diff --git "a/multi_student_5/checkpoint-10800/trainer_state.json" "b/multi_student_5/checkpoint-10800/trainer_state.json" new file mode 100644--- /dev/null +++ "b/multi_student_5/checkpoint-10800/trainer_state.json" @@ -0,0 +1,7594 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 6.070826306913997, + "eval_steps": 500, + "global_step": 10800, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.005621135469364812, + "grad_norm": 19.072066040018505, + "learning_rate": 7.223113964686999e-07, + "loss": 8.2549, + "step": 10 + }, + { + "epoch": 0.011242270938729624, + "grad_norm": 19.552394701284502, + "learning_rate": 1.5248796147672553e-06, + "loss": 8.1791, + "step": 20 + }, + { + "epoch": 0.016863406408094434, + "grad_norm": 9.93392906727828, + "learning_rate": 2.3274478330658108e-06, + "loss": 7.6759, + "step": 30 + }, + { + "epoch": 0.022484541877459248, + "grad_norm": 7.426576059473498, + "learning_rate": 3.1300160513643665e-06, + "loss": 6.913, + "step": 40 + }, + { + "epoch": 0.028105677346824058, + "grad_norm": 3.5646171014288366, + "learning_rate": 3.932584269662922e-06, + "loss": 6.3931, + "step": 50 + }, + { + "epoch": 0.03372681281618887, + "grad_norm": 2.8779163477648666, + "learning_rate": 4.735152487961477e-06, + "loss": 6.0533, + "step": 60 + }, + { + "epoch": 0.03934794828555368, + "grad_norm": 3.3997232487978883, + "learning_rate": 5.537720706260032e-06, + "loss": 5.7246, + "step": 70 + }, + { + "epoch": 0.044969083754918496, + "grad_norm": 4.059106771548473, + "learning_rate": 6.340288924558588e-06, + "loss": 5.2465, + "step": 80 + }, + { + "epoch": 0.050590219224283306, + "grad_norm": 4.7307302237777895, + "learning_rate": 7.142857142857143e-06, + "loss": 4.6393, + "step": 90 + }, + { + "epoch": 0.056211354693648116, + "grad_norm": 5.242698651813807, + "learning_rate": 7.9454253611557e-06, + "loss": 4.0362, + "step": 100 + }, + { + "epoch": 0.061832490163012926, + "grad_norm": 4.723996169899083, + "learning_rate": 8.747993579454254e-06, + "loss": 3.549, + "step": 110 + }, + { + "epoch": 0.06745362563237774, + "grad_norm": 2.885143750378497, + "learning_rate": 9.550561797752809e-06, + "loss": 3.2171, + "step": 120 + }, + { + "epoch": 0.07307476110174255, + "grad_norm": 6.825878781249646, + "learning_rate": 1.0353130016051364e-05, + "loss": 3.027, + "step": 130 + }, + { + "epoch": 0.07869589657110736, + "grad_norm": 6.737230278162123, + "learning_rate": 1.115569823434992e-05, + "loss": 2.9055, + "step": 140 + }, + { + "epoch": 0.08431703204047218, + "grad_norm": 6.716587811023919, + "learning_rate": 1.1958266452648475e-05, + "loss": 2.8089, + "step": 150 + }, + { + "epoch": 0.08993816750983699, + "grad_norm": 7.201415197148535, + "learning_rate": 1.276083467094703e-05, + "loss": 2.7296, + "step": 160 + }, + { + "epoch": 0.0955593029792018, + "grad_norm": 7.301380585480233, + "learning_rate": 1.3563402889245585e-05, + "loss": 2.6609, + "step": 170 + }, + { + "epoch": 0.10118043844856661, + "grad_norm": 11.575326761105941, + "learning_rate": 1.4365971107544143e-05, + "loss": 2.5979, + "step": 180 + }, + { + "epoch": 0.10680157391793142, + "grad_norm": 8.619354697957048, + "learning_rate": 1.5168539325842698e-05, + "loss": 2.5341, + "step": 190 + }, + { + "epoch": 0.11242270938729623, + "grad_norm": 7.390257991673745, + "learning_rate": 1.5971107544141253e-05, + "loss": 2.4767, + "step": 200 + }, + { + "epoch": 0.11804384485666104, + "grad_norm": 13.39500588081804, + "learning_rate": 1.6773675762439807e-05, + "loss": 2.4379, + "step": 210 + }, + { + "epoch": 0.12366498032602585, + "grad_norm": 27.117334976761317, + "learning_rate": 1.7576243980738362e-05, + "loss": 2.3867, + "step": 220 + }, + { + "epoch": 0.12928611579539068, + "grad_norm": 11.695868441564258, + "learning_rate": 1.8378812199036917e-05, + "loss": 2.3207, + "step": 230 + }, + { + "epoch": 0.13490725126475547, + "grad_norm": 10.785553590242758, + "learning_rate": 1.9181380417335472e-05, + "loss": 2.2455, + "step": 240 + }, + { + "epoch": 0.1405283867341203, + "grad_norm": 25.384076838053424, + "learning_rate": 1.998394863563403e-05, + "loss": 2.1723, + "step": 250 + }, + { + "epoch": 0.1461495222034851, + "grad_norm": 16.42978149349179, + "learning_rate": 2.0786516853932585e-05, + "loss": 2.1322, + "step": 260 + }, + { + "epoch": 0.15177065767284992, + "grad_norm": 15.299053481717623, + "learning_rate": 2.158908507223114e-05, + "loss": 2.0356, + "step": 270 + }, + { + "epoch": 0.15739179314221471, + "grad_norm": 16.598966653865332, + "learning_rate": 2.2391653290529695e-05, + "loss": 1.9785, + "step": 280 + }, + { + "epoch": 0.16301292861157954, + "grad_norm": 8.000521357822016, + "learning_rate": 2.319422150882825e-05, + "loss": 1.8885, + "step": 290 + }, + { + "epoch": 0.16863406408094436, + "grad_norm": 12.337240959757104, + "learning_rate": 2.3996789727126808e-05, + "loss": 1.7929, + "step": 300 + }, + { + "epoch": 0.17425519955030916, + "grad_norm": 9.313580975453032, + "learning_rate": 2.4799357945425362e-05, + "loss": 1.7252, + "step": 310 + }, + { + "epoch": 0.17987633501967398, + "grad_norm": 9.713971076745432, + "learning_rate": 2.5601926163723917e-05, + "loss": 1.6366, + "step": 320 + }, + { + "epoch": 0.18549747048903878, + "grad_norm": 6.604247664050388, + "learning_rate": 2.6404494382022472e-05, + "loss": 1.5532, + "step": 330 + }, + { + "epoch": 0.1911186059584036, + "grad_norm": 8.06631111933664, + "learning_rate": 2.7207062600321027e-05, + "loss": 1.5014, + "step": 340 + }, + { + "epoch": 0.1967397414277684, + "grad_norm": 5.369237324186266, + "learning_rate": 2.8009630818619585e-05, + "loss": 1.4407, + "step": 350 + }, + { + "epoch": 0.20236087689713322, + "grad_norm": 6.309483928633083, + "learning_rate": 2.881219903691814e-05, + "loss": 1.398, + "step": 360 + }, + { + "epoch": 0.20798201236649802, + "grad_norm": 6.173807170921828, + "learning_rate": 2.9614767255216698e-05, + "loss": 1.3462, + "step": 370 + }, + { + "epoch": 0.21360314783586284, + "grad_norm": 7.913665828353373, + "learning_rate": 3.0417335473515253e-05, + "loss": 1.2942, + "step": 380 + }, + { + "epoch": 0.21922428330522767, + "grad_norm": 6.320887435288008, + "learning_rate": 3.1219903691813805e-05, + "loss": 1.2581, + "step": 390 + }, + { + "epoch": 0.22484541877459246, + "grad_norm": 4.001174978157139, + "learning_rate": 3.202247191011236e-05, + "loss": 1.2068, + "step": 400 + }, + { + "epoch": 0.2304665542439573, + "grad_norm": 5.833384870789841, + "learning_rate": 3.2825040128410914e-05, + "loss": 1.1774, + "step": 410 + }, + { + "epoch": 0.23608768971332209, + "grad_norm": 5.418298506902902, + "learning_rate": 3.362760834670947e-05, + "loss": 1.1285, + "step": 420 + }, + { + "epoch": 0.2417088251826869, + "grad_norm": 6.822807280672466, + "learning_rate": 3.4430176565008024e-05, + "loss": 1.0971, + "step": 430 + }, + { + "epoch": 0.2473299606520517, + "grad_norm": 5.1633853492959725, + "learning_rate": 3.523274478330658e-05, + "loss": 1.0631, + "step": 440 + }, + { + "epoch": 0.25295109612141653, + "grad_norm": 4.794453623656116, + "learning_rate": 3.6035313001605133e-05, + "loss": 1.0233, + "step": 450 + }, + { + "epoch": 0.25857223159078135, + "grad_norm": 6.493147309773226, + "learning_rate": 3.683788121990369e-05, + "loss": 1.0093, + "step": 460 + }, + { + "epoch": 0.2641933670601461, + "grad_norm": 7.38273852542595, + "learning_rate": 3.764044943820225e-05, + "loss": 0.9745, + "step": 470 + }, + { + "epoch": 0.26981450252951095, + "grad_norm": 5.953745454074908, + "learning_rate": 3.8443017656500805e-05, + "loss": 0.9609, + "step": 480 + }, + { + "epoch": 0.27543563799887577, + "grad_norm": 5.92674703602574, + "learning_rate": 3.924558587479936e-05, + "loss": 0.9295, + "step": 490 + }, + { + "epoch": 0.2810567734682406, + "grad_norm": 4.500282226524396, + "learning_rate": 4.0048154093097914e-05, + "loss": 0.8992, + "step": 500 + }, + { + "epoch": 0.2866779089376054, + "grad_norm": 4.320018263048497, + "learning_rate": 4.085072231139647e-05, + "loss": 0.8855, + "step": 510 + }, + { + "epoch": 0.2922990444069702, + "grad_norm": 6.008360095853525, + "learning_rate": 4.1653290529695024e-05, + "loss": 0.8589, + "step": 520 + }, + { + "epoch": 0.297920179876335, + "grad_norm": 5.252075871228091, + "learning_rate": 4.245585874799358e-05, + "loss": 0.8272, + "step": 530 + }, + { + "epoch": 0.30354131534569984, + "grad_norm": 6.427285353419691, + "learning_rate": 4.3258426966292134e-05, + "loss": 0.808, + "step": 540 + }, + { + "epoch": 0.30916245081506466, + "grad_norm": 5.854800753365741, + "learning_rate": 4.4060995184590695e-05, + "loss": 0.7901, + "step": 550 + }, + { + "epoch": 0.31478358628442943, + "grad_norm": 7.046341886366144, + "learning_rate": 4.486356340288925e-05, + "loss": 0.7701, + "step": 560 + }, + { + "epoch": 0.32040472175379425, + "grad_norm": 5.594440118602691, + "learning_rate": 4.5666131621187805e-05, + "loss": 0.7462, + "step": 570 + }, + { + "epoch": 0.3260258572231591, + "grad_norm": 5.653552049437021, + "learning_rate": 4.646869983948636e-05, + "loss": 0.7284, + "step": 580 + }, + { + "epoch": 0.3316469926925239, + "grad_norm": 5.196665356454571, + "learning_rate": 4.7271268057784915e-05, + "loss": 0.712, + "step": 590 + }, + { + "epoch": 0.3372681281618887, + "grad_norm": 6.1381022060931025, + "learning_rate": 4.807383627608347e-05, + "loss": 0.6868, + "step": 600 + }, + { + "epoch": 0.3428892636312535, + "grad_norm": 5.052546164435208, + "learning_rate": 4.8876404494382024e-05, + "loss": 0.6672, + "step": 610 + }, + { + "epoch": 0.3485103991006183, + "grad_norm": 4.916726449591455, + "learning_rate": 4.9678972712680586e-05, + "loss": 0.6501, + "step": 620 + }, + { + "epoch": 0.35413153456998314, + "grad_norm": 4.832632041097204, + "learning_rate": 5.0481540930979134e-05, + "loss": 0.6313, + "step": 630 + }, + { + "epoch": 0.35975267003934797, + "grad_norm": 4.751573877646501, + "learning_rate": 5.1284109149277696e-05, + "loss": 0.6165, + "step": 640 + }, + { + "epoch": 0.36537380550871273, + "grad_norm": 5.912237963361695, + "learning_rate": 5.2086677367576244e-05, + "loss": 0.5907, + "step": 650 + }, + { + "epoch": 0.37099494097807756, + "grad_norm": 4.086167788806448, + "learning_rate": 5.2889245585874805e-05, + "loss": 0.5768, + "step": 660 + }, + { + "epoch": 0.3766160764474424, + "grad_norm": 4.582982284092926, + "learning_rate": 5.369181380417335e-05, + "loss": 0.5579, + "step": 670 + }, + { + "epoch": 0.3822372119168072, + "grad_norm": 4.620402866105723, + "learning_rate": 5.4494382022471915e-05, + "loss": 0.5379, + "step": 680 + }, + { + "epoch": 0.38785834738617203, + "grad_norm": 3.8404903506330834, + "learning_rate": 5.5296950240770476e-05, + "loss": 0.5112, + "step": 690 + }, + { + "epoch": 0.3934794828555368, + "grad_norm": 3.2956728880451784, + "learning_rate": 5.6099518459069025e-05, + "loss": 0.502, + "step": 700 + }, + { + "epoch": 0.3991006183249016, + "grad_norm": 4.455851011958456, + "learning_rate": 5.6902086677367586e-05, + "loss": 0.4889, + "step": 710 + }, + { + "epoch": 0.40472175379426645, + "grad_norm": 3.3081515499225693, + "learning_rate": 5.7704654895666134e-05, + "loss": 0.4703, + "step": 720 + }, + { + "epoch": 0.41034288926363127, + "grad_norm": 3.5766535302434037, + "learning_rate": 5.8507223113964696e-05, + "loss": 0.4591, + "step": 730 + }, + { + "epoch": 0.41596402473299604, + "grad_norm": 3.1816986592066154, + "learning_rate": 5.9309791332263244e-05, + "loss": 0.437, + "step": 740 + }, + { + "epoch": 0.42158516020236086, + "grad_norm": 2.746088460027587, + "learning_rate": 6.0112359550561805e-05, + "loss": 0.429, + "step": 750 + }, + { + "epoch": 0.4272062956717257, + "grad_norm": 2.7638950111486777, + "learning_rate": 6.0914927768860353e-05, + "loss": 0.4056, + "step": 760 + }, + { + "epoch": 0.4328274311410905, + "grad_norm": 2.621320210943319, + "learning_rate": 6.171749598715892e-05, + "loss": 0.3951, + "step": 770 + }, + { + "epoch": 0.43844856661045534, + "grad_norm": 2.9039501807988635, + "learning_rate": 6.252006420545746e-05, + "loss": 0.3752, + "step": 780 + }, + { + "epoch": 0.4440697020798201, + "grad_norm": 2.672708782303941, + "learning_rate": 6.332263242375602e-05, + "loss": 0.3732, + "step": 790 + }, + { + "epoch": 0.44969083754918493, + "grad_norm": 2.534243622485374, + "learning_rate": 6.412520064205457e-05, + "loss": 0.3544, + "step": 800 + }, + { + "epoch": 0.45531197301854975, + "grad_norm": 2.2060854376474297, + "learning_rate": 6.492776886035313e-05, + "loss": 0.337, + "step": 810 + }, + { + "epoch": 0.4609331084879146, + "grad_norm": 2.8950080552663353, + "learning_rate": 6.573033707865169e-05, + "loss": 0.3278, + "step": 820 + }, + { + "epoch": 0.46655424395727935, + "grad_norm": 2.8077239251031183, + "learning_rate": 6.653290529695024e-05, + "loss": 0.3188, + "step": 830 + }, + { + "epoch": 0.47217537942664417, + "grad_norm": 2.4109713129957173, + "learning_rate": 6.73354735152488e-05, + "loss": 0.3023, + "step": 840 + }, + { + "epoch": 0.477796514896009, + "grad_norm": 2.4108771548493864, + "learning_rate": 6.813804173354735e-05, + "loss": 0.2907, + "step": 850 + }, + { + "epoch": 0.4834176503653738, + "grad_norm": 2.381688936764209, + "learning_rate": 6.894060995184591e-05, + "loss": 0.28, + "step": 860 + }, + { + "epoch": 0.48903878583473864, + "grad_norm": 2.226240645538486, + "learning_rate": 6.974317817014446e-05, + "loss": 0.265, + "step": 870 + }, + { + "epoch": 0.4946599213041034, + "grad_norm": 2.36477519460133, + "learning_rate": 7.054574638844302e-05, + "loss": 0.2563, + "step": 880 + }, + { + "epoch": 0.5002810567734682, + "grad_norm": 2.1877534753089325, + "learning_rate": 7.134831460674157e-05, + "loss": 0.2458, + "step": 890 + }, + { + "epoch": 0.5059021922428331, + "grad_norm": 1.9910110516875703, + "learning_rate": 7.215088282504013e-05, + "loss": 0.237, + "step": 900 + }, + { + "epoch": 0.5115233277121979, + "grad_norm": 2.6541637478760203, + "learning_rate": 7.295345104333868e-05, + "loss": 0.2274, + "step": 910 + }, + { + "epoch": 0.5171444631815627, + "grad_norm": 2.1481697620944273, + "learning_rate": 7.375601926163724e-05, + "loss": 0.2125, + "step": 920 + }, + { + "epoch": 0.5227655986509275, + "grad_norm": 2.2194085230343283, + "learning_rate": 7.455858747993579e-05, + "loss": 0.2086, + "step": 930 + }, + { + "epoch": 0.5283867341202922, + "grad_norm": 1.8687719324053211, + "learning_rate": 7.536115569823435e-05, + "loss": 0.2002, + "step": 940 + }, + { + "epoch": 0.5340078695896571, + "grad_norm": 1.809218029026303, + "learning_rate": 7.616372391653292e-05, + "loss": 0.19, + "step": 950 + }, + { + "epoch": 0.5396290050590219, + "grad_norm": 1.544075304452749, + "learning_rate": 7.696629213483147e-05, + "loss": 0.1824, + "step": 960 + }, + { + "epoch": 0.5452501405283867, + "grad_norm": 1.7029729768183128, + "learning_rate": 7.776886035313003e-05, + "loss": 0.1766, + "step": 970 + }, + { + "epoch": 0.5508712759977515, + "grad_norm": 1.9321640691371915, + "learning_rate": 7.857142857142858e-05, + "loss": 0.1674, + "step": 980 + }, + { + "epoch": 0.5564924114671164, + "grad_norm": 1.6567308102186273, + "learning_rate": 7.937399678972713e-05, + "loss": 0.1616, + "step": 990 + }, + { + "epoch": 0.5621135469364812, + "grad_norm": 1.75914588466869, + "learning_rate": 8.017656500802569e-05, + "loss": 0.1496, + "step": 1000 + }, + { + "epoch": 0.567734682405846, + "grad_norm": 1.6033430089086067, + "learning_rate": 8.097913322632424e-05, + "loss": 0.1411, + "step": 1010 + }, + { + "epoch": 0.5733558178752108, + "grad_norm": 1.5592637787510577, + "learning_rate": 8.17817014446228e-05, + "loss": 0.1349, + "step": 1020 + }, + { + "epoch": 0.5789769533445756, + "grad_norm": 1.490862273444445, + "learning_rate": 8.258426966292135e-05, + "loss": 0.1321, + "step": 1030 + }, + { + "epoch": 0.5845980888139404, + "grad_norm": 1.103403083662274, + "learning_rate": 8.338683788121991e-05, + "loss": 0.1237, + "step": 1040 + }, + { + "epoch": 0.5902192242833052, + "grad_norm": 1.3322879027309737, + "learning_rate": 8.418940609951846e-05, + "loss": 0.1175, + "step": 1050 + }, + { + "epoch": 0.59584035975267, + "grad_norm": 1.1302428773890956, + "learning_rate": 8.499197431781702e-05, + "loss": 0.1134, + "step": 1060 + }, + { + "epoch": 0.6014614952220348, + "grad_norm": 1.043524404705153, + "learning_rate": 8.579454253611557e-05, + "loss": 0.1079, + "step": 1070 + }, + { + "epoch": 0.6070826306913997, + "grad_norm": 1.0807715777082445, + "learning_rate": 8.659711075441413e-05, + "loss": 0.1027, + "step": 1080 + }, + { + "epoch": 0.6127037661607645, + "grad_norm": 1.0613560487233085, + "learning_rate": 8.739967897271268e-05, + "loss": 0.1014, + "step": 1090 + }, + { + "epoch": 0.6183249016301293, + "grad_norm": 1.7970050911102724, + "learning_rate": 8.820224719101124e-05, + "loss": 0.0956, + "step": 1100 + }, + { + "epoch": 0.6239460370994941, + "grad_norm": 1.2518666492303272, + "learning_rate": 8.900481540930979e-05, + "loss": 0.0934, + "step": 1110 + }, + { + "epoch": 0.6295671725688589, + "grad_norm": 1.1953063290224126, + "learning_rate": 8.980738362760836e-05, + "loss": 0.1004, + "step": 1120 + }, + { + "epoch": 0.6351883080382237, + "grad_norm": 0.9616154833979669, + "learning_rate": 9.06099518459069e-05, + "loss": 0.0843, + "step": 1130 + }, + { + "epoch": 0.6408094435075885, + "grad_norm": 0.8282021805405166, + "learning_rate": 9.141252006420547e-05, + "loss": 0.0788, + "step": 1140 + }, + { + "epoch": 0.6464305789769533, + "grad_norm": 0.7645874993850531, + "learning_rate": 9.221508828250401e-05, + "loss": 0.0741, + "step": 1150 + }, + { + "epoch": 0.6520517144463182, + "grad_norm": 1.0333771317137108, + "learning_rate": 9.301765650080258e-05, + "loss": 0.0708, + "step": 1160 + }, + { + "epoch": 0.657672849915683, + "grad_norm": 0.6189773458607581, + "learning_rate": 9.382022471910112e-05, + "loss": 0.0722, + "step": 1170 + }, + { + "epoch": 0.6632939853850478, + "grad_norm": 0.5553259399184006, + "learning_rate": 9.462279293739969e-05, + "loss": 0.0654, + "step": 1180 + }, + { + "epoch": 0.6689151208544126, + "grad_norm": 0.6418699826913336, + "learning_rate": 9.542536115569823e-05, + "loss": 0.0625, + "step": 1190 + }, + { + "epoch": 0.6745362563237775, + "grad_norm": 0.692560675949567, + "learning_rate": 9.62279293739968e-05, + "loss": 0.0591, + "step": 1200 + }, + { + "epoch": 0.6801573917931422, + "grad_norm": 0.4228488541088058, + "learning_rate": 9.703049759229535e-05, + "loss": 0.0565, + "step": 1210 + }, + { + "epoch": 0.685778527262507, + "grad_norm": 0.4504043674769784, + "learning_rate": 9.783306581059391e-05, + "loss": 0.0546, + "step": 1220 + }, + { + "epoch": 0.6913996627318718, + "grad_norm": 0.44916766960088017, + "learning_rate": 9.863563402889246e-05, + "loss": 0.0518, + "step": 1230 + }, + { + "epoch": 0.6970207982012366, + "grad_norm": 0.46844229890271083, + "learning_rate": 9.943820224719102e-05, + "loss": 0.0504, + "step": 1240 + }, + { + "epoch": 0.7026419336706015, + "grad_norm": 0.4911174405862922, + "learning_rate": 9.999998231913558e-05, + "loss": 0.0482, + "step": 1250 + }, + { + "epoch": 0.7082630691399663, + "grad_norm": 0.3927574276223742, + "learning_rate": 9.999966799300477e-05, + "loss": 0.0469, + "step": 1260 + }, + { + "epoch": 0.7138842046093311, + "grad_norm": 0.3446568196906315, + "learning_rate": 9.999896076161874e-05, + "loss": 0.0447, + "step": 1270 + }, + { + "epoch": 0.7195053400786959, + "grad_norm": 0.6708037379046565, + "learning_rate": 9.999786063053502e-05, + "loss": 0.0437, + "step": 1280 + }, + { + "epoch": 0.7251264755480608, + "grad_norm": 0.5265622160752478, + "learning_rate": 9.999636760839859e-05, + "loss": 0.0428, + "step": 1290 + }, + { + "epoch": 0.7307476110174255, + "grad_norm": 0.3741304206840176, + "learning_rate": 9.999448170694188e-05, + "loss": 0.0418, + "step": 1300 + }, + { + "epoch": 0.7363687464867903, + "grad_norm": 0.3488892086457768, + "learning_rate": 9.999220294098459e-05, + "loss": 0.0401, + "step": 1310 + }, + { + "epoch": 0.7419898819561551, + "grad_norm": 0.364156745530358, + "learning_rate": 9.998953132843362e-05, + "loss": 0.0392, + "step": 1320 + }, + { + "epoch": 0.7476110174255199, + "grad_norm": 0.316984179052284, + "learning_rate": 9.998646689028292e-05, + "loss": 0.038, + "step": 1330 + }, + { + "epoch": 0.7532321528948848, + "grad_norm": 0.4122286636140672, + "learning_rate": 9.998300965061335e-05, + "loss": 0.0369, + "step": 1340 + }, + { + "epoch": 0.7588532883642496, + "grad_norm": 0.2828777162977292, + "learning_rate": 9.997915963659241e-05, + "loss": 0.0358, + "step": 1350 + }, + { + "epoch": 0.7644744238336144, + "grad_norm": 0.35038412611293795, + "learning_rate": 9.997491687847413e-05, + "loss": 0.0344, + "step": 1360 + }, + { + "epoch": 0.7700955593029792, + "grad_norm": 0.3359568867412438, + "learning_rate": 9.997028140959878e-05, + "loss": 0.0348, + "step": 1370 + }, + { + "epoch": 0.7757166947723441, + "grad_norm": 0.3008434778248411, + "learning_rate": 9.99652532663926e-05, + "loss": 0.0331, + "step": 1380 + }, + { + "epoch": 0.7813378302417088, + "grad_norm": 0.24375407955159972, + "learning_rate": 9.995983248836749e-05, + "loss": 0.0321, + "step": 1390 + }, + { + "epoch": 0.7869589657110736, + "grad_norm": 0.2580356766574075, + "learning_rate": 9.995401911812084e-05, + "loss": 0.0305, + "step": 1400 + }, + { + "epoch": 0.7925801011804384, + "grad_norm": 0.22377682987678044, + "learning_rate": 9.994781320133499e-05, + "loss": 0.0297, + "step": 1410 + }, + { + "epoch": 0.7982012366498032, + "grad_norm": 0.3192038760895579, + "learning_rate": 9.9941214786777e-05, + "loss": 0.0295, + "step": 1420 + }, + { + "epoch": 0.8038223721191681, + "grad_norm": 0.42209606301924135, + "learning_rate": 9.99342239262983e-05, + "loss": 0.0283, + "step": 1430 + }, + { + "epoch": 0.8094435075885329, + "grad_norm": 0.2740277370912408, + "learning_rate": 9.992684067483411e-05, + "loss": 0.0276, + "step": 1440 + }, + { + "epoch": 0.8150646430578977, + "grad_norm": 0.2403536361378785, + "learning_rate": 9.991906509040323e-05, + "loss": 0.027, + "step": 1450 + }, + { + "epoch": 0.8206857785272625, + "grad_norm": 0.25177256736722925, + "learning_rate": 9.99108972341074e-05, + "loss": 0.026, + "step": 1460 + }, + { + "epoch": 0.8263069139966274, + "grad_norm": 0.2611127538329995, + "learning_rate": 9.990233717013092e-05, + "loss": 0.0254, + "step": 1470 + }, + { + "epoch": 0.8319280494659921, + "grad_norm": 0.26600699519251914, + "learning_rate": 9.989338496574013e-05, + "loss": 0.0246, + "step": 1480 + }, + { + "epoch": 0.8375491849353569, + "grad_norm": 0.2511518661704276, + "learning_rate": 9.988404069128286e-05, + "loss": 0.0241, + "step": 1490 + }, + { + "epoch": 0.8431703204047217, + "grad_norm": 0.24275092866635561, + "learning_rate": 9.987430442018787e-05, + "loss": 0.0231, + "step": 1500 + }, + { + "epoch": 0.8487914558740866, + "grad_norm": 0.2117797837985066, + "learning_rate": 9.986417622896433e-05, + "loss": 0.0222, + "step": 1510 + }, + { + "epoch": 0.8544125913434514, + "grad_norm": 0.23706464693081078, + "learning_rate": 9.985365619720114e-05, + "loss": 0.0224, + "step": 1520 + }, + { + "epoch": 0.8600337268128162, + "grad_norm": 0.21681891631644465, + "learning_rate": 9.984274440756639e-05, + "loss": 0.0216, + "step": 1530 + }, + { + "epoch": 0.865654862282181, + "grad_norm": 0.21895203687881423, + "learning_rate": 9.983144094580662e-05, + "loss": 0.0205, + "step": 1540 + }, + { + "epoch": 0.8712759977515458, + "grad_norm": 0.20958525636292016, + "learning_rate": 9.981974590074623e-05, + "loss": 0.0197, + "step": 1550 + }, + { + "epoch": 0.8768971332209107, + "grad_norm": 0.2092568671749629, + "learning_rate": 9.980765936428668e-05, + "loss": 0.0192, + "step": 1560 + }, + { + "epoch": 0.8825182686902754, + "grad_norm": 0.20242771809585056, + "learning_rate": 9.979518143140591e-05, + "loss": 0.0186, + "step": 1570 + }, + { + "epoch": 0.8881394041596402, + "grad_norm": 0.2067518737526698, + "learning_rate": 9.97823122001575e-05, + "loss": 0.0184, + "step": 1580 + }, + { + "epoch": 0.893760539629005, + "grad_norm": 0.2131831404574167, + "learning_rate": 9.976905177166984e-05, + "loss": 0.0179, + "step": 1590 + }, + { + "epoch": 0.8993816750983699, + "grad_norm": 0.19975090191123349, + "learning_rate": 9.975540025014551e-05, + "loss": 0.0171, + "step": 1600 + }, + { + "epoch": 0.9050028105677347, + "grad_norm": 0.1840388312874417, + "learning_rate": 9.974135774286027e-05, + "loss": 0.0167, + "step": 1610 + }, + { + "epoch": 0.9106239460370995, + "grad_norm": 0.16080844246323525, + "learning_rate": 9.972692436016237e-05, + "loss": 0.0162, + "step": 1620 + }, + { + "epoch": 0.9162450815064643, + "grad_norm": 0.7406842420136835, + "learning_rate": 9.971210021547163e-05, + "loss": 0.0164, + "step": 1630 + }, + { + "epoch": 0.9218662169758292, + "grad_norm": 0.6842280048869254, + "learning_rate": 9.969688542527847e-05, + "loss": 0.0185, + "step": 1640 + }, + { + "epoch": 0.927487352445194, + "grad_norm": 1.580555984265757, + "learning_rate": 9.968128010914313e-05, + "loss": 0.0316, + "step": 1650 + }, + { + "epoch": 0.9331084879145587, + "grad_norm": 0.37860426611015985, + "learning_rate": 9.966528438969462e-05, + "loss": 0.0249, + "step": 1660 + }, + { + "epoch": 0.9387296233839235, + "grad_norm": 0.23262955055081527, + "learning_rate": 9.964889839262983e-05, + "loss": 0.0186, + "step": 1670 + }, + { + "epoch": 0.9443507588532883, + "grad_norm": 0.20273477480045216, + "learning_rate": 9.963212224671249e-05, + "loss": 0.0161, + "step": 1680 + }, + { + "epoch": 0.9499718943226532, + "grad_norm": 0.13881800111147724, + "learning_rate": 9.961495608377221e-05, + "loss": 0.0148, + "step": 1690 + }, + { + "epoch": 0.955593029792018, + "grad_norm": 0.17108269648499383, + "learning_rate": 9.959740003870339e-05, + "loss": 0.0144, + "step": 1700 + }, + { + "epoch": 0.9612141652613828, + "grad_norm": 0.16536831484992337, + "learning_rate": 9.95794542494642e-05, + "loss": 0.0137, + "step": 1710 + }, + { + "epoch": 0.9668353007307476, + "grad_norm": 0.17400543803234356, + "learning_rate": 9.956111885707546e-05, + "loss": 0.0131, + "step": 1720 + }, + { + "epoch": 0.9724564362001125, + "grad_norm": 0.14754960398432485, + "learning_rate": 9.954239400561958e-05, + "loss": 0.0126, + "step": 1730 + }, + { + "epoch": 0.9780775716694773, + "grad_norm": 0.13872283525841225, + "learning_rate": 9.952327984223938e-05, + "loss": 0.012, + "step": 1740 + }, + { + "epoch": 0.983698707138842, + "grad_norm": 0.15345071148869308, + "learning_rate": 9.950377651713698e-05, + "loss": 0.0119, + "step": 1750 + }, + { + "epoch": 0.9893198426082068, + "grad_norm": 0.15868096773505136, + "learning_rate": 9.948388418357255e-05, + "loss": 0.0117, + "step": 1760 + }, + { + "epoch": 0.9949409780775716, + "grad_norm": 0.13964728528048598, + "learning_rate": 9.94636029978632e-05, + "loss": 0.0111, + "step": 1770 + }, + { + "epoch": 1.0005621135469365, + "grad_norm": 0.12990131986586648, + "learning_rate": 9.944293311938164e-05, + "loss": 0.0111, + "step": 1780 + }, + { + "epoch": 1.0061832490163012, + "grad_norm": 0.15109117991919174, + "learning_rate": 9.942187471055507e-05, + "loss": 0.0106, + "step": 1790 + }, + { + "epoch": 1.0118043844856661, + "grad_norm": 0.15023868490530576, + "learning_rate": 9.940042793686378e-05, + "loss": 0.0101, + "step": 1800 + }, + { + "epoch": 1.0174255199550308, + "grad_norm": 0.13748040160742883, + "learning_rate": 9.937859296683984e-05, + "loss": 0.0098, + "step": 1810 + }, + { + "epoch": 1.0230466554243958, + "grad_norm": 0.13694721938981147, + "learning_rate": 9.935636997206592e-05, + "loss": 0.0099, + "step": 1820 + }, + { + "epoch": 1.0286677908937605, + "grad_norm": 0.13775398006820014, + "learning_rate": 9.933375912717379e-05, + "loss": 0.0096, + "step": 1830 + }, + { + "epoch": 1.0342889263631254, + "grad_norm": 0.12873175645779594, + "learning_rate": 9.931076060984306e-05, + "loss": 0.0094, + "step": 1840 + }, + { + "epoch": 1.0399100618324901, + "grad_norm": 0.13491704932600562, + "learning_rate": 9.928737460079964e-05, + "loss": 0.0093, + "step": 1850 + }, + { + "epoch": 1.045531197301855, + "grad_norm": 0.13052217132802058, + "learning_rate": 9.926360128381452e-05, + "loss": 0.0089, + "step": 1860 + }, + { + "epoch": 1.0511523327712198, + "grad_norm": 0.15139988915587, + "learning_rate": 9.923944084570214e-05, + "loss": 0.0089, + "step": 1870 + }, + { + "epoch": 1.0567734682405847, + "grad_norm": 0.11412503222119044, + "learning_rate": 9.9214893476319e-05, + "loss": 0.0085, + "step": 1880 + }, + { + "epoch": 1.0623946037099494, + "grad_norm": 0.13727989640704788, + "learning_rate": 9.918995936856224e-05, + "loss": 0.0085, + "step": 1890 + }, + { + "epoch": 1.0680157391793141, + "grad_norm": 0.1249207256562971, + "learning_rate": 9.916463871836796e-05, + "loss": 0.0084, + "step": 1900 + }, + { + "epoch": 1.073636874648679, + "grad_norm": 0.12406222038106689, + "learning_rate": 9.913893172470982e-05, + "loss": 0.0081, + "step": 1910 + }, + { + "epoch": 1.0792580101180438, + "grad_norm": 0.13127597326529594, + "learning_rate": 9.911283858959746e-05, + "loss": 0.0083, + "step": 1920 + }, + { + "epoch": 1.0848791455874087, + "grad_norm": 0.1368675589547782, + "learning_rate": 9.908635951807477e-05, + "loss": 0.0082, + "step": 1930 + }, + { + "epoch": 1.0905002810567734, + "grad_norm": 0.1172965848549566, + "learning_rate": 9.905949471821851e-05, + "loss": 0.008, + "step": 1940 + }, + { + "epoch": 1.0961214165261384, + "grad_norm": 0.11798553949395943, + "learning_rate": 9.903224440113649e-05, + "loss": 0.008, + "step": 1950 + }, + { + "epoch": 1.101742551995503, + "grad_norm": 0.13016996611164902, + "learning_rate": 9.900460878096598e-05, + "loss": 0.0078, + "step": 1960 + }, + { + "epoch": 1.107363687464868, + "grad_norm": 0.12122199750464839, + "learning_rate": 9.897658807487202e-05, + "loss": 0.0077, + "step": 1970 + }, + { + "epoch": 1.1129848229342327, + "grad_norm": 0.11846619937347036, + "learning_rate": 9.894818250304576e-05, + "loss": 0.0075, + "step": 1980 + }, + { + "epoch": 1.1186059584035974, + "grad_norm": 0.13622498242403422, + "learning_rate": 9.891939228870261e-05, + "loss": 0.0076, + "step": 1990 + }, + { + "epoch": 1.1242270938729624, + "grad_norm": 0.1132057504919735, + "learning_rate": 9.889021765808064e-05, + "loss": 0.0076, + "step": 2000 + }, + { + "epoch": 1.129848229342327, + "grad_norm": 0.12782091049281527, + "learning_rate": 9.886065884043868e-05, + "loss": 0.0076, + "step": 2010 + }, + { + "epoch": 1.135469364811692, + "grad_norm": 0.1327283957213651, + "learning_rate": 9.883071606805454e-05, + "loss": 0.0072, + "step": 2020 + }, + { + "epoch": 1.1410905002810567, + "grad_norm": 0.11466155852542638, + "learning_rate": 9.880038957622327e-05, + "loss": 0.0074, + "step": 2030 + }, + { + "epoch": 1.1467116357504217, + "grad_norm": 0.1248526409625004, + "learning_rate": 9.876967960325518e-05, + "loss": 0.0071, + "step": 2040 + }, + { + "epoch": 1.1523327712197864, + "grad_norm": 0.12458128006542332, + "learning_rate": 9.873858639047409e-05, + "loss": 0.0069, + "step": 2050 + }, + { + "epoch": 1.1579539066891513, + "grad_norm": 0.11664744118592543, + "learning_rate": 9.870711018221535e-05, + "loss": 0.0066, + "step": 2060 + }, + { + "epoch": 1.163575042158516, + "grad_norm": 0.12050522669971526, + "learning_rate": 9.867525122582394e-05, + "loss": 0.0068, + "step": 2070 + }, + { + "epoch": 1.1691961776278808, + "grad_norm": 0.13065303841446785, + "learning_rate": 9.864300977165259e-05, + "loss": 0.0067, + "step": 2080 + }, + { + "epoch": 1.1748173130972457, + "grad_norm": 0.10805422192260604, + "learning_rate": 9.861038607305968e-05, + "loss": 0.0066, + "step": 2090 + }, + { + "epoch": 1.1804384485666104, + "grad_norm": 0.11507644557756966, + "learning_rate": 9.857738038640737e-05, + "loss": 0.0063, + "step": 2100 + }, + { + "epoch": 1.1860595840359753, + "grad_norm": 0.10761515703888919, + "learning_rate": 9.854399297105957e-05, + "loss": 0.0061, + "step": 2110 + }, + { + "epoch": 1.19168071950534, + "grad_norm": 0.12594601810241263, + "learning_rate": 9.851022408937979e-05, + "loss": 0.0067, + "step": 2120 + }, + { + "epoch": 1.197301854974705, + "grad_norm": 0.12986420025775458, + "learning_rate": 9.847607400672925e-05, + "loss": 0.0065, + "step": 2130 + }, + { + "epoch": 1.2029229904440697, + "grad_norm": 0.10577243861744662, + "learning_rate": 9.844154299146466e-05, + "loss": 0.0062, + "step": 2140 + }, + { + "epoch": 1.2085441259134346, + "grad_norm": 0.11560521707742587, + "learning_rate": 9.840663131493617e-05, + "loss": 0.0063, + "step": 2150 + }, + { + "epoch": 1.2141652613827993, + "grad_norm": 0.12172276197577853, + "learning_rate": 9.837133925148523e-05, + "loss": 0.006, + "step": 2160 + }, + { + "epoch": 1.219786396852164, + "grad_norm": 0.11989705951251413, + "learning_rate": 9.833566707844241e-05, + "loss": 0.0061, + "step": 2170 + }, + { + "epoch": 1.225407532321529, + "grad_norm": 0.12073736751173343, + "learning_rate": 9.829961507612526e-05, + "loss": 0.0062, + "step": 2180 + }, + { + "epoch": 1.2310286677908937, + "grad_norm": 0.10581806093891255, + "learning_rate": 9.82631835278361e-05, + "loss": 0.006, + "step": 2190 + }, + { + "epoch": 1.2366498032602586, + "grad_norm": 0.12350567107631101, + "learning_rate": 9.822637271985977e-05, + "loss": 0.0058, + "step": 2200 + }, + { + "epoch": 1.2422709387296234, + "grad_norm": 0.10706106863986312, + "learning_rate": 9.818918294146136e-05, + "loss": 0.0055, + "step": 2210 + }, + { + "epoch": 1.2478920741989883, + "grad_norm": 0.11497496683839335, + "learning_rate": 9.815161448488403e-05, + "loss": 0.0059, + "step": 2220 + }, + { + "epoch": 1.253513209668353, + "grad_norm": 0.1230577251222716, + "learning_rate": 9.81136676453466e-05, + "loss": 0.0056, + "step": 2230 + }, + { + "epoch": 1.259134345137718, + "grad_norm": 0.11408509041104613, + "learning_rate": 9.80753427210413e-05, + "loss": 0.0058, + "step": 2240 + }, + { + "epoch": 1.2647554806070826, + "grad_norm": 0.13501233611835498, + "learning_rate": 9.803664001313142e-05, + "loss": 0.0057, + "step": 2250 + }, + { + "epoch": 1.2703766160764474, + "grad_norm": 0.09995034213120912, + "learning_rate": 9.799755982574894e-05, + "loss": 0.0057, + "step": 2260 + }, + { + "epoch": 1.2759977515458123, + "grad_norm": 0.10016312243592881, + "learning_rate": 9.79581024659921e-05, + "loss": 0.0055, + "step": 2270 + }, + { + "epoch": 1.281618887015177, + "grad_norm": 0.12356597886645763, + "learning_rate": 9.791826824392305e-05, + "loss": 0.0056, + "step": 2280 + }, + { + "epoch": 1.287240022484542, + "grad_norm": 0.10251488515673644, + "learning_rate": 9.787805747256536e-05, + "loss": 0.0055, + "step": 2290 + }, + { + "epoch": 1.2928611579539067, + "grad_norm": 0.10694624192420116, + "learning_rate": 9.783747046790162e-05, + "loss": 0.0054, + "step": 2300 + }, + { + "epoch": 1.2984822934232714, + "grad_norm": 0.0837306764941164, + "learning_rate": 9.779650754887086e-05, + "loss": 0.0052, + "step": 2310 + }, + { + "epoch": 1.3041034288926363, + "grad_norm": 0.12536680825139915, + "learning_rate": 9.775516903736616e-05, + "loss": 0.0053, + "step": 2320 + }, + { + "epoch": 1.3097245643620012, + "grad_norm": 0.09801113540564874, + "learning_rate": 9.771345525823208e-05, + "loss": 0.0053, + "step": 2330 + }, + { + "epoch": 1.315345699831366, + "grad_norm": 0.3194597362808477, + "learning_rate": 9.767136653926199e-05, + "loss": 0.0055, + "step": 2340 + }, + { + "epoch": 1.3209668353007307, + "grad_norm": 1.8141673380574084, + "learning_rate": 9.762890321119572e-05, + "loss": 0.0263, + "step": 2350 + }, + { + "epoch": 1.3265879707700956, + "grad_norm": 2.0373228278291986, + "learning_rate": 9.758606560771675e-05, + "loss": 0.0595, + "step": 2360 + }, + { + "epoch": 1.3322091062394603, + "grad_norm": 0.24089249295422568, + "learning_rate": 9.754285406544969e-05, + "loss": 0.0223, + "step": 2370 + }, + { + "epoch": 1.3378302417088253, + "grad_norm": 0.14596942738787602, + "learning_rate": 9.749926892395768e-05, + "loss": 0.0095, + "step": 2380 + }, + { + "epoch": 1.34345137717819, + "grad_norm": 0.11444722263474487, + "learning_rate": 9.745531052573961e-05, + "loss": 0.0072, + "step": 2390 + }, + { + "epoch": 1.3490725126475547, + "grad_norm": 0.0997521525083771, + "learning_rate": 9.741097921622746e-05, + "loss": 0.0063, + "step": 2400 + }, + { + "epoch": 1.3546936481169196, + "grad_norm": 0.10164810149706918, + "learning_rate": 9.736627534378364e-05, + "loss": 0.0058, + "step": 2410 + }, + { + "epoch": 1.3603147835862845, + "grad_norm": 0.10637858866117042, + "learning_rate": 9.732119925969823e-05, + "loss": 0.0055, + "step": 2420 + }, + { + "epoch": 1.3659359190556493, + "grad_norm": 0.1154184870469501, + "learning_rate": 9.727575131818617e-05, + "loss": 0.0052, + "step": 2430 + }, + { + "epoch": 1.371557054525014, + "grad_norm": 0.09812174226964852, + "learning_rate": 9.722993187638454e-05, + "loss": 0.0051, + "step": 2440 + }, + { + "epoch": 1.377178189994379, + "grad_norm": 0.08717591400124859, + "learning_rate": 9.718374129434971e-05, + "loss": 0.0051, + "step": 2450 + }, + { + "epoch": 1.3827993254637436, + "grad_norm": 0.08608962077541132, + "learning_rate": 9.713717993505456e-05, + "loss": 0.0049, + "step": 2460 + }, + { + "epoch": 1.3884204609331086, + "grad_norm": 0.10182986875302574, + "learning_rate": 9.709024816438555e-05, + "loss": 0.0047, + "step": 2470 + }, + { + "epoch": 1.3940415964024733, + "grad_norm": 0.10115287527791918, + "learning_rate": 9.704294635113993e-05, + "loss": 0.0049, + "step": 2480 + }, + { + "epoch": 1.399662731871838, + "grad_norm": 0.0784538159472712, + "learning_rate": 9.69952748670228e-05, + "loss": 0.0048, + "step": 2490 + }, + { + "epoch": 1.405283867341203, + "grad_norm": 0.0776387928340622, + "learning_rate": 9.694723408664414e-05, + "loss": 0.0044, + "step": 2500 + }, + { + "epoch": 1.4109050028105679, + "grad_norm": 0.0837732573729813, + "learning_rate": 9.689882438751601e-05, + "loss": 0.0046, + "step": 2510 + }, + { + "epoch": 1.4165261382799326, + "grad_norm": 0.09489114181755515, + "learning_rate": 9.68500461500494e-05, + "loss": 0.0046, + "step": 2520 + }, + { + "epoch": 1.4221472737492973, + "grad_norm": 0.101494695984547, + "learning_rate": 9.680089975755139e-05, + "loss": 0.0045, + "step": 2530 + }, + { + "epoch": 1.4277684092186622, + "grad_norm": 0.08821477058203668, + "learning_rate": 9.675138559622205e-05, + "loss": 0.0043, + "step": 2540 + }, + { + "epoch": 1.433389544688027, + "grad_norm": 0.07881075966273893, + "learning_rate": 9.670150405515147e-05, + "loss": 0.0044, + "step": 2550 + }, + { + "epoch": 1.4390106801573919, + "grad_norm": 0.09899683557846178, + "learning_rate": 9.665125552631662e-05, + "loss": 0.0043, + "step": 2560 + }, + { + "epoch": 1.4446318156267566, + "grad_norm": 0.08629556264118085, + "learning_rate": 9.660064040457837e-05, + "loss": 0.0044, + "step": 2570 + }, + { + "epoch": 1.4502529510961213, + "grad_norm": 0.09415390850669993, + "learning_rate": 9.654965908767827e-05, + "loss": 0.0044, + "step": 2580 + }, + { + "epoch": 1.4558740865654862, + "grad_norm": 0.09596874126004533, + "learning_rate": 9.649831197623556e-05, + "loss": 0.0043, + "step": 2590 + }, + { + "epoch": 1.4614952220348512, + "grad_norm": 0.09072868361315219, + "learning_rate": 9.64465994737439e-05, + "loss": 0.0041, + "step": 2600 + }, + { + "epoch": 1.4671163575042159, + "grad_norm": 0.08557262009476094, + "learning_rate": 9.639452198656828e-05, + "loss": 0.0041, + "step": 2610 + }, + { + "epoch": 1.4727374929735806, + "grad_norm": 0.08310295695036411, + "learning_rate": 9.634207992394178e-05, + "loss": 0.0041, + "step": 2620 + }, + { + "epoch": 1.4783586284429455, + "grad_norm": 0.09055939067871042, + "learning_rate": 9.628927369796236e-05, + "loss": 0.0041, + "step": 2630 + }, + { + "epoch": 1.4839797639123102, + "grad_norm": 0.09604647791927998, + "learning_rate": 9.623610372358969e-05, + "loss": 0.0042, + "step": 2640 + }, + { + "epoch": 1.4896008993816752, + "grad_norm": 0.08577660993704818, + "learning_rate": 9.618257041864176e-05, + "loss": 0.0043, + "step": 2650 + }, + { + "epoch": 1.4952220348510399, + "grad_norm": 0.09123279036915018, + "learning_rate": 9.61286742037917e-05, + "loss": 0.0041, + "step": 2660 + }, + { + "epoch": 1.5008431703204046, + "grad_norm": 0.08260166738909931, + "learning_rate": 9.607441550256445e-05, + "loss": 0.0039, + "step": 2670 + }, + { + "epoch": 1.5064643057897695, + "grad_norm": 0.07923250866368316, + "learning_rate": 9.601979474133343e-05, + "loss": 0.004, + "step": 2680 + }, + { + "epoch": 1.5120854412591345, + "grad_norm": 0.08239031912656816, + "learning_rate": 9.596481234931713e-05, + "loss": 0.0038, + "step": 2690 + }, + { + "epoch": 1.5177065767284992, + "grad_norm": 0.07610828082112323, + "learning_rate": 