6935666399740574, + "grad_norm": 0.11106012761592865, + "learning_rate": 4.533994385732706e-06, + "loss": 0.003768979012966156, + "step": 13100, + "token_acc": 0.9984486037433691 + }, + { + "epoch": 0.6940960801572438, + "grad_norm": 0.10040315240621567, + "learning_rate": 4.51964322799667e-06, + "loss": 0.004118700325489044, + "step": 13110, + "token_acc": 0.9984066204503175 + }, + { + "epoch": 0.69462552034043, + "grad_norm": 0.13968099653720856, + "learning_rate": 4.505308184511544e-06, + "loss": 0.003689059242606163, + "step": 13120, + "token_acc": 0.9985563964986357 + }, + { + "epoch": 0.6951549605236164, + "grad_norm": 0.09743746370077133, + "learning_rate": 4.4909892974275885e-06, + "loss": 0.003868892788887024, + "step": 13130, + "token_acc": 0.9983906608296852 + }, + { + "epoch": 0.6956844007068026, + "grad_norm": 0.07281576842069626, + "learning_rate": 4.476686608847564e-06, + "loss": 0.0035872265696525575, + "step": 13140, + "token_acc": 0.9985145127100963 + }, + { + "epoch": 0.6962138408899889, + "grad_norm": 0.08898575603961945, + "learning_rate": 4.4624001608265995e-06, + "loss": 0.003576020523905754, + "step": 13150, + "token_acc": 0.9986560149591378 + }, + { + "epoch": 0.6967432810731753, + "grad_norm": 0.10269992798566818, + "learning_rate": 4.448129995372072e-06, + "loss": 0.0035270024091005325, + "step": 13160, + "token_acc": 0.9985493351119263 + }, + { + "epoch": 0.6972727212563615, + "grad_norm": 0.0980374738574028, + "learning_rate": 4.4338761544434824e-06, + "loss": 0.003872380033135414, + "step": 13170, + "token_acc": 0.9983739295035898 + }, + { + "epoch": 0.6978021614395479, + "grad_norm": 0.08846528828144073, + "learning_rate": 4.419638679952331e-06, + "loss": 0.003935432434082032, + "step": 13180, + "token_acc": 0.9982729563316591 + }, + { + "epoch": 0.6983316016227341, + "grad_norm": 0.10639336705207825, + "learning_rate": 4.405417613761988e-06, + "loss": 0.003411738947033882, + "step": 13190, + "token_acc": 0.9986399098835996 + }, + { + "epoch": 0.6988610418059205, + "grad_norm": 0.07377149164676666, + "learning_rate": 4.391212997687586e-06, + "loss": 0.0037175446748733522, + "step": 13200, + "token_acc": 0.9985235112071339 + }, + { + "epoch": 0.6993904819891068, + "grad_norm": 0.07839785516262054, + "learning_rate": 4.377024873495885e-06, + "loss": 0.003614959865808487, + "step": 13210, + "token_acc": 0.998532171302281 + }, + { + "epoch": 0.699919922172293, + "grad_norm": 0.09231901168823242, + "learning_rate": 4.362853282905154e-06, + "loss": 0.0037563085556030274, + "step": 13220, + "token_acc": 0.9983479900545664 + }, + { + "epoch": 0.7004493623554794, + "grad_norm": 0.07816437631845474, + "learning_rate": 4.348698267585047e-06, + "loss": 0.0038487941026687624, + "step": 13230, + "token_acc": 0.998414367499541 + }, + { + "epoch": 0.7009788025386656, + "grad_norm": 0.11418303847312927, + "learning_rate": 4.334559869156481e-06, + "loss": 0.004073138162493706, + "step": 13240, + "token_acc": 0.9983395911556112 + }, + { + "epoch": 0.701508242721852, + "grad_norm": 0.07303052395582199, + "learning_rate": 4.320438129191506e-06, + "loss": 0.003499014303088188, + "step": 13250, + "token_acc": 0.9985226855406804 + }, + { + "epoch": 0.7020376829050383, + "grad_norm": 0.09707936644554138, + "learning_rate": 4.306333089213201e-06, + "loss": 0.0035513103008270263, + "step": 13260, + "token_acc": 0.998590339236114 + }, + { + "epoch": 0.7025671230882246, + "grad_norm": 0.09590703248977661, + "learning_rate": 4.292244790695537e-06, + "loss": 0.003596171736717224, + "step": 13270, + "token_acc": 0.998523055356219 + }, + { + "epoch": 0.7030965632714109, + "grad_norm": 0.06643738597631454, + "learning_rate": 4.278173275063254e-06, + "loss": 0.0037018552422523497, + "step": 13280, + "token_acc": 0.9984891864143636 + }, + { + "epoch": 0.7036260034545972, + "grad_norm": 0.07913632690906525, + "learning_rate": 4.264118583691753e-06, + "loss": 0.003918538987636566, + "step": 13290, + "token_acc": 0.9984735626047861 + }, + { + "epoch": 0.7041554436377835, + "grad_norm": 0.09850672632455826, + "learning_rate": 4.250080757906958e-06, + "loss": 0.003775758296251297, + "step": 13300, + "token_acc": 0.9984227260967895 + }, + { + "epoch": 0.7046848838209698, + "grad_norm": 0.09037099033594131, + "learning_rate": 4.236059838985206e-06, + "loss": 0.0036908917129039763, + "step": 13310, + "token_acc": 0.9984902449806902 + }, + { + "epoch": 0.7052143240041561, + "grad_norm": 0.086330346763134, + "learning_rate": 4.222055868153118e-06, + "loss": 0.003663964569568634, + "step": 13320, + "token_acc": 0.9984898796909677 + }, + { + "epoch": 0.7057437641873424, + "grad_norm": 0.07633336633443832, + "learning_rate": 4.20806888658749e-06, + "loss": 0.003997672349214554, + "step": 13330, + "token_acc": 0.998415412201326 + }, + { + "epoch": 0.7062732043705288, + "grad_norm": 0.07574186474084854, + "learning_rate": 4.194098935415155e-06, + "loss": 0.003832285106182098, + "step": 13340, + "token_acc": 0.9983480727515435 + }, + { + "epoch": 0.706802644553715, + "grad_norm": 0.10792764276266098, + "learning_rate": 4.180146055712867e-06, + "loss": 0.003378502279520035, + "step": 13350, + "token_acc": 0.9986236925078825 + }, + { + "epoch": 0.7073320847369013, + "grad_norm": 0.06305970996618271, + "learning_rate": 4.166210288507195e-06, + "loss": 0.0036281347274780273, + "step": 13360, + "token_acc": 0.9984895268296754 + }, + { + "epoch": 0.7078615249200876, + "grad_norm": 0.08588401973247528, + "learning_rate": 4.152291674774384e-06, + "loss": 0.0035809896886348723, + "step": 13370, + "token_acc": 0.998623967975982 + }, + { + "epoch": 0.7083909651032739, + "grad_norm": 0.07186521589756012, + "learning_rate": 4.138390255440242e-06, + "loss": 0.0034058429300785065, + "step": 13380, + "token_acc": 0.998598586908466 + }, + { + "epoch": 0.7089204052864603, + "grad_norm": 0.06955961883068085, + "learning_rate": 4.124506071380021e-06, + "loss": 0.003599441051483154, + "step": 13390, + "token_acc": 0.998548354802109 + }, + { + "epoch": 0.7094498454696465, + "grad_norm": 0.11671196669340134, + "learning_rate": 4.1106391634182945e-06, + "loss": 0.003571125864982605, + "step": 13400, + "token_acc": 0.9984649653365814 + }, + { + "epoch": 0.7099792856528329, + "grad_norm": 0.09935661405324936, + "learning_rate": 4.096789572328838e-06, + "loss": 0.003690510615706444, + "step": 13410, + "token_acc": 0.9985400972712333 + }, + { + "epoch": 0.7105087258360191, + "grad_norm": 0.07537563145160675, + "learning_rate": 4.082957338834506e-06, + "loss": 0.003518431261181831, + "step": 13420, + "token_acc": 0.9987156927336563 + }, + { + "epoch": 0.7110381660192054, + "grad_norm": 0.09148381650447845, + "learning_rate": 4.0691425036071204e-06, + "loss": 0.003819703683257103, + "step": 13430, + "token_acc": 0.9983560317778223 + }, + { + "epoch": 0.7115676062023918, + "grad_norm": 0.06931379437446594, + "learning_rate": 4.055345107267347e-06, + "loss": 0.0038111522793769837, + "step": 13440, + "token_acc": 0.9984317781800285 + }, + { + "epoch": 0.712097046385578, + "grad_norm": 0.08167047053575516, + "learning_rate": 4.041565190384562e-06, + "loss": 0.0034765325486660003, + "step": 13450, + "token_acc": 0.9985558783264884 + }, + { + "epoch": 0.7126264865687644, + "grad_norm": 0.1034456342458725, + "learning_rate": 4.027802793476765e-06, + "loss": 0.0032634101808071136, + "step": 13460, + "token_acc": 0.9987320230905269 + }, + { + "epoch": 0.7131559267519506, + "grad_norm": 0.08099982142448425, + "learning_rate": 4.014057957010421e-06, + "loss": 0.0032812949270009995, + "step": 13470, + "token_acc": 0.9986569567139651 + }, + { + "epoch": 0.713685366935137, + "grad_norm": 0.09634703397750854, + "learning_rate": 4.000330721400374e-06, + "loss": 0.0034518830478191377, + "step": 13480, + "token_acc": 0.9986479832416688 + }, + { + "epoch": 0.7142148071183233, + "grad_norm": 0.08419275283813477, + "learning_rate": 3.986621127009712e-06, + "loss": 0.0034293245524168016, + "step": 13490, + "token_acc": 0.9986480735047443 + }, + { + "epoch": 0.7147442473015095, + "grad_norm": 0.0962294489145279, + "learning_rate": 3.972929214149651e-06, + "loss": 0.0036505814641714096, + "step": 13500, + "token_acc": 0.9984900938477581 + }, + { + "epoch": 0.7152736874846959, + "grad_norm": 0.08934230357408524, + "learning_rate": 3.959255023079416e-06, + "loss": 0.003594578802585602, + "step": 13510, + "token_acc": 0.9985155780905998 + }, + { + "epoch": 0.7158031276678821, + "grad_norm": 0.07983361184597015, + "learning_rate": 3.945598594006124e-06, + "loss": 0.0036843284964561463, + "step": 13520, + "token_acc": 0.9984732442308655 + }, + { + "epoch": 0.7163325678510685, + "grad_norm": 0.07887694984674454, + "learning_rate": 3.931959967084665e-06, + "loss": 0.003663025051355362, + "step": 13530, + "token_acc": 0.9985735616746887 + }, + { + "epoch": 0.7168620080342548, + "grad_norm": 0.09390752017498016, + "learning_rate": 3.918339182417584e-06, + "loss": 0.0036802031099796295, + "step": 13540, + "token_acc": 0.9985487180342636 + }, + { + "epoch": 0.717391448217441, + "grad_norm": 0.07416029274463654, + "learning_rate": 3.904736280054966e-06, + "loss": 0.003518931195139885, + "step": 13550, + "token_acc": 0.9985661767770655 + }, + { + "epoch": 0.7179208884006274, + "grad_norm": 0.0881822258234024, + "learning_rate": 3.891151299994314e-06, + "loss": 0.003579546511173248, + "step": 13560, + "token_acc": 0.9984978594497158 + }, + { + "epoch": 0.7184503285838136, + "grad_norm": 0.08102105557918549, + "learning_rate": 3.877584282180425e-06, + "loss": 0.003566005080938339, + "step": 13570, + "token_acc": 0.9986232446369121 + }, + { + "epoch": 0.718979768767, + "grad_norm": 0.06895655393600464, + "learning_rate": 3.864035266505294e-06, + "loss": 0.00344962440431118, + "step": 13580, + "token_acc": 0.9985155038113188 + }, + { + "epoch": 0.7195092089501863, + "grad_norm": 0.07847443968057632, + "learning_rate": 3.850504292807974e-06, + "loss": 0.0031574346125125886, + "step": 13590, + "token_acc": 0.9988322044643155 + }, + { + "epoch": 0.7200386491333726, + "grad_norm": 0.07478877902030945, + "learning_rate": 3.83699140087447e-06, + "loss": 0.0033171877264976503, + "step": 13600, + "token_acc": 0.9986143341291174 + }, + { + "epoch": 0.7205680893165589, + "grad_norm": 0.07074201852083206, + "learning_rate": 3.823496630437623e-06, + "loss": 0.0036013219505548477, + "step": 13610, + "token_acc": 0.9984570089326672 + }, + { + "epoch": 0.7210975294997453, + "grad_norm": 0.0868447795510292, + "learning_rate": 3.8100200211769855e-06, + "loss": 0.0037518560886383055, + "step": 13620, + "token_acc": 0.9985304922850845 + }, + { + "epoch": 0.7216269696829315, + "grad_norm": 0.06650970131158829, + "learning_rate": 3.7965616127187133e-06, + "loss": 0.0032344207167625426, + "step": 13630, + "token_acc": 0.998706187761371 + }, + { + "epoch": 0.7221564098661178, + "grad_norm": 0.10092959553003311, + "learning_rate": 3.783121444635444e-06, + "loss": 0.003444017469882965, + "step": 13640, + "token_acc": 0.9985808498205192 + }, + { + "epoch": 0.7226858500493041, + "grad_norm": 0.08295388519763947, + "learning_rate": 3.769699556446179e-06, + "loss": 0.004010315984487534, + "step": 13650, + "token_acc": 0.9983987323297611 + }, + { + "epoch": 0.7232152902324904, + "grad_norm": 0.08131997287273407, + "learning_rate": 3.7562959876161787e-06, + "loss": 0.0036193259060382844, + "step": 13660, + "token_acc": 0.9984640555611206 + }, + { + "epoch": 0.7237447304156768, + "grad_norm": 0.07519711554050446, + "learning_rate": 3.742910777556824e-06, + "loss": 0.0032779626548290254, + "step": 13670, + "token_acc": 0.9987153177503045 + }, + { + "epoch": 0.724274170598863, + "grad_norm": 0.08185270428657532, + "learning_rate": 3.729543965625526e-06, + "loss": 0.0033484019339084627, + "step": 13680, + "token_acc": 0.9986646413727487 + }, + { + "epoch": 0.7248036107820494, + "grad_norm": 0.10115562379360199, + "learning_rate": 3.716195591125594e-06, + "loss": 0.00359770692884922, + "step": 13690, + "token_acc": 0.9986471816283925 + }, + { + "epoch": 0.7253330509652356, + "grad_norm": 0.1143842488527298, + "learning_rate": 3.7028656933061267e-06, + "loss": 0.0034013006836175917, + "step": 13700, + "token_acc": 0.998639455782313 + }, + { + "epoch": 0.7258624911484219, + "grad_norm": 0.09851127862930298, + "learning_rate": 3.6895543113618916e-06, + "loss": 0.003480953723192215, + "step": 13710, + "token_acc": 0.9986058820091995 + }, + { + "epoch": 0.7263919313316083, + "grad_norm": 0.08221554011106491, + "learning_rate": 3.6762614844332145e-06, + "loss": 0.0034946896135807036, + "step": 13720, + "token_acc": 0.9985234865737381 + }, + { + "epoch": 0.7269213715147945, + "grad_norm": 0.08676858991384506, + "learning_rate": 3.662987251605866e-06, + "loss": 0.0033679164946079252, + "step": 13730, + "token_acc": 0.998539146694715 + }, + { + "epoch": 0.7274508116979809, + "grad_norm": 0.0642840713262558, + "learning_rate": 3.6497316519109337e-06, + "loss": 0.003327426686882973, + "step": 13740, + "token_acc": 0.998673308468297 + }, + { + "epoch": 0.7279802518811671, + "grad_norm": 0.07299963384866714, + "learning_rate": 3.636494724324725e-06, + "loss": 0.0031659990549087525, + "step": 13750, + "token_acc": 0.9987065440526061 + }, + { + "epoch": 0.7285096920643535, + "grad_norm": 0.08553624153137207, + "learning_rate": 3.623276507768645e-06, + "loss": 0.0037304766476154326, + "step": 13760, + "token_acc": 0.9984661042382207 + }, + { + "epoch": 0.7290391322475398, + "grad_norm": 0.06790778785943985, + "learning_rate": 3.6100770411090767e-06, + "loss": 0.003315257281064987, + "step": 13770, + "token_acc": 0.9986745139884624 + }, + { + "epoch": 0.729568572430726, + "grad_norm": 0.10107589513063431, + "learning_rate": 3.5968963631572774e-06, + "loss": 0.0032653950154781343, + "step": 13780, + "token_acc": 0.9985815011055947 + }, + { + "epoch": 0.7300980126139124, + "grad_norm": 0.09684211760759354, + "learning_rate": 3.5837345126692504e-06, + "loss": 0.0036232031881809233, + "step": 13790, + "token_acc": 0.9986659440530288 + }, + { + "epoch": 0.7306274527970986, + "grad_norm": 0.07079575210809708, + "learning_rate": 3.570591528345645e-06, + "loss": 0.0031457334756851197, + "step": 13800, + "token_acc": 0.9987567585608438 + }, + { + "epoch": 0.731156892980285, + "grad_norm": 0.09268027544021606, + "learning_rate": 3.5574674488316384e-06, + "loss": 0.0033174045383930206, + "step": 13810, + "token_acc": 0.9986483554991907 + }, + { + "epoch": 0.7316863331634713, + "grad_norm": 0.10290899127721786, + "learning_rate": 3.544362312716818e-06, + "loss": 0.0037267588078975676, + "step": 13820, + "token_acc": 0.9985153180805899 + }, + { + "epoch": 