9.59094687585759e-05, + "loss": 0.0039, + "step": 2700 + }, + { + "epoch": 1.523327712197864, + "grad_norm": 0.11252793520644211, + "learning_rate": 9.585376440400833e-05, + "loss": 0.0041, + "step": 2710 + }, + { + "epoch": 1.5289488476672288, + "grad_norm": 0.09330378622902911, + "learning_rate": 9.579769972334802e-05, + "loss": 0.004, + "step": 2720 + }, + { + "epoch": 1.5345699831365935, + "grad_norm": 0.07857369281450616, + "learning_rate": 9.574127515716004e-05, + "loss": 0.0039, + "step": 2730 + }, + { + "epoch": 1.5401911186059585, + "grad_norm": 0.07210626541279375, + "learning_rate": 9.56844911488375e-05, + "loss": 0.0038, + "step": 2740 + }, + { + "epoch": 1.5458122540753232, + "grad_norm": 0.07707927197752867, + "learning_rate": 9.562734814459811e-05, + "loss": 0.0038, + "step": 2750 + }, + { + "epoch": 1.551433389544688, + "grad_norm": 0.08248750325983276, + "learning_rate": 9.556984659348052e-05, + "loss": 0.0039, + "step": 2760 + }, + { + "epoch": 1.5570545250140528, + "grad_norm": 0.08508167021142606, + "learning_rate": 9.551198694734102e-05, + "loss": 0.0039, + "step": 2770 + }, + { + "epoch": 1.5626756604834178, + "grad_norm": 0.08080512980956885, + "learning_rate": 9.545376966084977e-05, + "loss": 0.0038, + "step": 2780 + }, + { + "epoch": 1.5682967959527825, + "grad_norm": 0.09346422370365523, + "learning_rate": 9.539519519148738e-05, + "loss": 0.0037, + "step": 2790 + }, + { + "epoch": 1.5739179314221472, + "grad_norm": 0.08917779155675297, + "learning_rate": 9.533626399954126e-05, + "loss": 0.0036, + "step": 2800 + }, + { + "epoch": 1.5795390668915121, + "grad_norm": 0.06721304489449485, + "learning_rate": 9.527697654810195e-05, + "loss": 0.0035, + "step": 2810 + }, + { + "epoch": 1.5851602023608768, + "grad_norm": 0.07427558428573001, + "learning_rate": 9.52173333030596e-05, + "loss": 0.0036, + "step": 2820 + }, + { + "epoch": 1.5907813378302418, + "grad_norm": 0.08659228174058138, + "learning_rate": 9.515733473310019e-05, + "loss": 0.0036, + "step": 2830 + }, + { + "epoch": 1.5964024732996065, + "grad_norm": 0.0719969223939997, + "learning_rate": 9.509698130970193e-05, + "loss": 0.0036, + "step": 2840 + }, + { + "epoch": 1.6020236087689712, + "grad_norm": 0.08181084829875987, + "learning_rate": 9.50362735071315e-05, + "loss": 0.0034, + "step": 2850 + }, + { + "epoch": 1.6076447442383361, + "grad_norm": 0.0894284800559242, + "learning_rate": 9.497521180244038e-05, + "loss": 0.0035, + "step": 2860 + }, + { + "epoch": 1.613265879707701, + "grad_norm": 0.08282207277574022, + "learning_rate": 9.491379667546104e-05, + "loss": 0.0034, + "step": 2870 + }, + { + "epoch": 1.6188870151770658, + "grad_norm": 0.145986818674443, + "learning_rate": 9.485202860880321e-05, + "loss": 0.0037, + "step": 2880 + }, + { + "epoch": 1.6245081506464305, + "grad_norm": 0.08546744479627431, + "learning_rate": 9.478990808785006e-05, + "loss": 0.0036, + "step": 2890 + }, + { + "epoch": 1.6301292861157954, + "grad_norm": 0.08759751416458819, + "learning_rate": 9.472743560075446e-05, + "loss": 0.0034, + "step": 2900 + }, + { + "epoch": 1.6357504215851602, + "grad_norm": 0.07818496950657797, + "learning_rate": 9.466461163843497e-05, + "loss": 0.0033, + "step": 2910 + }, + { + "epoch": 1.641371557054525, + "grad_norm": 0.07598460655098531, + "learning_rate": 9.460143669457221e-05, + "loss": 0.0034, + "step": 2920 + }, + { + "epoch": 1.6469926925238898, + "grad_norm": 0.07204405112584633, + "learning_rate": 9.453791126560482e-05, + "loss": 0.0033, + "step": 2930 + }, + { + "epoch": 1.6526138279932545, + "grad_norm": 0.07548100465060122, + "learning_rate": 9.44740358507256e-05, + "loss": 0.0033, + "step": 2940 + }, + { + "epoch": 1.6582349634626194, + "grad_norm": 0.07627624404632878, + "learning_rate": 9.440981095187764e-05, + "loss": 0.0034, + "step": 2950 + }, + { + "epoch": 1.6638560989319844, + "grad_norm": 0.07801223336358269, + "learning_rate": 9.434523707375024e-05, + "loss": 0.0033, + "step": 2960 + }, + { + "epoch": 1.669477234401349, + "grad_norm": 0.07014240339623043, + "learning_rate": 9.428031472377516e-05, + "loss": 0.0031, + "step": 2970 + }, + { + "epoch": 1.6750983698707138, + "grad_norm": 0.0725952595345933, + "learning_rate": 9.421504441212238e-05, + "loss": 0.0032, + "step": 2980 + }, + { + "epoch": 1.6807195053400787, + "grad_norm": 0.08432049334212283, + "learning_rate": 9.414942665169631e-05, + "loss": 0.0032, + "step": 2990 + }, + { + "epoch": 1.6863406408094435, + "grad_norm": 0.07105741598976323, + "learning_rate": 9.408346195813159e-05, + "loss": 0.0034, + "step": 3000 + }, + { + "epoch": 1.6919617762788084, + "grad_norm": 0.13906184784395983, + "learning_rate": 9.401715084978922e-05, + "loss": 0.0052, + "step": 3010 + }, + { + "epoch": 1.697582911748173, + "grad_norm": 0.11601157873012202, + "learning_rate": 9.395049384775225e-05, + "loss": 0.0042, + "step": 3020 + }, + { + "epoch": 1.7032040472175378, + "grad_norm": 0.09015124532353191, + "learning_rate": 9.388349147582194e-05, + "loss": 0.0037, + "step": 3030 + }, + { + "epoch": 1.7088251826869028, + "grad_norm": 0.07000264697695573, + "learning_rate": 9.381614426051344e-05, + "loss": 0.0034, + "step": 3040 + }, + { + "epoch": 1.7144463181562677, + "grad_norm": 0.08113336516882994, + "learning_rate": 9.374845273105177e-05, + "loss": 0.0033, + "step": 3050 + }, + { + "epoch": 1.7200674536256324, + "grad_norm": 0.08261811708445045, + "learning_rate": 9.368041741936762e-05, + "loss": 0.0033, + "step": 3060 + }, + { + "epoch": 1.7256885890949971, + "grad_norm": 0.07838609207688987, + "learning_rate": 9.361203886009316e-05, + "loss": 0.0033, + "step": 3070 + }, + { + "epoch": 1.731309724564362, + "grad_norm": 0.08665499355732291, + "learning_rate": 9.354331759055787e-05, + "loss": 0.0033, + "step": 3080 + }, + { + "epoch": 1.7369308600337268, + "grad_norm": 0.07918424347748348, + "learning_rate": 9.347425415078427e-05, + "loss": 0.0034, + "step": 3090 + }, + { + "epoch": 1.7425519955030917, + "grad_norm": 0.08191128168223484, + "learning_rate": 9.34048490834838e-05, + "loss": 0.0031, + "step": 3100 + }, + { + "epoch": 1.7481731309724564, + "grad_norm": 0.0759010615813787, + "learning_rate": 9.333510293405231e-05, + "loss": 0.003, + "step": 3110 + }, + { + "epoch": 1.7537942664418211, + "grad_norm": 0.08477221828549904, + "learning_rate": 9.326501625056606e-05, + "loss": 0.0031, + "step": 3120 + }, + { + "epoch": 1.759415401911186, + "grad_norm": 0.08622296059752232, + "learning_rate": 9.319458958377723e-05, + "loss": 0.0032, + "step": 3130 + }, + { + "epoch": 1.765036537380551, + "grad_norm": 0.07276169675225246, + "learning_rate": 9.312382348710962e-05, + "loss": 0.0031, + "step": 3140 + }, + { + "epoch": 1.7706576728499157, + "grad_norm": 0.07298764401064435, + "learning_rate": 9.305271851665438e-05, + "loss": 0.0031, + "step": 3150 + }, + { + "epoch": 1.7762788083192804, + "grad_norm": 0.0765870165752984, + "learning_rate": 9.298127523116555e-05, + "loss": 0.003, + "step": 3160 + }, + { + "epoch": 1.7818999437886454, + "grad_norm": 0.0963853743897375, + "learning_rate": 9.290949419205567e-05, + "loss": 0.0032, + "step": 3170 + }, + { + "epoch": 1.78752107925801, + "grad_norm": 0.08167306430005795, + "learning_rate": 9.283737596339147e-05, + "loss": 0.003, + "step": 3180 + }, + { + "epoch": 1.793142214727375, + "grad_norm": 0.07386683946543225, + "learning_rate": 9.276492111188933e-05, + "loss": 0.0029, + "step": 3190 + }, + { + "epoch": 1.7987633501967397, + "grad_norm": 0.079496491310148, + "learning_rate": 9.269213020691083e-05, + "loss": 0.003, + "step": 3200 + }, + { + "epoch": 1.8043844856661044, + "grad_norm": 0.08006049495470177, + "learning_rate": 9.261900382045841e-05, + "loss": 0.003, + "step": 3210 + }, + { + "epoch": 1.8100056211354694, + "grad_norm": 0.06946915908105838, + "learning_rate": 9.254554252717067e-05, + "loss": 0.003, + "step": 3220 + }, + { + "epoch": 1.8156267566048343, + "grad_norm": 0.06692045669453084, + "learning_rate": 9.247174690431803e-05, + "loss": 0.0028, + "step": 3230 + }, + { + "epoch": 1.821247892074199, + "grad_norm": 0.06938244252654852, + "learning_rate": 9.23976175317981e-05, + "loss": 0.0029, + "step": 3240 + }, + { + "epoch": 1.8268690275435637, + "grad_norm": 0.0677124057287239, + "learning_rate": 9.232315499213114e-05, + "loss": 0.0028, + "step": 3250 + }, + { + "epoch": 1.8324901630129287, + "grad_norm": 0.0659706764858958, + "learning_rate": 9.224835987045549e-05, + "loss": 0.0027, + "step": 3260 + }, + { + "epoch": 1.8381112984822934, + "grad_norm": 0.07797163812538724, + "learning_rate": 9.217323275452301e-05, + "loss": 0.0029, + "step": 3270 + }, + { + "epoch": 1.8437324339516583, + "grad_norm": 0.06371412754471345, + "learning_rate": 9.209777423469436e-05, + "loss": 0.0027, + "step": 3280 + }, + { + "epoch": 1.849353569421023, + "grad_norm": 0.07195198600776227, + "learning_rate": 9.202198490393448e-05, + "loss": 0.0027, + "step": 3290 + }, + { + "epoch": 1.8549747048903877, + "grad_norm": 0.09032730876019279, + "learning_rate": 9.194586535780785e-05, + "loss": 0.0027, + "step": 3300 + }, + { + "epoch": 1.8605958403597527, + "grad_norm": 0.07184863572635682, + "learning_rate": 9.186941619447382e-05, + "loss": 0.0028, + "step": 3310 + }, + { + "epoch": 1.8662169758291176, + "grad_norm": 0.07972046752123726, + "learning_rate": 9.179263801468196e-05, + "loss": 0.0028, + "step": 3320 + }, + { + "epoch": 1.8718381112984823, + "grad_norm": 0.06462786356243252, + "learning_rate": 9.171553142176725e-05, + "loss": 0.0026, + "step": 3330 + }, + { + "epoch": 1.877459246767847, + "grad_norm": 0.0748752457223882, + "learning_rate": 9.163809702164547e-05, + "loss": 0.0028, + "step": 3340 + }, + { + "epoch": 1.883080382237212, + "grad_norm": 0.06644669684356859, + "learning_rate": 9.156033542280829e-05, + "loss": 0.0027, + "step": 3350 + }, + { + "epoch": 1.8887015177065767, + "grad_norm": 0.06481410104589969, + "learning_rate": 9.148224723631858e-05, + "loss": 0.0028, + "step": 3360 + }, + { + "epoch": 1.8943226531759416, + "grad_norm": 0.07493396927909493, + "learning_rate": 9.14038330758056e-05, + "loss": 0.0028, + "step": 3370 + }, + { + "epoch": 1.8999437886453063, + "grad_norm": 0.07930938338708406, + "learning_rate": 9.132509355746016e-05, + "loss": 0.0028, + "step": 3380 + }, + { + "epoch": 1.905564924114671, + "grad_norm": 0.0703227155299206, + "learning_rate": 9.124602930002977e-05, + "loss": 0.0027, + "step": 3390 + }, + { + "epoch": 1.911186059584036, + "grad_norm": 0.06415414989508922, + "learning_rate": 9.11666409248138e-05, + "loss": 0.0027, + "step": 3400 + }, + { + "epoch": 1.916807195053401, + "grad_norm": 0.05989446878848808, + "learning_rate": 9.10869290556586e-05, + "loss": 0.0026, + "step": 3410 + }, + { + "epoch": 1.9224283305227656, + "grad_norm": 0.06596944372327754, + "learning_rate": 9.100689431895257e-05, + "loss": 0.0026, + "step": 3420 + }, + { + "epoch": 1.9280494659921303, + "grad_norm": 0.06839374667585654, + "learning_rate": 9.092653734362126e-05, + "loss": 0.0025, + "step": 3430 + }, + { + "epoch": 1.9336706014614953, + "grad_norm": 0.05902981335451462, + "learning_rate": 9.084585876112241e-05, + "loss": 0.0025, + "step": 3440 + }, + { + "epoch": 1.93929173693086, + "grad_norm": 0.06610787957287352, + "learning_rate": 9.076485920544105e-05, + "loss": 0.0027, + "step": 3450 + }, + { + "epoch": 1.944912872400225, + "grad_norm": 0.08257300358198327, + "learning_rate": 9.068353931308437e-05, + "loss": 0.0027, + "step": 3460 + }, + { + "epoch": 1.9505340078695896, + "grad_norm": 0.0643606698153111, + "learning_rate": 9.060189972307691e-05, + "loss": 0.0025, + "step": 3470 + }, + { + "epoch": 1.9561551433389543, + "grad_norm": 0.06558027003590208, + "learning_rate": 9.051994107695541e-05, + "loss": 0.0025, + "step": 3480 + }, + { + "epoch": 1.9617762788083193, + "grad_norm": 0.0752177480017892, + "learning_rate": 9.043766401876376e-05, + "loss": 0.0024, + "step": 3490 + }, + { + "epoch": 1.9673974142776842, + "grad_norm": 0.07854251356309212, + "learning_rate": 9.03550691950481e-05, + "loss": 0.0026, + "step": 3500 + }, + { + "epoch": 1.973018549747049, + "grad_norm": 0.07298249855242128, + "learning_rate": 9.027215725485148e-05, + "loss": 0.0026, + "step": 3510 + }, + { + "epoch": 1.9786396852164136, + "grad_norm": 0.08010026435433024, + "learning_rate": 9.018892884970899e-05, + "loss": 0.0026, + "step": 3520 + }, + { + "epoch": 1.9842608206857786, + "grad_norm": 0.06202515462439953, + "learning_rate": 9.010538463364256e-05, + "loss": 0.0024, + "step": 3530 + }, + { + "epoch": 1.9898819561551433, + "grad_norm": 0.0638298747620675, + "learning_rate": 9.002152526315571e-05, + "loss": 0.0025, + "step": 3540 + }, + { + "epoch": 1.9955030916245082, + "grad_norm": 0.05989999228664472, + "learning_rate": 8.993735139722862e-05, + "loss": 0.0024, + "step": 3550 + }, + { + "epoch": 2.001124227093873, + "grad_norm": 0.058109118290538206, + "learning_rate": 8.985286369731275e-05, + "loss": 0.0024, + "step": 3560 + }, + { + "epoch": 2.0067453625632377, + "grad_norm": 0.057457616331219645, + "learning_rate": 8.976806282732572e-05, + "loss": 0.0019, + "step": 3570 + }, + { + "epoch": 2.0123664980326024, + "grad_norm": 0.05651975998498543, + "learning_rate": 8.968294945364613e-05, + "loss": 0.002, + "step": 3580 + }, + { + "epoch": 2.0179876335019675, + "grad_norm": 0.05599584314470216, + "learning_rate": 8.959752424510824e-05, + "loss": 0.0021, + "step": 3590 + }, + { + "epoch": 2.0236087689713322, + "grad_norm": 0.06816841901562025, + "learning_rate": 8.951178787299679e-05, + "loss": 0.0018, + "step": 3600 + }, + { + "epoch": 2.029229904440697, + "grad_norm": 0.07476474513368925, + "learning_rate": 8.942574101104164e-05, + "loss": 0.0019, + "step": 3610 + }, + { + "epoch": 2.0348510399100617, + "grad_norm": 0.058566412874928996, + "learning_rate": 8.933938433541262e-05, + "loss": 0.0019, + "step": 3620 + }, + { + "epoch": 2.040472175379427, + "grad_norm": 0.054106525795158976, + "learning_rate": 8.925271852471403e-05, + "loss": 0.0019, + "step": 3630 + }, + { + "epoch": 2.0460933108487915, + "grad_norm": 0.0594202177574867, + "learning_rate": 8.916574425997946e-05, + "loss": 0.0018, + "step": 3640 + }, + { + "epoch": 2.0517144463181562, + "grad_norm": 0.053841873095984816, + "learning_rate": 8.907846222466634e-05, + "loss": 0.0017, + "step": 3650 + }, + { + "epoch": 2.057335581787521, + "grad_norm": 0.052808507853994574, + "learning_rate": 8.899087310465066e-05, + "loss": 0.0016, + "step": 3660 + }, + { + "epoch": 2.0629567172568857, + "grad_norm": 0.057001105318726125, + "learning_rate": 8.890297758822148e-05, + "loss": 0.0018, + "step": 3670 + }, + { + "epoch": 2.068577852726251, + "grad_norm": 0.057273072509835384, + "learning_rate": 8.881477636607562e-05, + "loss": 0.0018, + "step": 3680 + }, + { + "epoch": 2.0741989881956155, + "grad_norm": 0.0546263439951423, + "learning_rate": 8.872627013131211e-05, + "loss": 0.0017, + "step": 3690 + }, + { + "epoch": 2.0798201236649803, + "grad_norm": 0.06711361958170248, + "learning_rate": 8.863745957942691e-05, + "loss": 0.0018, + "step": 3700 + }, + { + "epoch": 2.085441259134345, + "grad_norm": 0.050819332117840434, + "learning_rate": 8.854834540830728e-05, + "loss": 0.0017, + "step": 3710 + }, + { + "epoch": 2.09106239460371, + "grad_norm": 0.06830828572411692, + "learning_rate": 8.845892831822639e-05, + "loss": 0.0017, + "step": 3720 + }, + { + "epoch": 2.096683530073075, + "grad_norm": 0.061987972777272506, + "learning_rate": 8.83692090118378e-05, + "loss": 0.0017, + "step": 3730 + }, + { + "epoch": 2.1023046655424396, + "grad_norm": 0.0669128652733702, + "learning_rate": 8.827918819416996e-05, + "loss": 0.0018, + "step": 3740 + }, + { + "epoch": 2.1079258010118043, + "grad_norm": 0.07873101299778994, + "learning_rate": 8.818886657262058e-05, + "loss": 0.0019, + "step": 3750 + }, + { + "epoch": 2.1135469364811694, + "grad_norm": 0.07345024781427896, + "learning_rate": 8.809824485695118e-05, + "loss": 0.0017, + "step": 3760 + }, + { + "epoch": 2.119168071950534, + "grad_norm": 0.06739076371941057, + "learning_rate": 8.800732375928146e-05, + "loss": 0.0018, + "step": 3770 + }, + { + "epoch": 2.124789207419899, + "grad_norm": 0.07680093288435995, + "learning_rate": 8.791610399408369e-05, + "loss": 0.002, + "step": 3780 + }, + { + "epoch": 2.1304103428892636, + "grad_norm": 0.0884538226879485, + "learning_rate": 8.782458627817716e-05, + "loss": 0.0019, + "step": 3790 + }, + { + "epoch": 2.1360314783586283, + "grad_norm": 0.06512081179717873, + "learning_rate": 8.773277133072244e-05, + "loss": 0.0019, + "step": 3800 + }, + { + "epoch": 2.1416526138279934, + "grad_norm": 0.06562103971194272, + "learning_rate": 8.764065987321586e-05, + "loss": 0.002, + "step": 3810 + }, + { + "epoch": 2.147273749297358, + "grad_norm": 0.06091852571490337, + "learning_rate": 8.754825262948371e-05, + "loss": 0.002, + "step": 3820 + }, + { + "epoch": 2.152894884766723, + "grad_norm": 0.05792467251440998, + "learning_rate": 8.745555032567667e-05, + "loss": 0.0019, + "step": 3830 + }, + { + "epoch": 2.1585160202360876, + "grad_norm": 0.053473661297281636, + "learning_rate": 8.736255369026403e-05, + "loss": 0.0017, + "step": 3840 + }, + { + "epoch": 2.1641371557054523, + "grad_norm": 0.08632132178323625, + "learning_rate": 8.726926345402796e-05, + "loss": 0.0018, + "step": 3850 + }, + { + "epoch": 2.1697582911748174, + "grad_norm": 0.05824894493699371, + "learning_rate": 8.71756803500578e-05, + "loss": 0.0018, + "step": 3860 + }, + { + "epoch": 2.175379426644182, + "grad_norm": 0.05568698678649655, + "learning_rate": 8.708180511374432e-05, + "loss": 0.0018, + "step": 3870 + }, + { + "epoch": 2.181000562113547, + "grad_norm": 0.056802968519803224, + "learning_rate": 8.69876384827739e-05, + "loss": 0.0017, + "step": 3880 + }, + { + "epoch": 2.1866216975829116, + "grad_norm": 0.06697937497255127, + "learning_rate": 8.689318119712269e-05, + "loss": 0.0018, + "step": 3890 + }, + { + "epoch": 2.1922428330522767, + "grad_norm": 0.1352898553987361, + "learning_rate": 8.679843399905096e-05, + "loss": 0.002, + "step": 3900 + }, + { + "epoch": 2.1978639685216415, + "grad_norm": 0.06346871111752156, + "learning_rate": 8.670339763309706e-05, + "loss": 0.0019, + "step": 3910 + }, + { + "epoch": 2.203485103991006, + "grad_norm": 0.07365846761282421, + "learning_rate": 8.660807284607171e-05, + "loss": 0.002, + "step": 3920 + }, + { + "epoch": 2.209106239460371, + "grad_norm": 0.07340874139355225, + "learning_rate": 8.651246038705206e-05, + "loss": 0.002, + "step": 3930 + }, + { + "epoch": 2.214727374929736, + "grad_norm": 0.07574517541710105, + "learning_rate": 8.641656100737588e-05, + "loss": 0.0021, + "step": 3940 + }, + { + "epoch": 2.2203485103991007, + "grad_norm": 0.06886932484491658, + "learning_rate": 8.632037546063556e-05, + "loss": 0.0018, + "step": 3950 + }, + { + "epoch": 2.2259696458684655, + "grad_norm": 0.04966512197542119, + "learning_rate": 8.622390450267227e-05, + "loss": 0.0018, + "step": 3960 + }, + { + "epoch": 2.23159078133783, + "grad_norm": 0.06298016565071278, + "learning_rate": 8.612714889156996e-05, + "loss": 0.0018, + "step": 3970 + }, + { + "epoch": 2.237211916807195, + "grad_norm": 0.09602521126309173, + "learning_rate": 8.603010938764944e-05, + "loss": 0.0017, + "step": 3980 + }, + { + "epoch": 2.24283305227656, + "grad_norm": 0.06179034901494223, + "learning_rate": 8.593278675346239e-05, + "loss": 0.0018, + "step": 3990 + }, + { + "epoch": 2.2484541877459248, + "grad_norm": 0.05768935033293568, + "learning_rate": 8.58351817537854e-05, + "loss": 0.0017, + "step": 4000 + }, + { + "epoch": 2.2540753232152895, + "grad_norm": 0.05638492675248429, + "learning_rate": 8.573729515561387e-05, + "loss": 0.0019, + "step": 4010 + }, + { + "epoch": 2.259696458684654, + "grad_norm": 0.06657528366005455, + "learning_rate": 8.563912772815612e-05, + "loss": 0.0018, + "step": 4020 + }, + { + "epoch": 2.265317594154019, + "grad_norm": 0.06296651541934814, + "learning_rate": 8.554068024282722e-05, + "loss": 0.0018, + "step": 4030 + }, + { + "epoch": 2.270938729623384, + "grad_norm": 0.06738339509707009, + "learning_rate": 8.5441953473243e-05, + "loss": 0.0017, + "step": 4040 + }, + { + "epoch": 2.2765598650927488, + "grad_norm": 0.07704835455024767, + "learning_rate": 8.534294819521394e-05, + "loss": 0.0016, + "step": 4050 + }, + { + "epoch": 2.2821810005621135, + "grad_norm": 