0.7322157733466576, + "grad_norm": 0.08426038920879364, + "learning_rate": 3.531276158535071e-06, + "loss": 0.0035196110606193543, + "step": 13830, + "token_acc": 0.9986987312629814 + }, + { + "epoch": 0.7327452135298439, + "grad_norm": 0.06399177759885788, + "learning_rate": 3.518209024764472e-06, + "loss": 0.003399909287691116, + "step": 13840, + "token_acc": 0.9986226930107931 + }, + { + "epoch": 0.7332746537130301, + "grad_norm": 0.1005530059337616, + "learning_rate": 3.5051609498271654e-06, + "loss": 0.003392476588487625, + "step": 13850, + "token_acc": 0.9986976883969045 + }, + { + "epoch": 0.7338040938962165, + "grad_norm": 0.07391153275966644, + "learning_rate": 3.492131972089261e-06, + "loss": 0.0032870016992092134, + "step": 13860, + "token_acc": 0.9987068677834879 + }, + { + "epoch": 0.7343335340794028, + "grad_norm": 0.09027525782585144, + "learning_rate": 3.4791221298607125e-06, + "loss": 0.0034948129206895827, + "step": 13870, + "token_acc": 0.9986400687474449 + }, + { + "epoch": 0.7348629742625891, + "grad_norm": 0.0985957607626915, + "learning_rate": 3.46613146139521e-06, + "loss": 0.0034495133906602858, + "step": 13880, + "token_acc": 0.9986645076206534 + }, + { + "epoch": 0.7353924144457754, + "grad_norm": 0.08017542213201523, + "learning_rate": 3.4531600048900584e-06, + "loss": 0.0031513679772615434, + "step": 13890, + "token_acc": 0.9987315998531326 + }, + { + "epoch": 0.7359218546289616, + "grad_norm": 0.07839474081993103, + "learning_rate": 3.440207798486083e-06, + "loss": 0.003453974053263664, + "step": 13900, + "token_acc": 0.9986404090450334 + }, + { + "epoch": 0.736451294812148, + "grad_norm": 0.059698570519685745, + "learning_rate": 3.4272748802675003e-06, + "loss": 0.003311881795525551, + "step": 13910, + "token_acc": 0.998774887698039 + }, + { + "epoch": 0.7369807349953343, + "grad_norm": 0.09852812439203262, + "learning_rate": 3.414361288261816e-06, + "loss": 0.003312372416257858, + "step": 13920, + "token_acc": 0.9986230493198698 + }, + { + "epoch": 0.7375101751785206, + "grad_norm": 0.07676622271537781, + "learning_rate": 3.401467060439707e-06, + "loss": 0.0032824233174324037, + "step": 13930, + "token_acc": 0.9986566429423691 + }, + { + "epoch": 0.7380396153617069, + "grad_norm": 0.07956480234861374, + "learning_rate": 3.3885922347149136e-06, + "loss": 0.0035583797842264177, + "step": 13940, + "token_acc": 0.9986228654414342 + }, + { + "epoch": 0.7385690555448933, + "grad_norm": 0.0711434930562973, + "learning_rate": 3.3757368489441266e-06, + "loss": 0.0030506592243909835, + "step": 13950, + "token_acc": 0.9987736919380673 + }, + { + "epoch": 0.7390984957280795, + "grad_norm": 0.1118825227022171, + "learning_rate": 3.362900940926875e-06, + "loss": 0.003380221873521805, + "step": 13960, + "token_acc": 0.9986563514517246 + }, + { + "epoch": 0.7396279359112659, + "grad_norm": 0.07906787842512131, + "learning_rate": 3.350084548405421e-06, + "loss": 0.0031727030873298645, + "step": 13970, + "token_acc": 0.9987993396368002 + }, + { + "epoch": 0.7401573760944521, + "grad_norm": 0.0896146222949028, + "learning_rate": 3.3372877090646405e-06, + "loss": 0.003206241875886917, + "step": 13980, + "token_acc": 0.9987822475040244 + }, + { + "epoch": 0.7406868162776384, + "grad_norm": 0.07737207412719727, + "learning_rate": 3.32451046053191e-06, + "loss": 0.00357636958360672, + "step": 13990, + "token_acc": 0.9985559385981753 + }, + { + "epoch": 0.7412162564608248, + "grad_norm": 0.08291522413492203, + "learning_rate": 3.3117528403770116e-06, + "loss": 0.0034991003572940826, + "step": 14000, + "token_acc": 0.9985577804825183 + }, + { + "epoch": 0.741745696644011, + "grad_norm": 0.09278791397809982, + "learning_rate": 3.299014886112013e-06, + "loss": 0.003487355634570122, + "step": 14010, + "token_acc": 0.9985982244175956 + }, + { + "epoch": 0.7422751368271974, + "grad_norm": 0.06723111122846603, + "learning_rate": 3.286296635191144e-06, + "loss": 0.0034265875816345214, + "step": 14020, + "token_acc": 0.9986408398442386 + }, + { + "epoch": 0.7428045770103836, + "grad_norm": 0.08533704280853271, + "learning_rate": 3.273598125010714e-06, + "loss": 0.0031130878254771233, + "step": 14030, + "token_acc": 0.9987405541561712 + }, + { + "epoch": 0.74333401719357, + "grad_norm": 0.0779648870229721, + "learning_rate": 3.260919392908981e-06, + "loss": 0.0028697635978460314, + "step": 14040, + "token_acc": 0.9989404481820761 + }, + { + "epoch": 0.7438634573767563, + "grad_norm": 0.10078655183315277, + "learning_rate": 3.2482604761660486e-06, + "loss": 0.003422523662447929, + "step": 14050, + "token_acc": 0.9985649924912398 + }, + { + "epoch": 0.7443928975599425, + "grad_norm": 0.09953100979328156, + "learning_rate": 3.2356214120037567e-06, + "loss": 0.0034061163663864137, + "step": 14060, + "token_acc": 0.998655880314908 + }, + { + "epoch": 0.7449223377431289, + "grad_norm": 0.08661240339279175, + "learning_rate": 3.2230022375855707e-06, + "loss": 0.003185698390007019, + "step": 14070, + "token_acc": 0.9986479945252124 + }, + { + "epoch": 0.7454517779263151, + "grad_norm": 0.09516514837741852, + "learning_rate": 3.2104029900164734e-06, + "loss": 0.003453529626131058, + "step": 14080, + "token_acc": 0.9985310809908526 + }, + { + "epoch": 0.7459812181095015, + "grad_norm": 0.07555476576089859, + "learning_rate": 3.197823706342855e-06, + "loss": 0.0032791893929243086, + "step": 14090, + "token_acc": 0.9987392607559425 + }, + { + "epoch": 0.7465106582926878, + "grad_norm": 0.09716933220624924, + "learning_rate": 3.1852644235524087e-06, + "loss": 0.003528785705566406, + "step": 14100, + "token_acc": 0.9985317059741213 + }, + { + "epoch": 0.747040098475874, + "grad_norm": 0.08044704794883728, + "learning_rate": 3.1727251785740077e-06, + "loss": 0.003041881322860718, + "step": 14110, + "token_acc": 0.9987569556258186 + }, + { + "epoch": 0.7475695386590604, + "grad_norm": 0.07473593950271606, + "learning_rate": 3.1602060082776142e-06, + "loss": 0.0031833186745643615, + "step": 14120, + "token_acc": 0.9988402072608031 + }, + { + "epoch": 0.7480989788422466, + "grad_norm": 0.07337811589241028, + "learning_rate": 3.147706949474164e-06, + "loss": 0.0034189064055681227, + "step": 14130, + "token_acc": 0.9986569118984259 + }, + { + "epoch": 0.748628419025433, + "grad_norm": 0.06130555272102356, + "learning_rate": 3.1352280389154553e-06, + "loss": 0.0027870308607816697, + "step": 14140, + "token_acc": 0.998873413390525 + }, + { + "epoch": 0.7491578592086193, + "grad_norm": 0.07504958659410477, + "learning_rate": 3.122769313294044e-06, + "loss": 0.0029739052057266234, + "step": 14150, + "token_acc": 0.9987816276119901 + }, + { + "epoch": 0.7496872993918056, + "grad_norm": 0.0803057923913002, + "learning_rate": 3.110330809243134e-06, + "loss": 0.003081291355192661, + "step": 14160, + "token_acc": 0.9987240111086092 + }, + { + "epoch": 0.7502167395749919, + "grad_norm": 0.09009984880685806, + "learning_rate": 3.0979125633364728e-06, + "loss": 0.003174242377281189, + "step": 14170, + "token_acc": 0.9988062841210745 + }, + { + "epoch": 0.7507461797581781, + "grad_norm": 0.07801827788352966, + "learning_rate": 3.0855146120882363e-06, + "loss": 0.0033581145107746124, + "step": 14180, + "token_acc": 0.9987312504695208 + }, + { + "epoch": 0.7512756199413645, + "grad_norm": 0.07267160713672638, + "learning_rate": 3.0731369919529374e-06, + "loss": 0.003075828216969967, + "step": 14190, + "token_acc": 0.9987824810700824 + }, + { + "epoch": 0.7518050601245508, + "grad_norm": 0.06993144005537033, + "learning_rate": 3.0607797393252914e-06, + "loss": 0.003212468698620796, + "step": 14200, + "token_acc": 0.998757038581856 + }, + { + "epoch": 0.7523345003077371, + "grad_norm": 0.09156052768230438, + "learning_rate": 3.0484428905401396e-06, + "loss": 0.0033934835344552996, + "step": 14210, + "token_acc": 0.9986477010918562 + }, + { + "epoch": 0.7528639404909234, + "grad_norm": 0.07862966507673264, + "learning_rate": 3.0361264818723256e-06, + "loss": 0.0030300889164209368, + "step": 14220, + "token_acc": 0.9987654938858582 + }, + { + "epoch": 0.7533933806741097, + "grad_norm": 0.08523651212453842, + "learning_rate": 3.0238305495365872e-06, + "loss": 0.0035031594336032866, + "step": 14230, + "token_acc": 0.9986475881989548 + }, + { + "epoch": 0.753922820857296, + "grad_norm": 0.0946173369884491, + "learning_rate": 3.0115551296874603e-06, + "loss": 0.0032571140676736833, + "step": 14240, + "token_acc": 0.9987573806585048 + }, + { + "epoch": 0.7544522610404824, + "grad_norm": 0.07246138900518417, + "learning_rate": 2.999300258419161e-06, + "loss": 0.0030216820538043975, + "step": 14250, + "token_acc": 0.9988244514106583 + }, + { + "epoch": 0.7549817012236686, + "grad_norm": 0.0825478658080101, + "learning_rate": 2.9870659717654883e-06, + "loss": 0.0028778720647096635, + "step": 14260, + "token_acc": 0.9988571047209083 + }, + { + "epoch": 0.7555111414068549, + "grad_norm": 0.08581020683050156, + "learning_rate": 2.9748523056997146e-06, + "loss": 0.0030291680246591567, + "step": 14270, + "token_acc": 0.9988734415942053 + }, + { + "epoch": 0.7560405815900413, + "grad_norm": 0.07391929626464844, + "learning_rate": 2.9626592961344802e-06, + "loss": 0.0031259000301361085, + "step": 14280, + "token_acc": 0.9987734565995544 + }, + { + "epoch": 0.7565700217732275, + "grad_norm": 0.09388487040996552, + "learning_rate": 2.9504869789216895e-06, + "loss": 0.0032490063458681107, + "step": 14290, + "token_acc": 0.9986816634403578 + }, + { + "epoch": 0.7570994619564139, + "grad_norm": 0.10506483167409897, + "learning_rate": 2.938335389852397e-06, + "loss": 0.0032929468899965285, + "step": 14300, + "token_acc": 0.9987234042553191 + }, + { + "epoch": 0.7576289021396001, + "grad_norm": 0.11294104158878326, + "learning_rate": 2.926204564656717e-06, + "loss": 0.00308716744184494, + "step": 14310, + "token_acc": 0.9988488200405415 + }, + { + "epoch": 0.7581583423227864, + "grad_norm": 0.09168960154056549, + "learning_rate": 2.9140945390037045e-06, + "loss": 0.003040589392185211, + "step": 14320, + "token_acc": 0.9987649478858077 + }, + { + "epoch": 0.7586877825059728, + "grad_norm": 0.10604191571474075, + "learning_rate": 2.902005348501259e-06, + "loss": 0.0032668627798557283, + "step": 14330, + "token_acc": 0.9987326679840249 + }, + { + "epoch": 0.759217222689159, + "grad_norm": 0.11078047007322311, + "learning_rate": 2.8899370286960173e-06, + "loss": 0.0032401148229837416, + "step": 14340, + "token_acc": 0.9987231698768234 + }, + { + "epoch": 0.7597466628723454, + "grad_norm": 0.08638836443424225, + "learning_rate": 2.8778896150732494e-06, + "loss": 0.003432933986186981, + "step": 14350, + "token_acc": 0.9986557680907733 + }, + { + "epoch": 0.7602761030555316, + "grad_norm": 0.07859744876623154, + "learning_rate": 2.8658631430567506e-06, + "loss": 0.003356776386499405, + "step": 14360, + "token_acc": 0.9986308460369672 + }, + { + "epoch": 0.760805543238718, + "grad_norm": 0.0754460021853447, + "learning_rate": 2.8538576480087443e-06, + "loss": 0.002875025570392609, + "step": 14370, + "token_acc": 0.9989401386999591 + }, + { + "epoch": 0.7613349834219043, + "grad_norm": 0.0649600401520729, + "learning_rate": 2.8418731652297706e-06, + "loss": 0.0031451087445020674, + "step": 14380, + "token_acc": 0.9987478609290872 + }, + { + "epoch": 0.7618644236050905, + "grad_norm": 0.09046947211027145, + "learning_rate": 2.8299097299585853e-06, + "loss": 0.003486282378435135, + "step": 14390, + "token_acc": 0.9984986112153742 + }, + { + "epoch": 0.7623938637882769, + "grad_norm": 0.08914913237094879, + "learning_rate": 2.8179673773720593e-06, + "loss": 0.0030184250324964525, + "step": 14400, + "token_acc": 0.9987895382714607 + }, + { + "epoch": 0.7629233039714631, + "grad_norm": 0.09421832859516144, + "learning_rate": 2.806046142585076e-06, + "loss": 0.0031730350106954574, + "step": 14410, + "token_acc": 0.9987482579923728 + }, + { + "epoch": 0.7634527441546495, + "grad_norm": 0.06963995844125748, + "learning_rate": 2.7941460606504144e-06, + "loss": 0.0031274013221263886, + "step": 14420, + "token_acc": 0.9986890666488536 + }, + { + "epoch": 0.7639821843378358, + "grad_norm": 0.09054435044527054, + "learning_rate": 2.782267166558664e-06, + "loss": 0.003320121765136719, + "step": 14430, + "token_acc": 0.9986651983848901 + }, + { + "epoch": 0.7645116245210221, + "grad_norm": 0.07607497274875641, + "learning_rate": 2.7704094952381133e-06, + "loss": 0.0032065901905298233, + "step": 14440, + "token_acc": 0.9987571629951538 + }, + { + "epoch": 0.7650410647042084, + "grad_norm": 0.09331351518630981, + "learning_rate": 2.7585730815546484e-06, + "loss": 0.0030787473544478415, + "step": 14450, + "token_acc": 0.99876510246312 + }, + { + "epoch": 0.7655705048873946, + "grad_norm": 0.06382590532302856, + "learning_rate": 2.746757960311649e-06, + "loss": 0.003117814287543297, + "step": 14460, + "token_acc": 0.9987984078904548 + }, + { + "epoch": 0.766099945070581, + "grad_norm": 0.09759918600320816, + "learning_rate": 2.734964166249888e-06, + "loss": 0.003050769120454788, + "step": 14470, + "token_acc": 0.9987898211439111 + }, + { + "epoch": 0.7666293852537673, + "grad_norm": 0.10626062750816345, + "learning_rate": 2.723191734047428e-06, + "loss": 0.0030577316880226134, + "step": 14480, + "token_acc": 0.9988321265307485 + }, + { + "epoch": 0.7671588254369536, + "grad_norm": 0.10445170104503632, + "learning_rate": 2.7114406983195195e-06, + "loss": 0.003257005289196968, + "step": 14490, + "token_acc": 0.9986560373972202 + }, + { + "epoch": 0.7676882656201399, + "grad_norm": 0.10006985813379288, + "learning_rate": 2.699711093618501e-06, + "loss": 0.002853582799434662, + "step": 14500, + "token_acc": 0.9988746530180139 + }, + { + "epoch": 0.7682177058033262, + "grad_norm": 0.08097430318593979, + "learning_rate": 2.6880029544336993e-06, + "loss": 0.0030025636777281763, + "step": 14510, + "token_acc": 0.998807498582341 + }, + { + "epoch": 0.7687471459865125, + "grad_norm": 0.09099474549293518, + "learning_rate": 2.676316315191313e-06, + "loss": 0.0030976399779319764, + "step": 14520, + "token_acc": 0.9987815869412825 + }, + { + "epoch": 0.7692765861696989, + "grad_norm": 0.0832013487815857, + "learning_rate": 2.664651210254333e-06, + "loss": 0.002869490534067154, + "step": 14530, + "token_acc": 0.9988492808005003 + }, + { + "epoch": 0.7698060263528851, + "grad_norm": 0.08816047012805939, + "learning_rate": 2.6530076739224308e-06, + "loss": 0.0030502840876579286, + "step": 14540, + "token_acc": 0.9988568567733322 + }, + { + "epoch": 0.7703354665360714, + "grad_norm": 0.09318000823259354, + "learning_rate": 2.6413857404318565e-06, + "loss": 0.002840893715620041, + "step": 14550, + "token_acc": 0.9988983475212819 + }, + { + "epoch": 0.7708649067192577, + "grad_norm": 0.08808407187461853, + "learning_rate": 2.6297854439553427e-06, + "loss": 0.002807438001036644, + "step": 14560, + "token_acc": 0.9988140476886458 + }, + { + "epoch": 0.771394346902444, + "grad_norm": 0.09155762940645218, + "learning_rate": 2.618206818601993e-06, + "loss": 0.0034898769110441206, + "step": 14570, + "token_acc": 0.9986734412935199 + }, + { + "epoch": 0.7719237870856304, + "grad_norm": 0.08833259344100952, + "learning_rate": 2.6066498984171973e-06, + "loss": 0.003583289682865143, + "step": 14580, + "token_acc": 0.9986730205890453 + }, + { + "epoch": 0.7724532272688166, + "grad_norm": 0.09406397491693497, + "learning_rate": 2.595114717382524e-06, + "loss": 0.00266465675085783, + "step": 14590, + "token_acc": 0.9989821543287641 + }, + { + "epoch": 0.772982667452003, + "grad_norm": 0.10119790583848953, + "learning_rate": 2.583601309415619e-06, + "loss": 0.0027526970952749252, + "step": 14600, + "token_acc": 0.9989986565308455 + }, + { + "epoch": 0.7735121076351893, + "grad_norm": 0.07635380327701569, + "learning_rate": 2.5721097083701085e-06, + "loss": 0.0027423959225416185, + "step": 14610, + "token_acc": 0.9988239121187098 + }, + { + "epoch": 0.7740415478183755, + "grad_norm": 0.08291584998369217, + "learning_rate": 2.560639948035496e-06, + "loss": 0.002922051027417183, + "step": 14620, + "token_acc": 0.9988736389804347 + }, + { + "epoch": 0.7745709880015619, + "grad_norm": 0.09461184591054916, + "learning_rate": 2.54919206213707e-06, + "loss": 0.00283438004553318, + "step": 14630, + "token_acc": 0.9988643101101452 + }, + { + "epoch": 0.7751004281847481, + "grad_norm": 0.07952001690864563, + "learning_rate": 2.537766084335791e-06, + "loss": 0.0028042174875736237, + "step": 14640, + "token_acc": 0.9989237443684298 + }, + { + "epoch": 0.7756298683679345, + "grad_norm": 0.06798510998487473, + "learning_rate": 2.5263620482282123e-06, + "loss": 0.0030165232717990877, + "step": 14650, + "token_acc": 0.9988065929480492 + }, + { + "epoch": 0.7761593085511208, + "grad_norm": 0.0819244533777237, + "learning_rate": 2.5149799873463645e-06, + "loss": 0.0027458954602479933, + "step": 14660, + "token_acc": 0.9989149305555556 + }, + { + "epoch": 0.776688748734307, + "grad_norm": 0.09735994040966034, + "learning_rate": 2.503619935157665e-06, + "loss": 0.0029595118016004563, + "step": 14670, + "token_acc": 0.9987981070185542 + }, + { + "epoch": 0.7772181889174934, + "grad_norm": 0.08358258754014969, + "learning_rate": 2.492281925064817e-06, + "loss": 0.002777286246418953, + "step": 14680, + "token_acc": 0.9990070921985815 + }, + { + "epoch": 0.7777476291006796, + "grad_norm": 0.06106926128268242, + "learning_rate": 2.4809659904057115e-06, + "loss": 0.0029251063242554665, + "step": 14690, + "token_acc": 0.9988903721007842 + }, + { + "epoch": 0.778277069283866, + "grad_norm": 0.07477682828903198, + "learning_rate": 2.4696721644533304e-06, + "loss": 0.0028445301577448845, + "step": 14700, + "token_acc": 0.99894851916449 + }, + { + "epoch": 0.7788065094670523, + "grad_norm": 0.08938727527856827, + "learning_rate": 2.458400480415648e-06, + "loss": 0.002863958477973938, + "step": 14710, + "token_acc": 0.9988407392580732 + }, + { + "epoch": 0.7793359496502386, + "grad_norm": 0.11173029989004135, + "learning_rate": 2.4471509714355335e-06, + "loss": 0.0031489934772253036, + "step": 14720, + "token_acc": 0.9986978514549005 + }, + { + "epoch": 0.7798653898334249, + "grad_norm": 0.09802668541669846, + "learning_rate": 2.435923670590655e-06, + "loss": 0.0028875328600406645, + "step": 14730, + "token_acc": 0.9988825511191167 + }, + { + "epoch": 0.7803948300166111, + "grad_norm": 0.08157936483621597, + "learning_rate": 2.424718610893373e-06, + "loss": 0.00284363217651844, + "step": 14740, + "token_acc": 0.9988643859751668 + }, + { + "epoch": 0.7809242701997975, + "grad_norm": 0.0791802629828453, + "learning_rate": 2.413535825290657e-06, + "loss": 0.002671162039041519, + "step": 14750, + "token_acc": 0.9989652870493992 + }, + { + "epoch": 0.7814537103829838, + "grad_norm": 0.08670476078987122, + "learning_rate": 2.4023753466639855e-06, + "loss": 0.0028258666396141054, + "step": 14760, + "token_acc": 0.9989241034195163 + }, + { + "epoch": 0.7819831505661701, + "grad_norm": 0.07970444113016129, + "learning_rate": 2.3912372078292402e-06, + "loss": 0.003194136917591095, + "step": 14770, + "token_acc": 0.9986820043543907 + }, + { + "epoch": 0.7825125907493564, + "grad_norm": 0.0908869206905365, + "learning_rate": 2.380121441536618e-06, + "loss": 0.002976713702082634, + "step": 14780, + "token_acc": 0.9987992395183616 + }, + { + "epoch": 0.7830420309325427, + "grad_norm": 0.10501644760370255, + "learning_rate": 2.3690280804705335e-06, + "loss": 0.003120335377752781, + "step": 14790, + "token_acc": 0.9987496248874662 + }, + { + "epoch": 0.783571471115729, + "grad_norm": 0.07744579762220383, + "learning_rate": 2.3579571572495207e-06, + "loss": 0.0026989962905645372, + "step": 14800, + "token_acc": 0.9989735886844411 + }, + { + "epoch": 0.7841009112989153, + "grad_norm": 0.08961227536201477, + "learning_rate": 2.346908704426136e-06, + "loss": 0.0029162347316741945, + "step": 14810, + "token_acc": 0.9987489783323047 + }, + { + "epoch": 0.7846303514821016, + "grad_norm": 0.07339447736740112, + "learning_rate": 2.33588275448687e-06, + "loss": 0.0028346633538603784, + "step": 14820, + "token_acc": 0.9989235557706589 + }, + { + "epoch": 0.7851597916652879, + "grad_norm": 0.11385338008403778, + "learning_rate": 2.324879339852044e-06, + "loss": 0.0028848091140389444, + "step": 14830, + "token_acc": 0.9989822986703148 + }, + { + "epoch": 0.7856892318484742, + "grad_norm": 0.10979268699884415, + "learning_rate": 2.3138984928757122e-06, + "loss": 0.0029622877016663553, + "step": 14840, + "token_acc": 0.9988822901374617 + }, + { + "epoch": 0.7862186720316605, + "grad_norm": 0.09100823849439621, + "learning_rate": 2.302940245845581e-06, + "loss": 0.0027660321444272994, + "step": 14850, + "token_acc": 0.9988735825914276 + }, + { + "epoch": 0.7867481122148469, + "grad_norm": 0.08152920007705688, + "learning_rate": 2.2920046309828935e-06, + "loss": 0.0028596729040145873, + "step": 14860, + "token_acc": 0.9988062442607897 + }, + { + "epoch": 0.7872775523980331, + "grad_norm": 0.11900486797094345, + "learning_rate": 2.2810916804423566e-06, + "loss": 0.0031838081777095795, + "step": 14870, + "token_acc": 0.9986650981570011 + }, + { + "epoch": 0.7878069925812194, + "grad_norm": 0.09837418794631958, + "learning_rate": 2.2702014263120298e-06, + "loss": 0.002691994421184063, + "step": 14880, + "token_acc": 0.9989410842630113 + }, + { + "epoch": 0.7883364327644057, + "grad_norm": 0.10275392979383469, + "learning_rate": 2.25933390061324e-06, + "loss": 0.0028628962114453315, + "step": 14890, + "token_acc": 0.9989314277127544 + }, + { + "epoch": 0.788865872947592, + "grad_norm": 0.0804351419210434, + "learning_rate": 2.2484891353004814e-06, + "loss": 0.003006833419203758, + "step": 14900, + "token_acc": 0.9987307949231797 + }, + { + "epoch": 0.7893953131307784, + "grad_norm": 0.08549032360315323, + "learning_rate": 2.2376671622613256e-06, + "loss": 0.0027782201766967775, + "step": 14910, + "token_acc": 0.9989647514568618 + }, + { + "epoch": 0.7899247533139646, + "grad_norm": 0.10040626674890518, + "learning_rate": 2.226868013316328e-06, + "loss": 0.002863612398505211, + "step": 14920, + "token_acc": 0.9988475991649269 + }, + { + "epoch": 0.790454193497151, + "grad_norm": 0.08848872780799866, + "learning_rate": 2.216091720218931e-06, + "loss": 0.0029786832630634306, + "step": 14930, + "token_acc": 0.9988078365985827 + }, + { + "epoch": 0.7909836336803373, + "grad_norm": 0.07284176349639893, + "learning_rate": 2.2053383146553733e-06, + "loss": 0.0027898747473955156, + "step": 14940, + "token_acc": 0.9990566195254713 + }, + { + "epoch": 0.7915130738635235, + "grad_norm": 0.07434673607349396, + "learning_rate": 2.194607828244596e-06, + "loss": 0.002725965715944767, + "step": 14950, + "token_acc": 0.998956271970475 + }, + { + "epoch": 0.7920425140467099, + "grad_norm": 0.09228596091270447, + "learning_rate": 2.183900292538147e-06, + "loss": 0.002767041325569153, + "step": 14960, + "token_acc": 0.9989570731300321 + }, + { + "epoch": 0.7925719542298961, + "grad_norm": 0.09298758953809738, + "learning_rate": 2.1732157390200927e-06, + "loss": 0.002995377965271473, + "step": 14970, + "token_acc": 0.9988563128193141 + }, + { + "epoch": 0.7931013944130825, + "grad_norm": 0.07650390267372131, + "learning_rate": 2.162554199106924e-06, + "loss": 0.002613643556833267, + "step": 14980, + "token_acc": 0.9989565159319147 + }, + { + "epoch": 0.7936308345962688, + "grad_norm": 0.0731302946805954, + "learning_rate": 2.151915704147465e-06, + "loss": 0.002704036980867386, + "step": 14990, + "token_acc": 0.9988579336100969 + }, + { + "epoch": 0.794160274779455, + "grad_norm": 0.07136200368404388, + "learning_rate": 2.1413002854227747e-06, + "loss": 0.002697054296731949, + "step": 15000, + "token_acc": 0.9989397936354226 + }, + { + "epoch": 0.794160274779455, + "eval_loss": 0.004724802915006876, + "eval_runtime": 712.0674, + "eval_samples_per_second": 16.821, + "eval_steps_per_second": 1.052, + "eval_token_acc": 0.9983419047670015, + "step": 15000 + }, + { + "epoch": 0.7946897149626414, + "grad_norm": 0.1689254492521286, + "learning_rate": 2.1307079741460633e-06, + "loss": 0.002980345115065575, + "step": 15010, + "token_acc": 0.9987986184111729 + }, + { + "epoch": 0.7952191551458276, + "grad_norm": 0.07408677786588669, + "learning_rate": 2.1201388014625955e-06, + "loss": 0.002685100771486759, + "step": 15020, + "token_acc": 0.9989148490388067 + }, + { + "epoch": 0.795748595329014, + "grad_norm": 0.09993043541908264, + "learning_rate": 2.1095927984496e-06, + "loss": 0.0030009247362613677, + "step": 15030, + "token_acc": 0.9988070409610411 + }, + { + "epoch": 0.7962780355122003, + "grad_norm": 0.09249048680067062, + "learning_rate": 2.0990699961161764e-06, + "loss": 0.0030381496995687484, + "step": 15040, + "token_acc": 0.9988073792367227 + }, + { + "epoch": 0.7968074756953866, + "grad_norm": 0.07976478338241577, + "learning_rate": 2.0885704254032122e-06, + "loss": 0.002795400097966194, + "step": 15050, + "token_acc": 0.9989329776592197 + }, + { + "epoch": 0.7973369158785729, + "grad_norm": 0.08819480240345001, + "learning_rate": 2.0780941171832757e-06, + "loss": 0.002814822271466255, + "step": 15060, + "token_acc": 0.9989152293455495 + }, + { + "epoch": 0.7978663560617592, + "grad_norm": 0.0817602202296257, + "learning_rate": 2.0676411022605415e-06, + "loss": 0.0027516849339008333, + "step": 15070, + "token_acc": 0.9988740335454599 + }, + { + "epoch": 0.7983957962449455, + "grad_norm": 0.07159268856048584, + "learning_rate": 2.057211411370692e-06, + "loss": 0.0028716757893562316, + "step": 15080, + "token_acc": 0.9988731218697829 + }, + { + "epoch": 0.7989252364281318, + "grad_norm": 0.07562939822673798, + "learning_rate": 2.0468050751808287e-06, + "loss": 0.00299276877194643, + "step": 15090, + "token_acc": 0.9988986416580449 + }, + { + "epoch": 0.7994546766113181, + "grad_norm": 0.08248212188482285, + "learning_rate": 2.0364221242893823e-06, + "loss": 0.0027312519028782844, + "step": 15100, + "token_acc": 0.9989486332231902 + }, + { + "epoch": 0.7999841167945044, + "grad_norm": 0.10717691481113434, + "learning_rate": 2.0260625892260232e-06, + "loss": 0.0025301286950707436, + "step": 15110, + "token_acc": 0.9990580114872581 + }, + { + "epoch": 0.8005135569776907, + "grad_norm": 0.07399527728557587, + "learning_rate": 2.0157265004515647e-06, + "loss": 0.0028338510543107986, + "step": 15120, + "token_acc": 0.9988900943002587 + }, + { + "epoch": 0.801042997160877, + "grad_norm": 0.08512876182794571, + "learning_rate": 2.0054138883578844e-06, + "loss": 0.00275920070707798, + "step": 15130, + "token_acc": 0.9988901035625172 + }, + { + "epoch": 0.8015724373440634, + "grad_norm": 0.10306844115257263, + "learning_rate": 1.99512478326783e-06, + "loss": 0.0032227423042058946, + "step": 15140, + "token_acc": 0.9987233083836082 + }, + { + "epoch": 0.8021018775272496, + "grad_norm": 0.07537955790758133, + "learning_rate": 1.9848592154351287e-06, + "loss": 0.0027443328872323036, + "step": 15150, + "token_acc": 0.9989653906484665 + }, + { + "epoch": 0.8026313177104359, + "grad_norm": 0.0753602460026741, + "learning_rate": 1.974617215044301e-06, + "loss": 0.002537006512284279, + "step": 15160, + "token_acc": 0.9990318892347753 + }, + { + "epoch": 0.8031607578936222, + "grad_norm": 0.08275630325078964, + "learning_rate": 1.9643988122105717e-06, + "loss": 0.002800140529870987, + "step": 15170, + "token_acc": 0.9989072952638339 + }, + { + "epoch": 0.8036901980768085, + "grad_norm": 0.06321652978658676, + "learning_rate": 1.954204036979772e-06, + "loss": 0.0024375002831220627, + "step": 15180, + "token_acc": 0.9989322834120218 + }, + { + "epoch": 0.8042196382599949, + "grad_norm": 0.061290886253118515, + "learning_rate": 1.9440329193282693e-06, + "loss": 0.002595100924372673, + "step": 15190, + "token_acc": 0.9988985589479569 + }, + { + "epoch": 0.8047490784431811, + "grad_norm": 0.08478914201259613, + "learning_rate": 1.9338854891628633e-06, + "loss": 0.0025872129946947096, + "step": 15200, + "token_acc": 0.9989988904369009 + }, + { + "epoch": 0.8052785186263675, + "grad_norm": 0.09350861608982086, + "learning_rate": 1.9237617763207074e-06, + "loss": 0.0025544418022036552, + "step": 15210, + "token_acc": 0.9989898315272745 + }, + { + "epoch": 0.8058079588095537, + "grad_norm": 0.09016543626785278, + "learning_rate": 1.913661810569214e-06, + "loss": 0.002735462039709091, + "step": 15220, + "token_acc": 0.9989407133086444 + }, + { + "epoch": 0.80633739899274, + "grad_norm": 0.09619557112455368, + "learning_rate": 1.9035856216059724e-06, + "loss": 0.002473337948322296, + "step": 15230, + "token_acc": 0.9989655632674853 + }, + { + "epoch": 0.8068668391759264, + "grad_norm": 0.11192234605550766, + "learning_rate": 1.8935332390586604e-06, + "loss": 0.002525945380330086, + "step": 15240, + "token_acc": 0.9990321477142833 + }, + { + "epoch": 0.8073962793591126, + "grad_norm": 0.09001670032739639, + "learning_rate": 1.8835046924849532e-06, + "loss": 0.002306840941309929, + "step": 15250, + "token_acc": 0.9991409078018917 + }, + { + "epoch": 0.807925719542299, + "grad_norm": 0.10716260224580765, + "learning_rate": 1.873500011372441e-06, + "loss": 0.002586829289793968, + "step": 15260, + "token_acc": 0.9990155920212899 + }, + { + "epoch": 0.8084551597254853, + "grad_norm": 0.0925392135977745, + "learning_rate": 1.8635192251385447e-06, + "loss": 0.0026578862220048905, + "step": 15270, + "token_acc": 0.9989988403233746 + }, + { + "epoch": 0.8089845999086716, + "grad_norm": 0.08346688747406006, + "learning_rate": 1.8535623631304177e-06, + "loss": 0.0029551561921834944, + "step": 15280, + "token_acc": 0.9988896958768481 + }, + { + "epoch": 0.8095140400918579, + "grad_norm": 0.08865714073181152, + "learning_rate": 1.8436294546248712e-06, + "loss": 0.0026189696043729784, + "step": 15290, + "token_acc": 0.9989737428349729 + }, + { + "epoch": 0.8100434802750441, + "grad_norm": 0.09471502155065536, + "learning_rate": 1.8337205288282877e-06, + "loss": 0.0028749682009220124, + "step": 15300, + "token_acc": 0.9988075483026325 + }, + { + "epoch": 0.8105729204582305, + "grad_norm": 0.09338246285915375, + "learning_rate": 1.8238356148765269e-06, + "loss": 0.002410345524549484, + "step": 15310, + "token_acc": 0.9990732236787175 + }, + { + "epoch": 0.8111023606414168, + "grad_norm": 0.09444420784711838, + "learning_rate": 1.8139747418348463e-06, + "loss": 0.0023157740011811256, + "step": 15320, + "token_acc": 0.9990990164262653 + }, + { + "epoch": 0.8116318008246031, + "grad_norm": 0.09008310735225677, + "learning_rate": 1.8041379386978175e-06, + "loss": 0.0026003452017903327, + "step": 15330, + "token_acc": 0.9989985311790627 + }, + { + "epoch": 0.8121612410077894, + "grad_norm": 0.08330412954092026, + "learning_rate": 1.7943252343892326e-06, + "loss": 0.0027289748191833495, + "step": 15340, + "token_acc": 0.9989574210767755 + }, + { + "epoch": 0.8126906811909757, + "grad_norm": 0.08764860033988953, + "learning_rate": 1.7845366577620293e-06, + "loss": 0.0029014809057116508, + "step": 15350, + "token_acc": 0.998857543134001 + }, + { + "epoch": 0.813220121374162, + "grad_norm": 0.09109850972890854, + "learning_rate": 1.7747722375981979e-06, + "loss": 0.0028282759711146356, + "step": 15360, + "token_acc": 0.9990236412340507 + }, + { + "epoch": 0.8137495615573483, + "grad_norm": 0.11446291208267212, + "learning_rate": 1.7650320026086987e-06, + "loss": 0.002674074098467827, + "step": 15370, + "token_acc": 0.9990156248696537 + }, + { + "epoch": 0.8142790017405346, + "grad_norm": 0.07126311212778091, + "learning_rate": 1.755315981433382e-06, + "loss": 0.0022654183208942412, + "step": 15380, + "token_acc": 0.9990661374779042 + }, + { + "epoch": 0.8148084419237209, + "grad_norm": 0.07957299798727036, + "learning_rate": 1.745624202640902e-06, + "loss": 0.0024698738008737562, + "step": 15390, + "token_acc": 0.9991331599653264 + }, + { + "epoch": 0.8153378821069072, + "grad_norm": 0.07216495275497437, + "learning_rate": 1.7359566947286243e-06, + "loss": 0.002627863921225071, + "step": 15400, + "token_acc": 0.9990320265692018 + }, + { + "epoch": 0.8158673222900935, + "grad_norm": 0.069883793592453, + "learning_rate": 1.726313486122555e-06, + "loss": 0.0028042053803801537, + "step": 15410, + "token_acc": 0.998898485417449 + }, + { + "epoch": 0.8163967624732799, + "grad_norm": 0.09146000444889069, + "learning_rate": 1.7166946051772504e-06, + "loss": 0.0028740085661411287, + "step": 15420, + "token_acc": 0.9988488776557143 + }, + { + "epoch": 0.8169262026564661, + "grad_norm": 0.08396581560373306, + "learning_rate": 1.7071000801757354e-06, + "loss": 0.0025847548618912696, + "step": 15430, + "token_acc": 0.999007489699578 + }, + { + "epoch": 0.8174556428396524, + "grad_norm": 0.09808298200368881, + "learning_rate": 1.6975299393294176e-06, + "loss": 0.0027008550241589546, + "step": 15440, + "token_acc": 0.9989741793434748 + }, + { + "epoch": 0.8179850830228387, + "grad_norm": 0.08761489391326904, + "learning_rate": 1.6879842107780075e-06, + "loss": 0.0024102041497826576, + "step": 15450, + "token_acc": 0.999065818118124 + }, + { + "epoch": 0.818514523206025, + "grad_norm": 0.10248196870088577, + "learning_rate": 1.6784629225894334e-06, + "loss": 0.0027877837419509888, + "step": 15460, + "token_acc": 0.9989235198397797 + }, + { + "epoch": 0.8190439633892114, + "grad_norm": 0.09514966607093811, + "learning_rate": 1.6689661027597615e-06, + "loss": 0.0026691924780607224, + "step": 15470, + "token_acc": 0.9989912463526469 + }, + { + "epoch": 0.8195734035723976, + "grad_norm": 0.10945416241884232, + "learning_rate": 1.6594937792131128e-06, + "loss": 0.003082018345594406, + "step": 15480, + "token_acc": 0.9987986284341289 + }, + { + "epoch": 0.820102843755584, + "grad_norm": 0.07468570023775101, + "learning_rate": 1.6500459798015756e-06, + "loss": 0.002528471499681473, + "step": 15490, + "token_acc": 0.9989327641408752 + }, + { + "epoch": 0.8206322839387702, + "grad_norm": 0.08511780947446823, + "learning_rate": 1.6406227323051327e-06, + "loss": 0.002537349611520767, + "step": 15500, + "token_acc": 0.9990067441239316 + }, + { + "epoch": 0.8211617241219565, + "grad_norm": 0.06801439821720123, + "learning_rate": 1.6312240644315737e-06, + "loss": 0.0027577653527259827, + "step": 15510, + "token_acc": 0.9990327207231247 + }, + { + "epoch": 0.8216911643051429, + "grad_norm": 0.09126472473144531, + "learning_rate": 1.6218500038164175e-06, + "loss": 0.002791179530322552, + "step": 15520, + "token_acc": 0.9989818315348472 + }, + { + "epoch": 0.8222206044883291, + "grad_norm": 0.07962197065353394, + "learning_rate": 1.6125005780228241e-06, + "loss": 0.0026240047067403794, + "step": 15530, + "token_acc": 0.9989154284474775 + }, + { + "epoch": 0.8227500446715155, + "grad_norm": 0.09373387694358826, + "learning_rate": 1.6031758145415222e-06, + "loss": 0.0025842390954494477, + "step": 15540, + "token_acc": 0.9989986063938981 + }, + { + "epoch": 0.8232794848547018, + "grad_norm": 0.07656331360340118, + "learning_rate": 1.5938757407907224e-06, + "loss": 0.002850542590022087, + "step": 15550, + "token_acc": 0.9988907515366844 + }, + { + "epoch": 0.823808925037888, + "grad_norm": 0.08031673729419708, + "learning_rate": 1.5846003841160397e-06, + "loss": 0.002816171199083328, + "step": 15560, + "token_acc": 0.9989493082946272 + }, + { + "epoch": 0.8243383652210744, + "grad_norm": 0.08136258274316788, + "learning_rate": 1.575349771790411e-06, + "loss": 0.00322466678917408, + "step": 15570, + "token_acc": 0.9986484569883951 + }, + { + "epoch": 0.8248678054042606, + "grad_norm": 0.07025915384292603, + "learning_rate": 1.5661239310140163e-06, + "loss": 0.0027570284903049467, + "step": 15580, + "token_acc": 0.9989495535602632 + }, + { + "epoch": 0.825397245587447, + "grad_norm": 0.08101579546928406, + "learning_rate": 1.5569228889141951e-06, + "loss": 0.0024117318913340568, + "step": 15590, + "token_acc": 0.9990823392008009 + }, + { + "epoch": 0.8259266857706333, + "grad_norm": 0.0845760703086853, + "learning_rate": 1.5477466725453737e-06, + "loss": 0.002349299564957619, + "step": 15600, + "token_acc": 0.9991074928891372 + }, + { + "epoch": 0.8264561259538196, + "grad_norm": 0.09509969502687454, + "learning_rate": 1.5385953088889805e-06, + "loss": 0.0029516143724322317, + "step": 15610, + "token_acc": 0.998873977195953 + }, + { + "epoch": 0.8269855661370059, + "grad_norm": 0.10121852904558182, + "learning_rate": 1.5294688248533674e-06, + "loss": 0.002662522718310356, + "step": 15620, + "token_acc": 0.9989231514099204 + }, + { + "epoch": 0.8275150063201921, + "grad_norm": 0.10379631072282791, + "learning_rate": 1.5203672472737318e-06, + "loss": 0.0025891119614243506, + "step": 15630, + "token_acc": 0.9990235108540524 + }, + { + "epoch": 0.8280444465033785, + "grad_norm": 0.06680166721343994, + "learning_rate": 1.5112906029120377e-06, + "loss": 0.0025252636522054674, + "step": 15640, + "token_acc": 0.9990654049633673 + }, + { + "epoch": 0.8285738866865648, + "grad_norm": 0.06837759166955948, + "learning_rate": 1.5022389184569331e-06, + "loss": 0.0025278031826019287, + "step": 15650, + "token_acc": 0.9990318245933246 + }, + { + "epoch": 0.8291033268697511, + "grad_norm": 0.07174143195152283, + "learning_rate": 1.4932122205236809e-06, + "loss": 0.002458263747394085, + "step": 15660, + "token_acc": 0.9990158957850316 + }, + { + "epoch": 0.8296327670529374, + "grad_norm": 0.08632295578718185, + "learning_rate": 1.4842105356540648e-06, + "loss": 0.0028199270367622376, + "step": 15670, + "token_acc": 0.998956681412236 + }, + { + "epoch": 0.8301622072361237, + "grad_norm": 0.0951794907450676, + "learning_rate": 1.47523389031633e-06, + "loss": 0.002957139350473881, + "step": 15680, + "token_acc": 0.9988736107866368 + }, + { + "epoch": 0.83069164741931, + "grad_norm": 0.09027064591646194, + "learning_rate": 1.4662823109050916e-06, + "loss": 0.002741934359073639, + "step": 15690, + "token_acc": 0.9989742135637322 + }, + { + "epoch": 0.8312210876024964, + "grad_norm": 0.11203402280807495, + "learning_rate": 1.4573558237412677e-06, + "loss": 0.002680276706814766, + "step": 15700, + "token_acc": 0.9989156726999749 + }, + { + "epoch": 0.8317505277856826, + "grad_norm": 0.07407877594232559, + "learning_rate": 1.4484544550719837e-06, + "loss": 0.0027659183368086817, + "step": 15710, + "token_acc": 0.9989061273568363 + }, + { + "epoch": 0.8322799679688689, + "grad_norm": 0.09242725372314453, + "learning_rate": 1.4395782310705209e-06, + "loss": 0.002823355421423912, + "step": 15720, + "token_acc": 0.998940368618224 + }, + { + "epoch": 0.8328094081520552, + "grad_norm": 0.07496926188468933, + "learning_rate": 1.4307271778362175e-06, + "loss": 0.0023126501590013506, + "step": 15730, + "token_acc": 0.9991072173550272 + }, + { + "epoch": 0.8333388483352415, + "grad_norm": 0.09402763843536377, + "learning_rate": 1.4219013213944045e-06, + "loss": 0.002599596418440342, + "step": 15740, + "token_acc": 0.9990162321692081 + }, + { + "epoch": 0.8338682885184279, + "grad_norm": 0.07234492152929306, + "learning_rate": 1.413100687696325e-06, + "loss": 0.0025988150388002395, + "step": 15750, + "token_acc": 0.9989979792581707 + }, + { + "epoch": 0.8343977287016141, + "grad_norm": 0.08698644489049911, + "learning_rate": 1.404325302619055e-06, + "loss": 0.002579846233129501, + "step": 15760, + "token_acc": 0.9989657703342897 + }, + { + "epoch": 0.8349271688848005, + "grad_norm": 0.077992744743824, + "learning_rate": 1.3955751919654337e-06, + "loss": 0.0025061095133423806, + "step": 15770, + "token_acc": 0.9990739427513078 + }, + { + "epoch": 0.8354566090679867, + "grad_norm": 0.1002521812915802, + "learning_rate": 1.3868503814639834e-06, + "loss": 0.0023947948589920997, + "step": 15780, + "token_acc": 0.9990574457614253 + }, + { + "epoch": 0.835986049251173, + "grad_norm": 0.0880778580904007, + "learning_rate": 1.3781508967688316e-06, + "loss": 0.002555733360350132, + "step": 15790, + "token_acc": 0.9989823071597194 + }, + { + "epoch": 0.8365154894343594, + "grad_norm": 0.10500499606132507, + "learning_rate": 1.369476763459646e-06, + "loss": 0.0025547750294208525, + "step": 15800, + "token_acc": 0.9989910276508456 + }, + { + "epoch": 0.8370449296175456, + "grad_norm": 0.08560687303543091, + "learning_rate": 1.3608280070415425e-06, + "loss": 0.0023888822644948958, + "step": 15810, + "token_acc": 0.9990659427723152 + }, + { + "epoch": 0.837574369800732, + "grad_norm": 0.08371592313051224, + "learning_rate": 1.352204652945026e-06, + "loss": 0.002433671988546848, + "step": 15820, + "token_acc": 0.9991067179816835 + }, + { + "epoch": 0.8381038099839182, + "grad_norm": 0.0834755226969719, + "learning_rate": 1.3436067265259066e-06, + "loss": 0.00223463773727417, + "step": 15830, + "token_acc": 0.9991740294846445 + }, + { + "epoch": 0.8386332501671045, + "grad_norm": 0.059335678815841675, + "learning_rate": 1.3350342530652317e-06, + "loss": 0.0022908492013812066, + "step": 15840, + "token_acc": 0.9991662984484814 + }, + { + "epoch": 0.8391626903502909, + "grad_norm": 0.09442520141601562, + "learning_rate": 1.3264872577692022e-06, + "loss": 0.0021867575123906136, + "step": 15850, + "token_acc": 0.9990575558168823 + }, + { + "epoch": 0.8396921305334771, + "grad_norm": 0.1053178459405899, + "learning_rate": 1.317965765769107e-06, + "loss": 0.0022815696895122526, + "step": 15860, + "token_acc": 0.9990992343491968 + }, + { + "epoch": 0.8402215707166635, + "grad_norm": 0.10747252404689789, + "learning_rate": 1.309469802121246e-06, + "loss": 0.0025641445070505144, + "step": 15870, + "token_acc": 0.9989817465550482 + }, + { + "epoch": 0.8407510108998498, + "grad_norm": 0.09118606895208359, + "learning_rate": 1.3009993918068564e-06, + "loss": 0.0026136547327041625, + "step": 15880, + "token_acc": 0.99894851916449 + }, + { + "epoch": 0.8412804510830361, + "grad_norm": 0.09919490665197372, + "learning_rate": 1.2925545597320388e-06, + "loss": 0.0026182178407907486, + "step": 15890, + "token_acc": 0.9989986565308455 + }, + { + "epoch": 0.8418098912662224, + "grad_norm": 0.09187951683998108, + "learning_rate": 1.284135330727685e-06, + "loss": 0.002405889332294464, + "step": 15900, + "token_acc": 0.9990152551991187 + }, + { + "epoch": 0.8423393314494086, + "grad_norm": 0.09006477892398834, + "learning_rate": 1.2757417295494012e-06, + "loss": 0.00229044109582901, + "step": 15910, + "token_acc": 0.9991488083316643 + }, + { + "epoch": 0.842868771632595, + "grad_norm": 0.08589550852775574, + "learning_rate": 1.2673737808774434e-06, + "loss": 0.0028909342363476753, + "step": 15920, + "token_acc": 0.998948992784752 + }, + { + "epoch": 0.8433982118157813, + "grad_norm": 0.06804647296667099, + "learning_rate": 1.2590315093166361e-06, + "loss": 0.0025324176996946333, + "step": 15930, + "token_acc": 0.999057760137417 + }, + { + "epoch": 0.8439276519989676, + "grad_norm": 0.09280627220869064, + "learning_rate": 1.2507149393963081e-06, + "loss": 0.0023559173569083214, + "step": 15940, + "token_acc": 0.9992320726526047 + }, + { + "epoch": 0.8444570921821539, + "grad_norm": 0.06362074613571167, + "learning_rate": 1.2424240955702082e-06, + "loss": 0.0023885246366262434, + "step": 15950, + "token_acc": 0.9991239121909704 + }, + { + "epoch": 0.8449865323653402, + "grad_norm": 0.1041622906923294, + "learning_rate": 1.2341590022164496e-06, + "loss": 0.0026021456345915795, + "step": 15960, + "token_acc": 0.9990328497582125 + }, + { + "epoch": 0.8455159725485265, + "grad_norm": 0.08286857604980469, + "learning_rate": 1.2259196836374244e-06, + "loss": 0.0025524552911520004, + "step": 15970, + "token_acc": 0.9991073514198953 + }, + { + "epoch": 0.8460454127317129, + "grad_norm": 0.08514775335788727, + "learning_rate": 1.2177061640597399e-06, + "loss": 0.0024499766528606414, + "step": 15980, + "token_acc": 0.9990988585541686 + }, + { + "epoch": 0.8465748529148991, + "grad_norm": 0.07216134667396545, + "learning_rate": 1.209518467634143e-06, + "loss": 0.0022013064473867415, + "step": 15990, + "token_acc": 0.9991075524417199 + }, + { + "epoch": 0.8471042930980854, + "grad_norm": 0.0996067002415657, + "learning_rate": 1.2013566184354542e-06, + "loss": 