0.0567808801938328, + "learning_rate": 8.524366518673907e-05, + "loss": 0.0017, + "step": 4060 + }, + { + "epoch": 2.287802136031478, + "grad_norm": 0.06033970611341973, + "learning_rate": 8.514410522799992e-05, + "loss": 0.0017, + "step": 4070 + }, + { + "epoch": 2.2934232715008434, + "grad_norm": 0.06295828590005369, + "learning_rate": 8.504426910135429e-05, + "loss": 0.0016, + "step": 4080 + }, + { + "epoch": 2.299044406970208, + "grad_norm": 0.06724936488740965, + "learning_rate": 8.494415759133014e-05, + "loss": 0.0017, + "step": 4090 + }, + { + "epoch": 2.304665542439573, + "grad_norm": 0.06118776673887835, + "learning_rate": 8.48437714846195e-05, + "loss": 0.0016, + "step": 4100 + }, + { + "epoch": 2.3102866779089375, + "grad_norm": 0.09003973689457073, + "learning_rate": 8.474311157007218e-05, + "loss": 0.0018, + "step": 4110 + }, + { + "epoch": 2.3159078133783026, + "grad_norm": 0.06239353412217714, + "learning_rate": 8.46421786386896e-05, + "loss": 0.0018, + "step": 4120 + }, + { + "epoch": 2.3215289488476674, + "grad_norm": 0.05655954006890423, + "learning_rate": 8.45409734836186e-05, + "loss": 0.0016, + "step": 4130 + }, + { + "epoch": 2.327150084317032, + "grad_norm": 0.0521687512636754, + "learning_rate": 8.443949690014527e-05, + "loss": 0.0016, + "step": 4140 + }, + { + "epoch": 2.332771219786397, + "grad_norm": 0.05228400769684489, + "learning_rate": 8.433774968568849e-05, + "loss": 0.0016, + "step": 4150 + }, + { + "epoch": 2.3383923552557615, + "grad_norm": 0.054685428841644984, + "learning_rate": 8.423573263979392e-05, + "loss": 0.0016, + "step": 4160 + }, + { + "epoch": 2.3440134907251267, + "grad_norm": 0.05279358062852916, + "learning_rate": 8.413344656412754e-05, + "loss": 0.0017, + "step": 4170 + }, + { + "epoch": 2.3496346261944914, + "grad_norm": 0.0619594130610791, + "learning_rate": 8.40308922624694e-05, + "loss": 0.0015, + "step": 4180 + }, + { + "epoch": 2.355255761663856, + "grad_norm": 0.05549436621885542, + "learning_rate": 8.392807054070738e-05, + "loss": 0.0015, + "step": 4190 + }, + { + "epoch": 2.360876897133221, + "grad_norm": 0.05844615472881089, + "learning_rate": 8.382498220683068e-05, + "loss": 0.0017, + "step": 4200 + }, + { + "epoch": 2.3664980326025855, + "grad_norm": 0.05785480991010937, + "learning_rate": 8.37216280709237e-05, + "loss": 0.0016, + "step": 4210 + }, + { + "epoch": 2.3721191680719507, + "grad_norm": 0.07481921771572901, + "learning_rate": 8.361800894515946e-05, + "loss": 0.0016, + "step": 4220 + }, + { + "epoch": 2.3777403035413154, + "grad_norm": 0.05822829929196074, + "learning_rate": 8.351412564379336e-05, + "loss": 0.0016, + "step": 4230 + }, + { + "epoch": 2.38336143901068, + "grad_norm": 0.05597231845386826, + "learning_rate": 8.340997898315671e-05, + "loss": 0.0017, + "step": 4240 + }, + { + "epoch": 2.388982574480045, + "grad_norm": 0.05518800111012908, + "learning_rate": 8.330556978165036e-05, + "loss": 0.0017, + "step": 4250 + }, + { + "epoch": 2.39460370994941, + "grad_norm": 0.06684935678037553, + "learning_rate": 8.320089885973826e-05, + "loss": 0.0018, + "step": 4260 + }, + { + "epoch": 2.4002248454187747, + "grad_norm": 0.072573592930173, + "learning_rate": 8.309596703994094e-05, + "loss": 0.0017, + "step": 4270 + }, + { + "epoch": 2.4058459808881394, + "grad_norm": 0.06709976923029225, + "learning_rate": 8.299077514682915e-05, + "loss": 0.0017, + "step": 4280 + }, + { + "epoch": 2.411467116357504, + "grad_norm": 0.06006598574157663, + "learning_rate": 8.288532400701735e-05, + "loss": 0.0018, + "step": 4290 + }, + { + "epoch": 2.4170882518268693, + "grad_norm": 0.06053465610650408, + "learning_rate": 8.277961444915716e-05, + "loss": 0.0017, + "step": 4300 + }, + { + "epoch": 2.422709387296234, + "grad_norm": 0.0633679769120732, + "learning_rate": 8.267364730393093e-05, + "loss": 0.0016, + "step": 4310 + }, + { + "epoch": 2.4283305227655987, + "grad_norm": 0.04905130321561388, + "learning_rate": 8.256742340404515e-05, + "loss": 0.0016, + "step": 4320 + }, + { + "epoch": 2.4339516582349634, + "grad_norm": 0.05084581543250133, + "learning_rate": 8.246094358422393e-05, + "loss": 0.0015, + "step": 4330 + }, + { + "epoch": 2.439572793704328, + "grad_norm": 0.06929941783453905, + "learning_rate": 8.235420868120245e-05, + "loss": 0.0015, + "step": 4340 + }, + { + "epoch": 2.4451939291736933, + "grad_norm": 0.0495981785149641, + "learning_rate": 8.224721953372035e-05, + "loss": 0.0015, + "step": 4350 + }, + { + "epoch": 2.450815064643058, + "grad_norm": 0.06154487277663696, + "learning_rate": 8.213997698251519e-05, + "loss": 0.0015, + "step": 4360 + }, + { + "epoch": 2.4564362001124227, + "grad_norm": 0.06702227730246357, + "learning_rate": 8.203248187031577e-05, + "loss": 0.0016, + "step": 4370 + }, + { + "epoch": 2.4620573355817874, + "grad_norm": 0.056017350947715615, + "learning_rate": 8.192473504183566e-05, + "loss": 0.0015, + "step": 4380 + }, + { + "epoch": 2.467678471051152, + "grad_norm": 0.11198849203086247, + "learning_rate": 8.181673734376634e-05, + "loss": 0.0016, + "step": 4390 + }, + { + "epoch": 2.4732996065205173, + "grad_norm": 0.06329875901984851, + "learning_rate": 8.170848962477068e-05, + "loss": 0.0018, + "step": 4400 + }, + { + "epoch": 2.478920741989882, + "grad_norm": 0.06331119073409751, + "learning_rate": 8.159999273547636e-05, + "loss": 0.0016, + "step": 4410 + }, + { + "epoch": 2.4845418774592467, + "grad_norm": 0.052348939039535504, + "learning_rate": 8.149124752846894e-05, + "loss": 0.0016, + "step": 4420 + }, + { + "epoch": 2.4901630129286114, + "grad_norm": 0.054010497916272666, + "learning_rate": 8.138225485828537e-05, + "loss": 0.0015, + "step": 4430 + }, + { + "epoch": 2.4957841483979766, + "grad_norm": 0.0654068401724483, + "learning_rate": 8.12730155814072e-05, + "loss": 0.0016, + "step": 4440 + }, + { + "epoch": 2.5014052838673413, + "grad_norm": 0.06774434911902472, + "learning_rate": 8.116353055625385e-05, + "loss": 0.0016, + "step": 4450 + }, + { + "epoch": 2.507026419336706, + "grad_norm": 0.058645050551856606, + "learning_rate": 8.105380064317587e-05, + "loss": 0.0015, + "step": 4460 + }, + { + "epoch": 2.5126475548060707, + "grad_norm": 0.06378242762268096, + "learning_rate": 8.094382670444817e-05, + "loss": 0.0014, + "step": 4470 + }, + { + "epoch": 2.518268690275436, + "grad_norm": 0.07035140206345245, + "learning_rate": 8.083360960426327e-05, + "loss": 0.0015, + "step": 4480 + }, + { + "epoch": 2.5238898257448006, + "grad_norm": 0.054138269993327354, + "learning_rate": 8.072315020872448e-05, + "loss": 0.0014, + "step": 4490 + }, + { + "epoch": 2.5295109612141653, + "grad_norm": 0.0520380194907875, + "learning_rate": 8.061244938583907e-05, + "loss": 0.0015, + "step": 4500 + }, + { + "epoch": 2.53513209668353, + "grad_norm": 0.0652285964745877, + "learning_rate": 8.050150800551155e-05, + "loss": 0.0014, + "step": 4510 + }, + { + "epoch": 2.5407532321528947, + "grad_norm": 0.06146797888465651, + "learning_rate": 8.039032693953672e-05, + "loss": 0.0015, + "step": 4520 + }, + { + "epoch": 2.54637436762226, + "grad_norm": 0.04413018291772843, + "learning_rate": 8.02789070615929e-05, + "loss": 0.0015, + "step": 4530 + }, + { + "epoch": 2.5519955030916246, + "grad_norm": 0.04939992283587218, + "learning_rate": 8.016724924723501e-05, + "loss": 0.0015, + "step": 4540 + }, + { + "epoch": 2.5576166385609893, + "grad_norm": 0.043663817829164675, + "learning_rate": 8.00553543738877e-05, + "loss": 0.0016, + "step": 4550 + }, + { + "epoch": 2.563237774030354, + "grad_norm": 0.054145213065889916, + "learning_rate": 7.994322332083852e-05, + "loss": 0.0016, + "step": 4560 + }, + { + "epoch": 2.5688589094997187, + "grad_norm": 0.05699514555987272, + "learning_rate": 7.983085696923087e-05, + "loss": 0.0019, + "step": 4570 + }, + { + "epoch": 2.574480044969084, + "grad_norm": 0.0640274235075573, + "learning_rate": 7.971825620205726e-05, + "loss": 0.0018, + "step": 4580 + }, + { + "epoch": 2.5801011804384486, + "grad_norm": 0.05404722214829411, + "learning_rate": 7.96054219041522e-05, + "loss": 0.0016, + "step": 4590 + }, + { + "epoch": 2.5857223159078133, + "grad_norm": 0.060989090707734825, + "learning_rate": 7.94923549621854e-05, + "loss": 0.0015, + "step": 4600 + }, + { + "epoch": 2.591343451377178, + "grad_norm": 0.05089975294543041, + "learning_rate": 7.937905626465458e-05, + "loss": 0.0015, + "step": 4610 + }, + { + "epoch": 2.5969645868465427, + "grad_norm": 0.047753743345937585, + "learning_rate": 7.926552670187881e-05, + "loss": 0.0014, + "step": 4620 + }, + { + "epoch": 2.602585722315908, + "grad_norm": 0.04387585711392661, + "learning_rate": 7.915176716599122e-05, + "loss": 0.0015, + "step": 4630 + }, + { + "epoch": 2.6082068577852726, + "grad_norm": 0.05324929700338459, + "learning_rate": 7.903777855093213e-05, + "loss": 0.0014, + "step": 4640 + }, + { + "epoch": 2.6138279932546373, + "grad_norm": 0.04210024748842374, + "learning_rate": 7.892356175244202e-05, + "loss": 0.0014, + "step": 4650 + }, + { + "epoch": 2.6194491287240025, + "grad_norm": 0.09552386039463091, + "learning_rate": 7.880911766805444e-05, + "loss": 0.0014, + "step": 4660 + }, + { + "epoch": 2.625070264193367, + "grad_norm": 0.059038210930734324, + "learning_rate": 7.869444719708903e-05, + "loss": 0.0013, + "step": 4670 + }, + { + "epoch": 2.630691399662732, + "grad_norm": 0.04816659963124649, + "learning_rate": 7.857955124064435e-05, + "loss": 0.0012, + "step": 4680 + }, + { + "epoch": 2.6363125351320966, + "grad_norm": 0.05442452499531946, + "learning_rate": 7.846443070159096e-05, + "loss": 0.0014, + "step": 4690 + }, + { + "epoch": 2.6419336706014613, + "grad_norm": 0.05419618414147171, + "learning_rate": 7.834908648456408e-05, + "loss": 0.0013, + "step": 4700 + }, + { + "epoch": 2.6475548060708265, + "grad_norm": 0.05046575142656625, + "learning_rate": 7.823351949595676e-05, + "loss": 0.0013, + "step": 4710 + }, + { + "epoch": 2.653175941540191, + "grad_norm": 0.046485397857171505, + "learning_rate": 7.811773064391257e-05, + "loss": 0.0013, + "step": 4720 + }, + { + "epoch": 2.658797077009556, + "grad_norm": 0.043494442406102805, + "learning_rate": 7.800172083831849e-05, + "loss": 0.0012, + "step": 4730 + }, + { + "epoch": 2.6644182124789206, + "grad_norm": 0.05197640361791803, + "learning_rate": 7.788549099079787e-05, + "loss": 0.0013, + "step": 4740 + }, + { + "epoch": 2.6700393479482853, + "grad_norm": 0.050588815857194186, + "learning_rate": 7.77690420147031e-05, + "loss": 0.0013, + "step": 4750 + }, + { + "epoch": 2.6756604834176505, + "grad_norm": 0.055277287421162696, + "learning_rate": 7.765237482510857e-05, + "loss": 0.0012, + "step": 4760 + }, + { + "epoch": 2.681281618887015, + "grad_norm": 0.047387945885664126, + "learning_rate": 7.753549033880336e-05, + "loss": 0.0012, + "step": 4770 + }, + { + "epoch": 2.68690275435638, + "grad_norm": 0.055314706331344304, + "learning_rate": 7.74183894742842e-05, + "loss": 0.0013, + "step": 4780 + }, + { + "epoch": 2.6925238898257446, + "grad_norm": 0.054991171970991085, + "learning_rate": 7.730107315174808e-05, + "loss": 0.0015, + "step": 4790 + }, + { + "epoch": 2.6981450252951094, + "grad_norm": 0.07142443615869051, + "learning_rate": 7.718354229308511e-05, + "loss": 0.0014, + "step": 4800 + }, + { + "epoch": 2.7037661607644745, + "grad_norm": 0.06880613464299018, + "learning_rate": 7.706579782187127e-05, + "loss": 0.0014, + "step": 4810 + }, + { + "epoch": 2.7093872962338392, + "grad_norm": 0.10332828544462125, + "learning_rate": 7.694784066336114e-05, + "loss": 0.0015, + "step": 4820 + }, + { + "epoch": 2.715008431703204, + "grad_norm": 0.06452511979597532, + "learning_rate": 7.68296717444806e-05, + "loss": 0.0015, + "step": 4830 + }, + { + "epoch": 2.720629567172569, + "grad_norm": 0.05595821471841563, + "learning_rate": 7.671129199381963e-05, + "loss": 0.0015, + "step": 4840 + }, + { + "epoch": 2.726250702641934, + "grad_norm": 0.04384518983020577, + "learning_rate": 7.659270234162491e-05, + "loss": 0.0015, + "step": 4850 + }, + { + "epoch": 2.7318718381112985, + "grad_norm": 0.04831975704039876, + "learning_rate": 7.647390371979257e-05, + "loss": 0.0014, + "step": 4860 + }, + { + "epoch": 2.7374929735806632, + "grad_norm": 0.059347623006389266, + "learning_rate": 7.635489706186092e-05, + "loss": 0.0012, + "step": 4870 + }, + { + "epoch": 2.743114109050028, + "grad_norm": 0.05694220115635119, + "learning_rate": 7.623568330300294e-05, + "loss": 0.0013, + "step": 4880 + }, + { + "epoch": 2.748735244519393, + "grad_norm": 0.0468828291912844, + "learning_rate": 7.611626338001912e-05, + "loss": 0.0013, + "step": 4890 + }, + { + "epoch": 2.754356379988758, + "grad_norm": 0.04787035904833046, + "learning_rate": 7.599663823132998e-05, + "loss": 0.0014, + "step": 4900 + }, + { + "epoch": 2.7599775154581225, + "grad_norm": 0.05657663969108575, + "learning_rate": 7.587680879696876e-05, + "loss": 0.0013, + "step": 4910 + }, + { + "epoch": 2.7655986509274872, + "grad_norm": 0.04466607117951967, + "learning_rate": 7.575677601857402e-05, + "loss": 0.0013, + "step": 4920 + }, + { + "epoch": 2.771219786396852, + "grad_norm": 0.06027260612305898, + "learning_rate": 7.563654083938221e-05, + "loss": 0.0013, + "step": 4930 + }, + { + "epoch": 2.776840921866217, + "grad_norm": 0.05081088548969293, + "learning_rate": 7.551610420422026e-05, + "loss": 0.0013, + "step": 4940 + }, + { + "epoch": 2.782462057335582, + "grad_norm": 0.05428661275304198, + "learning_rate": 7.53954670594982e-05, + "loss": 0.0012, + "step": 4950 + }, + { + "epoch": 2.7880831928049465, + "grad_norm": 0.0482700567695289, + "learning_rate": 7.52746303532017e-05, + "loss": 0.0012, + "step": 4960 + }, + { + "epoch": 2.7937043282743113, + "grad_norm": 0.04707913457424893, + "learning_rate": 7.515359503488457e-05, + "loss": 0.0012, + "step": 4970 + }, + { + "epoch": 2.799325463743676, + "grad_norm": 0.06312636376317708, + "learning_rate": 7.503236205566142e-05, + "loss": 0.0011, + "step": 4980 + }, + { + "epoch": 2.804946599213041, + "grad_norm": 0.039101096639687, + "learning_rate": 7.491093236820003e-05, + "loss": 0.0011, + "step": 4990 + }, + { + "epoch": 2.810567734682406, + "grad_norm": 0.04203545515495065, + "learning_rate": 7.478930692671403e-05, + "loss": 0.0013, + "step": 5000 + }, + { + "epoch": 2.8161888701517706, + "grad_norm": 0.05994425738017975, + "learning_rate": 7.466748668695522e-05, + "loss": 0.0012, + "step": 5010 + }, + { + "epoch": 2.8218100056211357, + "grad_norm": 0.053168584599147396, + "learning_rate": 7.454547260620623e-05, + "loss": 0.0013, + "step": 5020 + }, + { + "epoch": 2.8274311410905004, + "grad_norm": 0.050877693867899004, + "learning_rate": 7.442326564327291e-05, + "loss": 0.0013, + "step": 5030 + }, + { + "epoch": 2.833052276559865, + "grad_norm": 0.06380365444654545, + "learning_rate": 7.430086675847676e-05, + "loss": 0.0012, + "step": 5040 + }, + { + "epoch": 2.83867341202923, + "grad_norm": 0.04679933950214477, + "learning_rate": 7.417827691364751e-05, + "loss": 0.0011, + "step": 5050 + }, + { + "epoch": 2.8442945474985946, + "grad_norm": 0.04377109248658636, + "learning_rate": 7.40554970721154e-05, + "loss": 0.0012, + "step": 5060 + }, + { + "epoch": 2.8499156829679597, + "grad_norm": 0.0608311362918617, + "learning_rate": 7.393252819870375e-05, + "loss": 0.0014, + "step": 5070 + }, + { + "epoch": 2.8555368184373244, + "grad_norm": 0.0943118819372207, + "learning_rate": 7.380937125972133e-05, + "loss": 0.0014, + "step": 5080 + }, + { + "epoch": 2.861157953906689, + "grad_norm": 0.0403091749249693, + "learning_rate": 7.368602722295471e-05, + "loss": 0.0013, + "step": 5090 + }, + { + "epoch": 2.866779089376054, + "grad_norm": 0.055971258310376075, + "learning_rate": 7.356249705766075e-05, + "loss": 0.0013, + "step": 5100 + }, + { + "epoch": 2.8724002248454186, + "grad_norm": 0.08519796434852828, + "learning_rate": 7.34387817345589e-05, + "loss": 0.0013, + "step": 5110 + }, + { + "epoch": 2.8780213603147837, + "grad_norm": 0.06853625802370912, + "learning_rate": 7.331488222582364e-05, + "loss": 0.0012, + "step": 5120 + }, + { + "epoch": 2.8836424957841484, + "grad_norm": 0.04218234679052211, + "learning_rate": 7.319079950507682e-05, + "loss": 0.0013, + "step": 5130 + }, + { + "epoch": 2.889263631253513, + "grad_norm": 0.03601990309692009, + "learning_rate": 7.306653454737992e-05, + "loss": 0.0012, + "step": 5140 + }, + { + "epoch": 2.894884766722878, + "grad_norm": 0.05970290400064692, + "learning_rate": 7.294208832922655e-05, + "loss": 0.0011, + "step": 5150 + }, + { + "epoch": 2.9005059021922426, + "grad_norm": 0.06211859449356491, + "learning_rate": 7.281746182853469e-05, + "loss": 0.0017, + "step": 5160 + }, + { + "epoch": 2.9061270376616077, + "grad_norm": 0.04244606481235483, + "learning_rate": 7.269265602463895e-05, + "loss": 0.0013, + "step": 5170 + }, + { + "epoch": 2.9117481731309725, + "grad_norm": 0.04796952285255335, + "learning_rate": 7.256767189828297e-05, + "loss": 0.0012, + "step": 5180 + }, + { + "epoch": 2.917369308600337, + "grad_norm": 0.04891082555346954, + "learning_rate": 7.24425104316117e-05, + "loss": 0.0012, + "step": 5190 + }, + { + "epoch": 2.9229904440697023, + "grad_norm": 0.04272228076487222, + "learning_rate": 7.231717260816361e-05, + "loss": 0.0011, + "step": 5200 + }, + { + "epoch": 2.928611579539067, + "grad_norm": 0.04454348084467995, + "learning_rate": 7.219165941286303e-05, + "loss": 0.0012, + "step": 5210 + }, + { + "epoch": 2.9342327150084317, + "grad_norm": 0.05868816541252896, + "learning_rate": 7.20659718320124e-05, + "loss": 0.0012, + "step": 5220 + }, + { + "epoch": 2.9398538504777965, + "grad_norm": 0.052032777228837916, + "learning_rate": 7.194011085328448e-05, + "loss": 0.0013, + "step": 5230 + }, + { + "epoch": 2.945474985947161, + "grad_norm": 0.049334668563237004, + "learning_rate": 7.181407746571466e-05, + "loss": 0.0011, + "step": 5240 + }, + { + "epoch": 2.9510961214165263, + "grad_norm": 0.05111824349302509, + "learning_rate": 7.16878726596931e-05, + "loss": 0.0012, + "step": 5250 + }, + { + "epoch": 2.956717256885891, + "grad_norm": 0.047943356040951235, + "learning_rate": 7.156149742695703e-05, + "loss": 0.0012, + "step": 5260 + }, + { + "epoch": 2.9623383923552558, + "grad_norm": 0.04666570165091871, + "learning_rate": 7.143495276058292e-05, + "loss": 0.0013, + "step": 5270 + }, + { + "epoch": 2.9679595278246205, + "grad_norm": 0.04633872778424716, + "learning_rate": 7.130823965497863e-05, + "loss": 0.0012, + "step": 5280 + }, + { + "epoch": 2.973580663293985, + "grad_norm": 0.04728600175386692, + "learning_rate": 7.11813591058757e-05, + "loss": 0.001, + "step": 5290 + }, + { + "epoch": 2.9792017987633503, + "grad_norm": 0.04331703035403907, + "learning_rate": 7.105431211032142e-05, + "loss": 0.001, + "step": 5300 + }, + { + "epoch": 2.984822934232715, + "grad_norm": 0.0411566396160935, + "learning_rate": 7.092709966667107e-05, + "loss": 0.001, + "step": 5310 + }, + { + "epoch": 2.9904440697020798, + "grad_norm": 0.045270747800133036, + "learning_rate": 7.079972277458007e-05, + "loss": 0.0011, + "step": 5320 + }, + { + "epoch": 2.9960652051714445, + "grad_norm": 0.03583276636219469, + "learning_rate": 7.0672182434996e-05, + "loss": 0.0009, + "step": 5330 + }, + { + "epoch": 3.0016863406408096, + "grad_norm": 0.049772188634408326, + "learning_rate": 7.054447965015099e-05, + "loss": 0.001, + "step": 5340 + }, + { + "epoch": 3.0073074761101743, + "grad_norm": 0.04417115124033103, + "learning_rate": 7.041661542355356e-05, + "loss": 0.0008, + "step": 5350 + }, + { + "epoch": 3.012928611579539, + "grad_norm": 0.04016838015205063, + "learning_rate": 7.028859075998093e-05, + "loss": 0.0007, + "step": 5360 + }, + { + "epoch": 3.0185497470489038, + "grad_norm": 0.03429961640956788, + "learning_rate": 7.016040666547105e-05, + "loss": 0.0008, + "step": 5370 + }, + { + "epoch": 3.0241708825182685, + "grad_norm": 0.03891109220250057, + "learning_rate": 7.003206414731467e-05, + "loss": 0.0007, + "step": 5380 + }, + { + "epoch": 3.0297920179876336, + "grad_norm": 0.04617952784664274, + "learning_rate": 6.990356421404751e-05, + "loss": 0.0008, + "step": 5390 + }, + { + "epoch": 3.0354131534569984, + "grad_norm": 0.04730962147871257, + "learning_rate": 6.977490787544224e-05, + "loss": 0.0008, + "step": 5400 + }, + { + "epoch": 3.041034288926363, + "grad_norm": 0.03603163741293914, + "learning_rate": 6.964609614250062e-05, + "loss": 0.0007, + "step": 5410 + }, + { + "epoch": 3.046655424395728, + "grad_norm": 0.0530910315560429, + "learning_rate": 6.95171300274455e-05, + "loss": 0.0007, + "step": 5420 + }, + { + "epoch": 3.052276559865093, + "grad_norm": 0.1931933934780385, + "learning_rate": 6.938801054371288e-05, + "loss": 0.0041, + "step": 5430 + }, + { + "epoch": 3.0578976953344577, + "grad_norm": 0.06655505125657529, + "learning_rate": 6.9258738705944e-05, + "loss": 0.0021, + "step": 5440 + }, + { + "epoch": 3.0635188308038224, + "grad_norm": 0.05910110421982745, + "learning_rate": 6.912931552997726e-05, + "loss": 0.0012, + "step": 5450 + }, + { + "epoch": 3.069139966273187, + "grad_norm": 0.041328431110691684, + "learning_rate": 6.899974203284036e-05, + "loss": 0.001, + "step": 5460 + }, + { + "epoch": 3.074761101742552, + "grad_norm": 0.06500092697262157, + "learning_rate": 6.887001923274224e-05, + "loss": 0.001, + "step": 5470 + }, + { + "epoch": 3.080382237211917, + "grad_norm": 0.034993472788750396, + "learning_rate": 6.874014814906499e-05, + "loss": 0.0008, + "step": 5480 + }, + { + "epoch": 3.0860033726812817, + "grad_norm": 0.03408391423635107, + "learning_rate": 6.861012980235609e-05, + "loss": 0.0008, + "step": 5490 + }, + { + "epoch": 3.0916245081506464, + "grad_norm": 0.04550774505878738, + "learning_rate": 6.847996521432014e-05, + "loss": 0.0008, + "step": 5500 + }, + { + "epoch": 3.097245643620011, + "grad_norm": 0.03383144826061609, + "learning_rate": 6.834965540781091e-05, + "loss": 0.0007, + "step": 5510 + }, + { + "epoch": 3.1028667790893762, + "grad_norm": 0.03491993329718346, + "learning_rate": 6.821920140682341e-05, + "loss": 0.0008, + "step": 5520 + }, + { + "epoch": 3.108487914558741, + "grad_norm": 0.030436033090870892, + "learning_rate": 6.808860423648564e-05, + "loss": 0.0008, + "step": 5530 + }, + { + "epoch": 3.1141090500281057, + "grad_norm": 0.027596654141064453, + "learning_rate": 6.795786492305076e-05, + "loss": 0.0008, + "step": 5540 + }, + { + "epoch": 3.1197301854974704, + "grad_norm": 0.029799149160684817, + "learning_rate": 6.782698449388882e-05, + "loss": 0.0008, + "step": 5550 + }, + { + "epoch": 3.125351320966835, + "grad_norm": 0.043563499361036456, + "learning_rate": 6.769596397747882e-05, + "loss": 0.0007, + "step": 5560 + }, + { + "epoch": 3.1309724564362003, + "grad_norm": 0.04096658745685941, + "learning_rate": 6.756480440340061e-05, + "loss": 0.0008, + "step": 5570 + }, + { + "epoch": 3.136593591905565, + "grad_norm": 0.035312765105403834, + "learning_rate": 6.743350680232672e-05, + "loss": 0.0008, + "step": 5580 + }, + { + "epoch": 3.1422147273749297, + "grad_norm": 0.03769833433929328, + "learning_rate": 6.730207220601437e-05, + "loss": 0.0007, + "step": 5590 + }, + { + "epoch": 3.1478358628442944, + "grad_norm": 0.03561680116741662, + "learning_rate": 6.717050164729732e-05, + "loss": 0.0006, + "step": 5600 + }, + { + "epoch": 3.1534569983136596, + "grad_norm": 0.03702862585894474, + "learning_rate": 6.703879616007765e-05, + "loss": 0.0007, + "step": 5610 + }, + { + "epoch": 3.1590781337830243, + "grad_norm": 0.044959795997499, + "learning_rate": 6.690695677931784e-05, + "loss": 0.0007, + "step": 5620 + }, + { + "epoch": 3.164699269252389, + "grad_norm": 0.05316506407712143, + "learning_rate": 6.677498454103246e-05, + "loss": 0.0007, + "step": 5630 + }, + { + "epoch": 3.1703204047217537, + "grad_norm": 0.037153559869762896, + "learning_rate": 6.664288048228012e-05, + "loss": 0.0008, + "step": 5640 + }, + { + "epoch": 3.1759415401911184, + "grad_norm": 0.03522308312685447, + "learning_rate": 6.651064564115531e-05, + "loss": 0.0007, + "step": 5650 + }, + { + "epoch": 3.1815626756604836, + "grad_norm": 0.035498055293177955, + "learning_rate": 6.637828105678016e-05, + "loss": 0.0007, + "step": 5660 + }, + { + "epoch": 3.1871838111298483, + "grad_norm": 0.04849511803110099, + "learning_rate": 6.624578776929646e-05, + "loss": 0.0008, + "step": 5670 + }, + { + "epoch": 3.192804946599213, + "grad_norm": 0.04529042982293002, + "learning_rate": 6.611316681985727e-05, + "loss": 0.0007, + "step": 5680 + }, + { + "epoch": 3.1984260820685777, + "grad_norm": 0.03406066309616036, + "learning_rate": 6.598041925061884e-05, + "loss": 0.0007, + "step": 5690 + }, + { + "epoch": 3.204047217537943, + "grad_norm": 0.06054225227012086, + "learning_rate": 6.584754610473249e-05, + "loss": 0.0007, + "step": 5700 + }, + { + "epoch": 3.2096683530073076, + "grad_norm": 0.03991536097817506, + "learning_rate": 6.571454842633629e-05, + "loss": 0.0007, + "step": 5710 + }, + { + "epoch": 3.2152894884766723, + "grad_norm": 0.024707780489681547, + "learning_rate": 6.55814272605469e-05, + "loss": 0.0007, + "step": 5720 + }, + { + "epoch": 3.220910623946037, + "grad_norm": 0.039996092561845845, + "learning_rate": 6.54481836534514e-05, + "loss": 0.0006, + "step": 5730 + }, + { + "epoch": 3.2265317594154017, + "grad_norm": 0.030295813415196538, + "learning_rate": 6.531481865209898e-05, + "loss": 0.0006, + "step": 5740 + }, + { + "epoch": 3.232152894884767, + "grad_norm": 0.046703107162872695, + "learning_rate": 6.518133330449285e-05, + "loss": 0.0007, + "step": 5750 + }, + { + "epoch": 3.2377740303541316, + "grad_norm": 0.03962853217344156, + "learning_rate": 6.504772865958183e-05, + "loss": 0.0007, + "step": 5760 + }, + { + "epoch": 3.2433951658234963, + "grad_norm": 0.037683309509755764, + "learning_rate": 6.491400576725222e-05, + "loss": 0.0006, + "step": 5770 + }, + { + "epoch": 3.249016301292861, + "grad_norm": 0.04406593602094287, + "learning_rate": 6.478016567831957e-05, + "loss": 0.0007, + "step": 5780 + }, + { + "epoch": 3.254637436762226, + "grad_norm": 0.03686328673212962, + "learning_rate": 6.464620944452034e-05, + "loss": 0.0007, + "step": 5790 + }, + { + "epoch": 3.260258572231591, + "grad_norm": 0.03081692559338337, + "learning_rate": 6.451213811850368e-05, + "loss": 0.0006, + "step": 5800 + }, + { + "epoch": 3.2658797077009556, + "grad_norm": 0.05702037792999332, + "learning_rate": 6.437795275382314e-05, + "loss": 0.0007, + "step": 5810 + }, + { + "epoch": 3.2715008431703203, + "grad_norm": 0.030331479613686176, + "learning_rate": 6.424365440492845e-05, + "loss": 0.0006, + "step": 5820 + }, + { + "epoch": 3.277121978639685, + "grad_norm": 0.019931759430157396, + "learning_rate": 6.410924412715716e-05, + "loss": 0.0007, + "step": 5830 + }, + { + "epoch": 3.28274311410905, + "grad_norm": 0.0416836370295991, + "learning_rate": 6.397472297672636e-05, + "loss": 0.0007, + "step": 5840 + }, + { + "epoch": 3.288364249578415, + "grad_norm": 0.04291211828866574, + "learning_rate": 6.384009201072441e-05, + "loss": 0.0007, + "step": 5850 + }, + { + "epoch": 3.2939853850477796, + "grad_norm": 0.08999379021592732, + "learning_rate": 6.370535228710262e-05, + "loss": 0.0012, + "step": 5860 + }, + { + "epoch": 3.2996065205171443, + "grad_norm": 0.04144860128233685, + "learning_rate": 6.357050486466691e-05, + "loss": 0.0008, + "step": 5870 + }, + { + "epoch": 3.305227655986509, + "grad_norm": 0.026993119294154346, + "learning_rate": 6.343555080306957e-05, + "loss": 0.0007, + "step": 5880 + }, + { + "epoch": 3.310848791455874, + "grad_norm": 0.04100005257492785, + "learning_rate": 6.330049116280081e-05, + "loss": 0.0006, + "step": 5890 + }, + { + "epoch": 3.316469926925239, + "grad_norm": 0.03402996703596254, + "learning_rate": 6.316532700518054e-05, + "loss": 0.0006, + "step": 5900 + }, + { + "epoch": 3.3220910623946036, + "grad_norm": 0.039306464782034437, + "learning_rate": 6.303005939234998e-05, + "loss": 0.0007, + "step": 5910 + }, + { + "epoch": 3.3277121978639688, + "grad_norm": 0.03156560284235597, + "learning_rate": 6.28946893872633e-05, + "loss": 0.0006, + "step": 5920 + }, + { + "epoch": 3.3333333333333335, + "grad_norm": 0.04328187887722745, + "learning_rate": 6.275921805367928e-05, + "loss": 0.0006, + "step": 5930 + }, + { + "epoch": 3.338954468802698, + "grad_norm": 0.023736302261381972, + "learning_rate": 6.262364645615297e-05, + "loss": 0.0006, + "step": 5940 + }, + { + "epoch": 3.344575604272063, + "grad_norm": 0.03407885997096527, + "learning_rate": 6.248797566002732e-05, + "loss": 0.0006, + "step": 5950 + }, + { + "epoch": 3.3501967397414276, + "grad_norm": 0.038738114932414844, + "learning_rate": 6.235220673142477e-05, + "loss": 0.0006, + "step": 5960 + }, + { + "epoch": 3.3558178752107928, + "grad_norm": 0.046977880490892035, + "learning_rate": 6.221634073723894e-05, + "loss": 0.0007, + "step": 5970 + }, + { + "epoch": 3.3614390106801575, + "grad_norm": 0.03952714211660955, + "learning_rate": 6.208037874512614e-05, + "loss": 0.0006, + "step": 5980 + }, + { + "epoch": 3.367060146149522, + "grad_norm": 0.07702445396642703, + "learning_rate": 6.194432182349712e-05, + "loss": 0.0007, + "step": 5990 + }, + { + "epoch": 3.372681281618887, + "grad_norm": 0.0428731028419256, + "learning_rate": 6.180817104150855e-05, + "loss": 0.0007, + "step": 6000 + }, + { + "epoch": 3.3783024170882516, + "grad_norm": 0.03421296119870643, + "learning_rate": 6.167192746905472e-05, + "loss": 0.0009, + "step": 6010 + }, + { + "epoch": 3.383923552557617, + "grad_norm": 0.03773661861403026, + "learning_rate": 6.153559217675903e-05, + "loss": 0.0007, + "step": 6020 + }, + { + "epoch": 3.3895446880269815, + "grad_norm": 0.04853637866330848, + "learning_rate": 6.139916623596564e-05, + "loss": 0.0007, + "step": 6030 + }, + { + "epoch": 3.395165823496346, + "grad_norm": 0.03360065772987571, + "learning_rate": 6.126265071873108e-05, + "loss": 0.0007, + "step": 6040 + }, + { + "epoch": 3.400786958965711, + "grad_norm": 0.04399350597808129, + "learning_rate": 6.112604669781572e-05, + "loss": 0.0007, + "step": 6050 + }, + { + "epoch": 3.4064080944350756, + "grad_norm": 0.048232748136132014, + "learning_rate": 6.0989355246675475e-05, + "loss": 0.0007, + "step": 6060 + }, + { + "epoch": 3.412029229904441, + "grad_norm": 0.04624519302109482, + "learning_rate": 6.0852577439453254e-05, + "loss": 0.0007, + "step": 6070 + }, + { + "epoch": 3.4176503653738055, + "grad_norm": 0.051780902316214385, + "learning_rate": 6.0715714350970565e-05, + "loss": 0.0008, + "step": 6080 + }, + { + "epoch": 3.4232715008431702, + "grad_norm": 0.041918827401852814, + "learning_rate": 6.057876705671911e-05, + "loss": 0.0007, + "step": 6090 + }, + { + "epoch": 3.4288926363125354, + "grad_norm": 0.038383410348652475, + "learning_rate": 6.044173663285224e-05, + "loss": 0.0007, + "step": 6100 + }, + { + "epoch": 3.4345137717819, + "grad_norm": 0.04090534757383718, + "learning_rate": 6.03046241561766e-05, + "loss": 0.0007, + "step": 6110 + }, + { + "epoch": 3.440134907251265, + "grad_norm": 0.039848883792442086, + "learning_rate": 6.01674307041436e-05, + "loss": 0.0006, + "step": 6120 + }, + { + "epoch": 3.4457560427206295, + "grad_norm": 0.05882632590382963, + "learning_rate": 6.003015735484094e-05, + "loss": 0.0007, + "step": 6130 + }, + { + "epoch": 3.4513771781899942, + "grad_norm": 0.03761834122685572, + "learning_rate": 5.9892805186984214e-05, + "loss": 0.0007, + "step": 6140 + }, + { + "epoch": 3.4569983136593594, + "grad_norm": 0.04310262197686143, + "learning_rate": 5.9755375279908345e-05, + "loss": 0.0006, + "step": 6150 + }, + { + "epoch": 3.462619449128724, + "grad_norm": 0.03732240584723843, + "learning_rate": 5.961786871355915e-05, + "loss": 0.0006, + "step": 6160 + }, + { + "epoch": 3.468240584598089, + "grad_norm": 0.03720731608238794, + "learning_rate": 5.948028656848489e-05, + "loss": 0.0006, + "step": 6170 + }, + { + "epoch": 3.4738617200674535, + "grad_norm": 0.02766017890535317, + "learning_rate": 5.934262992582766e-05, + "loss": 0.0006, + "step": 6180 + }, + { + "epoch": 3.4794828555368182, + "grad_norm": 0.03351173455826141, + "learning_rate": 5.9204899867315046e-05, + "loss": 0.0006, + "step": 6190 + }, + { + "epoch": 3.4851039910061834, + "grad_norm": 0.036377311347947, + "learning_rate": 5.9067097475251485e-05, + "loss": 0.0006, + "step": 6200 + }, + { + "epoch": 3.490725126475548, + "grad_norm": 0.05687008128806063, + "learning_rate": 5.892922383250985e-05, + "loss": 0.0006, + "step": 6210 + }, + { + "epoch": 3.496346261944913, + "grad_norm": 0.029548269683333687, + "learning_rate": 5.8791280022522924e-05, + "loss": 0.0005, + "step": 6220 + }, + { + "epoch": 3.5019673974142775, + "grad_norm": 0.04429407677490239, + "learning_rate": 5.865326712927484e-05, + "loss": 0.0006, + "step": 6230 + }, + { + "epoch": 3.5075885328836423, + "grad_norm": 0.043026212285912684, + "learning_rate": 5.851518623729264e-05, + "loss": 0.0006, + "step": 6240 + }, + { + "epoch": 3.5132096683530074, + "grad_norm": 0.03660719492014762, + "learning_rate": 5.83770384316377e-05, + "loss": 0.0006, + "step": 6250 + }, + { + "epoch": 3.518830803822372, + "grad_norm": 0.04859733678008268, + "learning_rate": 5.823882479789717e-05, + "loss": 0.0006, + "step": 6260 + }, + { + "epoch": 3.524451939291737, + "grad_norm": 0.04160837330988752, + "learning_rate": 5.810054642217556e-05, + "loss": 0.0006, + "step": 6270 + }, + { + "epoch": 3.530073074761102, + "grad_norm": 0.037355611112239645, + "learning_rate": 5.7962204391086086e-05, + "loss": 0.0006, + "step": 6280 + }, + { + "epoch": 3.5356942102304667, + "grad_norm": 0.04616794076981165, + "learning_rate": 5.7823799791742164e-05, + "loss": 0.0006, + "step": 6290 + }, + { + "epoch": 3.5413153456998314, + "grad_norm": 0.037466830670271706, + "learning_rate": 5.768533371174892e-05, + "loss": 0.0006, + "step": 6300 + }, + { + "epoch": 3.546936481169196, + "grad_norm": 0.03575800311749461, + "learning_rate": 5.75468072391946e-05, + "loss": 0.0006, + "step": 6310 + }, + { + "epoch": 3.552557616638561, + "grad_norm": 0.039153566708862844, + "learning_rate": 5.7408221462642e-05, + "loss": 0.0006, + "step": 6320 + }, + { + "epoch": 3.558178752107926, + "grad_norm": 0.031531672897527285, + "learning_rate": 5.7269577471119964e-05, + "loss": 0.0005, + "step": 6330 + }, + { + "epoch": 3.5637998875772907, + "grad_norm": 0.04597524128573998, + "learning_rate": 5.713087635411475e-05, + "loss": 0.0005, + "step": 6340 + }, + { + "epoch": 3.5694210230466554, + "grad_norm": 0.036368453070765874, + "learning_rate": 5.699211920156158e-05, + "loss": 0.0006, + "step": 6350 + }, + { + "epoch": 3.57504215851602, + "grad_norm": 0.06259482931042772, + "learning_rate": 5.6853307103835986e-05, + "loss": 0.0008, + "step": 6360 + }, + { + "epoch": 3.580663293985385, + "grad_norm": 0.03688253114136713, + "learning_rate": 5.671444115174525e-05, + "loss": 0.0007, + "step": 6370 + }, + { + "epoch": 3.58628442945475, + "grad_norm": 0.039443958201916825, + "learning_rate": 5.657552243651988e-05, + "loss": 0.0007, + "step": 6380 + }, + { + "epoch": 3.5919055649241147, + "grad_norm": 0.05006353041003688, + "learning_rate": 5.643655204980496e-05, + "loss": 0.0007, + "step": 6390 + }, + { + "epoch": 3.5975267003934794, + "grad_norm": 0.032151940329149434, + "learning_rate": 5.629753108365168e-05, + "loss": 0.0007, + "step": 6400 + }, + { + "epoch": 3.603147835862844, + "grad_norm": 0.04630562022563597, + "learning_rate": 5.6158460630508644e-05, + "loss": 0.0006, + "step": 6410 + }, + { + "epoch": 3.608768971332209, + "grad_norm": 0.03487301295105969, + "learning_rate": 5.6019341783213344e-05, + "loss": 0.0006, + "step": 6420 + }, + { + "epoch": 3.614390106801574, + "grad_norm": 0.04439396337405047, + "learning_rate": 5.588017563498357e-05, + "loss": 0.0006, + "step": 6430 + }, + { + "epoch": 3.6200112422709387, + "grad_norm": 0.04938064167191991, + "learning_rate": 5.5740963279408786e-05, + "loss": 0.0005, + "step": 6440 + }, + { + "epoch": 3.6256323777403034, + "grad_norm": 0.029652798637612207, + "learning_rate": 5.5601705810441605e-05, + "loss": 0.0006, + "step": 6450 + }, + { + "epoch": 3.6312535132096686, + "grad_norm": 0.0460551768203557, + "learning_rate": 5.546240432238912e-05, + "loss": 0.0007, + "step": 6460 + }, + { + "epoch": 3.6368746486790333, + "grad_norm": 0.03638213764726579, + "learning_rate": 5.5323059909904286e-05, + "loss": 0.0006, + "step": 6470 + }, + { + "epoch": 3.642495784148398, + "grad_norm": 0.02818147464170146, + "learning_rate": 5.5183673667977485e-05, + "loss": 0.0006, + "step": 6480 + }, + { + "epoch": 3.6481169196177627, + "grad_norm": 0.07028405895811617, + "learning_rate": 5.504424669192766e-05, + "loss": 0.0005, + "step": 6490 + }, + { + "epoch": 3.6537380550871275, + "grad_norm": 0.04710056052143868, + "learning_rate": 5.490478007739397e-05, + "loss": 0.0006, + "step": 6500 + }, + { + "epoch": 3.6593591905564926, + "grad_norm": 0.04233353011790259, + "learning_rate": 5.476527492032699e-05, + "loss": 0.0009, + "step": 6510 + }, + { + "epoch": 3.6649803260258573, + "grad_norm": 0.040515386163587984, + "learning_rate": 5.462573231698018e-05, + "loss": 0.001, + "step": 6520 + }, + { + "epoch": 3.670601461495222, + "grad_norm": 0.045421912466151994, + "learning_rate": 5.4486153363901283e-05, + "loss": 0.0009, + "step": 6530 + }, + { + "epoch": 3.6762225969645868, + "grad_norm": 0.04643960314747107, + "learning_rate": 5.434653915792364e-05, + "loss": 0.001, + "step": 6540 + }, + { + "epoch": 3.6818437324339515, + "grad_norm": 0.028825835758280453, + "learning_rate": 5.420689079615764e-05, + "loss": 0.0007, + "step": 6550 + }, + { + "epoch": 3.6874648679033166, + "grad_norm": 0.03952904479542909, + "learning_rate": 5.406720937598212e-05, + "loss": 0.0007, + "step": 6560 + }, + { + "epoch": 3.6930860033726813, + "grad_norm": 0.02396144669457461, + "learning_rate": 5.3927495995035585e-05, + "loss": 0.0007, + "step": 6570 + }, + { + "epoch": 3.698707138842046, + "grad_norm": 0.03481894355127607, + "learning_rate": 5.378775175120782e-05, + "loss": 0.0006, + "step": 6580 + }, + { + "epoch": 3.7043282743114108, + "grad_norm": 0.03865436978734628, + "learning_rate": 5.3647977742631016e-05, + "loss": 0.0005, + "step": 6590 + }, + { + "epoch": 3.7099494097807755, + "grad_norm": 0.027451562875999622, + "learning_rate": 5.350817506767135e-05, + "loss": 0.0006, + "step": 6600 + }, + { + "epoch": 3.7155705452501406, + "grad_norm": 0.035603323472503834, + "learning_rate": 5.336834482492023e-05, + "loss": 0.0006, + "step": 6610 + }, + { + "epoch": 3.7211916807195053, + "grad_norm": 0.03906213829811842, + "learning_rate": 5.322848811318568e-05, + "loss": 0.0005, + "step": 6620 + }, + { + "epoch": 3.72681281618887, + "grad_norm": 0.04311732491532985, + "learning_rate": 5.3088606031483756e-05, + "loss": 0.0008, + "step": 6630 + }, + { + "epoch": 3.732433951658235, + "grad_norm": 0.04769978225065517, + "learning_rate": 5.2948699679029845e-05, + "loss": 0.0008, + "step": 6640 + }, + { + "epoch": 3.7380550871276, + "grad_norm": 0.03559861111356046, + "learning_rate": 5.280877015523007e-05, + "loss": 0.0007, + "step": 6650 + }, + { + "epoch": 3.7436762225969646, + "grad_norm": 0.08570526963619465, + "learning_rate": 5.266881855967265e-05, + "loss": 0.001, + "step": 6660 + }, + { + "epoch": 3.7492973580663294, + "grad_norm": 0.03292192914921317, + "learning_rate": 5.2528845992119216e-05, + "loss": 0.0009, + "step": 6670 + }, + { + "epoch": 3.754918493535694, + "grad_norm": 0.03860199621128338, + "learning_rate": 5.2388853552496234e-05, + "loss": 0.0006, + "step": 6680 + }, + { + "epoch": 3.7605396290050592, + "grad_norm": 0.04102952846412716, + "learning_rate": 5.2248842340886304e-05, + "loss": 0.0006, + "step": 6690 + }, + { + "epoch": 3.766160764474424, + "grad_norm": 0.032060366311920385, + "learning_rate": 5.210881345751955e-05, + "loss": 0.0006, + "step": 6700 + }, + { + "epoch": 3.7717818999437887, + "grad_norm": 0.03063083487664102, + "learning_rate": 5.1968768002764975e-05, + "loss": 0.0006, + "step": 6710 + }, + { + "epoch": 3.7774030354131534, + "grad_norm": 0.028410174140144077, + "learning_rate": 5.182870707712179e-05, + "loss": 0.0005, + "step": 6720 + }, + { + "epoch": 3.783024170882518, + "grad_norm": 0.03070821415807422, + "learning_rate": 5.168863178121074e-05, + "loss": 0.0006, + "step": 6730 + }, + { + "epoch": 3.7886453063518832, + "grad_norm": 0.03408205460688081, + "learning_rate": 5.154854321576558e-05, + "loss": 0.0006, + "step": 6740 + }, + { + "epoch": 3.794266441821248, + "grad_norm": 0.032449076862453664, + "learning_rate": 5.1408442481624244e-05, + "loss": 0.0005, + "step": 6750 + }, + { + "epoch": 3.7998875772906127, + "grad_norm": 0.028003698332145076, + "learning_rate": 5.1268330679720386e-05, + "loss": 0.0005, + "step": 6760 + }, + { + "epoch": 3.8055087127599774, + "grad_norm": 0.021492309010048873, + "learning_rate": 5.112820891107455e-05, + "loss": 0.0005, + "step": 6770 + }, + { + "epoch": 3.811129848229342, + "grad_norm": 0.032580045120800843, + "learning_rate": 5.098807827678565e-05, + "loss": 0.0006, + "step": 6780 + }, + { + "epoch": 3.8167509836987072, + "grad_norm": 0.03806972540518316, + "learning_rate": 5.084793987802222e-05, + "loss": 0.0005, + "step": 6790 + }, + { + "epoch": 3.822372119168072, + "grad_norm": 0.02075268836005831, + "learning_rate": 5.0707794816013875e-05, + "loss": 0.0005, + "step": 6800 + }, + { + "epoch": 3.8279932546374367, + "grad_norm": 0.03094738730365526, + "learning_rate": 5.056764419204255e-05, + "loss": 0.0006, + "step": 6810 + }, + { + "epoch": 3.833614390106802, + "grad_norm": 0.03590767955890068, + "learning_rate": 5.0427489107433865e-05, + "loss": 0.0006, + "step": 6820 + }, + { + "epoch": 3.8392355255761665, + "grad_norm": 0.027827293745374723, + "learning_rate": 5.028733066354854e-05, + "loss": 0.0005, + "step": 6830 + }, + { + "epoch": 3.8448566610455313, + "grad_norm": 0.034298482666518935, + "learning_rate": 5.0147169961773685e-05, + "loss": 0.0006, + "step": 6840 + }, + { + "epoch": 3.850477796514896, + "grad_norm": 0.03072255397717621, + "learning_rate": 5.000700810351411e-05, + "loss": 0.0005, + "step": 6850 + }, + { + "epoch": 3.8560989319842607, + "grad_norm": 0.03126460291678688, + "learning_rate": 4.9866846190183746e-05, + "loss": 0.0005, + "step": 6860 + }, + { + "epoch": 3.861720067453626, + "grad_norm": 0.04799478576299796, + "learning_rate": 4.9726685323197e-05, + "loss": 0.0005, + "step": 6870 + }, + { + "epoch": 3.8673412029229906, + "grad_norm": 0.0325995767253313, + "learning_rate": 4.958652660395994e-05, + "loss": 0.0005, + "step": 6880 + }, + { + "epoch": 3.8729623383923553, + "grad_norm": 0.02987925906659042, + "learning_rate": 4.944637113386186e-05, + "loss": 0.0006, + "step": 6890 + }, + { + "epoch": 3.87858347386172, + "grad_norm": 0.03885877608701623, + "learning_rate": 4.930622001426649e-05, + "loss": 0.0005, + "step": 6900 + }, + { + "epoch": 3.8842046093310847, + "grad_norm": 0.028901959147139254, + "learning_rate": 4.916607434650336e-05, + "loss": 0.0006, + "step": 6910 + }, + { + "epoch": 3.88982574480045, + "grad_norm": 0.05739190448168031, + "learning_rate": 4.902593523185921e-05, + "loss": 0.0005, + "step": 6920 + }, + { + "epoch": 3.8954468802698146, + "grad_norm": 0.02826179650208044, + "learning_rate": 4.888580377156918e-05, + "loss": 0.0005, + "step": 6930 + }, + { + "epoch": 3.9010680157391793, + "grad_norm": 0.026006769806075185, + "learning_rate": 4.8745681066808366e-05, + "loss": 0.0004, + "step": 6940 + }, + { + "epoch": 3.906689151208544, + "grad_norm": 0.043865660139696915, + "learning_rate": 4.860556821868302e-05, + "loss": 0.0005, + "step": 6950 + }, + { + "epoch": 3.9123102866779087, + "grad_norm": 0.036999155281970994, + "learning_rate": 4.846546632822192e-05, + "loss": 0.0005, + "step": 6960 + }, + { + "epoch": 3.917931422147274, + "grad_norm": 0.03149861758052067, + "learning_rate": 4.83253764963678e-05, + "loss": 0.0005, + "step": 6970 + }, + { + "epoch": 3.9235525576166386, + "grad_norm": 0.03002234501301141, + "learning_rate": 4.818529982396853e-05, + "loss": 0.0005, + "step": 6980 + }, + { + "epoch": 3.9291736930860033, + "grad_norm": 0.02107866007385802, + "learning_rate": 4.804523741176868e-05, + "loss": 0.0005, + "step": 6990 + }, + { + "epoch": 3.9347948285553684, + "grad_norm": 0.030837081657031844, + "learning_rate": 4.7905190360400704e-05, + "loss": 0.0005, + "step": 7000 + }, + { + "epoch": 3.940415964024733, + "grad_norm": 0.027598674486161757, + "learning_rate": 4.776515977037635e-05, + "loss": 0.0005, + "step": 7010 + }, + { + "epoch": 3.946037099494098, + "grad_norm": 0.023441740322168624, + "learning_rate": 4.762514674207804e-05, + "loss": 0.0005, + "step": 7020 + }, + { + "epoch": 3.9516582349634626, + "grad_norm": 0.030551916257378967, + "learning_rate": 4.7485152375750144e-05, + "loss": 0.0004, + "step": 7030 + }, + { + "epoch": 3.9572793704328273, + "grad_norm": 0.029677542566248414, + "learning_rate": 4.7345177771490415e-05, + "loss": 0.0004, + "step": 7040 + }, + { + "epoch": 3.9629005059021924, + "grad_norm": 0.03408068861198226, + "learning_rate": 4.7205224029241325e-05, + "loss": 0.0004, + "step": 7050 + }, + { + "epoch": 3.968521641371557, + "grad_norm": 0.026854348880128696, + "learning_rate": 4.706529224878139e-05, + "loss": 0.0004, + "step": 7060 + }, + { + "epoch": 3.974142776840922, + "grad_norm": 0.036054258072901336, + "learning_rate": 4.692538352971655e-05, + "loss": 0.0005, + "step": 7070 + }, + { + "epoch": 3.9797639123102866, + "grad_norm": 0.03269259176981357, + "learning_rate": 4.678549897147152e-05, + "loss": 0.0004, + "step": 7080 + }, + { + "epoch": 3.9853850477796513, + "grad_norm": 0.04940817332978318, + "learning_rate": 4.6645639673281146e-05, + "loss": 0.0005, + "step": 7090 + }, + { + "epoch": 3.9910061832490165, + "grad_norm": 0.0322068167245232, + "learning_rate": 4.650580673418182e-05, + "loss": 0.0004, + "step": 7100 + }, + { + "epoch": 3.996627318718381, + "grad_norm": 0.03316436922530186, + "learning_rate": 4.636600125300275e-05, + "loss": 0.0005, + "step": 7110 + }, + { + "epoch": 4.002248454187746, + "grad_norm": 0.053765664720441635, + "learning_rate": 4.62262243283574e-05, + "loss": 0.0004, + "step": 7120 + }, + { + "epoch": 4.007869589657111, + "grad_norm": 0.023247303094544765, + "learning_rate": 4.608647705863486e-05, + "loss": 0.0003, + "step": 7130 + }, + { + "epoch": 4.013490725126475, + "grad_norm": 0.023042727766823867, + "learning_rate": 4.5946760541991096e-05, + "loss": 0.0003, + "step": 7140 + }, + { + "epoch": 4.0191118605958405, + "grad_norm": 0.02148899673828703, + "learning_rate": 4.580707587634051e-05, + "loss": 0.0003, + "step": 7150 + }, + { + "epoch": 4.024732996065205, + "grad_norm": 0.026716752800634017, + "learning_rate": 4.5667424159347174e-05, + "loss": 0.0003, + "step": 7160 + }, + { + "epoch": 4.03035413153457, + "grad_norm": 0.015244763608302027, + "learning_rate": 4.5527806488416224e-05, + "loss": 0.0003, + "step": 7170 + }, + { + "epoch": 4.035975267003935, + "grad_norm": 0.020311939018056193, + "learning_rate": 4.5388223960685324e-05, + "loss": 0.0003, + "step": 7180 + }, + { + "epoch": 4.041596402473299, + "grad_norm": 0.02932819570359388, + "learning_rate": 4.5248677673015866e-05, + "loss": 0.0003, + "step": 7190 + }, + { + "epoch": 4.0472175379426645, + "grad_norm": 0.017932745700789614, + "learning_rate": 4.510916872198457e-05, + "loss": 0.0003, + "step": 7200 + }, + { + "epoch": 4.05283867341203, + "grad_norm": 0.019477917595088934, + "learning_rate": 4.49696982038747e-05, + "loss": 0.0003, + "step": 7210 + }, + { + "epoch": 4.058459808881394, + "grad_norm": 0.022874657820461344, + "learning_rate": 4.483026721466753e-05, + "loss": 0.0003, + "step": 7220 + }, + { + "epoch": 4.064080944350759, + "grad_norm": 0.026809086252421507, + "learning_rate": 4.469087685003373e-05, + "loss": 0.0003, + "step": 7230 + }, + { + "epoch": 4.069702079820123, + "grad_norm": 0.02214151959206777, + "learning_rate": 4.4551528205324654e-05, + "loss": 0.0003, + "step": 7240 + }, + { + "epoch": 4.0753232152894885, + "grad_norm": 0.02068953515769804, + "learning_rate": 4.44122223755639e-05, + "loss": 0.0003, + "step": 7250 + }, + { + "epoch": 4.080944350758854, + "grad_norm": 0.024301040033901152, + "learning_rate": 4.4272960455438596e-05, + "loss": 0.0002, + "step": 7260 + }, + { + "epoch": 4.086565486228218, + "grad_norm": 0.0333079113323935, + "learning_rate": 4.413374353929079e-05, + "loss": 0.0003, + "step": 7270 + }, + { + "epoch": 4.092186621697583, + "grad_norm": 0.009036306225206955, + "learning_rate": 4.399457272110895e-05, + "loss": 0.0002, + "step": 7280 + }, + { + "epoch": 4.097807757166947, + "grad_norm": 0.020419373089518276, + "learning_rate": 4.3855449094519195e-05, + "loss": 0.0003, + "step": 7290 + }, + { + "epoch": 4.1034288926363125, + "grad_norm": 0.04107952045105247, + "learning_rate": 4.371637375277689e-05, + "loss": 0.0003, + "step": 7300 + }, + { + "epoch": 4.109050028105678, + "grad_norm": 0.034278953941617805, + "learning_rate": 4.3577347788757964e-05, + "loss": 0.0003, + "step": 7310 + }, + { + "epoch": 4.114671163575042, + "grad_norm": 0.02409972544680746, + "learning_rate": 4.343837229495029e-05, + "loss": 0.0004, + "step": 7320 + }, + { + "epoch": 4.120292299044407, + "grad_norm": 0.03868630473010241, + "learning_rate": 4.3299448363445174e-05, + "loss": 0.0003, + "step": 7330 + }, + { + "epoch": 4.125913434513771, + "grad_norm": 0.03293230943453663, + "learning_rate": 4.3160577085928686e-05, + "loss": 0.0003, + "step": 7340 + }, + { + "epoch": 4.1315345699831365, + "grad_norm": 0.011750767586229857, + "learning_rate": 4.3021759553673186e-05, + "loss": 0.0002, + "step": 7350 + }, + { + "epoch": 4.137155705452502, + "grad_norm": 0.021442299144220203, + "learning_rate": 4.2882996857528693e-05, + "loss": 0.0003, + "step": 7360 + }, + { + "epoch": 4.142776840921866, + "grad_norm": 0.030386010781675597, + "learning_rate": 4.274429008791427e-05, + "loss": 0.0003, + "step": 7370 + }, + { + "epoch": 4.148397976391231, + "grad_norm": 0.04094857681247711, + "learning_rate": 4.260564033480956e-05, + "loss": 0.0002, + "step": 7380 + }, + { + "epoch": 4.154019111860596, + "grad_norm": 0.020828602865352885, + "learning_rate": 4.24670486877461e-05, + "loss": 0.0002, + "step": 7390 + }, + { + "epoch": 4.1596402473299605, + "grad_norm": 0.014617406286699521, + "learning_rate": 4.232851623579886e-05, + "loss": 0.0002, + "step": 7400 + }, + { + "epoch": 4.165261382799326, + "grad_norm": 0.020223957935634106, + "learning_rate": 4.2190044067577645e-05, + "loss": 0.0003, + "step": 7410 + }, + { + "epoch": 4.17088251826869, + "grad_norm": 0.01951832075190107, + "learning_rate": 4.2051633271218526e-05, + "loss": 0.0002, + "step": 7420 + }, + { + "epoch": 4.176503653738055, + "grad_norm": 0.009388850983315674, + "learning_rate": 4.1913284934375326e-05, + "loss": 0.0002, + "step": 7430 + }, + { + "epoch": 4.18212478920742, + "grad_norm": 0.016383661209162084, + "learning_rate": 4.177500014421101e-05, + "loss": 0.0002, + "step": 7440 + }, + { + "epoch": 4.1877459246767845, + "grad_norm": 0.0217820471689951, + "learning_rate": 4.163677998738921e-05, + "loss": 0.0002, + "step": 7450 + }, + { + "epoch": 4.19336706014615, + "grad_norm": 0.017064282471927743, + "learning_rate": 4.1498625550065674e-05, + "loss": 0.0002, + "step": 7460 + }, + { + "epoch": 4.198988195615514, + "grad_norm": 0.023521168473738936, + "learning_rate": 4.13605379178797e-05, + "loss": 0.0003, + "step": 7470 + }, + { + "epoch": 4.204609331084879, + "grad_norm": 0.025558567013910526, + "learning_rate": 4.1222518175945626e-05, + "loss": 0.0002, + "step": 7480 + }, + { + "epoch": 4.210230466554244, + "grad_norm": 0.029888597839622374, + "learning_rate": 4.108456740884426e-05, + "loss": 0.0002, + "step": 7490 + }, + { + "epoch": 4.2158516020236085, + "grad_norm": 0.019140673956923263, + "learning_rate": 4.0946686700614467e-05, + "loss": 0.0003, + "step": 7500 + }, + { + "epoch": 4.221472737492974, + "grad_norm": 0.026351365791585438, + "learning_rate": 4.08088771347445e-05, + "loss": 0.0002, + "step": 7510 + }, + { + "epoch": 4.227093872962339, + "grad_norm": 0.02774699536317935, + "learning_rate": 4.0671139794163635e-05, + "loss": 0.0002, + "step": 7520 + }, + { + "epoch": 4.232715008431703, + "grad_norm": 0.014482991869498418, + "learning_rate": 4.0533475761233564e-05, + "loss": 0.0002, + "step": 7530 + }, + { + "epoch": 4.238336143901068, + "grad_norm": 0.022558554606976047, + "learning_rate": 4.0395886117739876e-05, + "loss": 0.0002, + "step": 7540 + }, + { + "epoch": 4.2439572793704325, + "grad_norm": 0.02440775332622257, + "learning_rate": 4.0258371944883645e-05, + "loss": 0.0003, + "step": 7550 + }, + { + "epoch": 4.249578414839798, + "grad_norm": 0.031717020522322864, + "learning_rate": 4.0120934323272866e-05, + "loss": 0.0003, + "step": 7560 + }, + { + "epoch": 4.255199550309163, + "grad_norm": 0.02639783936630194, + "learning_rate": 3.9983574332914004e-05, + "loss": 0.0002, + "step": 7570 + }, + { + "epoch": 4.260820685778527, + "grad_norm": 0.04309011076809027, + "learning_rate": 3.984629305320347e-05, + "loss": 0.0003, + "step": 7580 + }, + { + "epoch": 4.266441821247892, + "grad_norm": 0.028085306019164806, + "learning_rate": 3.9709091562919114e-05, + "loss": 0.0002, + "step": 7590 + }, + { + "epoch": 4.2720629567172566, + "grad_norm": 0.019253775967156658, + "learning_rate": 3.957197094021186e-05, + "loss": 0.0002, + "step": 7600 + }, + { + "epoch": 4.277684092186622, + "grad_norm": 0.017414959622448697, + "learning_rate": 3.9434932262597105e-05, + "loss": 0.0003, + "step": 7610 + }, + { + "epoch": 4.283305227655987, + "grad_norm": 0.022870972349351766, + "learning_rate": 3.9297976606946354e-05, + "loss": 0.0003, + "step": 7620 + }, + { + "epoch": 4.288926363125351, + "grad_norm": 0.02244497574425279, + "learning_rate": 3.9161105049478705e-05, + "loss": 0.0002, + "step": 7630 + }, + { + "epoch": 4.294547498594716, + "grad_norm": 0.010990955466765762, + "learning_rate": 3.902431866575233e-05, + "loss": 0.0002, + "step": 7640 + }, + { + "epoch": 4.300168634064081, + "grad_norm": 0.015419696703903574, + "learning_rate": 3.888761853065619e-05, + "loss": 0.0002, + "step": 7650 + }, + { + "epoch": 4.305789769533446, + "grad_norm": 0.01707129456470333, + "learning_rate": 3.875100571840142e-05, + "loss": 0.0002, + "step": 7660 + }, + { + "epoch": 4.311410905002811, + "grad_norm": 0.027400843679214623, + "learning_rate": 3.8614481302513015e-05, + "loss": 0.0003, + "step": 7670 + }, + { + "epoch": 4.317032040472175, + "grad_norm": 0.016533221629184323, + "learning_rate": 3.847804635582131e-05, + "loss": 0.0002, + "step": 7680 + }, + { + "epoch": 4.32265317594154, + "grad_norm": 0.026100801339734025, + "learning_rate": 3.834170195045354e-05, + "loss": 0.0002, + "step": 7690 + }, + { + "epoch": 4.328274311410905, + "grad_norm": 0.03614875173358743, + "learning_rate": 3.82054491578255e-05, + "loss": 0.0003, + "step": 7700 + }, + { + "epoch": 4.33389544688027, + "grad_norm": 0.029613527995187076, + "learning_rate": 3.806928904863306e-05, + "loss": 0.0002, + "step": 7710 + }, + { + "epoch": 4.339516582349635, + "grad_norm": 0.017247793474072175, + "learning_rate": 3.7933222692843804e-05, + "loss": 0.0002, + "step": 7720 + }, + { + "epoch": 4.345137717818999, + "grad_norm": 0.021120407652614248, + "learning_rate": 3.779725115968854e-05, + "loss": 0.0002, + "step": 7730 + }, + { + "epoch": 4.350758853288364, + "grad_norm": 0.011389481138970219, + "learning_rate": 3.7661375517652954e-05, + "loss": 0.0002, + "step": 7740 + }, + { + "epoch": 4.3563799887577295, + "grad_norm": 0.02036372485836203, + "learning_rate": 3.752559683446922e-05, + "loss": 0.0002, + "step": 7750 + }, + { + "epoch": 4.362001124227094, + "grad_norm": 0.010573966374116898, + "learning_rate": 3.738991617710758e-05, + "loss": 0.0002, + "step": 7760 + }, + { + "epoch": 4.367622259696459, + "grad_norm": 0.023544005907481107, + "learning_rate": 3.725433461176796e-05, + "loss": 0.0002, + "step": 7770 + }, + { + "epoch": 4.373243395165823, + "grad_norm": 0.021221672309639533, + "learning_rate": 3.7118853203871695e-05, + "loss": 0.0003, + "step": 7780 + }, + { + "epoch": 4.378864530635188, + "grad_norm": 0.03340584396262711, + "learning_rate": 3.698347301805292e-05, + "loss": 0.0003, + "step": 7790 + }, + { + "epoch": 4.3844856661045535, + "grad_norm": 0.02002293114687527, + "learning_rate": 3.684819511815046e-05, + "loss": 0.0002, + "step": 7800 + }, + { + "epoch": 4.390106801573918, + "grad_norm": 0.008890609179159856, + "learning_rate": 3.671302056719933e-05, + "loss": 0.0003, + "step": 7810 + }, + { + "epoch": 4.395727937043283, + "grad_norm": 0.009077695184372852, + "learning_rate": 3.6577950427422404e-05, + "loss": 0.0002, + "step": 7820 + }, + { + "epoch": 4.401349072512647, + "grad_norm": 0.023883211445300135, + "learning_rate": 3.644298576022212e-05, + "loss": 0.0003, + "step": 7830 + }, + { + "epoch": 4.406970207982012, + "grad_norm": 0.0355821012404783, + "learning_rate": 3.6308127626172006e-05, + "loss": 0.0002, + "step": 7840 + }, + { + "epoch": 4.4125913434513775, + "grad_norm": 0.033969175109265615, + "learning_rate": 3.617337708500852e-05, + "loss": 0.0002, + "step": 7850 + }, + { + "epoch": 4.418212478920742, + "grad_norm": 0.01570570599111059, + "learning_rate": 3.6038735195622605e-05, + "loss": 0.0003, + "step": 7860 + }, + { + "epoch": 4.423833614390107, + "grad_norm": 0.026627640310681842, + "learning_rate": 3.59042030160514e-05, + "loss": 0.0003, + "step": 7870 + }, + { + "epoch": 4.429454749859472, + "grad_norm": 0.01980774304334287, + "learning_rate": 3.5769781603469964e-05, + "loss": 0.0002, + "step": 7880 + }, + { + "epoch": 4.435075885328836, + "grad_norm": 0.017110373429014062, + "learning_rate": 3.563547201418283e-05, + "loss": 0.0002, + "step": 7890 + }, + { + "epoch": 4.4406970207982015, + "grad_norm": 0.016565269343143407, + "learning_rate": 3.550127530361594e-05, + "loss": 0.0002, + "step": 7900 + }, + { + "epoch": 4.446318156267566, + "grad_norm": 0.023226558772809765, + "learning_rate": 3.536719252630814e-05, + "loss": 0.0002, + "step": 7910 + }, + { + "epoch": 4.451939291736931, + "grad_norm": 0.023364636917571364, + "learning_rate": 3.523322473590295e-05, + "loss": 0.0002, + "step": 7920 + }, + { + "epoch": 4.457560427206296, + "grad_norm": 0.02718470992103199, + "learning_rate": 3.509937298514039e-05, + "loss": 0.0002, + "step": 7930 + }, + { + "epoch": 4.46318156267566, + "grad_norm": 0.026564888434526944, + "learning_rate": 3.496563832584851e-05, + "loss": 0.0002, + "step": 7940 + }, + { + "epoch": 4.4688026981450255, + "grad_norm": 0.03194986325566669, + "learning_rate": 3.483202180893534e-05, + "loss": 0.0002, + "step": 7950 + }, + { + "epoch": 4.47442383361439, + "grad_norm": 0.02133279413758573, + "learning_rate": 3.469852448438048e-05, + "loss": 0.0002, + "step": 7960 + }, + { + "epoch": 4.480044969083755, + "grad_norm": 0.026321245881711555, + "learning_rate": 3.45651474012269e-05, + "loss": 0.0002, + "step": 7970 + }, + { + "epoch": 4.48566610455312, + "grad_norm": 0.04748817942053944, + "learning_rate": 3.443189160757269e-05, + "loss": 0.0003, + "step": 7980 + }, + { + "epoch": 4.491287240022484, + "grad_norm": 0.019255570276199837, + "learning_rate": 3.4298758150562826e-05, + "loss": 0.0003, + "step": 7990 + }, + { + "epoch": 4.4969083754918495, + "grad_norm": 0.024066640941954424, + "learning_rate": 3.4165748076380974e-05, + "loss": 0.0003, + "step": 8000 + }, + { + "epoch": 4.502529510961214, + "grad_norm": 0.02313277536928503, + "learning_rate": 3.403286243024122e-05, + "loss": 0.0002, + "step": 8010 + }, + { + "epoch": 4.508150646430579, + "grad_norm": 0.03411629334389642, + "learning_rate": 3.3900102256379854e-05, + "loss": 0.0003, + "step": 8020 + }, + { + "epoch": 4.513771781899944, + "grad_norm": 0.03474632268783524, + "learning_rate": 3.3767468598047223e-05, + "loss": 0.0002, + "step": 8030 + }, + { + "epoch": 4.519392917369308, + "grad_norm": 0.030969585523263173, + "learning_rate": 3.363496249749946e-05, + "loss": 0.0003, + "step": 8040 + }, + { + "epoch": 4.5250140528386735, + "grad_norm": 0.015735323763602975, + "learning_rate": 3.350258499599033e-05, + "loss": 0.0002, + "step": 8050 + }, + { + "epoch": 4.530635188308038, + "grad_norm": 0.02012160827942359, + "learning_rate": 3.337033713376309e-05, + "loss": 0.0002, + "step": 8060 + }, + { + "epoch": 4.536256323777403, + "grad_norm": 0.015021640409973227, + "learning_rate": 3.323821995004224e-05, + "loss": 0.0002, + "step": 8070 + }, + { + "epoch": 4.541877459246768, + "grad_norm": 0.018997055625676443, + "learning_rate": 3.310623448302538e-05, + "loss": 0.0002, + "step": 8080 + }, + { + "epoch": 4.547498594716132, + "grad_norm": 0.030249040868316847, + "learning_rate": 3.2974381769875065e-05, + "loss": 0.0002, + "step": 8090 + }, + { + "epoch": 4.5531197301854975, + "grad_norm": 0.023284994637939597, + "learning_rate": 3.284266284671064e-05, + "loss": 0.0003, + "step": 8100 + }, + { + "epoch": 4.558740865654862, + "grad_norm": 0.029041349577412928, + "learning_rate": 3.2711078748600156e-05, + "loss": 0.0002, + "step": 8110 + }, + { + "epoch": 