0.002331162244081497, + "step": 16000, + "token_acc": 0.9990908408471028 + }, + { + "epoch": 0.8476337332812717, + "grad_norm": 0.08986614644527435, + "learning_rate": 1.1932206404624901e-06, + "loss": 0.002367625944316387, + "step": 16010, + "token_acc": 0.9990906512272038 + }, + { + "epoch": 0.848163173464458, + "grad_norm": 0.08185578137636185, + "learning_rate": 1.1851105576380006e-06, + "loss": 0.0024272631853818893, + "step": 16020, + "token_acc": 0.9991075301104327 + }, + { + "epoch": 0.8486926136476444, + "grad_norm": 0.06554555892944336, + "learning_rate": 1.1770263938085879e-06, + "loss": 0.002320902980864048, + "step": 16030, + "token_acc": 0.9991158193268549 + }, + { + "epoch": 0.8492220538308306, + "grad_norm": 0.10153020173311234, + "learning_rate": 1.1689681727446477e-06, + "loss": 0.002306465245783329, + "step": 16040, + "token_acc": 0.9991075152222871 + }, + { + "epoch": 0.849751494014017, + "grad_norm": 0.092742420732975, + "learning_rate": 1.1609359181402958e-06, + "loss": 0.0021018244326114653, + "step": 16050, + "token_acc": 0.9990989938764954 + }, + { + "epoch": 0.8502809341972032, + "grad_norm": 0.08784468472003937, + "learning_rate": 1.1529296536132939e-06, + "loss": 0.002349836379289627, + "step": 16060, + "token_acc": 0.9991657768286172 + }, + { + "epoch": 0.8508103743803895, + "grad_norm": 0.12630102038383484, + "learning_rate": 1.1449494027049835e-06, + "loss": 0.0025905279442667963, + "step": 16070, + "token_acc": 0.9990239751739326 + }, + { + "epoch": 0.8513398145635759, + "grad_norm": 0.11619700491428375, + "learning_rate": 1.1369951888802188e-06, + "loss": 0.002401433512568474, + "step": 16080, + "token_acc": 0.9990907801773395 + }, + { + "epoch": 0.8518692547467621, + "grad_norm": 0.10642553120851517, + "learning_rate": 1.1290670355272938e-06, + "loss": 0.0025503624230623244, + "step": 16090, + "token_acc": 0.9991240802842986 + }, + { + "epoch": 0.8523986949299485, + "grad_norm": 0.0960921123623848, + "learning_rate": 1.1211649659578761e-06, + "loss": 0.0025045568123459815, + "step": 16100, + "token_acc": 0.9990404512382351 + }, + { + "epoch": 0.8529281351131347, + "grad_norm": 0.12831172347068787, + "learning_rate": 1.1132890034069354e-06, + "loss": 0.002685297280550003, + "step": 16110, + "token_acc": 0.9989324080870088 + }, + { + "epoch": 0.853457575296321, + "grad_norm": 0.08403823524713516, + "learning_rate": 1.1054391710326816e-06, + "loss": 0.0022169895470142364, + "step": 16120, + "token_acc": 0.9990986554944459 + }, + { + "epoch": 0.8539870154795074, + "grad_norm": 0.09647926688194275, + "learning_rate": 1.097615491916485e-06, + "loss": 0.0023500382900238036, + "step": 16130, + "token_acc": 0.9991235758106923 + }, + { + "epoch": 0.8545164556626936, + "grad_norm": 0.08204001188278198, + "learning_rate": 1.0898179890628234e-06, + "loss": 0.0024673786014318467, + "step": 16140, + "token_acc": 0.9991321255413784 + }, + { + "epoch": 0.85504589584588, + "grad_norm": 0.10058377683162689, + "learning_rate": 1.0820466853992028e-06, + "loss": 0.002559250220656395, + "step": 16150, + "token_acc": 0.9990319134730938 + }, + { + "epoch": 0.8555753360290662, + "grad_norm": 0.0797661691904068, + "learning_rate": 1.0743016037760946e-06, + "loss": 0.0020648933947086333, + "step": 16160, + "token_acc": 0.9992909221349083 + }, + { + "epoch": 0.8561047762122526, + "grad_norm": 0.12094248831272125, + "learning_rate": 1.0665827669668682e-06, + "loss": 0.0026639699935913086, + "step": 16170, + "token_acc": 0.9990402830747661 + }, + { + "epoch": 0.8566342163954389, + "grad_norm": 0.08476892113685608, + "learning_rate": 1.0588901976677235e-06, + "loss": 0.002490963228046894, + "step": 16180, + "token_acc": 0.9991071577577143 + }, + { + "epoch": 0.8571636565786251, + "grad_norm": 0.08236558735370636, + "learning_rate": 1.0512239184976259e-06, + "loss": 0.0025233741849660873, + "step": 16190, + "token_acc": 0.9989989488963411 + }, + { + "epoch": 0.8576930967618115, + "grad_norm": 0.08283514529466629, + "learning_rate": 1.0435839519982349e-06, + "loss": 0.002327500842511654, + "step": 16200, + "token_acc": 0.9990574221747689 + }, + { + "epoch": 0.8582225369449978, + "grad_norm": 0.08618906885385513, + "learning_rate": 1.0359703206338435e-06, + "loss": 0.002279461920261383, + "step": 16210, + "token_acc": 0.9991401117019945 + }, + { + "epoch": 0.8587519771281841, + "grad_norm": 0.0749441608786583, + "learning_rate": 1.0283830467913103e-06, + "loss": 0.0021953023970127106, + "step": 16220, + "token_acc": 0.9991322341632735 + }, + { + "epoch": 0.8592814173113704, + "grad_norm": 0.08628387749195099, + "learning_rate": 1.02082215277999e-06, + "loss": 0.002403458207845688, + "step": 16230, + "token_acc": 0.999099864979747 + }, + { + "epoch": 0.8598108574945567, + "grad_norm": 0.10228714346885681, + "learning_rate": 1.0132876608316754e-06, + "loss": 0.0021017735823988916, + "step": 16240, + "token_acc": 0.9991571463143928 + }, + { + "epoch": 0.860340297677743, + "grad_norm": 0.09213738888502121, + "learning_rate": 1.005779593100522e-06, + "loss": 0.0024308063089847564, + "step": 16250, + "token_acc": 0.99903228497539 + }, + { + "epoch": 0.8608697378609294, + "grad_norm": 0.0972374975681305, + "learning_rate": 9.982979716629936e-07, + "loss": 0.0023803522810339926, + "step": 16260, + "token_acc": 0.999123729407641 + }, + { + "epoch": 0.8613991780441156, + "grad_norm": 0.08680824935436249, + "learning_rate": 9.908428185177898e-07, + "loss": 0.0022451061755418777, + "step": 16270, + "token_acc": 0.9991241241241241 + }, + { + "epoch": 0.8619286182273019, + "grad_norm": 0.09000048041343689, + "learning_rate": 9.834141555857847e-07, + "loss": 0.002094823308289051, + "step": 16280, + "token_acc": 0.9991825635807051 + }, + { + "epoch": 0.8624580584104882, + "grad_norm": 0.0871637836098671, + "learning_rate": 9.760120047099608e-07, + "loss": 0.002291351556777954, + "step": 16290, + "token_acc": 0.9991989319092123 + }, + { + "epoch": 0.8629874985936745, + "grad_norm": 0.08163755387067795, + "learning_rate": 9.68636387655345e-07, + "loss": 0.0022896572947502134, + "step": 16300, + "token_acc": 0.9990490252508821 + }, + { + "epoch": 0.8635169387768609, + "grad_norm": 0.07211092859506607, + "learning_rate": 9.612873261089483e-07, + "loss": 0.0024365322664380074, + "step": 16310, + "token_acc": 0.9990741358601362 + }, + { + "epoch": 0.8640463789600471, + "grad_norm": 0.06685639172792435, + "learning_rate": 9.539648416796933e-07, + "loss": 0.0020786356180906296, + "step": 16320, + "token_acc": 0.9992077787784782 + }, + { + "epoch": 0.8645758191432334, + "grad_norm": 0.0951664075255394, + "learning_rate": 9.46668955898361e-07, + "loss": 0.0023359864950180053, + "step": 16330, + "token_acc": 0.9990825535038116 + }, + { + "epoch": 0.8651052593264197, + "grad_norm": 0.06530774384737015, + "learning_rate": 9.393996902175218e-07, + "loss": 0.0022738704457879065, + "step": 16340, + "token_acc": 0.9991494471406415 + }, + { + "epoch": 0.865634699509606, + "grad_norm": 0.09316892921924591, + "learning_rate": 9.321570660114687e-07, + "loss": 0.0024407083168625832, + "step": 16350, + "token_acc": 0.9990323011211959 + }, + { + "epoch": 0.8661641396927924, + "grad_norm": 0.07008277624845505, + "learning_rate": 9.249411045761658e-07, + "loss": 0.0021681353449821473, + "step": 16360, + "token_acc": 0.999174649225921 + }, + { + "epoch": 0.8666935798759786, + "grad_norm": 0.10108017921447754, + "learning_rate": 9.177518271291741e-07, + "loss": 0.002104620262980461, + "step": 16370, + "token_acc": 0.9992244441849992 + }, + { + "epoch": 0.867223020059165, + "grad_norm": 0.07476262003183365, + "learning_rate": 9.105892548095951e-07, + "loss": 0.0023061862215399743, + "step": 16380, + "token_acc": 0.9991159004470541 + }, + { + "epoch": 0.8677524602423512, + "grad_norm": 0.09893724322319031, + "learning_rate": 9.0345340867801e-07, + "loss": 0.00253312848508358, + "step": 16390, + "token_acc": 0.9991572588383523 + }, + { + "epoch": 0.8682819004255375, + "grad_norm": 0.12122004479169846, + "learning_rate": 8.963443097164126e-07, + "loss": 0.002156557887792587, + "step": 16400, + "token_acc": 0.9990740740740741 + }, + { + "epoch": 0.8688113406087239, + "grad_norm": 0.08731667697429657, + "learning_rate": 8.892619788281487e-07, + "loss": 0.0019489720463752747, + "step": 16410, + "token_acc": 0.9992409772209757 + }, + { + "epoch": 0.8693407807919101, + "grad_norm": 0.06583050638437271, + "learning_rate": 8.822064368378613e-07, + "loss": 0.002591731399297714, + "step": 16420, + "token_acc": 0.9990235353029544 + }, + { + "epoch": 0.8698702209750965, + "grad_norm": 0.08148118108510971, + "learning_rate": 8.7517770449142e-07, + "loss": 0.0021829381585121155, + "step": 16430, + "token_acc": 0.9991653311965812 + }, + { + "epoch": 0.8703996611582827, + "grad_norm": 0.09586796164512634, + "learning_rate": 8.681758024558661e-07, + "loss": 0.0024335684254765512, + "step": 16440, + "token_acc": 0.9990574693469013 + }, + { + "epoch": 0.8709291013414691, + "grad_norm": 0.0914069190621376, + "learning_rate": 8.612007513193477e-07, + "loss": 0.002287931181490421, + "step": 16450, + "token_acc": 0.9990906132937319 + }, + { + "epoch": 0.8714585415246554, + "grad_norm": 0.10264045745134354, + "learning_rate": 8.542525715910644e-07, + "loss": 0.0021370701491832734, + "step": 16460, + "token_acc": 0.9992991882263622 + }, + { + "epoch": 0.8719879817078416, + "grad_norm": 0.08214473724365234, + "learning_rate": 8.473312837012027e-07, + "loss": 0.0020849965512752533, + "step": 16470, + "token_acc": 0.9992575166015951 + }, + { + "epoch": 0.872517421891028, + "grad_norm": 0.08767972886562347, + "learning_rate": 8.404369080008768e-07, + "loss": 0.0021775305271148683, + "step": 16480, + "token_acc": 0.9992077325305023 + }, + { + "epoch": 0.8730468620742142, + "grad_norm": 0.09107723832130432, + "learning_rate": 8.335694647620729e-07, + "loss": 0.002438681572675705, + "step": 16490, + "token_acc": 0.9991160773849233 + }, + { + "epoch": 0.8735763022574006, + "grad_norm": 0.07920833677053452, + "learning_rate": 8.267289741775775e-07, + "loss": 0.0022123508155345918, + "step": 16500, + "token_acc": 0.9992072499082079 + }, + { + "epoch": 0.8741057424405869, + "grad_norm": 0.0976443886756897, + "learning_rate": 8.199154563609357e-07, + "loss": 0.0025723207741975783, + "step": 16510, + "token_acc": 0.9990654673497655 + }, + { + "epoch": 0.8746351826237732, + "grad_norm": 0.08808494359254837, + "learning_rate": 8.131289313463764e-07, + "loss": 0.00221429243683815, + "step": 16520, + "token_acc": 0.99912405836274 + }, + { + "epoch": 0.8751646228069595, + "grad_norm": 0.1260966658592224, + "learning_rate": 8.06369419088765e-07, + "loss": 0.0023735411465168, + "step": 16530, + "token_acc": 0.9990490649138319 + }, + { + "epoch": 0.8756940629901458, + "grad_norm": 0.09164728969335556, + "learning_rate": 7.996369394635351e-07, + "loss": 0.002113846689462662, + "step": 16540, + "token_acc": 0.9991741948399689 + }, + { + "epoch": 0.8762235031733321, + "grad_norm": 0.08544573187828064, + "learning_rate": 7.929315122666359e-07, + "loss": 0.002081836760044098, + "step": 16550, + "token_acc": 0.9992069521082552 + }, + { + "epoch": 0.8767529433565184, + "grad_norm": 0.07350046187639236, + "learning_rate": 7.862531572144749e-07, + "loss": 0.0025009023025631904, + "step": 16560, + "token_acc": 0.9990902717500167 + }, + { + "epoch": 0.8772823835397047, + "grad_norm": 0.1262260377407074, + "learning_rate": 7.796018939438533e-07, + "loss": 0.002323727682232857, + "step": 16570, + "token_acc": 0.9991408361416035 + }, + { + "epoch": 0.877811823722891, + "grad_norm": 0.0857425257563591, + "learning_rate": 7.729777420119155e-07, + "loss": 0.0024500522762537003, + "step": 16580, + "token_acc": 0.9990995047276002 + }, + { + "epoch": 0.8783412639060774, + "grad_norm": 0.09190305322408676, + "learning_rate": 7.663807208960883e-07, + "loss": 0.0020844219252467155, + "step": 16590, + "token_acc": 0.9991741741741742 + }, + { + "epoch": 0.8788707040892636, + "grad_norm": 0.10032513737678528, + "learning_rate": 7.598108499940237e-07, + "loss": 0.0021449126303195953, + "step": 16600, + "token_acc": 0.9991818401916831 + }, + { + "epoch": 0.87940014427245, + "grad_norm": 0.13777637481689453, + "learning_rate": 7.532681486235416e-07, + "loss": 0.0021868335083127023, + "step": 16610, + "token_acc": 0.9992073491251637 + }, + { + "epoch": 0.8799295844556362, + "grad_norm": 0.1032845675945282, + "learning_rate": 7.467526360225752e-07, + "loss": 0.002194729819893837, + "step": 16620, + "token_acc": 0.9991071354066706 + }, + { + "epoch": 0.8804590246388225, + "grad_norm": 0.09758809208869934, + "learning_rate": 7.402643313491109e-07, + "loss": 0.0019295927137136458, + "step": 16630, + "token_acc": 0.9991988918003237 + }, + { + "epoch": 0.8809884648220089, + "grad_norm": 0.08712895959615707, + "learning_rate": 7.338032536811357e-07, + "loss": 0.002124910242855549, + "step": 16640, + "token_acc": 0.9991238025301244 + }, + { + "epoch": 0.8815179050051951, + "grad_norm": 0.11778214573860168, + "learning_rate": 7.273694220165762e-07, + "loss": 0.002427450940012932, + "step": 16650, + "token_acc": 0.9990075393648252 + }, + { + "epoch": 0.8820473451883815, + "grad_norm": 0.1225259006023407, + "learning_rate": 7.209628552732506e-07, + "loss": 0.0022301387041807176, + "step": 16660, + "token_acc": 0.9990658882402002 + }, + { + "epoch": 0.8825767853715677, + "grad_norm": 0.08817021548748016, + "learning_rate": 7.145835722888028e-07, + "loss": 0.0023319533094763754, + "step": 16670, + "token_acc": 0.999115140283656 + }, + { + "epoch": 0.883106225554754, + "grad_norm": 0.1040254533290863, + "learning_rate": 7.082315918206561e-07, + "loss": 0.0021027131006121635, + "step": 16680, + "token_acc": 0.9991653869266208 + }, + { + "epoch": 0.8836356657379404, + "grad_norm": 0.08091525733470917, + "learning_rate": 7.019069325459538e-07, + "loss": 0.002465756982564926, + "step": 16690, + "token_acc": 0.9990238530273071 + }, + { + "epoch": 0.8841651059211266, + "grad_norm": 0.13280081748962402, + "learning_rate": 6.956096130615042e-07, + "loss": 0.0022840695455670357, + "step": 16700, + "token_acc": 0.9990988059178412 + }, + { + "epoch": 0.884694546104313, + "grad_norm": 0.058297161012887955, + "learning_rate": 6.89339651883727e-07, + "loss": 0.0021291492506861688, + "step": 16710, + "token_acc": 0.9992325146823278 + }, + { + "epoch": 0.8852239862874992, + "grad_norm": 0.08407409489154816, + "learning_rate": 6.830970674485982e-07, + "loss": 0.0023063525557518007, + "step": 16720, + "token_acc": 0.9990820558610733 + }, + { + "epoch": 0.8857534264706856, + "grad_norm": 0.08805397897958755, + "learning_rate": 6.768818781115982e-07, + "loss": 0.002531985193490982, + "step": 16730, + "token_acc": 0.9990401869548888 + }, + { + "epoch": 0.8862828666538719, + "grad_norm": 0.0804048478603363, + "learning_rate": 6.706941021476521e-07, + "loss": 0.0022350046783685684, + "step": 16740, + "token_acc": 0.9991162321474725 + }, + { + "epoch": 0.8868123068370581, + "grad_norm": 0.06452525407075882, + "learning_rate": 6.645337577510847e-07, + "loss": 0.002026721462607384, + "step": 16750, + "token_acc": 0.9992661162538571 + }, + { + "epoch": 0.8873417470202445, + "grad_norm": 0.09755001962184906, + "learning_rate": 6.584008630355543e-07, + "loss": 0.0020289402455091476, + "step": 16760, + "token_acc": 0.9992827356130108 + }, + { + "epoch": 0.8878711872034307, + "grad_norm": 0.08147627115249634, + "learning_rate": 6.522954360340139e-07, + "loss": 0.0022091187536716463, + "step": 16770, + "token_acc": 0.9991750068749428 + }, + { + "epoch": 0.8884006273866171, + "grad_norm": 0.10032638162374496, + "learning_rate": 6.462174946986511e-07, + "loss": 0.002504967525601387, + "step": 16780, + "token_acc": 0.9989487121724112 + }, + { + "epoch": 0.8889300675698034, + "grad_norm": 0.12397350370883942, + "learning_rate": 6.401670569008311e-07, + "loss": 0.002355308458209038, + "step": 16790, + "token_acc": 0.9990569738039006 + }, + { + "epoch": 0.8894595077529897, + "grad_norm": 0.07949100434780121, + "learning_rate": 6.341441404310522e-07, + "loss": 0.002355266734957695, + "step": 16800, + "token_acc": 0.9991070310870018 + }, + { + "epoch": 0.889988947936176, + "grad_norm": 0.07372858375310898, + "learning_rate": 6.281487629988881e-07, + "loss": 0.002205265313386917, + "step": 16810, + "token_acc": 0.9991406641081262 + }, + { + "epoch": 0.8905183881193622, + "grad_norm": 0.08473403751850128, + "learning_rate": 6.221809422329406e-07, + "loss": 0.0020689476281404496, + "step": 16820, + "token_acc": 0.9991904928813446 + }, + { + "epoch": 0.8910478283025486, + "grad_norm": 0.0972638949751854, + "learning_rate": 6.16240695680781e-07, + "loss": 0.0021564757451415064, + "step": 16830, + "token_acc": 0.9992156339179921 + }, + { + "epoch": 0.8915772684857349, + "grad_norm": 0.09670817106962204, + "learning_rate": 6.103280408089052e-07, + "loss": 0.00235123485326767, + "step": 16840, + "token_acc": 0.9992069123846893 + }, + { + "epoch": 0.8921067086689212, + "grad_norm": 0.10003198683261871, + "learning_rate": 6.044429950026786e-07, + "loss": 0.002288242429494858, + "step": 16850, + "token_acc": 0.9991404991780501 + }, + { + "epoch": 0.8926361488521075, + "grad_norm": 0.059302497655153275, + "learning_rate": 5.985855755662851e-07, + "loss": 0.0017237659543752671, + "step": 16860, + "token_acc": 0.9992821788376305 + }, + { + "epoch": 0.8931655890352939, + "grad_norm": 0.07633254677057266, + "learning_rate": 5.927557997226785e-07, + "loss": 0.0020890194922685624, + "step": 16870, + "token_acc": 0.9992154047760148 + }, + { + "epoch": 0.8936950292184801, + "grad_norm": 0.07453097403049469, + "learning_rate": 5.869536846135282e-07, + "loss": 0.0019986411556601525, + "step": 16880, + "token_acc": 0.9992408315814062 + }, + { + "epoch": 0.8942244694016664, + "grad_norm": 0.10289797186851501, + "learning_rate": 5.811792472991718e-07, + "loss": 0.0026891566812992096, + "step": 16890, + "token_acc": 0.9990237064109946 + }, + { + "epoch": 0.8947539095848527, + "grad_norm": 0.08072052896022797, + "learning_rate": 5.754325047585651e-07, + "loss": 0.0019872715696692468, + "step": 16900, + "token_acc": 0.9992492492492493 + }, + { + "epoch": 0.895283349768039, + "grad_norm": 0.09491624683141708, + "learning_rate": 5.697134738892318e-07, + "loss": 0.0020303145051002503, + "step": 16910, + "token_acc": 0.9991575610976728 + }, + { + "epoch": 0.8958127899512254, + "grad_norm": 0.08324210345745087, + "learning_rate": 5.640221715072091e-07, + "loss": 0.002317227050662041, + "step": 16920, + "token_acc": 0.9990651007103565 + }, + { + "epoch": 0.8963422301344116, + "grad_norm": 0.08922237902879715, + "learning_rate": 5.583586143470076e-07, + "loss": 0.0023006686940789224, + "step": 16930, + "token_acc": 0.9990493820983639 + }, + { + "epoch": 0.896871670317598, + "grad_norm": 0.08643261343240738, + "learning_rate": 5.527228190615541e-07, + "loss": 0.0018477454781532288, + "step": 16940, + "token_acc": 0.9993409472006941 + }, + { + "epoch": 0.8974011105007842, + "grad_norm": 0.06402246654033661, + "learning_rate": 5.471148022221429e-07, + "loss": 0.001960694417357445, + "step": 16950, + "token_acc": 0.9992992233058307 + }, + { + "epoch": 0.8979305506839705, + "grad_norm": 0.07398195564746857, + "learning_rate": 5.415345803183958e-07, + "loss": 0.0020707461982965468, + "step": 16960, + "token_acc": 0.9992740986725184 + }, + { + "epoch": 0.8984599908671569, + "grad_norm": 0.06887250393629074, + "learning_rate": 5.359821697582024e-07, + "loss": 0.001969555765390396, + "step": 16970, + "token_acc": 0.9993240649565232 + }, + { + "epoch": 0.8989894310503431, + "grad_norm": 0.09057144820690155, + "learning_rate": 5.304575868676754e-07, + "loss": 0.0019350048154592515, + "step": 16980, + "token_acc": 0.9993244764692637 + }, + { + "epoch": 0.8995188712335295, + "grad_norm": 0.08465085178613663, + "learning_rate": 5.249608478911095e-07, + "loss": 0.001894519105553627, + "step": 16990, + "token_acc": 0.9992825261752806 + }, + { + "epoch": 0.9000483114167157, + "grad_norm": 0.08330614864826202, + "learning_rate": 5.19491968990925e-07, + "loss": 0.0022463735193014147, + "step": 17000, + "token_acc": 0.9991155242187826 + }, + { + "epoch": 0.900577751599902, + "grad_norm": 0.100593701004982, + "learning_rate": 5.140509662476223e-07, + "loss": 0.0023477278649806977, + "step": 17010, + "token_acc": 0.999107708729444 + }, + { + "epoch": 0.9011071917830884, + "grad_norm": 0.0832865908741951, + "learning_rate": 5.086378556597383e-07, + "loss": 0.0018615100532770156, + "step": 17020, + "token_acc": 0.9993740349705796 + }, + { + "epoch": 0.9016366319662746, + "grad_norm": 0.08051163703203201, + "learning_rate": 5.032526531437954e-07, + "loss": 0.0018396107479929924, + "step": 17030, + "token_acc": 0.9993160511126681 + }, + { + "epoch": 0.902166072149461, + "grad_norm": 0.09023325145244598, + "learning_rate": 4.978953745342574e-07, + "loss": 0.00224925484508276, + "step": 17040, + "token_acc": 0.9991068000066782 + }, + { + "epoch": 0.9026955123326472, + "grad_norm": 0.06281016021966934, + "learning_rate": 4.925660355834783e-07, + "loss": 0.00189041830599308, + "step": 17050, + "token_acc": 0.9992661346142621 + }, + { + "epoch": 0.9032249525158336, + "grad_norm": 0.12031037360429764, + "learning_rate": 4.87264651961662e-07, + "loss": 0.002138815447688103, + "step": 17060, + "token_acc": 0.9992242242242242 + }, + { + "epoch": 0.9037543926990199, + "grad_norm": 0.08921172469854355, + "learning_rate": 4.819912392568126e-07, + "loss": 0.002419743686914444, + "step": 17070, + "token_acc": 0.9991077459327391 + }, + { + "epoch": 0.9042838328822062, + "grad_norm": 0.05873069912195206, + "learning_rate": 4.767458129746916e-07, + "loss": 0.0022126238793134688, + "step": 17080, + "token_acc": 0.9992152868794296 + }, + { + "epoch": 0.9048132730653925, + "grad_norm": 0.05552367866039276, + "learning_rate": 4.715283885387678e-07, + "loss": 0.0018749767914414406, + "step": 17090, + "token_acc": 0.9992910049379421 + }, + { + "epoch": 0.9053427132485787, + "grad_norm": 0.09197156131267548, + "learning_rate": 4.663389812901742e-07, + "loss": 0.001953360065817833, + "step": 17100, + "token_acc": 0.9992076466270768 + }, + { + "epoch": 0.9058721534317651, + "grad_norm": 0.08967173099517822, + "learning_rate": 4.6117760648766276e-07, + "loss": 0.0020676851272583008, + "step": 17110, + "token_acc": 0.9991908372748735 + }, + { + "epoch": 0.9064015936149514, + "grad_norm": 0.061723340302705765, + "learning_rate": 4.5604427930756366e-07, + "loss": 0.002228090725839138, + "step": 17120, + "token_acc": 0.9991320241364058 + }, + { + "epoch": 0.9069310337981377, + "grad_norm": 0.09508956968784332, + "learning_rate": 4.50939014843732e-07, + "loss": 0.0018032921478152275, + "step": 17130, + "token_acc": 0.99937474989996 + }, + { + "epoch": 0.907460473981324, + "grad_norm": 0.10391975939273834, + "learning_rate": 4.458618281075111e-07, + "loss": 0.0021069152280688286, + "step": 17140, + "token_acc": 0.9991241168177912 + }, + { + "epoch": 0.9079899141645102, + "grad_norm": 0.08783455193042755, + "learning_rate": 4.4081273402768464e-07, + "loss": 0.002133310027420521, + "step": 17150, + "token_acc": 0.9992074417052518 + }, + { + "epoch": 0.9085193543476966, + "grad_norm": 0.10606454312801361, + "learning_rate": 4.3579174745043253e-07, + "loss": 0.0020396407693624497, + "step": 17160, + "token_acc": 0.9991994596352538 + }, + { + "epoch": 0.9090487945308829, + "grad_norm": 0.08260182291269302, + "learning_rate": 4.3079888313929175e-07, + "loss": 0.002501970902085304, + "step": 17170, + "token_acc": 0.9990064123973015 + }, + { + "epoch": 0.9095782347140692, + "grad_norm": 0.10310252755880356, + "learning_rate": 4.2583415577510554e-07, + "loss": 0.0022511731833219526, + "step": 17180, + "token_acc": 0.9991405924071757 + }, + { + "epoch": 0.9101076748972555, + "grad_norm": 0.07963304221630096, + "learning_rate": 4.2089757995598777e-07, + "loss": 0.0019296742975711823, + "step": 17190, + "token_acc": 0.9992244959223495 + }, + { + "epoch": 0.9106371150804419, + "grad_norm": 0.1135302186012268, + "learning_rate": 4.1598917019727293e-07, + "loss": 0.002195674553513527, + "step": 17200, + "token_acc": 0.9991146820789938 + }, + { + "epoch": 0.9111665552636281, + "grad_norm": 0.08967743068933487, + "learning_rate": 4.1110894093147837e-07, + "loss": 0.0022018466144800184, + "step": 17210, + "token_acc": 0.9991733674006162 + }, + { + "epoch": 0.9116959954468145, + "grad_norm": 0.10516800731420517, + "learning_rate": 4.062569065082622e-07, + "loss": 0.0022115685045719148, + "step": 17220, + "token_acc": 0.9991405493808618 + }, + { + "epoch": 0.9122254356300007, + "grad_norm": 0.09111347049474716, + "learning_rate": 4.0143308119437783e-07, + "loss": 0.002279951795935631, + "step": 17230, + "token_acc": 0.9991733466933868 + }, + { + "epoch": 0.912754875813187, + "grad_norm": 0.1059509739279747, + "learning_rate": 3.966374791736338e-07, + "loss": 0.0022253010421991347, + "step": 17240, + "token_acc": 0.9992411164761118 + }, + { + "epoch": 0.9132843159963734, + "grad_norm": 0.08143110573291779, + "learning_rate": 3.9187011454685174e-07, + "loss": 0.002197651192545891, + "step": 17250, + "token_acc": 0.999107522666422 + }, + { + "epoch": 0.9138137561795596, + "grad_norm": 0.088469497859478, + "learning_rate": 3.8713100133182524e-07, + "loss": 0.0020544635131955146, + "step": 17260, + "token_acc": 0.99926567532836 + }, + { + "epoch": 0.914343196362746, + "grad_norm": 0.0857153907418251, + "learning_rate": 3.8242015346327877e-07, + "loss": 0.002419964596629143, + "step": 17270, + "token_acc": 0.9991988717443733 + }, + { + "epoch": 0.9148726365459322, + "grad_norm": 0.09507952630519867, + "learning_rate": 3.777375847928244e-07, + "loss": 0.0024507462978363035, + "step": 17280, + "token_acc": 0.9990906815717027 + }, + { + "epoch": 0.9154020767291186, + "grad_norm": 0.09115224331617355, + "learning_rate": 3.7308330908892855e-07, + "loss": 0.0023792386054992676, + "step": 17290, + "token_acc": 0.999107708729444 + }, + { + "epoch": 0.9159315169123049, + "grad_norm": 0.09789731353521347, + "learning_rate": 3.684573400368574e-07, + "loss": 0.0020947787910699845, + "step": 17300, + "token_acc": 0.9991822771079311 + }, + { + "epoch": 0.9164609570954911, + "grad_norm": 0.08335071057081223, + "learning_rate": 3.638596912386516e-07, + "loss": 0.001968991383910179, + "step": 17310, + "token_acc": 0.9992744376892091 + }, + { + "epoch": 0.9169903972786775, + "grad_norm": 0.10355088114738464, + "learning_rate": 3.5929037621307947e-07, + "loss": 0.0020446885377168656, + "step": 17320, + "token_acc": 0.9992573493211839 + }, + { + "epoch": 0.9175198374618637, + "grad_norm": 0.09731577336788177, + "learning_rate": 3.547494083955938e-07, + "loss": 0.0020982807502150534, + "step": 17330, + "token_acc": 0.9992573803047244 + }, + { + "epoch": 0.9180492776450501, + "grad_norm": 0.10574337840080261, + "learning_rate": 3.502368011383006e-07, + "loss": 0.0022320859134197235, + "step": 17340, + "token_acc": 0.9990990840688033 + }, + { + "epoch": 0.9185787178282364, + "grad_norm": 0.07314448058605194, + "learning_rate": 3.4575256770991274e-07, + "loss": 0.0021878771483898165, + "step": 17350, + "token_acc": 0.9992909457948915 + }, + { + "epoch": 0.9191081580114226, + "grad_norm": 0.056934453547000885, + "learning_rate": 3.412967212957152e-07, + "loss": 0.002494687959551811, + "step": 17360, + "token_acc": 0.9990567376478543 + }, + { + "epoch": 0.919637598194609, + "grad_norm": 0.09077022969722748, + "learning_rate": 3.3686927499752443e-07, + "loss": 0.0018930362537503242, + "step": 17370, + "token_acc": 0.9992157975089057 + }, + { + "epoch": 0.9201670383777952, + "grad_norm": 0.0928955003619194, + "learning_rate": 3.324702418336501e-07, + "loss": 0.001938376948237419, + "step": 17380, + "token_acc": 0.9991906212190746 + }, + { + "epoch": 0.9206964785609816, + "grad_norm": 0.09506716579198837, + "learning_rate": 3.2809963473885543e-07, + "loss": 0.0017499210312962531, + "step": 17390, + "token_acc": 0.9993326771325136 + }, + { + "epoch": 0.9212259187441679, + "grad_norm": 0.09124301373958588, + "learning_rate": 3.2375746656432284e-07, + "loss": 0.0019548801705241204, + "step": 17400, + "token_acc": 0.9992149395336407 + }, + { + "epoch": 0.9217553589273542, + "grad_norm": 0.05927102640271187, + "learning_rate": 3.1944375007761374e-07, + "loss": 0.0017473045736551285, + "step": 17410, + "token_acc": 0.9993576910243577 + }, + { + "epoch": 0.9222847991105405, + "grad_norm": 0.05175984278321266, + "learning_rate": 3.151584979626288e-07, + "loss": 0.0019814418628811836, + "step": 17420, + "token_acc": 0.99924894226035 + }, + { + "epoch": 0.9228142392937267, + "grad_norm": 0.08383145928382874, + "learning_rate": 3.1090172281957543e-07, + "loss": 0.0018943574279546738, + "step": 17430, + "token_acc": 0.9992992291520673 + }, + { + "epoch": 0.9233436794769131, + "grad_norm": 0.09110525250434875, + "learning_rate": 3.0667343716492717e-07, + "loss": 0.002185625582933426, + "step": 17440, + "token_acc": 0.9991491633439549 + }, + { + "epoch": 0.9238731196600994, + "grad_norm": 0.09756845235824585, + "learning_rate": 3.024736534313888e-07, + "loss": 0.0018680911511182786, + "step": 17450, + "token_acc": 0.9993078141939788 + }, + { + "epoch": 0.9244025598432857, + "grad_norm": 0.09847177565097809, + "learning_rate": 2.9830238396785784e-07, + "loss": 0.0021465420722961426, + "step": 17460, + "token_acc": 0.9991997599279784 + }, + { + "epoch": 0.924932000026472, + "grad_norm": 0.06098906323313713, + "learning_rate": 2.941596410393921e-07, + "loss": 0.00216008760035038, + "step": 17470, + "token_acc": 0.9991743670147113 + }, + { + "epoch": 0.9254614402096583, + "grad_norm": 0.08452929556369781, + "learning_rate": 2.900454368271677e-07, + "loss": 0.001861930638551712, + "step": 17480, + "token_acc": 0.9992489610628036 + }, + { + "epoch": 0.9259908803928446, + "grad_norm": 0.10778235644102097, + "learning_rate": 2.8595978342844886e-07, + "loss": 0.0020772457122802735, + "step": 17490, + "token_acc": 0.9992152082620203 + }, + { + "epoch": 0.926520320576031, + "grad_norm": 0.05830886960029602, + "learning_rate": 2.819026928565483e-07, + "loss": 0.002205232158303261, + "step": 17500, + "token_acc": 0.9991820382271931 + }, + { + "epoch": 0.9270497607592172, + "grad_norm": 0.06828165799379349, + "learning_rate": 2.7787417704079554e-07, + "loss": 0.001999301463365555, + "step": 17510, + "token_acc": 0.9992327643001893 + }, + { + "epoch": 0.9275792009424035, + "grad_norm": 0.10267084836959839, + "learning_rate": 2.7387424782649664e-07, + "loss": 0.001958618499338627, + "step": 17520, + "token_acc": 0.999266299816575 + }, + { + "epoch": 0.9281086411255899, + "grad_norm": 0.12018513679504395, + "learning_rate": 2.699029169749068e-07, + "loss": 0.002154861204326153, + "step": 17530, + "token_acc": 0.999206866035499 + }, + { + "epoch": 0.9286380813087761, + "grad_norm": 0.1087157130241394, + "learning_rate": 2.659601961631886e-07, + "loss": 0.0019658215343952177, + "step": 17540, + "token_acc": 0.9993239351979367 + }, + { + "epoch": 0.9291675214919625, + "grad_norm": 0.06090916693210602, + "learning_rate": 2.620460969843808e-07, + "loss": 0.0017729105427861213, + "step": 17550, + "token_acc": 0.9993406886882209 + }, + { + "epoch": 0.9296969616751487, + "grad_norm": 0.08646995574235916, + "learning_rate": 2.581606309473661e-07, + "loss": 0.002241835743188858, + "step": 17560, + "token_acc": 0.9991153027584192 + }, + { + "epoch": 0.930226401858335, + "grad_norm": 0.09457089751958847, + "learning_rate": 2.5430380947683243e-07, + "loss": 0.001740437000989914, + "step": 17570, + "token_acc": 0.9992743105950653 + }, + { + "epoch": 0.9307558420415214, + "grad_norm": 0.12147544324398041, + "learning_rate": 2.5047564391324607e-07, + "loss": 0.0026052448898553847, + "step": 17580, + "token_acc": 0.9990906208806795 + }, + { + "epoch": 0.9312852822247076, + "grad_norm": 0.07668498158454895, + "learning_rate": 2.466761455128108e-07, + "loss": 0.002141008898615837, + "step": 17590, + "token_acc": 0.999115738191768 + }, + { + "epoch": 0.931814722407894, + "grad_norm": 0.10753285139799118, + "learning_rate": 2.4290532544744005e-07, + "loss": 0.0019825618714094163, + "step": 17600, + "token_acc": 0.999291105458488 + }, + { + "epoch": 0.9323441625910802, + "grad_norm": 0.09900170564651489, + "learning_rate": 2.391631948047224e-07, + "loss": 0.0022107940167188644, + "step": 17610, + "token_acc": 0.9991235758106923 + }, + { + "epoch": 0.9328736027742666, + "grad_norm": 0.10063713043928146, + "learning_rate": 2.3544976458788838e-07, + "loss": 0.002167546935379505, + "step": 17620, + "token_acc": 0.9990409153837172 + }, + { + "epoch": 0.9334030429574529, + "grad_norm": 0.10098527371883392, + "learning_rate": 2.3176504571578163e-07, + "loss": 0.0017058951780200005, + "step": 17630, + "token_acc": 0.9994158098544532 + }, + { + "epoch": 0.9339324831406391, + "grad_norm": 0.07803171128034592, + "learning_rate": 2.2810904902281882e-07, + "loss": 0.0021216347813606262, + "step": 17640, + "token_acc": 0.9992075144315793 + }, + { + "epoch": 0.9344619233238255, + "grad_norm": 0.08300700783729553, + "learning_rate": 2.2448178525896648e-07, + "loss": 0.0017645839601755143, + "step": 17650, + "token_acc": 0.9993740663155875 + }, + { + "epoch": 0.9349913635070117, + "grad_norm": 0.08854418247938156, + "learning_rate": 2.2088326508970525e-07, + "loss": 0.0019573625177145004, + "step": 17660, + "token_acc": 0.9992576961892291 + }, + { + "epoch": 0.9355208036901981, + "grad_norm": 0.06657784432172775, + "learning_rate": 2.173134990960002e-07, + "loss": 0.0019745051860809324, + "step": 17670, + "token_acc": 0.9993076063199693 + }, + { + "epoch": 0.9360502438733844, + "grad_norm": 0.09055756777524948, + "learning_rate": 2.1377249777426611e-07, + "loss": 0.0022365756332874296, + "step": 17680, + "token_acc": 0.9991403772325155 + }, + { + "epoch": 0.9365796840565707, + "grad_norm": 0.10941233485937119, + "learning_rate": 2.1026027153634e-07, + "loss": 0.002171015366911888, + "step": 17690, + "token_acc": 0.9991902901599385 + }, + { + "epoch": 0.937109124239757, + "grad_norm": 0.12454210221767426, + "learning_rate": 2.0677683070945197e-07, + "loss": 0.0020880803465843202, + "step": 17700, + "token_acc": 0.9991905334134454 + }, + { + "epoch": 0.9376385644229432, + "grad_norm": 0.10969654470682144, + "learning_rate": 2.0332218553618888e-07, + "loss": 0.0019478930160403251, + "step": 17710, + "token_acc": 0.999232425036293 + }, + { + "epoch": 0.9381680046061296, + "grad_norm": 0.09396298229694366, + "learning_rate": 1.998963461744696e-07, + "loss": 0.001985710486769676, + "step": 17720, + "token_acc": 0.9992491678276757 + }, + { + "epoch": 0.9386974447893159, + "grad_norm": 0.07106446474790573, + "learning_rate": 1.9649932269751537e-07, + "loss": 0.001944005861878395, + "step": 17730, + "token_acc": 0.9992735046303641 + }, + { + "epoch": 0.9392268849725022, + "grad_norm": 0.06245732679963112, + "learning_rate": 1.9313112509381394e-07, + "loss": 0.0017808031290769577, + "step": 17740, + "token_acc": 0.9992742379498474 + }, + { + "epoch": 0.9397563251556885, + "grad_norm": 0.07781852781772614, + "learning_rate": 1.897917632670976e-07, + "loss": 0.0023833632469177244, + "step": 17750, + "token_acc": 0.9991074258829811 + }, + { + "epoch": 0.9402857653388748, + "grad_norm": 0.10063618421554565, + "learning_rate": 1.8648124703630866e-07, + "loss": 0.0017320863902568818, + "step": 17760, + "token_acc": 0.9993828241632681 + }, + { + "epoch": 0.9408152055220611, + "grad_norm": 0.11506026983261108, + "learning_rate": 1.831995861355762e-07, + "loss": 0.0019873294979333876, + "step": 17770, + "token_acc": 0.9992237126568225 + }, + { + "epoch": 0.9413446457052475, + "grad_norm": 0.07472510635852814, + "learning_rate": 1.7994679021418048e-07, + "loss": 0.0018821466714143754, + "step": 17780, + "token_acc": 0.9992156797303273 + }, + { + "epoch": 0.9418740858884337, + "grad_norm": 0.07418346405029297, + "learning_rate": 1.7672286883653078e-07, + "loss": 0.002004493400454521, + "step": 17790, + "token_acc": 0.9991911071824678 + }, + { + "epoch": 0.94240352607162, + "grad_norm": 0.0904916301369667, + "learning_rate": 1.7352783148213315e-07, + "loss": 0.0018276162445545197, + "step": 17800, + "token_acc": 0.9994244388278571 + }, + { + "epoch": 0.9429329662548063, + "grad_norm": 0.10013597458600998, + "learning_rate": 1.7036168754556492e-07, + "loss": 0.002083142288029194, + "step": 17810, + "token_acc": 0.9992325851038095 + }, + { + "epoch": 0.9434624064379926, + "grad_norm": 0.0882093608379364, + "learning_rate": 1.6722444633644585e-07, + "loss": 0.0017276903614401816, + "step": 17820, + "token_acc": 0.9993076409743077 + }, + { + "epoch": 0.943991846621179, + "grad_norm": 0.12518928945064545, + "learning_rate": 1.6411611707941144e-07, + "loss": 0.002119935303926468, + "step": 17830, + "token_acc": 0.999282352548483 + }, + { + "epoch": 0.9445212868043652, + "grad_norm": 0.074955053627491, + "learning_rate": 1.6103670891408406e-07, + "loss": 0.0018457887694239617, + "step": 17840, + "token_acc": 0.9993075543319568 + }, + { + "epoch": 0.9450507269875515, + "grad_norm": 0.11051096767187119, + "learning_rate": 1.579862308950486e-07, + "loss": 0.0020054973661899566, + "step": 17850, + "token_acc": 0.9992158040511229 + }, + { + "epoch": 0.9455801671707379, + "grad_norm": 0.09825264662504196, + "learning_rate": 1.5496469199182685e-07, + "loss": 0.002380309998989105, + "step": 17860, + "token_acc": 0.9991495038772618 + }, + { + "epoch": 0.9461096073539241, + "grad_norm": 0.09114942699670792, + "learning_rate": 1.5197210108884531e-07, + "loss": 0.001965579017996788, + "step": 17870, + "token_acc": 0.999232373800584 + }, + { + "epoch": 0.9466390475371105, + "grad_norm": 0.07192118465900421, + "learning_rate": 1.4900846698541526e-07, + "loss": 0.001948327384889126, + "step": 17880, + "token_acc": 0.9991325671009391 + }, + { + "epoch": 0.9471684877202967, + "grad_norm": 0.08129381388425827, + "learning_rate": 1.4607379839570168e-07, + "loss": 0.001799514889717102, + "step": 17890, + "token_acc": 0.9992744376892091 + }, + { + "epoch": 0.9476979279034831, + "grad_norm": 0.09351889789104462, + "learning_rate": 1.4316810394870428e-07, + "loss": 0.001853940635919571, + "step": 17900, + "token_acc": 0.9993658581357898 + }, + { + "epoch": 0.9482273680866694, + "grad_norm": 0.09372225403785706, + "learning_rate": 1.402913921882254e-07, + "loss": 0.0021557532250881194, + "step": 17910, + "token_acc": 0.9991658742472015 + }, + { + "epoch": 0.9487568082698556, + "grad_norm": 0.0883384644985199, + "learning_rate": 1.3744367157284888e-07, + "loss": 0.0018237311393022536, + "step": 17920, + "token_acc": 0.9992743227011878 + }, + { + "epoch": 0.949286248453042, + "grad_norm": 0.08764438331127167, + "learning_rate": 1.346249504759134e-07, + "loss": 0.0021309252828359605, + "step": 17930, + "token_acc": 0.9991575400168492 + }, + { + "epoch": 0.9498156886362282, + "grad_norm": 0.10959600657224655, + "learning_rate": 1.3183523718548918e-07, + "loss": 0.002059393376111984, + "step": 17940, + "token_acc": 0.9992243471588587 + }, + { + "epoch": 0.9503451288194146, + "grad_norm": 0.1173255667090416, + "learning_rate": 1.290745399043525e-07, + "loss": 0.0018445458263158798, + "step": 17950, + "token_acc": 0.9992906025705225 + }, + { + "epoch": 0.9508745690026009, + "grad_norm": 0.08341265469789505, + "learning_rate": 1.2634286674995998e-07, + "loss": 0.0020848046988248825, + "step": 17960, + "token_acc": 0.9992325658992326 + }, + { + "epoch": 0.9514040091857872, + "grad_norm": 0.08953704684972763, + "learning_rate": 1.2364022575443002e-07, + "loss": 0.0020843427628278732, + "step": 17970, + "token_acc": 0.9992066739596991 + }, + { + "epoch": 0.9519334493689735, + "grad_norm": 0.06755571067333221, + "learning_rate": 1.2096662486451471e-07, + "loss": 0.0018383169546723366, + "step": 17980, + "token_acc": 0.9992571881650878 + }, + { + "epoch": 0.9524628895521597, + "grad_norm": 0.06749697029590607, + "learning_rate": 1.1832207194157674e-07, + "loss": 0.0017815949395298958, + "step": 17990, + "token_acc": 0.9992736926467641 + }, + { + "epoch": 0.9529923297353461, + "grad_norm": 0.10420604050159454, + "learning_rate": 1.157065747615671e-07, + "loss": 0.0016976058483123778, + "step": 18000, + "token_acc": 0.9993825254708243 + }, + { + "epoch": 0.9529923297353461, + "eval_loss": 0.003807932138442993, + "eval_runtime": 712.5255, + "eval_samples_per_second": 16.811, + "eval_steps_per_second": 1.051, + "eval_token_acc": 0.9987263259856257, + "step": 18000 + }, + { + "epoch": 0.9535217699185324, + "grad_norm": 0.09544046223163605, + "learning_rate": 1.1312014101500513e-07, + "loss": 0.002042551152408123, + "step": 18010, + "token_acc": 0.9992659387225665 + }, + { + "epoch": 0.9540512101017187, + "grad_norm": 0.06917744874954224, + "learning_rate": 1.105627783069485e-07, + "loss": 0.0016617190092802047, + "step": 18020, + "token_acc": 0.9993914080150731 + }, + { + "epoch": 0.954580650284905, + "grad_norm": 0.10581348836421967, + "learning_rate": 1.0803449415697886e-07, + "loss": 0.0020620428025722505, + "step": 18030, + "token_acc": 0.9992322456813819 + }, + { + "epoch": 0.9551100904680913, + "grad_norm": 0.06485496461391449, + "learning_rate": 1.0553529599917622e-07, + "loss": 0.0019014880061149597, + "step": 18040, + "token_acc": 0.9993239803370083 + }, + { + "epoch": 0.9556395306512776, + "grad_norm": 0.08478603512048721, + "learning_rate": 1.0306519118209457e-07, + "loss": 0.001971239224076271, + "step": 18050, + "token_acc": 0.9991574908241575 + }, + { + "epoch": 0.956168970834464, + "grad_norm": 0.08804220706224442, + "learning_rate": 1.006241869687452e-07, + "loss": 0.0019677083939313887, + "step": 18060, + "token_acc": 0.9992070315434505 + }, + { + "epoch": 0.9566984110176502, + "grad_norm": 0.0851602852344513, + "learning_rate": 9.821229053657233e-08, + "loss": 0.0019555501639842987, + "step": 18070, + "token_acc": 0.9991991858389363 + }, + { + "epoch": 0.9572278512008365, + "grad_norm": 0.10766731947660446, + "learning_rate": 9.582950897743082e-08, + "loss": 0.0020186502486467363, + "step": 18080, + "token_acc": 0.9992080231423974 + }, + { + "epoch": 0.9577572913840228, + "grad_norm": 0.07938266545534134, + "learning_rate": 9.34758492975707e-08, + "loss": 0.0019124969840049744, + "step": 18090, + "token_acc": 0.999282430391576 + }, + { + "epoch": 0.9582867315672091, + "grad_norm": 0.09312859177589417, + "learning_rate": 9.11513184176116e-08, + "loss": 0.0019601305946707727, + "step": 18100, + "token_acc": 0.9992910345057676 + }, + { + "epoch": 0.9588161717503955, + "grad_norm": 0.0860222727060318, + "learning_rate": 8.885592317252167e-08, + "loss": 0.0019595615565776827, + "step": 18110, + "token_acc": 0.9992743711216387 + }, + { + "epoch": 0.9593456119335817, + "grad_norm": 0.08982247859239578, + "learning_rate": 8.658967031160315e-08, + "loss": 0.002068643644452095, + "step": 18120, + "token_acc": 0.9992240106135323 + }, + { + "epoch": 0.959875052116768, + "grad_norm": 0.07292438298463821, + "learning_rate": 8.435256649846568e-08, + "loss": 0.001907522976398468, + "step": 18130, + "token_acc": 0.9992575351836558 + }, + { + "epoch": 0.9604044922999543, + "grad_norm": 0.09148246794939041, + "learning_rate": 8.214461831101417e-08, + "loss": 0.0017873343080282212, + "step": 18140, + "token_acc": 0.9993408702181803 + }, + { + "epoch": 0.9609339324831406, + "grad_norm": 0.1048528328537941, + "learning_rate": 7.996583224141985e-08, + "loss": 0.0019355010241270064, + "step": 18150, + "token_acc": 0.9993075081138356 + }, + { + "epoch": 0.961463372666327, + "grad_norm": 0.08234653621912003, + "learning_rate": 7.781621469611144e-08, + "loss": 0.001990067586302757, + "step": 18160, + "token_acc": 0.9992487980769231 + }, + { + "epoch": 0.9619928128495132, + "grad_norm": 0.0826045498251915, + "learning_rate": 7.569577199574963e-08, + "loss": 0.001684541255235672, + "step": 18170, + "token_acc": 0.9993826800030032 + }, + { + "epoch": 0.9625222530326996, + "grad_norm": 0.09972482919692993, + "learning_rate": 7.360451037521033e-08, + "loss": 0.001763300970196724, + "step": 18180, + "token_acc": 0.9993073058369916 + }, + { + "epoch": 0.9630516932158859, + "grad_norm": 0.07779534161090851, + "learning_rate": 7.154243598356703e-08, + "loss": 0.0021690007299184797, + "step": 18190, + "token_acc": 0.9991914038729254 + }, + { + "epoch": 0.9635811333990721, + "grad_norm": 0.07482873648405075, + "learning_rate": 6.950955488407185e-08, + "loss": 0.0019812822341918946, + "step": 18200, + "token_acc": 0.9992656324323422 + }, + { + "epoch": 0.9641105735822585, + "grad_norm": 0.08722586184740067, + "learning_rate": 6.750587305413669e-08, + "loss": 0.0021751241758465765, + "step": 18210, + "token_acc": 0.9992492617740778 + }, + { + "epoch": 0.9646400137654447, + "grad_norm": 0.08595967292785645, + "learning_rate": 6.553139638531881e-08, + "loss": 0.0017299339175224303, + "step": 18220, + "token_acc": 0.9993155715811965 + }, + { + "epoch": 0.9651694539486311, + "grad_norm": 0.07583785802125931, + "learning_rate": 6.358613068330077e-08, + "loss": 0.0022641865536570547, + "step": 18230, + "token_acc": 0.9991909182660628 + }, + { + "epoch": 0.9656988941318174, + "grad_norm": 0.1003444567322731, + "learning_rate": 6.167008166787502e-08, + "loss": 0.002081393450498581, + "step": 18240, + "token_acc": 0.9992489798643157 + }, + { + "epoch": 0.9662283343150037, + "grad_norm": 0.08482867479324341, + "learning_rate": 5.978325497292492e-08, + "loss": 0.0018806051462888717, + "step": 18250, + "token_acc": 0.9992740138688385 + }, + { + "epoch": 0.96675777449819, + "grad_norm": 0.07545541971921921, + "learning_rate": 5.792565614641033e-08, + "loss": 0.002113974653184414, + "step": 18260, + "token_acc": 0.9992317584381315 + }, + { + "epoch": 0.9672872146813762, + "grad_norm": 0.1279316246509552, + "learning_rate": 5.6097290650349855e-08, + "loss": 0.001985575444996357, + "step": 18270, + "token_acc": 0.9992661835708508 + }, + { + "epoch": 0.9678166548645626, + "grad_norm": 0.0832834541797638, + "learning_rate": 5.429816386080755e-08, + "loss": 0.0016820928081870078, + "step": 18280, + "token_acc": 0.9994241121729333 + }, + { + "epoch": 0.9683460950477489, + "grad_norm": 0.11210549622774124, + "learning_rate": 5.2528281067871776e-08, + "loss": 0.0021329956129193307, + "step": 18290, + "token_acc": 0.9992157975089057 + }, + { + "epoch": 0.9688755352309352, + "grad_norm": 0.06730542331933975, + "learning_rate": 5.078764747564524e-08, + "loss": 0.0019266251474618912, + "step": 18300, + "token_acc": 0.999274225840897 + }, + { + "epoch": 0.9694049754141215, + "grad_norm": 0.09534727036952972, + "learning_rate": 4.907626820222611e-08, + "loss": 0.0020238738507032394, + "step": 18310, + "token_acc": 0.9992328730571677 + }, + { + "epoch": 0.9699344155973078, + "grad_norm": 0.09610612690448761, + "learning_rate": 4.739414827969579e-08, + "loss": 0.002177560888230801, + "step": 18320, + "token_acc": 0.9991653869266208 + }, + { + "epoch": 0.9704638557804941, + "grad_norm": 0.08023147284984589, + "learning_rate": 4.57412926541001e-08, + "loss": 0.0018499422818422317, + "step": 18330, + "token_acc": 0.9992660978925334 + }, + { + "epoch": 0.9709932959636804, + "grad_norm": 0.08633679151535034, + "learning_rate": 4.4117706185438095e-08, + "loss": 0.002042420022189617, + "step": 18340, + "token_acc": 0.9992738381410257 + }, + { + "epoch": 0.9715227361468667, + "grad_norm": 0.09126947820186615, + "learning_rate": 4.252339364764879e-08, + "loss": 0.0019891276955604554, + "step": 18350, + "token_acc": 0.9993244426651988 + }, + { + "epoch": 0.972052176330053, + "grad_norm": 0.0719359889626503, + "learning_rate": 4.09583597285923e-08, + "loss": 0.002046186476945877, + "step": 18360, + "token_acc": 0.9992324762651628 + }, + { + "epoch": 0.9725816165132393, + "grad_norm": 0.09622064977884293, + "learning_rate": 3.9422609030039805e-08, + "loss": 0.0019241759553551673, + "step": 18370, + "token_acc": 0.9992650743277101 + }, + { + "epoch": 0.9731110566964256, + "grad_norm": 0.09742356836795807, + "learning_rate": 3.791614606766025e-08, + "loss": 0.0020631000399589537, + "step": 18380, + "token_acc": 0.9992410595142781 + }, + { + "epoch": 0.973640496879612, + "grad_norm": 0.048557527363300323, + "learning_rate": 3.6438975271003706e-08, + "loss": 0.0018245715647935868, + "step": 18390, + "token_acc": 0.99933253237218 + }, + { + "epoch": 0.9741699370627982, + "grad_norm": 0.08639136701822281, + "learning_rate": 3.4991100983493566e-08, + "loss": 0.002021871507167816, + "step": 18400, + "token_acc": 0.9992743650694358 + }, + { + "epoch": 0.9746993772459845, + "grad_norm": 0.09376675635576248, + "learning_rate": 3.35725274624088e-08, + "loss": 0.0019344331696629525, + "step": 18410, + "token_acc": 0.9992821848290598 + }, + { + "epoch": 0.9752288174291708, + "grad_norm": 0.07244961708784103, + "learning_rate": 3.218325887887286e-08, + "loss": 0.0019204583019018172, + "step": 18420, + "token_acc": 0.9992740926157697 + }, + { + "epoch": 0.9757582576123571, + "grad_norm": 0.1124308854341507, + "learning_rate": 3.082329931784256e-08, + "loss": 0.002000363916158676, + "step": 18430, + "token_acc": 0.999266256993488 + }, + { + "epoch": 0.9762876977955435, + "grad_norm": 0.0706089660525322, + "learning_rate": 2.9492652778098097e-08, + "loss": 0.002120649814605713, + "step": 18440, + "token_acc": 0.9991569985810868 + }, + { + "epoch": 0.9768171379787297, + "grad_norm": 0.10726562887430191, + "learning_rate": 2.8191323172224173e-08, + "loss": 0.0019288035109639168, + "step": 18450, + "token_acc": 0.9992741531787085 + }, + { + "epoch": 0.9773465781619161, + "grad_norm": 0.11050349473953247, + "learning_rate": 2.691931432660555e-08, + "loss": 0.0017303816974163055, + "step": 18460, + "token_acc": 0.9993405345843699 + }, + { + "epoch": 0.9778760183451023, + "grad_norm": 0.08911315351724625, + "learning_rate": 2.5676629981413738e-08, + "loss": 0.0020194873213768005, + "step": 18470, + "token_acc": 0.9992323866101526 + }, + { + "epoch": 0.9784054585282886, + "grad_norm": 0.09087201952934265, + "learning_rate": 2.4463273790594765e-08, + "loss": 0.002222846634685993, + "step": 18480, + "token_acc": 0.9992905848084996 + }, + { + "epoch": 0.978934898711475, + "grad_norm": 0.09747378528118134, + "learning_rate": 2.32792493218581e-08, + "loss": 0.002178752049803734, + "step": 18490, + "token_acc": 0.99914038440674 + }, + { + "epoch": 0.9794643388946612, + "grad_norm": 0.06852329522371292, + "learning_rate": 2.2124560056668853e-08, + "loss": 0.0020709581673145293, + "step": 18500, + "token_acc": 0.9992322841216329 + }, + { + "epoch": 0.9799937790778476, + "grad_norm": 0.07985907047986984, + "learning_rate": 2.0999209390235585e-08, + "loss": 0.0021202221512794496, + "step": 18510, + "token_acc": 0.9992320085480788 + }, + { + "epoch": 0.9805232192610339, + "grad_norm": 0.05642770975828171, + "learning_rate": 1.9903200631499197e-08, + "loss": 0.001705373078584671, + "step": 18520, + "token_acc": 0.9994327045809105 + }, + { + "epoch": 0.9810526594442202, + "grad_norm": 0.08196930587291718, + "learning_rate": 1.8836537003126266e-08, + "loss": 0.0019800242036581038, + "step": 18530, + "token_acc": 0.9992831003409441 + }, + { + "epoch": 0.9815820996274065, + "grad_norm": 0.08949355781078339, + "learning_rate": 1.7799221641497945e-08, + "loss": 0.002028035745024681, + "step": 18540, + "token_acc": 0.9992740017524095 + }, + { + "epoch": 0.9821115398105927, + "grad_norm": 0.11900454759597778, + "learning_rate": 1.6791257596699974e-08, + "loss": 0.002021810971200466, + "step": 18550, + "token_acc": 0.9992579128172631 + }, + { + "epoch": 0.9826409799937791, + "grad_norm": 0.10891634225845337, + "learning_rate": 1.5812647832514906e-08, + "loss": 0.0023721672594547273, + "step": 18560, + "token_acc": 0.9990407953891451 + }, + { + "epoch": 0.9831704201769654, + "grad_norm": 0.14587770402431488, + "learning_rate": 1.4863395226414334e-08, + "loss": 0.0021209102123975754, + "step": 18570, + "token_acc": 0.9991489432712285 + }, + { + "epoch": 0.9836998603601517, + "grad_norm": 0.0913902148604393, + "learning_rate": 1.3943502569545575e-08, + "loss": 0.0017486879602074622, + "step": 18580, + "token_acc": 0.9993578035212384 + }, + { + "epoch": 0.984229300543338, + "grad_norm": 0.09430061280727386, + "learning_rate": 1.3052972566729438e-08, + "loss": 0.0021475233137607575, + "step": 18590, + "token_acc": 0.9992070977272917 + }, + { + "epoch": 0.9847587407265243, + "grad_norm": 0.11087960004806519, + "learning_rate": 1.2191807836450242e-08, + "loss": 0.002006252482533455, + "step": 18600, + "token_acc": 0.9992071903661112 + }, + { + "epoch": 0.9852881809097106, + "grad_norm": 0.11041762679815292, + "learning_rate": 1.1360010910846931e-08, + "loss": 0.0019351227208971977, + "step": 18610, + "token_acc": 0.999265877484963 + }, + { + "epoch": 0.985817621092897, + "grad_norm": 0.06417790800333023, + "learning_rate": 1.0557584235705298e-08, + "loss": 0.001991066709160805, + "step": 18620, + "token_acc": 0.9992409518963691 + }, + { + "epoch": 0.9863470612760832, + "grad_norm": 0.0954044982790947, + "learning_rate": 9.784530170453554e-09, + "loss": 0.0020793016999959947, + "step": 18630, + "token_acc": 0.9991570970757111 + }, + { + "epoch": 0.9868765014592695, + "grad_norm": 0.12416848540306091, + "learning_rate": 9.040850988153437e-09, + "loss": 0.001730544865131378, + "step": 18640, + "token_acc": 0.9993578410114421 + }, + { + "epoch": 0.9874059416424558, + "grad_norm": 0.10388429462909698, + "learning_rate": 8.326548875493556e-09, + "loss": 0.002037663944065571, + "step": 18650, + "token_acc": 0.9992908688941726 + }, + { + "epoch": 0.9879353818256421, + "grad_norm": 0.10862120240926743, + "learning_rate": 7.641625932781616e-09, + "loss": 0.0016551494598388671, + "step": 18660, + "token_acc": 0.9994075038387076 + }, + { + "epoch": 0.9884648220088285, + "grad_norm": 0.08534073084592819, + "learning_rate": 6.986084173942198e-09, + "loss": 0.0018829140812158585, + "step": 18670, + "token_acc": 0.9992745163442295 + }, + { + "epoch": 0.9889942621920147, + "grad_norm": 0.07599177956581116, + "learning_rate": 6.359925526507882e-09, + "loss": 0.0020452138036489488, + "step": 18680, + "token_acc": 0.9992496686036333 + }, + { + "epoch": 0.989523702375201, + "grad_norm": 0.09368439018726349, + "learning_rate": 5.7631518316136896e-09, + "loss": 0.001770384982228279, + "step": 18690, + "token_acc": 0.999249349436178 + }, + { + "epoch": 0.9900531425583873, + "grad_norm": 0.09994275867938995, + "learning_rate": 5.1957648439915355e-09, + "loss": 0.0018109044060111046, + "step": 18700, + "token_acc": 0.9993075889914992 + }, + { + "epoch": 0.9905825827415736, + "grad_norm": 0.07695628702640533, + "learning_rate": 4.657766231968009e-09, + "loss": 0.001849803328514099, + "step": 18710, + "token_acc": 0.9993824687896389 + }, + { + "epoch": 0.99111202292476, + "grad_norm": 0.09553005546331406, + "learning_rate": 4.149157577453267e-09, + "loss": 0.0018735792487859725, + "step": 18720, + "token_acc": 0.9993071901971586 + }, + { + "epoch": 0.9916414631079462, + "grad_norm": 0.074878990650177, + "learning_rate": 3.6699403759443697e-09, + "loss": 0.0017944257706403732, + "step": 18730, + "token_acc": 0.999291211860944 + }, + { + "epoch": 0.9921709032911326, + "grad_norm": 0.11529374122619629, + "learning_rate": 3.220116036514176e-09, + "loss": 0.002326592244207859, + "step": 18740, + "token_acc": 0.9991741121705834 + }, + { + "epoch": 0.9927003434743188, + "grad_norm": 0.08639650791883469, + "learning_rate": 2.7996858818113425e-09, + "loss": 0.001996733248233795, + "step": 18750, + "token_acc": 0.9993074041005015 + }, + { + "epoch": 0.9932297836575051, + "grad_norm": 0.08170365542173386, + "learning_rate": 2.408651148052554e-09, + "loss": 0.0015399491414427758, + "step": 18760, + "token_acc": 0.9994077855349531 + }, + { + "epoch": 0.9937592238406915, + "grad_norm": 0.08576114475727081, + "learning_rate": 2.0470129850247434e-09, + "loss": 0.0020539745688438415, + "step": 18770, + "token_acc": 0.9992573245324899 + }, + { + "epoch": 0.9942886640238777, + "grad_norm": 0.09750602394342422, + "learning_rate": 1.7147724560739875e-09, + "loss": 0.0020490594208240507, + "step": 18780, + "token_acc": 0.9991909655034363 + }, + { + "epoch": 0.9948181042070641, + "grad_norm": 0.11412956565618515, + "learning_rate": 1.4119305381110616e-09, + "loss": 0.002065237984061241, + "step": 18790, + "token_acc": 0.9991660620616614 + }, + { + "epoch": 0.9953475443902503, + "grad_norm": 0.13405659794807434, + "learning_rate": 1.1384881216014444e-09, + "loss": 0.0024325214326381683, + "step": 18800, + "token_acc": 0.999149624417451 + }, + { + "epoch": 0.9958769845734367, + "grad_norm": 0.08538686484098434, + "learning_rate": 8.944460105642094e-10, + "loss": 0.0020084988325834276, + "step": 18810, + "token_acc": 0.9992492805605372 + }, + { + "epoch": 0.996406424756623, + "grad_norm": 0.07645708322525024, + "learning_rate": 6.798049225731351e-10, + "loss": 0.002058519423007965, + "step": 18820, + "token_acc": 0.999274752207004 + }, + { + "epoch": 0.9969358649398092, + "grad_norm": 0.08269505202770233, + "learning_rate": 4.945654887522633e-10, + "loss": 0.002136114425957203, + "step": 18830, + "token_acc": 0.999215502996111 + }, + { + "epoch": 0.9974653051229956, + "grad_norm": 0.08643733710050583, + "learning_rate": 3.387282537714587e-10, + "loss": 0.0020525578409433367, + "step": 18840, + "token_acc": 0.9992071374322937 + }, + { + "epoch": 0.9979947453061819, + "grad_norm": 0.08035251498222351, + "learning_rate": 2.1229367585084982e-10, + "loss": 0.0020373277366161346, + "step": 18850, + "token_acc": 0.9992071837497705 + }, + { + "epoch": 0.9985241854893682, + "grad_norm": 0.1004517674446106, + "learning_rate": 1.1526212675305737e-10, + "loss": 0.002096066251397133, + "step": 18860, + "token_acc": 0.9992569400960134 + }, + { + "epoch": 0.9990536256725545, + "grad_norm": 0.06819505989551544, + "learning_rate": 4.763389178652489e-11, + "loss": 0.0021103113889694215, + "step": 18870, + "token_acc": 0.9991994529595224 + }, + { + "epoch": 0.9995830658557407, + "grad_norm": 0.0742933377623558, + "learning_rate": 9.409169804408535e-12, + "loss": 0.0018262401223182677, + "step": 18880, + "token_acc": 0.9993242059419818 + }, + { + "epoch": 1.0, + "eval_loss": 0.0037917355075478554, + "eval_runtime": 710.867, + "eval_samples_per_second": 16.85, + "eval_steps_per_second": 1.054, + "eval_token_acc": 0.9987361372000221, + "step": 18888 + } + ], + "logging_steps": 10, + "max_steps": 18888, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 3000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 3.3644710538813165e+20, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}