4.564362001124227, + "grad_norm": 0.03263853813030728, + "learning_rate": 3.257963050955213e-05, + "loss": 0.0002, + "step": 8120 + }, + { + "epoch": 4.569983136593592, + "grad_norm": 0.03839784810100402, + "learning_rate": 3.244831916250751e-05, + "loss": 0.0002, + "step": 8130 + }, + { + "epoch": 4.575604272062956, + "grad_norm": 0.027377578185036756, + "learning_rate": 3.2317145739331514e-05, + "loss": 0.0002, + "step": 8140 + }, + { + "epoch": 4.5812254075323215, + "grad_norm": 0.010540688435854336, + "learning_rate": 3.21861112708055e-05, + "loss": 0.0002, + "step": 8150 + }, + { + "epoch": 4.586846543001687, + "grad_norm": 0.014459754377953717, + "learning_rate": 3.205521678661897e-05, + "loss": 0.0002, + "step": 8160 + }, + { + "epoch": 4.592467678471051, + "grad_norm": 0.02720330585491664, + "learning_rate": 3.192446331536135e-05, + "loss": 0.0002, + "step": 8170 + }, + { + "epoch": 4.598088813940416, + "grad_norm": 0.02225426332202112, + "learning_rate": 3.179385188451399e-05, + "loss": 0.0002, + "step": 8180 + }, + { + "epoch": 4.60370994940978, + "grad_norm": 0.020815014127148316, + "learning_rate": 3.166338352044208e-05, + "loss": 0.0002, + "step": 8190 + }, + { + "epoch": 4.609331084879146, + "grad_norm": 0.02638981083489274, + "learning_rate": 3.1533059248386495e-05, + "loss": 0.0002, + "step": 8200 + }, + { + "epoch": 4.614952220348511, + "grad_norm": 0.023957040422956453, + "learning_rate": 3.140288009245589e-05, + "loss": 0.0002, + "step": 8210 + }, + { + "epoch": 4.620573355817875, + "grad_norm": 0.021983337552939686, + "learning_rate": 3.127284707561855e-05, + "loss": 0.0002, + "step": 8220 + }, + { + "epoch": 4.62619449128724, + "grad_norm": 0.029882030566280155, + "learning_rate": 3.114296121969437e-05, + "loss": 0.0002, + "step": 8230 + }, + { + "epoch": 4.631815626756605, + "grad_norm": 0.020427298336691838, + "learning_rate": 3.101322354534685e-05, + "loss": 0.0002, + "step": 8240 + }, + { + "epoch": 4.63743676222597, + "grad_norm": 0.015040760982435529, + "learning_rate": 3.0883635072075e-05, + "loss": 0.0002, + "step": 8250 + }, + { + "epoch": 4.643057897695335, + "grad_norm": 0.02596420860121817, + "learning_rate": 3.075419681820545e-05, + "loss": 0.0002, + "step": 8260 + }, + { + "epoch": 4.648679033164699, + "grad_norm": 0.01210666288074982, + "learning_rate": 3.062490980088436e-05, + "loss": 0.0002, + "step": 8270 + }, + { + "epoch": 4.654300168634064, + "grad_norm": 0.02107502671994588, + "learning_rate": 3.0495775036069453e-05, + "loss": 0.0002, + "step": 8280 + }, + { + "epoch": 4.659921304103429, + "grad_norm": 0.020146862908233917, + "learning_rate": 3.0366793538522016e-05, + "loss": 0.0002, + "step": 8290 + }, + { + "epoch": 4.665542439572794, + "grad_norm": 0.017914524269681268, + "learning_rate": 3.0237966321798928e-05, + "loss": 0.0002, + "step": 8300 + }, + { + "epoch": 4.671163575042159, + "grad_norm": 0.01781063100049735, + "learning_rate": 3.0109294398244724e-05, + "loss": 0.0002, + "step": 8310 + }, + { + "epoch": 4.676784710511523, + "grad_norm": 0.011086967137251883, + "learning_rate": 2.9980778778983644e-05, + "loss": 0.0002, + "step": 8320 + }, + { + "epoch": 4.682405845980888, + "grad_norm": 0.01950112952407118, + "learning_rate": 2.9852420473911614e-05, + "loss": 0.0002, + "step": 8330 + }, + { + "epoch": 4.688026981450253, + "grad_norm": 0.020882001496958832, + "learning_rate": 2.9724220491688393e-05, + "loss": 0.0002, + "step": 8340 + }, + { + "epoch": 4.693648116919618, + "grad_norm": 0.007489495354854185, + "learning_rate": 2.9596179839729598e-05, + "loss": 0.0002, + "step": 8350 + }, + { + "epoch": 4.699269252388983, + "grad_norm": 0.02257449972040603, + "learning_rate": 2.9468299524198805e-05, + "loss": 0.0002, + "step": 8360 + }, + { + "epoch": 4.704890387858347, + "grad_norm": 0.032024296212818854, + "learning_rate": 2.9340580549999662e-05, + "loss": 0.0003, + "step": 8370 + }, + { + "epoch": 4.710511523327712, + "grad_norm": 0.014480377426817955, + "learning_rate": 2.9213023920767957e-05, + "loss": 0.0002, + "step": 8380 + }, + { + "epoch": 4.716132658797077, + "grad_norm": 0.02036274182188618, + "learning_rate": 2.908563063886376e-05, + "loss": 0.0002, + "step": 8390 + }, + { + "epoch": 4.721753794266442, + "grad_norm": 0.018936307100043633, + "learning_rate": 2.8958401705363458e-05, + "loss": 0.0002, + "step": 8400 + }, + { + "epoch": 4.727374929735807, + "grad_norm": 0.018078218900303245, + "learning_rate": 2.8831338120052064e-05, + "loss": 0.0002, + "step": 8410 + }, + { + "epoch": 4.732996065205171, + "grad_norm": 0.01629986677654801, + "learning_rate": 2.870444088141523e-05, + "loss": 0.0002, + "step": 8420 + }, + { + "epoch": 4.738617200674536, + "grad_norm": 0.01576508550334688, + "learning_rate": 2.857771098663138e-05, + "loss": 0.0002, + "step": 8430 + }, + { + "epoch": 4.744238336143901, + "grad_norm": 0.022533316248782954, + "learning_rate": 2.8451149431563995e-05, + "loss": 0.0002, + "step": 8440 + }, + { + "epoch": 4.749859471613266, + "grad_norm": 0.01919577657327632, + "learning_rate": 2.8324757210753638e-05, + "loss": 0.0002, + "step": 8450 + }, + { + "epoch": 4.755480607082631, + "grad_norm": 0.023440100905190362, + "learning_rate": 2.8198535317410313e-05, + "loss": 0.0002, + "step": 8460 + }, + { + "epoch": 4.761101742551995, + "grad_norm": 0.02646558773222747, + "learning_rate": 2.8072484743405458e-05, + "loss": 0.0002, + "step": 8470 + }, + { + "epoch": 4.76672287802136, + "grad_norm": 0.02249431881693691, + "learning_rate": 2.794660647926434e-05, + "loss": 0.0001, + "step": 8480 + }, + { + "epoch": 4.772344013490725, + "grad_norm": 0.02037372939946552, + "learning_rate": 2.7820901514158194e-05, + "loss": 0.0002, + "step": 8490 + }, + { + "epoch": 4.77796514896009, + "grad_norm": 0.019974127869670434, + "learning_rate": 2.7695370835896318e-05, + "loss": 0.0002, + "step": 8500 + }, + { + "epoch": 4.783586284429455, + "grad_norm": 0.014477867051889531, + "learning_rate": 2.7570015430918572e-05, + "loss": 0.0002, + "step": 8510 + }, + { + "epoch": 4.78920741989882, + "grad_norm": 0.013189373448917609, + "learning_rate": 2.7444836284287463e-05, + "loss": 0.0002, + "step": 8520 + }, + { + "epoch": 4.794828555368184, + "grad_norm": 0.018610397509619677, + "learning_rate": 2.7319834379680376e-05, + "loss": 0.0002, + "step": 8530 + }, + { + "epoch": 4.800449690837549, + "grad_norm": 0.01809790204419372, + "learning_rate": 2.719501069938198e-05, + "loss": 0.0001, + "step": 8540 + }, + { + "epoch": 4.806070826306914, + "grad_norm": 0.00950016605948751, + "learning_rate": 2.7070366224276345e-05, + "loss": 0.0002, + "step": 8550 + }, + { + "epoch": 4.811691961776279, + "grad_norm": 0.018276831065029895, + "learning_rate": 2.6945901933839403e-05, + "loss": 0.0002, + "step": 8560 + }, + { + "epoch": 4.817313097245644, + "grad_norm": 0.012470921784932046, + "learning_rate": 2.6821618806131078e-05, + "loss": 0.0001, + "step": 8570 + }, + { + "epoch": 4.822934232715008, + "grad_norm": 0.019057347526256048, + "learning_rate": 2.6697517817787786e-05, + "loss": 0.0001, + "step": 8580 + }, + { + "epoch": 4.828555368184373, + "grad_norm": 0.01660696062858163, + "learning_rate": 2.657359994401458e-05, + "loss": 0.0002, + "step": 8590 + }, + { + "epoch": 4.8341765036537385, + "grad_norm": 0.020830683127642444, + "learning_rate": 2.6449866158577598e-05, + "loss": 0.0002, + "step": 8600 + }, + { + "epoch": 4.839797639123103, + "grad_norm": 0.024340963139268566, + "learning_rate": 2.632631743379639e-05, + "loss": 0.0002, + "step": 8610 + }, + { + "epoch": 4.845418774592468, + "grad_norm": 0.03487684200908871, + "learning_rate": 2.6202954740536302e-05, + "loss": 0.0002, + "step": 8620 + }, + { + "epoch": 4.851039910061832, + "grad_norm": 0.016884589191415247, + "learning_rate": 2.6079779048200725e-05, + "loss": 0.0001, + "step": 8630 + }, + { + "epoch": 4.856661045531197, + "grad_norm": 0.013896439982841366, + "learning_rate": 2.5956791324723684e-05, + "loss": 0.0002, + "step": 8640 + }, + { + "epoch": 4.8622821810005625, + "grad_norm": 0.020685895284239898, + "learning_rate": 2.5833992536562014e-05, + "loss": 0.0001, + "step": 8650 + }, + { + "epoch": 4.867903316469927, + "grad_norm": 0.008740650028584812, + "learning_rate": 2.5711383648687893e-05, + "loss": 0.0001, + "step": 8660 + }, + { + "epoch": 4.873524451939292, + "grad_norm": 0.025397997859802884, + "learning_rate": 2.5588965624581275e-05, + "loss": 0.0002, + "step": 8670 + }, + { + "epoch": 4.879145587408656, + "grad_norm": 0.02152761900719379, + "learning_rate": 2.5466739426222264e-05, + "loss": 0.0001, + "step": 8680 + }, + { + "epoch": 4.884766722878021, + "grad_norm": 0.014583384791330915, + "learning_rate": 2.534470601408352e-05, + "loss": 0.0002, + "step": 8690 + }, + { + "epoch": 4.8903878583473865, + "grad_norm": 0.012147783679929098, + "learning_rate": 2.522286634712278e-05, + "loss": 0.0001, + "step": 8700 + }, + { + "epoch": 4.896008993816751, + "grad_norm": 0.019176395097341245, + "learning_rate": 2.510122138277533e-05, + "loss": 0.0002, + "step": 8710 + }, + { + "epoch": 4.901630129286116, + "grad_norm": 0.022718149676602763, + "learning_rate": 2.4979772076946446e-05, + "loss": 0.0001, + "step": 8720 + }, + { + "epoch": 4.90725126475548, + "grad_norm": 0.017201910632240528, + "learning_rate": 2.485851938400384e-05, + "loss": 0.0002, + "step": 8730 + }, + { + "epoch": 4.912872400224845, + "grad_norm": 0.02682265488324509, + "learning_rate": 2.4737464256770292e-05, + "loss": 0.0001, + "step": 8740 + }, + { + "epoch": 4.9184935356942105, + "grad_norm": 0.026565169672676452, + "learning_rate": 2.4616607646516e-05, + "loss": 0.0001, + "step": 8750 + }, + { + "epoch": 4.924114671163575, + "grad_norm": 0.015564650296789268, + "learning_rate": 2.4495950502951197e-05, + "loss": 0.0001, + "step": 8760 + }, + { + "epoch": 4.92973580663294, + "grad_norm": 0.018545556438106618, + "learning_rate": 2.437549377421872e-05, + "loss": 0.0002, + "step": 8770 + }, + { + "epoch": 4.935356942102304, + "grad_norm": 0.03647007562822495, + "learning_rate": 2.4255238406886505e-05, + "loss": 0.0001, + "step": 8780 + }, + { + "epoch": 4.940978077571669, + "grad_norm": 0.0138686229898494, + "learning_rate": 2.4135185345940126e-05, + "loss": 0.0002, + "step": 8790 + }, + { + "epoch": 4.946599213041035, + "grad_norm": 0.017438625657368284, + "learning_rate": 2.4015335534775396e-05, + "loss": 0.0002, + "step": 8800 + }, + { + "epoch": 4.952220348510399, + "grad_norm": 0.009248278898441906, + "learning_rate": 2.3895689915191004e-05, + "loss": 0.0001, + "step": 8810 + }, + { + "epoch": 4.957841483979764, + "grad_norm": 0.021511146929612292, + "learning_rate": 2.3776249427381063e-05, + "loss": 0.0001, + "step": 8820 + }, + { + "epoch": 4.963462619449128, + "grad_norm": 0.014026677868579224, + "learning_rate": 2.3657015009927675e-05, + "loss": 0.0001, + "step": 8830 + }, + { + "epoch": 4.969083754918493, + "grad_norm": 0.014821709992871065, + "learning_rate": 2.3537987599793688e-05, + "loss": 0.0001, + "step": 8840 + }, + { + "epoch": 4.974704890387859, + "grad_norm": 0.011458197800049738, + "learning_rate": 2.341916813231519e-05, + "loss": 0.0001, + "step": 8850 + }, + { + "epoch": 4.980326025857223, + "grad_norm": 0.005504111426471945, + "learning_rate": 2.3300557541194223e-05, + "loss": 0.0002, + "step": 8860 + }, + { + "epoch": 4.985947161326588, + "grad_norm": 0.022671899447444452, + "learning_rate": 2.318215675849149e-05, + "loss": 0.0002, + "step": 8870 + }, + { + "epoch": 4.991568296795953, + "grad_norm": 0.012596869556073446, + "learning_rate": 2.306396671461898e-05, + "loss": 0.0002, + "step": 8880 + }, + { + "epoch": 4.997189432265317, + "grad_norm": 0.01740550593180937, + "learning_rate": 2.294598833833263e-05, + "loss": 0.0001, + "step": 8890 + }, + { + "epoch": 5.002810567734683, + "grad_norm": 0.013182341708189877, + "learning_rate": 2.2828222556725066e-05, + "loss": 0.0001, + "step": 8900 + }, + { + "epoch": 5.008431703204047, + "grad_norm": 0.017452065111461423, + "learning_rate": 2.2710670295218345e-05, + "loss": 0.0001, + "step": 8910 + }, + { + "epoch": 5.014052838673412, + "grad_norm": 0.012485509948781557, + "learning_rate": 2.25933324775566e-05, + "loss": 0.0001, + "step": 8920 + }, + { + "epoch": 5.019673974142777, + "grad_norm": 0.019509664009862697, + "learning_rate": 2.2476210025798873e-05, + "loss": 0.0001, + "step": 8930 + }, + { + "epoch": 5.025295109612141, + "grad_norm": 0.009165722187332291, + "learning_rate": 2.2359303860311825e-05, + "loss": 0.0001, + "step": 8940 + }, + { + "epoch": 5.030916245081507, + "grad_norm": 0.010719463092017791, + "learning_rate": 2.224261489976248e-05, + "loss": 0.0001, + "step": 8950 + }, + { + "epoch": 5.036537380550871, + "grad_norm": 0.0063805967357909955, + "learning_rate": 2.212614406111103e-05, + "loss": 0.0001, + "step": 8960 + }, + { + "epoch": 5.042158516020236, + "grad_norm": 0.011903486479072889, + "learning_rate": 2.2009892259603654e-05, + "loss": 0.0001, + "step": 8970 + }, + { + "epoch": 5.047779651489601, + "grad_norm": 0.00756341072187003, + "learning_rate": 2.189386040876531e-05, + "loss": 0.0001, + "step": 8980 + }, + { + "epoch": 5.053400786958965, + "grad_norm": 0.019983321582894203, + "learning_rate": 2.1778049420392516e-05, + "loss": 0.0001, + "step": 8990 + }, + { + "epoch": 5.059021922428331, + "grad_norm": 0.009566664565644227, + "learning_rate": 2.166246020454622e-05, + "loss": 0.0001, + "step": 9000 + }, + { + "epoch": 5.064643057897696, + "grad_norm": 0.015816694581856614, + "learning_rate": 2.1547093669544688e-05, + "loss": 0.0001, + "step": 9010 + }, + { + "epoch": 5.07026419336706, + "grad_norm": 0.014169230553367947, + "learning_rate": 2.1431950721956268e-05, + "loss": 0.0001, + "step": 9020 + }, + { + "epoch": 5.075885328836425, + "grad_norm": 0.00931380565076678, + "learning_rate": 2.131703226659237e-05, + "loss": 0.0001, + "step": 9030 + }, + { + "epoch": 5.0815064643057895, + "grad_norm": 0.01731460881655287, + "learning_rate": 2.1202339206500305e-05, + "loss": 0.0001, + "step": 9040 + }, + { + "epoch": 5.087127599775155, + "grad_norm": 0.02249971547144122, + "learning_rate": 2.1087872442956163e-05, + "loss": 0.0001, + "step": 9050 + }, + { + "epoch": 5.09274873524452, + "grad_norm": 0.008602340370658544, + "learning_rate": 2.097363287545776e-05, + "loss": 0.0001, + "step": 9060 + }, + { + "epoch": 5.098369870713884, + "grad_norm": 0.016994095710031014, + "learning_rate": 2.08596214017176e-05, + "loss": 0.0001, + "step": 9070 + }, + { + "epoch": 5.103991006183249, + "grad_norm": 0.017938917921659477, + "learning_rate": 2.0745838917655798e-05, + "loss": 0.0001, + "step": 9080 + }, + { + "epoch": 5.1096121416526135, + "grad_norm": 0.011762561048709048, + "learning_rate": 2.0632286317392984e-05, + "loss": 0.0001, + "step": 9090 + }, + { + "epoch": 5.115233277121979, + "grad_norm": 0.026655746429423775, + "learning_rate": 2.051896449324334e-05, + "loss": 0.0001, + "step": 9100 + }, + { + "epoch": 5.120854412591344, + "grad_norm": 0.018619393276839705, + "learning_rate": 2.0405874335707613e-05, + "loss": 0.0001, + "step": 9110 + }, + { + "epoch": 5.126475548060708, + "grad_norm": 0.006233985913833896, + "learning_rate": 2.0293016733466026e-05, + "loss": 0.0001, + "step": 9120 + }, + { + "epoch": 5.132096683530073, + "grad_norm": 0.006966252351504392, + "learning_rate": 2.018039257337137e-05, + "loss": 0.0001, + "step": 9130 + }, + { + "epoch": 5.1377178189994375, + "grad_norm": 0.02465689191476909, + "learning_rate": 2.0068002740442054e-05, + "loss": 0.0001, + "step": 9140 + }, + { + "epoch": 5.143338954468803, + "grad_norm": 0.020056110290910097, + "learning_rate": 1.9955848117854993e-05, + "loss": 0.0001, + "step": 9150 + }, + { + "epoch": 5.148960089938168, + "grad_norm": 0.018033903209788756, + "learning_rate": 1.9843929586938874e-05, + "loss": 0.0001, + "step": 9160 + }, + { + "epoch": 5.154581225407532, + "grad_norm": 0.02243218376948561, + "learning_rate": 1.9732248027167144e-05, + "loss": 0.0001, + "step": 9170 + }, + { + "epoch": 5.160202360876897, + "grad_norm": 0.010805319539704424, + "learning_rate": 1.962080431615103e-05, + "loss": 0.0001, + "step": 9180 + }, + { + "epoch": 5.165823496346262, + "grad_norm": 0.008020693818493574, + "learning_rate": 1.9509599329632776e-05, + "loss": 0.0001, + "step": 9190 + }, + { + "epoch": 5.171444631815627, + "grad_norm": 0.016514857292656653, + "learning_rate": 1.9398633941478623e-05, + "loss": 0.0001, + "step": 9200 + }, + { + "epoch": 5.177065767284992, + "grad_norm": 0.011356503205489034, + "learning_rate": 1.9287909023672067e-05, + "loss": 0.0001, + "step": 9210 + }, + { + "epoch": 5.182686902754356, + "grad_norm": 0.01487977427000052, + "learning_rate": 1.9177425446306914e-05, + "loss": 0.0001, + "step": 9220 + }, + { + "epoch": 5.188308038223721, + "grad_norm": 0.015364464724645747, + "learning_rate": 1.9067184077580492e-05, + "loss": 0.0001, + "step": 9230 + }, + { + "epoch": 5.193929173693086, + "grad_norm": 0.0137386169697318, + "learning_rate": 1.8957185783786858e-05, + "loss": 0.0001, + "step": 9240 + }, + { + "epoch": 5.199550309162451, + "grad_norm": 0.012780205445714246, + "learning_rate": 1.884743142930983e-05, + "loss": 0.0001, + "step": 9250 + }, + { + "epoch": 5.205171444631816, + "grad_norm": 0.005694442859720592, + "learning_rate": 1.8737921876616433e-05, + "loss": 0.0001, + "step": 9260 + }, + { + "epoch": 5.21079258010118, + "grad_norm": 0.0011472692936280334, + "learning_rate": 1.8628657986249964e-05, + "loss": 0.0001, + "step": 9270 + }, + { + "epoch": 5.216413715570545, + "grad_norm": 0.008621356822639042, + "learning_rate": 1.851964061682323e-05, + "loss": 0.0001, + "step": 9280 + }, + { + "epoch": 5.22203485103991, + "grad_norm": 0.01005847049009601, + "learning_rate": 1.8410870625011878e-05, + "loss": 0.0001, + "step": 9290 + }, + { + "epoch": 5.227655986509275, + "grad_norm": 0.0058782054118399615, + "learning_rate": 1.8302348865547635e-05, + "loss": 0.0001, + "step": 9300 + }, + { + "epoch": 5.23327712197864, + "grad_norm": 0.010720866089406748, + "learning_rate": 1.819407619121154e-05, + "loss": 0.0001, + "step": 9310 + }, + { + "epoch": 5.238898257448004, + "grad_norm": 0.007597016070157031, + "learning_rate": 1.8086053452827274e-05, + "loss": 0.0001, + "step": 9320 + }, + { + "epoch": 5.244519392917369, + "grad_norm": 0.004591962543197543, + "learning_rate": 1.797828149925454e-05, + "loss": 0.0001, + "step": 9330 + }, + { + "epoch": 5.250140528386734, + "grad_norm": 0.00483959397426083, + "learning_rate": 1.7870761177382267e-05, + "loss": 0.0001, + "step": 9340 + }, + { + "epoch": 5.255761663856099, + "grad_norm": 0.003212679405200927, + "learning_rate": 1.7763493332122087e-05, + "loss": 0.0001, + "step": 9350 + }, + { + "epoch": 5.261382799325464, + "grad_norm": 0.009870040654337574, + "learning_rate": 1.765647880640158e-05, + "loss": 0.0001, + "step": 9360 + }, + { + "epoch": 5.267003934794829, + "grad_norm": 0.008321091552982733, + "learning_rate": 1.7549718441157735e-05, + "loss": 0.0001, + "step": 9370 + }, + { + "epoch": 5.272625070264193, + "grad_norm": 0.008812085420663805, + "learning_rate": 1.7443213075330274e-05, + "loss": 0.0001, + "step": 9380 + }, + { + "epoch": 5.278246205733558, + "grad_norm": 0.0047840481566283006, + "learning_rate": 1.7336963545855118e-05, + "loss": 0.0001, + "step": 9390 + }, + { + "epoch": 5.283867341202923, + "grad_norm": 0.0028731485160253018, + "learning_rate": 1.7230970687657826e-05, + "loss": 0.0001, + "step": 9400 + }, + { + "epoch": 5.289488476672288, + "grad_norm": 0.004755296673482084, + "learning_rate": 1.7125235333646877e-05, + "loss": 0.0001, + "step": 9410 + }, + { + "epoch": 5.295109612141653, + "grad_norm": 0.004520339233829479, + "learning_rate": 1.7019758314707347e-05, + "loss": 0.0001, + "step": 9420 + }, + { + "epoch": 5.300730747611017, + "grad_norm": 0.007508796579818705, + "learning_rate": 1.6914540459694252e-05, + "loss": 0.0, + "step": 9430 + }, + { + "epoch": 5.306351883080382, + "grad_norm": 0.005296297537101901, + "learning_rate": 1.680958259542602e-05, + "loss": 0.0001, + "step": 9440 + }, + { + "epoch": 5.311973018549747, + "grad_norm": 0.015279895446474633, + "learning_rate": 1.670488554667809e-05, + "loss": 0.0001, + "step": 9450 + }, + { + "epoch": 5.317594154019112, + "grad_norm": 0.018788880889209363, + "learning_rate": 1.66004501361763e-05, + "loss": 0.0001, + "step": 9460 + }, + { + "epoch": 5.323215289488477, + "grad_norm": 0.0030296194925682955, + "learning_rate": 1.6496277184590563e-05, + "loss": 0.0001, + "step": 9470 + }, + { + "epoch": 5.328836424957841, + "grad_norm": 0.005519827553427454, + "learning_rate": 1.6392367510528294e-05, + "loss": 0.0001, + "step": 9480 + }, + { + "epoch": 5.334457560427206, + "grad_norm": 0.004211436359007738, + "learning_rate": 1.6288721930528068e-05, + "loss": 0.0001, + "step": 9490 + }, + { + "epoch": 5.340078695896571, + "grad_norm": 0.010778026435103267, + "learning_rate": 1.6185341259053176e-05, + "loss": 0.0001, + "step": 9500 + }, + { + "epoch": 5.345699831365936, + "grad_norm": 0.004882940567633415, + "learning_rate": 1.608222630848515e-05, + "loss": 0.0001, + "step": 9510 + }, + { + "epoch": 5.351320966835301, + "grad_norm": 0.007766610345808007, + "learning_rate": 1.5979377889117503e-05, + "loss": 0.0001, + "step": 9520 + }, + { + "epoch": 5.356942102304665, + "grad_norm": 0.002932005753073856, + "learning_rate": 1.5876796809149313e-05, + "loss": 0.0001, + "step": 9530 + }, + { + "epoch": 5.36256323777403, + "grad_norm": 0.0017596947749754773, + "learning_rate": 1.5774483874678815e-05, + "loss": 0.0, + "step": 9540 + }, + { + "epoch": 5.368184373243395, + "grad_norm": 0.014538485830191395, + "learning_rate": 1.5672439889697166e-05, + "loss": 0.0001, + "step": 9550 + }, + { + "epoch": 5.37380550871276, + "grad_norm": 0.006894801101759357, + "learning_rate": 1.5570665656082018e-05, + "loss": 0.0001, + "step": 9560 + }, + { + "epoch": 5.379426644182125, + "grad_norm": 0.014125692899505632, + "learning_rate": 1.5469161973591346e-05, + "loss": 0.0001, + "step": 9570 + }, + { + "epoch": 5.385047779651489, + "grad_norm": 0.011600880614687736, + "learning_rate": 1.5367929639857005e-05, + "loss": 0.0001, + "step": 9580 + }, + { + "epoch": 5.390668915120854, + "grad_norm": 0.008464648881377617, + "learning_rate": 1.5266969450378643e-05, + "loss": 0.0001, + "step": 9590 + }, + { + "epoch": 5.39629005059022, + "grad_norm": 0.010138748417697779, + "learning_rate": 1.5166282198517279e-05, + "loss": 0.0001, + "step": 9600 + }, + { + "epoch": 5.401911186059584, + "grad_norm": 0.007045531646175274, + "learning_rate": 1.506586867548917e-05, + "loss": 0.0001, + "step": 9610 + }, + { + "epoch": 5.407532321528949, + "grad_norm": 0.016258357386644766, + "learning_rate": 1.4965729670359585e-05, + "loss": 0.0001, + "step": 9620 + }, + { + "epoch": 5.413153456998313, + "grad_norm": 0.004337566469948641, + "learning_rate": 1.486586597003659e-05, + "loss": 0.0001, + "step": 9630 + }, + { + "epoch": 5.4187745924676785, + "grad_norm": 0.011426518415032711, + "learning_rate": 1.4766278359264818e-05, + "loss": 0.0001, + "step": 9640 + }, + { + "epoch": 5.424395727937044, + "grad_norm": 0.009658496475203985, + "learning_rate": 1.4666967620619416e-05, + "loss": 0.0001, + "step": 9650 + }, + { + "epoch": 5.430016863406408, + "grad_norm": 0.01655319518096919, + "learning_rate": 1.4567934534499761e-05, + "loss": 0.0001, + "step": 9660 + }, + { + "epoch": 5.435637998875773, + "grad_norm": 0.021233474923124915, + "learning_rate": 1.4469179879123418e-05, + "loss": 0.0001, + "step": 9670 + }, + { + "epoch": 5.441259134345137, + "grad_norm": 0.01615159637096921, + "learning_rate": 1.4370704430520005e-05, + "loss": 0.0001, + "step": 9680 + }, + { + "epoch": 5.4468802698145025, + "grad_norm": 0.0058254457691027306, + "learning_rate": 1.4272508962525105e-05, + "loss": 0.0, + "step": 9690 + }, + { + "epoch": 5.452501405283868, + "grad_norm": 0.00466173148657289, + "learning_rate": 1.4174594246774142e-05, + "loss": 0.0001, + "step": 9700 + }, + { + "epoch": 5.458122540753232, + "grad_norm": 0.012301892695848636, + "learning_rate": 1.4076961052696324e-05, + "loss": 0.0001, + "step": 9710 + }, + { + "epoch": 5.463743676222597, + "grad_norm": 0.012103336289657901, + "learning_rate": 1.397961014750867e-05, + "loss": 0.0001, + "step": 9720 + }, + { + "epoch": 5.469364811691962, + "grad_norm": 0.010634620352713599, + "learning_rate": 1.3882542296209922e-05, + "loss": 0.0, + "step": 9730 + }, + { + "epoch": 5.4749859471613265, + "grad_norm": 0.010513895389970824, + "learning_rate": 1.3785758261574489e-05, + "loss": 0.0001, + "step": 9740 + }, + { + "epoch": 5.480607082630692, + "grad_norm": 0.0031957141319880005, + "learning_rate": 1.3689258804146566e-05, + "loss": 0.0, + "step": 9750 + }, + { + "epoch": 5.486228218100056, + "grad_norm": 0.004289473680201465, + "learning_rate": 1.3593044682234052e-05, + "loss": 0.0, + "step": 9760 + }, + { + "epoch": 5.491849353569421, + "grad_norm": 0.009026507229871182, + "learning_rate": 1.3497116651902642e-05, + "loss": 0.0001, + "step": 9770 + }, + { + "epoch": 5.497470489038786, + "grad_norm": 0.016479374961959867, + "learning_rate": 1.3401475466969893e-05, + "loss": 0.0001, + "step": 9780 + }, + { + "epoch": 5.5030916245081505, + "grad_norm": 0.00973726008723872, + "learning_rate": 1.33061218789993e-05, + "loss": 0.0001, + "step": 9790 + }, + { + "epoch": 5.508712759977516, + "grad_norm": 0.0048188340406655076, + "learning_rate": 1.3211056637294333e-05, + "loss": 0.0, + "step": 9800 + }, + { + "epoch": 5.51433389544688, + "grad_norm": 0.001940452008768415, + "learning_rate": 1.3116280488892608e-05, + "loss": 0.0001, + "step": 9810 + }, + { + "epoch": 5.519955030916245, + "grad_norm": 0.012397881082008299, + "learning_rate": 1.3021794178560025e-05, + "loss": 0.0001, + "step": 9820 + }, + { + "epoch": 5.52557616638561, + "grad_norm": 0.01167210422981983, + "learning_rate": 1.292759844878485e-05, + "loss": 0.0001, + "step": 9830 + }, + { + "epoch": 5.5311973018549745, + "grad_norm": 0.003723841389674529, + "learning_rate": 1.2833694039771953e-05, + "loss": 0.0, + "step": 9840 + }, + { + "epoch": 5.53681843732434, + "grad_norm": 0.011516792374199423, + "learning_rate": 1.2740081689436972e-05, + "loss": 0.0001, + "step": 9850 + }, + { + "epoch": 5.542439572793704, + "grad_norm": 0.009340743067464711, + "learning_rate": 1.2646762133400459e-05, + "loss": 0.0, + "step": 9860 + }, + { + "epoch": 5.548060708263069, + "grad_norm": 0.01609922281048168, + "learning_rate": 1.2553736104982144e-05, + "loss": 0.0001, + "step": 9870 + }, + { + "epoch": 5.553681843732434, + "grad_norm": 0.003589013326092523, + "learning_rate": 1.2461004335195214e-05, + "loss": 0.0001, + "step": 9880 + }, + { + "epoch": 5.5593029792017985, + "grad_norm": 0.008708616010894553, + "learning_rate": 1.2368567552740511e-05, + "loss": 0.0001, + "step": 9890 + }, + { + "epoch": 5.564924114671164, + "grad_norm": 0.0046745726441450525, + "learning_rate": 1.2276426484000797e-05, + "loss": 0.0, + "step": 9900 + }, + { + "epoch": 5.570545250140528, + "grad_norm": 0.003177696545995521, + "learning_rate": 1.2184581853035065e-05, + "loss": 0.0, + "step": 9910 + }, + { + "epoch": 5.576166385609893, + "grad_norm": 0.01331154050125917, + "learning_rate": 1.2093034381572899e-05, + "loss": 0.0001, + "step": 9920 + }, + { + "epoch": 5.581787521079258, + "grad_norm": 0.0009025285155647361, + "learning_rate": 1.2001784789008707e-05, + "loss": 0.0001, + "step": 9930 + }, + { + "epoch": 5.5874086565486225, + "grad_norm": 0.009260805527454553, + "learning_rate": 1.1910833792396147e-05, + "loss": 0.0001, + "step": 9940 + }, + { + "epoch": 5.593029792017988, + "grad_norm": 0.01256191913523711, + "learning_rate": 1.1820182106442473e-05, + "loss": 0.0, + "step": 9950 + }, + { + "epoch": 5.598650927487353, + "grad_norm": 0.003689925930113079, + "learning_rate": 1.1729830443502876e-05, + "loss": 0.0001, + "step": 9960 + }, + { + "epoch": 5.604272062956717, + "grad_norm": 0.01593166011622193, + "learning_rate": 1.1639779513574922e-05, + "loss": 0.0001, + "step": 9970 + }, + { + "epoch": 5.609893198426082, + "grad_norm": 0.008746198404705492, + "learning_rate": 1.1550030024292996e-05, + "loss": 0.0, + "step": 9980 + }, + { + "epoch": 5.6155143338954465, + "grad_norm": 0.01670774847204932, + "learning_rate": 1.1460582680922711e-05, + "loss": 0.0001, + "step": 9990 + }, + { + "epoch": 5.621135469364812, + "grad_norm": 0.0030732052076573486, + "learning_rate": 1.137143818635535e-05, + "loss": 0.0, + "step": 10000 + }, + { + "epoch": 5.626756604834177, + "grad_norm": 0.0009268577358015034, + "learning_rate": 1.1282597241102354e-05, + "loss": 0.0001, + "step": 10010 + }, + { + "epoch": 5.632377740303541, + "grad_norm": 0.002529733721835241, + "learning_rate": 1.1194060543289876e-05, + "loss": 0.0, + "step": 10020 + }, + { + "epoch": 5.637998875772906, + "grad_norm": 0.0015358334377357312, + "learning_rate": 1.1105828788653183e-05, + "loss": 0.0, + "step": 10030 + }, + { + "epoch": 5.643620011242271, + "grad_norm": 0.014191480991528633, + "learning_rate": 1.1017902670531294e-05, + "loss": 0.0, + "step": 10040 + }, + { + "epoch": 5.649241146711636, + "grad_norm": 0.009602844474403873, + "learning_rate": 1.0930282879861486e-05, + "loss": 0.0, + "step": 10050 + }, + { + "epoch": 5.654862282181001, + "grad_norm": 0.010379890553338624, + "learning_rate": 1.0842970105173849e-05, + "loss": 0.0, + "step": 10060 + }, + { + "epoch": 5.660483417650365, + "grad_norm": 0.0019316887963907402, + "learning_rate": 1.075596503258588e-05, + "loss": 0.0, + "step": 10070 + }, + { + "epoch": 5.66610455311973, + "grad_norm": 0.011383246448301267, + "learning_rate": 1.0669268345797156e-05, + "loss": 0.0, + "step": 10080 + }, + { + "epoch": 5.671725688589095, + "grad_norm": 0.004582652380958405, + "learning_rate": 1.0582880726083844e-05, + "loss": 0.0, + "step": 10090 + }, + { + "epoch": 5.67734682405846, + "grad_norm": 0.0060924394782012605, + "learning_rate": 1.0496802852293476e-05, + "loss": 0.0, + "step": 10100 + }, + { + "epoch": 5.682967959527825, + "grad_norm": 0.005185275224032428, + "learning_rate": 1.0411035400839497e-05, + "loss": 0.0, + "step": 10110 + }, + { + "epoch": 5.688589094997189, + "grad_norm": 0.005031727463284729, + "learning_rate": 1.032557904569606e-05, + "loss": 0.0, + "step": 10120 + }, + { + "epoch": 5.694210230466554, + "grad_norm": 0.019110082083012226, + "learning_rate": 1.0240434458392617e-05, + "loss": 0.0, + "step": 10130 + }, + { + "epoch": 5.699831365935919, + "grad_norm": 0.004194345137679298, + "learning_rate": 1.0155602308008732e-05, + "loss": 0.0, + "step": 10140 + }, + { + "epoch": 5.705452501405284, + "grad_norm": 0.01087180761754166, + "learning_rate": 1.0071083261168824e-05, + "loss": 0.0001, + "step": 10150 + }, + { + "epoch": 5.711073636874649, + "grad_norm": 0.000752290108658082, + "learning_rate": 9.986877982036797e-06, + "loss": 0.0, + "step": 10160 + }, + { + "epoch": 5.716694772344013, + "grad_norm": 0.0008118412633924574, + "learning_rate": 9.902987132311003e-06, + "loss": 0.0, + "step": 10170 + }, + { + "epoch": 5.722315907813378, + "grad_norm": 0.010368363878047909, + "learning_rate": 9.819411371218957e-06, + "loss": 0.0, + "step": 10180 + }, + { + "epoch": 5.7279370432827434, + "grad_norm": 0.006481962088390579, + "learning_rate": 9.73615135551209e-06, + "loss": 0.0, + "step": 10190 + }, + { + "epoch": 5.733558178752108, + "grad_norm": 0.0032518205075397944, + "learning_rate": 9.65320773946075e-06, + "loss": 0.0, + "step": 10200 + }, + { + "epoch": 5.739179314221473, + "grad_norm": 0.011067964708585305, + "learning_rate": 9.570581174848875e-06, + "loss": 0.0001, + "step": 10210 + }, + { + "epoch": 5.744800449690837, + "grad_norm": 0.010262125007067071, + "learning_rate": 9.488272310969038e-06, + "loss": 0.0, + "step": 10220 + }, + { + "epoch": 5.750421585160202, + "grad_norm": 0.001259623026545284, + "learning_rate": 9.406281794617205e-06, + "loss": 0.0001, + "step": 10230 + }, + { + "epoch": 5.7560427206295675, + "grad_norm": 0.01034390015833188, + "learning_rate": 9.324610270087774e-06, + "loss": 0.0, + "step": 10240 + }, + { + "epoch": 5.761663856098932, + "grad_norm": 0.026040788615119274, + "learning_rate": 9.243258379168446e-06, + "loss": 0.0, + "step": 10250 + }, + { + "epoch": 5.767284991568297, + "grad_norm": 0.0014373808597380373, + "learning_rate": 9.162226761135129e-06, + "loss": 0.0, + "step": 10260 + }, + { + "epoch": 5.772906127037661, + "grad_norm": 0.0016775320985968655, + "learning_rate": 9.081516052747047e-06, + "loss": 0.0, + "step": 10270 + }, + { + "epoch": 5.778527262507026, + "grad_norm": 0.009116805880355572, + "learning_rate": 9.001126888241651e-06, + "loss": 0.0, + "step": 10280 + }, + { + "epoch": 5.7841483979763915, + "grad_norm": 0.006368437122005412, + "learning_rate": 8.921059899329615e-06, + "loss": 0.0, + "step": 10290 + }, + { + "epoch": 5.789769533445756, + "grad_norm": 0.006370043278133735, + "learning_rate": 8.841315715189957e-06, + "loss": 0.0, + "step": 10300 + }, + { + "epoch": 5.795390668915121, + "grad_norm": 0.0008373835663513255, + "learning_rate": 8.761894962464995e-06, + "loss": 0.0001, + "step": 10310 + }, + { + "epoch": 5.801011804384486, + "grad_norm": 0.009848467204133933, + "learning_rate": 8.68279826525552e-06, + "loss": 0.0001, + "step": 10320 + }, + { + "epoch": 5.80663293985385, + "grad_norm": 0.0009984494423281116, + "learning_rate": 8.604026245115799e-06, + "loss": 0.0, + "step": 10330 + }, + { + "epoch": 5.8122540753232155, + "grad_norm": 0.005652129842337195, + "learning_rate": 8.525579521048766e-06, + "loss": 0.0, + "step": 10340 + }, + { + "epoch": 5.81787521079258, + "grad_norm": 0.006644396823546608, + "learning_rate": 8.44745870950112e-06, + "loss": 0.0, + "step": 10350 + }, + { + "epoch": 5.823496346261945, + "grad_norm": 0.006444551198979287, + "learning_rate": 8.369664424358464e-06, + "loss": 0.0, + "step": 10360 + }, + { + "epoch": 5.82911748173131, + "grad_norm": 0.0014485940727860442, + "learning_rate": 8.292197276940544e-06, + "loss": 0.0, + "step": 10370 + }, + { + "epoch": 5.834738617200674, + "grad_norm": 0.0011250507724268207, + "learning_rate": 8.215057875996401e-06, + "loss": 0.0, + "step": 10380 + }, + { + "epoch": 5.8403597526700395, + "grad_norm": 0.004649128708798416, + "learning_rate": 8.138246827699548e-06, + "loss": 0.0, + "step": 10390 + }, + { + "epoch": 5.845980888139405, + "grad_norm": 0.0019531718516151925, + "learning_rate": 8.061764735643306e-06, + "loss": 0.0, + "step": 10400 + }, + { + "epoch": 5.851602023608769, + "grad_norm": 0.008266429708395872, + "learning_rate": 7.985612200836e-06, + "loss": 0.0, + "step": 10410 + }, + { + "epoch": 5.857223159078134, + "grad_norm": 0.012736147760529011, + "learning_rate": 7.90978982169619e-06, + "loss": 0.0, + "step": 10420 + }, + { + "epoch": 5.862844294547498, + "grad_norm": 0.010240944420797708, + "learning_rate": 7.834298194048062e-06, + "loss": 0.0, + "step": 10430 + }, + { + "epoch": 5.8684654300168635, + "grad_norm": 0.017077585902826424, + "learning_rate": 7.759137911116732e-06, + "loss": 0.0, + "step": 10440 + }, + { + "epoch": 5.874086565486229, + "grad_norm": 0.0011400108361340135, + "learning_rate": 7.684309563523484e-06, + "loss": 0.0, + "step": 10450 + }, + { + "epoch": 5.879707700955593, + "grad_norm": 0.0013764780116337046, + "learning_rate": 7.609813739281274e-06, + "loss": 0.0, + "step": 10460 + }, + { + "epoch": 5.885328836424958, + "grad_norm": 0.01678962722416973, + "learning_rate": 7.535651023789986e-06, + "loss": 0.0, + "step": 10470 + }, + { + "epoch": 5.890949971894322, + "grad_norm": 0.0017520451490175162, + "learning_rate": 7.4618219998319136e-06, + "loss": 0.0, + "step": 10480 + }, + { + "epoch": 5.8965711073636875, + "grad_norm": 0.0014924681731851405, + "learning_rate": 7.388327247567134e-06, + "loss": 0.0, + "step": 10490 + }, + { + "epoch": 5.902192242833053, + "grad_norm": 0.01620711756198989, + "learning_rate": 7.315167344528967e-06, + "loss": 0.0001, + "step": 10500 + }, + { + "epoch": 5.907813378302417, + "grad_norm": 0.0011602429617979773, + "learning_rate": 7.242342865619467e-06, + "loss": 0.0, + "step": 10510 + }, + { + "epoch": 5.913434513771782, + "grad_norm": 0.012337594275431522, + "learning_rate": 7.169854383104796e-06, + "loss": 0.0, + "step": 10520 + }, + { + "epoch": 5.919055649241146, + "grad_norm": 0.0007826981137096052, + "learning_rate": 7.097702466610873e-06, + "loss": 0.0, + "step": 10530 + }, + { + "epoch": 5.9246767847105115, + "grad_norm": 0.0007883241748295494, + "learning_rate": 7.02588768311882e-06, + "loss": 0.0, + "step": 10540 + }, + { + "epoch": 5.930297920179877, + "grad_norm": 0.0006644432549583892, + "learning_rate": 6.954410596960476e-06, + "loss": 0.0, + "step": 10550 + }, + { + "epoch": 5.935919055649241, + "grad_norm": 0.000650703543680648, + "learning_rate": 6.883271769814048e-06, + "loss": 0.0, + "step": 10560 + }, + { + "epoch": 5.941540191118606, + "grad_norm": 0.012764105908429777, + "learning_rate": 6.812471760699618e-06, + "loss": 0.0, + "step": 10570 + }, + { + "epoch": 5.94716132658797, + "grad_norm": 0.011615813745783046, + "learning_rate": 6.742011125974785e-06, + "loss": 0.0, + "step": 10580 + }, + { + "epoch": 5.9527824620573355, + "grad_norm": 0.0004763158084680763, + "learning_rate": 6.671890419330307e-06, + "loss": 0.0, + "step": 10590 + }, + { + "epoch": 5.958403597526701, + "grad_norm": 0.01350431034579878, + "learning_rate": 6.6021101917857434e-06, + "loss": 0.0, + "step": 10600 + }, + { + "epoch": 5.964024732996065, + "grad_norm": 0.01586751333077214, + "learning_rate": 6.532670991685075e-06, + "loss": 0.0001, + "step": 10610 + }, + { + "epoch": 5.96964586846543, + "grad_norm": 0.0017659464508108288, + "learning_rate": 6.463573364692454e-06, + "loss": 0.0, + "step": 10620 + }, + { + "epoch": 5.975267003934794, + "grad_norm": 0.007968317210711666, + "learning_rate": 6.39481785378791e-06, + "loss": 0.0, + "step": 10630 + }, + { + "epoch": 5.9808881394041595, + "grad_norm": 0.0021138099661355794, + "learning_rate": 6.32640499926308e-06, + "loss": 0.0, + "step": 10640 + }, + { + "epoch": 5.986509274873525, + "grad_norm": 0.010556005361465481, + "learning_rate": 6.258335338716903e-06, + "loss": 0.0, + "step": 10650 + }, + { + "epoch": 5.992130410342889, + "grad_norm": 0.0007517944847116383, + "learning_rate": 6.190609407051501e-06, + "loss": 0.0, + "step": 10660 + }, + { + "epoch": 5.997751545812254, + "grad_norm": 0.01945580735544116, + "learning_rate": 6.12322773646789e-06, + "loss": 0.0, + "step": 10670 + }, + { + "epoch": 6.003372681281619, + "grad_norm": 0.0014983074140552343, + "learning_rate": 6.05619085646183e-06, + "loss": 0.0, + "step": 10680 + }, + { + "epoch": 6.0089938167509835, + "grad_norm": 0.0009223743664836045, + "learning_rate": 5.989499293819667e-06, + "loss": 0.0, + "step": 10690 + }, + { + "epoch": 6.014614952220349, + "grad_norm": 0.0069987640174615515, + "learning_rate": 5.923153572614204e-06, + "loss": 0.0, + "step": 10700 + }, + { + "epoch": 6.020236087689713, + "grad_norm": 0.014028010942940764, + "learning_rate": 5.85715421420055e-06, + "loss": 0.0, + "step": 10710 + }, + { + "epoch": 6.025857223159078, + "grad_norm": 0.009071220172560939, + "learning_rate": 5.7915017372120236e-06, + "loss": 0.0, + "step": 10720 + }, + { + "epoch": 6.031478358628443, + "grad_norm": 0.00031089934348448755, + "learning_rate": 5.726196657556132e-06, + "loss": 0.0, + "step": 10730 + }, + { + "epoch": 6.0370994940978076, + "grad_norm": 0.0004636519525957024, + "learning_rate": 5.661239488410475e-06, + "loss": 0.0, + "step": 10740 + }, + { + "epoch": 6.042720629567173, + "grad_norm": 0.0008492633467326528, + "learning_rate": 5.596630740218678e-06, + "loss": 0.0, + "step": 10750 + }, + { + "epoch": 6.048341765036537, + "grad_norm": 0.0006252622204856705, + "learning_rate": 5.532370920686464e-06, + "loss": 0.0, + "step": 10760 + }, + { + "epoch": 6.053962900505902, + "grad_norm": 0.0003542017082586836, + "learning_rate": 5.468460534777598e-06, + "loss": 0.0, + "step": 10770 + }, + { + "epoch": 6.059584035975267, + "grad_norm": 0.008654840872354522, + "learning_rate": 5.404900084709935e-06, + "loss": 0.0, + "step": 10780 + }, + { + "epoch": 6.065205171444632, + "grad_norm": 0.0013398280424392556, + "learning_rate": 5.341690069951494e-06, + "loss": 0.0, + "step": 10790 + }, + { + "epoch": 6.070826306913997, + "grad_norm": 0.0006325227630127055, + "learning_rate": 5.278830987216526e-06, + "loss": 0.0, + "step": 10800 + } + ], + "logging_steps": 10, + "max_steps": 12453, + "num_input_tokens_seen": 0, + "num_train_epochs": 7, + "save_steps": 1800, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 880742603685888.0, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +}