{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.8311420877554254, "eval_steps": 500, "global_step": 57800, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003168065895770632, "grad_norm": 72.35028076171875, "learning_rate": 4.2238648363252377e-07, "loss": 1.3479, "step": 100 }, { "epoch": 0.006336131791541264, "grad_norm": 60.59146499633789, "learning_rate": 8.447729672650475e-07, "loss": 1.3106, "step": 200 }, { "epoch": 0.009504197687311896, "grad_norm": 23.751380920410156, "learning_rate": 1.2671594508975712e-06, "loss": 1.3496, "step": 300 }, { "epoch": 0.012672263583082528, "grad_norm": 38.346038818359375, "learning_rate": 1.689545934530095e-06, "loss": 1.1463, "step": 400 }, { "epoch": 0.01584032947885316, "grad_norm": 153.24330139160156, "learning_rate": 2.111932418162619e-06, "loss": 0.7624, "step": 500 }, { "epoch": 0.01900839537462379, "grad_norm": 992.2465209960938, "learning_rate": 2.5343189017951423e-06, "loss": 0.6289, "step": 600 }, { "epoch": 0.022176461270394424, "grad_norm": 142.35829162597656, "learning_rate": 2.956705385427667e-06, "loss": 0.5052, "step": 700 }, { "epoch": 0.025344527166165056, "grad_norm": 110.47644805908203, "learning_rate": 3.37909186906019e-06, "loss": 0.5615, "step": 800 }, { "epoch": 0.02851259306193569, "grad_norm": 82.66803741455078, "learning_rate": 3.801478352692714e-06, "loss": 0.2871, "step": 900 }, { "epoch": 0.03168065895770632, "grad_norm": 10.453248977661133, "learning_rate": 4.223864836325238e-06, "loss": 0.4623, "step": 1000 }, { "epoch": 0.03484872485347695, "grad_norm": 151.24656677246094, "learning_rate": 4.646251319957762e-06, "loss": 0.5214, "step": 1100 }, { "epoch": 0.03801679074924758, "grad_norm": 1.533247947692871, "learning_rate": 5.068637803590285e-06, "loss": 0.4097, "step": 1200 }, { "epoch": 0.04118485664501822, "grad_norm": 8.166175842285156, "learning_rate": 5.49102428722281e-06, "loss": 0.4068, "step": 1300 }, { "epoch": 0.04435292254078885, "grad_norm": 0.7004966139793396, "learning_rate": 5.913410770855334e-06, "loss": 0.2873, "step": 1400 }, { "epoch": 0.04752098843655948, "grad_norm": 158.9846954345703, "learning_rate": 6.335797254487856e-06, "loss": 0.3133, "step": 1500 }, { "epoch": 0.05068905433233011, "grad_norm": 0.08742424100637436, "learning_rate": 6.75818373812038e-06, "loss": 0.1869, "step": 1600 }, { "epoch": 0.05385712022810074, "grad_norm": 0.5520446300506592, "learning_rate": 7.180570221752905e-06, "loss": 0.3004, "step": 1700 }, { "epoch": 0.05702518612387138, "grad_norm": 1.164171576499939, "learning_rate": 7.602956705385428e-06, "loss": 0.2164, "step": 1800 }, { "epoch": 0.060193252019642006, "grad_norm": 0.2625727355480194, "learning_rate": 8.025343189017952e-06, "loss": 0.293, "step": 1900 }, { "epoch": 0.06336131791541263, "grad_norm": 1.2144228219985962, "learning_rate": 8.447729672650476e-06, "loss": 0.17, "step": 2000 }, { "epoch": 0.06652938381118327, "grad_norm": 1.7134082317352295, "learning_rate": 8.870116156283e-06, "loss": 0.1669, "step": 2100 }, { "epoch": 0.0696974497069539, "grad_norm": 0.016615629196166992, "learning_rate": 9.292502639915524e-06, "loss": 0.1337, "step": 2200 }, { "epoch": 0.07286551560272454, "grad_norm": 3.4445388317108154, "learning_rate": 9.714889123548047e-06, "loss": 0.2076, "step": 2300 }, { "epoch": 0.07603358149849516, "grad_norm": 34.18547821044922, "learning_rate": 1.013727560718057e-05, "loss": 0.2348, "step": 2400 }, { "epoch": 0.0792016473942658, "grad_norm": 0.08411699533462524, "learning_rate": 1.0559662090813093e-05, "loss": 0.2016, "step": 2500 }, { "epoch": 0.08236971329003644, "grad_norm": 0.04223780706524849, "learning_rate": 1.098204857444562e-05, "loss": 0.1139, "step": 2600 }, { "epoch": 0.08553777918580706, "grad_norm": 80.90877532958984, "learning_rate": 1.1404435058078143e-05, "loss": 0.2098, "step": 2700 }, { "epoch": 0.0887058450815777, "grad_norm": 51.48539352416992, "learning_rate": 1.1826821541710668e-05, "loss": 0.1562, "step": 2800 }, { "epoch": 0.09187391097734833, "grad_norm": 0.06566716730594635, "learning_rate": 1.2249208025343192e-05, "loss": 0.1301, "step": 2900 }, { "epoch": 0.09504197687311897, "grad_norm": 0.04454626142978668, "learning_rate": 1.2671594508975712e-05, "loss": 0.1271, "step": 3000 }, { "epoch": 0.09821004276888959, "grad_norm": 0.03954063355922699, "learning_rate": 1.3093980992608237e-05, "loss": 0.1226, "step": 3100 }, { "epoch": 0.10137810866466022, "grad_norm": 0.00102125178091228, "learning_rate": 1.351636747624076e-05, "loss": 0.147, "step": 3200 }, { "epoch": 0.10454617456043086, "grad_norm": 0.031660884618759155, "learning_rate": 1.3938753959873287e-05, "loss": 0.1135, "step": 3300 }, { "epoch": 0.10771424045620148, "grad_norm": 6.420830249786377, "learning_rate": 1.436114044350581e-05, "loss": 0.1004, "step": 3400 }, { "epoch": 0.11088230635197212, "grad_norm": 1.1331299543380737, "learning_rate": 1.4783526927138331e-05, "loss": 0.0795, "step": 3500 }, { "epoch": 0.11405037224774275, "grad_norm": 0.06878916919231415, "learning_rate": 1.5205913410770856e-05, "loss": 0.1719, "step": 3600 }, { "epoch": 0.11721843814351339, "grad_norm": 0.5843459963798523, "learning_rate": 1.562829989440338e-05, "loss": 0.1445, "step": 3700 }, { "epoch": 0.12038650403928401, "grad_norm": 30.580446243286133, "learning_rate": 1.6050686378035904e-05, "loss": 0.1206, "step": 3800 }, { "epoch": 0.12355456993505465, "grad_norm": 0.07918854057788849, "learning_rate": 1.647307286166843e-05, "loss": 0.0458, "step": 3900 }, { "epoch": 0.12672263583082527, "grad_norm": 0.2442696988582611, "learning_rate": 1.6895459345300952e-05, "loss": 0.1122, "step": 4000 }, { "epoch": 0.12989070172659592, "grad_norm": 14.285125732421875, "learning_rate": 1.7317845828933475e-05, "loss": 0.0483, "step": 4100 }, { "epoch": 0.13305876762236654, "grad_norm": 1.1958580017089844, "learning_rate": 1.7740232312566e-05, "loss": 0.0493, "step": 4200 }, { "epoch": 0.13622683351813716, "grad_norm": 0.0011618838179856539, "learning_rate": 1.8162618796198523e-05, "loss": 0.0883, "step": 4300 }, { "epoch": 0.1393948994139078, "grad_norm": 1.1263456344604492, "learning_rate": 1.858500527983105e-05, "loss": 0.0926, "step": 4400 }, { "epoch": 0.14256296530967844, "grad_norm": 89.57861328125, "learning_rate": 1.900739176346357e-05, "loss": 0.1196, "step": 4500 }, { "epoch": 0.14573103120544909, "grad_norm": 1.9628229141235352, "learning_rate": 1.9429778247096094e-05, "loss": 0.0793, "step": 4600 }, { "epoch": 0.1488990971012197, "grad_norm": 0.13689815998077393, "learning_rate": 1.985216473072862e-05, "loss": 0.1418, "step": 4700 }, { "epoch": 0.15206716299699033, "grad_norm": 20.380517959594727, "learning_rate": 1.998554913294798e-05, "loss": 0.1341, "step": 4800 }, { "epoch": 0.15523522889276098, "grad_norm": 0.024527639150619507, "learning_rate": 1.9963317029791022e-05, "loss": 0.106, "step": 4900 }, { "epoch": 0.1584032947885316, "grad_norm": 29.77529525756836, "learning_rate": 1.9941084926634062e-05, "loss": 0.1032, "step": 5000 }, { "epoch": 0.16157136068430222, "grad_norm": 2.2095439434051514, "learning_rate": 1.9918852823477102e-05, "loss": 0.0789, "step": 5100 }, { "epoch": 0.16473942658007287, "grad_norm": 1.2087385654449463, "learning_rate": 1.9896620720320142e-05, "loss": 0.0513, "step": 5200 }, { "epoch": 0.1679074924758435, "grad_norm": 0.045304879546165466, "learning_rate": 1.9874388617163185e-05, "loss": 0.0244, "step": 5300 }, { "epoch": 0.17107555837161412, "grad_norm": 0.39902210235595703, "learning_rate": 1.9852156514006226e-05, "loss": 0.0621, "step": 5400 }, { "epoch": 0.17424362426738477, "grad_norm": 0.0002864371635951102, "learning_rate": 1.982992441084927e-05, "loss": 0.0301, "step": 5500 }, { "epoch": 0.1774116901631554, "grad_norm": 0.859196126461029, "learning_rate": 1.980769230769231e-05, "loss": 0.0456, "step": 5600 }, { "epoch": 0.18057975605892604, "grad_norm": 0.13034579157829285, "learning_rate": 1.9785460204535353e-05, "loss": 0.0444, "step": 5700 }, { "epoch": 0.18374782195469666, "grad_norm": 0.06543830782175064, "learning_rate": 1.9763228101378393e-05, "loss": 0.0573, "step": 5800 }, { "epoch": 0.18691588785046728, "grad_norm": 1.9895894527435303, "learning_rate": 1.9740995998221433e-05, "loss": 0.0635, "step": 5900 }, { "epoch": 0.19008395374623793, "grad_norm": 60.741798400878906, "learning_rate": 1.9718763895064473e-05, "loss": 0.1086, "step": 6000 }, { "epoch": 0.19325201964200855, "grad_norm": 0.4178132712841034, "learning_rate": 1.9696531791907516e-05, "loss": 0.1383, "step": 6100 }, { "epoch": 0.19642008553777918, "grad_norm": 0.05819793418049812, "learning_rate": 1.9674299688750556e-05, "loss": 0.1049, "step": 6200 }, { "epoch": 0.19958815143354983, "grad_norm": 0.0006314706988632679, "learning_rate": 1.96520675855936e-05, "loss": 0.0843, "step": 6300 }, { "epoch": 0.20275621732932045, "grad_norm": 0.1267533153295517, "learning_rate": 1.962983548243664e-05, "loss": 0.0458, "step": 6400 }, { "epoch": 0.20592428322509107, "grad_norm": 0.22171200811862946, "learning_rate": 1.960760337927968e-05, "loss": 0.059, "step": 6500 }, { "epoch": 0.20909234912086172, "grad_norm": 0.025029148906469345, "learning_rate": 1.9585371276122723e-05, "loss": 0.0269, "step": 6600 }, { "epoch": 0.21226041501663234, "grad_norm": 2.0198607444763184, "learning_rate": 1.9563139172965763e-05, "loss": 0.0417, "step": 6700 }, { "epoch": 0.21542848091240296, "grad_norm": 0.004318519961088896, "learning_rate": 1.9540907069808807e-05, "loss": 0.0593, "step": 6800 }, { "epoch": 0.21859654680817361, "grad_norm": 0.32084760069847107, "learning_rate": 1.9518674966651847e-05, "loss": 0.0534, "step": 6900 }, { "epoch": 0.22176461270394424, "grad_norm": 0.2787262797355652, "learning_rate": 1.949644286349489e-05, "loss": 0.0718, "step": 7000 }, { "epoch": 0.22493267859971489, "grad_norm": 0.10762894153594971, "learning_rate": 1.947421076033793e-05, "loss": 0.1301, "step": 7100 }, { "epoch": 0.2281007444954855, "grad_norm": 0.003302349941805005, "learning_rate": 1.945197865718097e-05, "loss": 0.0705, "step": 7200 }, { "epoch": 0.23126881039125613, "grad_norm": 0.007045666687190533, "learning_rate": 1.942974655402401e-05, "loss": 0.0492, "step": 7300 }, { "epoch": 0.23443687628702678, "grad_norm": 3.9791948795318604, "learning_rate": 1.9407514450867054e-05, "loss": 0.0908, "step": 7400 }, { "epoch": 0.2376049421827974, "grad_norm": 0.0009272198076359928, "learning_rate": 1.9385282347710094e-05, "loss": 0.0462, "step": 7500 }, { "epoch": 0.24077300807856802, "grad_norm": 0.009363643825054169, "learning_rate": 1.9363050244553138e-05, "loss": 0.0772, "step": 7600 }, { "epoch": 0.24394107397433867, "grad_norm": 9.51755428314209, "learning_rate": 1.9340818141396178e-05, "loss": 0.0906, "step": 7700 }, { "epoch": 0.2471091398701093, "grad_norm": 10.527387619018555, "learning_rate": 1.931858603823922e-05, "loss": 0.0739, "step": 7800 }, { "epoch": 0.2502772057658799, "grad_norm": 0.02538328990340233, "learning_rate": 1.9296353935082258e-05, "loss": 0.0325, "step": 7900 }, { "epoch": 0.25344527166165054, "grad_norm": 10.459150314331055, "learning_rate": 1.92741218319253e-05, "loss": 0.1081, "step": 8000 }, { "epoch": 0.2566133375574212, "grad_norm": 0.1348431557416916, "learning_rate": 1.925188972876834e-05, "loss": 0.0472, "step": 8100 }, { "epoch": 0.25978140345319184, "grad_norm": 0.001315874163992703, "learning_rate": 1.9229657625611385e-05, "loss": 0.0613, "step": 8200 }, { "epoch": 0.26294946934896246, "grad_norm": 0.011827241629362106, "learning_rate": 1.9207425522454425e-05, "loss": 0.0281, "step": 8300 }, { "epoch": 0.2661175352447331, "grad_norm": 0.468915730714798, "learning_rate": 1.918519341929747e-05, "loss": 0.0184, "step": 8400 }, { "epoch": 0.2692856011405037, "grad_norm": 2.4475345611572266, "learning_rate": 1.916296131614051e-05, "loss": 0.0447, "step": 8500 }, { "epoch": 0.2724536670362743, "grad_norm": 0.5275880694389343, "learning_rate": 1.9140729212983552e-05, "loss": 0.0609, "step": 8600 }, { "epoch": 0.275621732932045, "grad_norm": 0.11572447419166565, "learning_rate": 1.9118497109826592e-05, "loss": 0.0323, "step": 8700 }, { "epoch": 0.2787897988278156, "grad_norm": 0.04472646117210388, "learning_rate": 1.9096265006669632e-05, "loss": 0.0794, "step": 8800 }, { "epoch": 0.28195786472358625, "grad_norm": 0.0004972922033630311, "learning_rate": 1.9074032903512672e-05, "loss": 0.0477, "step": 8900 }, { "epoch": 0.28512593061935687, "grad_norm": 0.0010256944224238396, "learning_rate": 1.9051800800355716e-05, "loss": 0.0819, "step": 9000 }, { "epoch": 0.2882939965151275, "grad_norm": 0.0991922914981842, "learning_rate": 1.9029568697198756e-05, "loss": 0.0838, "step": 9100 }, { "epoch": 0.29146206241089817, "grad_norm": 0.15479788184165955, "learning_rate": 1.90073365940418e-05, "loss": 0.0512, "step": 9200 }, { "epoch": 0.2946301283066688, "grad_norm": 0.006891987752169371, "learning_rate": 1.898510449088484e-05, "loss": 0.0641, "step": 9300 }, { "epoch": 0.2977981942024394, "grad_norm": 1.7009308338165283, "learning_rate": 1.896287238772788e-05, "loss": 0.0549, "step": 9400 }, { "epoch": 0.30096626009821004, "grad_norm": 0.011433255858719349, "learning_rate": 1.8940640284570923e-05, "loss": 0.0328, "step": 9500 }, { "epoch": 0.30413432599398066, "grad_norm": 22.03217887878418, "learning_rate": 1.8918408181413963e-05, "loss": 0.0338, "step": 9600 }, { "epoch": 0.3073023918897513, "grad_norm": 0.015131734311580658, "learning_rate": 1.8896176078257006e-05, "loss": 0.0976, "step": 9700 }, { "epoch": 0.31047045778552196, "grad_norm": 0.03849724307656288, "learning_rate": 1.8873943975100046e-05, "loss": 0.0413, "step": 9800 }, { "epoch": 0.3136385236812926, "grad_norm": 0.00014183101302478462, "learning_rate": 1.885171187194309e-05, "loss": 0.0746, "step": 9900 }, { "epoch": 0.3168065895770632, "grad_norm": 17.10400390625, "learning_rate": 1.882947976878613e-05, "loss": 0.0683, "step": 10000 }, { "epoch": 0.3199746554728338, "grad_norm": 0.16147367656230927, "learning_rate": 1.880724766562917e-05, "loss": 0.0137, "step": 10100 }, { "epoch": 0.32314272136860445, "grad_norm": 0.0005398979410529137, "learning_rate": 1.878501556247221e-05, "loss": 0.0521, "step": 10200 }, { "epoch": 0.3263107872643751, "grad_norm": 0.00451250933110714, "learning_rate": 1.8762783459315253e-05, "loss": 0.0518, "step": 10300 }, { "epoch": 0.32947885316014575, "grad_norm": 0.0008030992466956377, "learning_rate": 1.8740551356158293e-05, "loss": 0.0764, "step": 10400 }, { "epoch": 0.33264691905591637, "grad_norm": 8.962018013000488, "learning_rate": 1.8718319253001337e-05, "loss": 0.0447, "step": 10500 }, { "epoch": 0.335814984951687, "grad_norm": 0.9032301902770996, "learning_rate": 1.8696087149844377e-05, "loss": 0.0698, "step": 10600 }, { "epoch": 0.3389830508474576, "grad_norm": 0.020043615251779556, "learning_rate": 1.867385504668742e-05, "loss": 0.0488, "step": 10700 }, { "epoch": 0.34215111674322823, "grad_norm": 0.6755151152610779, "learning_rate": 1.8651622943530457e-05, "loss": 0.0288, "step": 10800 }, { "epoch": 0.3453191826389989, "grad_norm": 0.13276956975460052, "learning_rate": 1.86293908403735e-05, "loss": 0.0155, "step": 10900 }, { "epoch": 0.34848724853476953, "grad_norm": 6.229333400726318, "learning_rate": 1.860715873721654e-05, "loss": 0.0443, "step": 11000 }, { "epoch": 0.35165531443054016, "grad_norm": 0.03049282357096672, "learning_rate": 1.8584926634059584e-05, "loss": 0.0451, "step": 11100 }, { "epoch": 0.3548233803263108, "grad_norm": 0.00785783026367426, "learning_rate": 1.8562694530902624e-05, "loss": 0.0735, "step": 11200 }, { "epoch": 0.3579914462220814, "grad_norm": 0.026226060464978218, "learning_rate": 1.8540462427745668e-05, "loss": 0.0245, "step": 11300 }, { "epoch": 0.3611595121178521, "grad_norm": 0.24636198580265045, "learning_rate": 1.8518230324588708e-05, "loss": 0.0311, "step": 11400 }, { "epoch": 0.3643275780136227, "grad_norm": 5.352351188659668, "learning_rate": 1.8495998221431748e-05, "loss": 0.0565, "step": 11500 }, { "epoch": 0.3674956439093933, "grad_norm": 4.2338571802247316e-05, "learning_rate": 1.847376611827479e-05, "loss": 0.0447, "step": 11600 }, { "epoch": 0.37066370980516394, "grad_norm": 8.307737350463867, "learning_rate": 1.845153401511783e-05, "loss": 0.0785, "step": 11700 }, { "epoch": 0.37383177570093457, "grad_norm": 0.002329683629795909, "learning_rate": 1.842930191196087e-05, "loss": 0.0509, "step": 11800 }, { "epoch": 0.3769998415967052, "grad_norm": 0.031433336436748505, "learning_rate": 1.8407069808803915e-05, "loss": 0.0496, "step": 11900 }, { "epoch": 0.38016790749247587, "grad_norm": 1.186360239982605, "learning_rate": 1.8384837705646955e-05, "loss": 0.0482, "step": 12000 }, { "epoch": 0.3833359733882465, "grad_norm": 0.4914741516113281, "learning_rate": 1.836260560249e-05, "loss": 0.0174, "step": 12100 }, { "epoch": 0.3865040392840171, "grad_norm": 9.85021686553955, "learning_rate": 1.834037349933304e-05, "loss": 0.0665, "step": 12200 }, { "epoch": 0.38967210517978773, "grad_norm": 14.743742942810059, "learning_rate": 1.831814139617608e-05, "loss": 0.0475, "step": 12300 }, { "epoch": 0.39284017107555835, "grad_norm": 0.09932620823383331, "learning_rate": 1.8295909293019122e-05, "loss": 0.01, "step": 12400 }, { "epoch": 0.39600823697132903, "grad_norm": 0.031565092504024506, "learning_rate": 1.8273677189862162e-05, "loss": 0.0345, "step": 12500 }, { "epoch": 0.39917630286709965, "grad_norm": 0.653196394443512, "learning_rate": 1.8251445086705205e-05, "loss": 0.027, "step": 12600 }, { "epoch": 0.4023443687628703, "grad_norm": 0.03125124052166939, "learning_rate": 1.8229212983548246e-05, "loss": 0.0364, "step": 12700 }, { "epoch": 0.4055124346586409, "grad_norm": 0.6245060563087463, "learning_rate": 1.8206980880391286e-05, "loss": 0.0226, "step": 12800 }, { "epoch": 0.4086805005544115, "grad_norm": 0.22413401305675507, "learning_rate": 1.818474877723433e-05, "loss": 0.1074, "step": 12900 }, { "epoch": 0.41184856645018214, "grad_norm": 0.12312527745962143, "learning_rate": 1.816251667407737e-05, "loss": 0.0179, "step": 13000 }, { "epoch": 0.4150166323459528, "grad_norm": 0.28867945075035095, "learning_rate": 1.814028457092041e-05, "loss": 0.0377, "step": 13100 }, { "epoch": 0.41818469824172344, "grad_norm": 0.02332826890051365, "learning_rate": 1.8118052467763453e-05, "loss": 0.0384, "step": 13200 }, { "epoch": 0.42135276413749406, "grad_norm": 0.5182498693466187, "learning_rate": 1.8095820364606493e-05, "loss": 0.0309, "step": 13300 }, { "epoch": 0.4245208300332647, "grad_norm": 0.0012676503974944353, "learning_rate": 1.8073588261449536e-05, "loss": 0.0277, "step": 13400 }, { "epoch": 0.4276888959290353, "grad_norm": 2.4484052658081055, "learning_rate": 1.8051356158292576e-05, "loss": 0.0196, "step": 13500 }, { "epoch": 0.43085696182480593, "grad_norm": 0.02983473800122738, "learning_rate": 1.802912405513562e-05, "loss": 0.0386, "step": 13600 }, { "epoch": 0.4340250277205766, "grad_norm": 0.19795529544353485, "learning_rate": 1.8006891951978656e-05, "loss": 0.0135, "step": 13700 }, { "epoch": 0.43719309361634723, "grad_norm": 8.084446744760498e-05, "learning_rate": 1.79846598488217e-05, "loss": 0.0375, "step": 13800 }, { "epoch": 0.44036115951211785, "grad_norm": 0.00272519257850945, "learning_rate": 1.796242774566474e-05, "loss": 0.0583, "step": 13900 }, { "epoch": 0.44352922540788847, "grad_norm": 0.020339926704764366, "learning_rate": 1.7940195642507783e-05, "loss": 0.0175, "step": 14000 }, { "epoch": 0.4466972913036591, "grad_norm": 44.699337005615234, "learning_rate": 1.7917963539350823e-05, "loss": 0.0366, "step": 14100 }, { "epoch": 0.44986535719942977, "grad_norm": 0.0036636805161833763, "learning_rate": 1.7895731436193867e-05, "loss": 0.0445, "step": 14200 }, { "epoch": 0.4530334230952004, "grad_norm": 0.16195251047611237, "learning_rate": 1.7873499333036907e-05, "loss": 0.014, "step": 14300 }, { "epoch": 0.456201488990971, "grad_norm": 0.19172553718090057, "learning_rate": 1.7851267229879947e-05, "loss": 0.0369, "step": 14400 }, { "epoch": 0.45936955488674164, "grad_norm": 0.010684460401535034, "learning_rate": 1.782903512672299e-05, "loss": 0.0109, "step": 14500 }, { "epoch": 0.46253762078251226, "grad_norm": 0.44754496216773987, "learning_rate": 1.780680302356603e-05, "loss": 0.0151, "step": 14600 }, { "epoch": 0.4657056866782829, "grad_norm": 0.0005041993572376668, "learning_rate": 1.778457092040907e-05, "loss": 0.0487, "step": 14700 }, { "epoch": 0.46887375257405356, "grad_norm": 2.0641252994537354, "learning_rate": 1.7762338817252114e-05, "loss": 0.0166, "step": 14800 }, { "epoch": 0.4720418184698242, "grad_norm": 0.9774904251098633, "learning_rate": 1.7740106714095154e-05, "loss": 0.0047, "step": 14900 }, { "epoch": 0.4752098843655948, "grad_norm": 0.87712562084198, "learning_rate": 1.7717874610938198e-05, "loss": 0.0247, "step": 15000 }, { "epoch": 0.4783779502613654, "grad_norm": 0.18599839508533478, "learning_rate": 1.7695642507781238e-05, "loss": 0.0198, "step": 15100 }, { "epoch": 0.48154601615713605, "grad_norm": 0.010839251801371574, "learning_rate": 1.7673410404624278e-05, "loss": 0.0492, "step": 15200 }, { "epoch": 0.4847140820529067, "grad_norm": 0.00023437423806171864, "learning_rate": 1.765117830146732e-05, "loss": 0.027, "step": 15300 }, { "epoch": 0.48788214794867735, "grad_norm": 0.8362918496131897, "learning_rate": 1.762894619831036e-05, "loss": 0.0368, "step": 15400 }, { "epoch": 0.49105021384444797, "grad_norm": 0.00038645800668746233, "learning_rate": 1.7606714095153405e-05, "loss": 0.0072, "step": 15500 }, { "epoch": 0.4942182797402186, "grad_norm": 0.05926353856921196, "learning_rate": 1.7584481991996445e-05, "loss": 0.0448, "step": 15600 }, { "epoch": 0.4973863456359892, "grad_norm": 0.0021166419610381126, "learning_rate": 1.7562249888839485e-05, "loss": 0.0334, "step": 15700 }, { "epoch": 0.5005544115317598, "grad_norm": 0.08251558989286423, "learning_rate": 1.7540017785682525e-05, "loss": 0.0401, "step": 15800 }, { "epoch": 0.5037224774275305, "grad_norm": 0.012086935341358185, "learning_rate": 1.751778568252557e-05, "loss": 0.0158, "step": 15900 }, { "epoch": 0.5068905433233011, "grad_norm": 0.0009838377591222525, "learning_rate": 1.749555357936861e-05, "loss": 0.0247, "step": 16000 }, { "epoch": 0.5100586092190718, "grad_norm": 0.006506912410259247, "learning_rate": 1.7473321476211652e-05, "loss": 0.0452, "step": 16100 }, { "epoch": 0.5132266751148424, "grad_norm": 0.01391199603676796, "learning_rate": 1.7451089373054692e-05, "loss": 0.0337, "step": 16200 }, { "epoch": 0.516394741010613, "grad_norm": 0.0006601364002563059, "learning_rate": 1.7428857269897736e-05, "loss": 0.0106, "step": 16300 }, { "epoch": 0.5195628069063837, "grad_norm": 0.01070409081876278, "learning_rate": 1.7406625166740776e-05, "loss": 0.0105, "step": 16400 }, { "epoch": 0.5227308728021542, "grad_norm": 0.010202988050878048, "learning_rate": 1.7384393063583816e-05, "loss": 0.0167, "step": 16500 }, { "epoch": 0.5258989386979249, "grad_norm": 0.11668938398361206, "learning_rate": 1.7362160960426856e-05, "loss": 0.0104, "step": 16600 }, { "epoch": 0.5290670045936956, "grad_norm": 0.0031198032665997744, "learning_rate": 1.73399288572699e-05, "loss": 0.022, "step": 16700 }, { "epoch": 0.5322350704894662, "grad_norm": 3.408655881881714, "learning_rate": 1.731769675411294e-05, "loss": 0.0591, "step": 16800 }, { "epoch": 0.5354031363852368, "grad_norm": 0.013084974139928818, "learning_rate": 1.7295464650955983e-05, "loss": 0.0227, "step": 16900 }, { "epoch": 0.5385712022810074, "grad_norm": 0.22683855891227722, "learning_rate": 1.7273232547799023e-05, "loss": 0.0503, "step": 17000 }, { "epoch": 0.5417392681767781, "grad_norm": 0.005859915167093277, "learning_rate": 1.7251000444642066e-05, "loss": 0.0424, "step": 17100 }, { "epoch": 0.5449073340725487, "grad_norm": 6.267719745635986, "learning_rate": 1.7228768341485106e-05, "loss": 0.0185, "step": 17200 }, { "epoch": 0.5480753999683193, "grad_norm": 0.04545474424958229, "learning_rate": 1.7206536238328146e-05, "loss": 0.0174, "step": 17300 }, { "epoch": 0.55124346586409, "grad_norm": 0.008304772898554802, "learning_rate": 1.7184304135171186e-05, "loss": 0.0086, "step": 17400 }, { "epoch": 0.5544115317598606, "grad_norm": 0.0038544738199561834, "learning_rate": 1.716207203201423e-05, "loss": 0.0292, "step": 17500 }, { "epoch": 0.5575795976556313, "grad_norm": 0.009957126341760159, "learning_rate": 1.713983992885727e-05, "loss": 0.0072, "step": 17600 }, { "epoch": 0.5607476635514018, "grad_norm": 0.00033597013680264354, "learning_rate": 1.7117607825700313e-05, "loss": 0.0633, "step": 17700 }, { "epoch": 0.5639157294471725, "grad_norm": 0.026435261592268944, "learning_rate": 1.7095375722543353e-05, "loss": 0.0425, "step": 17800 }, { "epoch": 0.5670837953429432, "grad_norm": 0.029516736045479774, "learning_rate": 1.7073143619386397e-05, "loss": 0.0208, "step": 17900 }, { "epoch": 0.5702518612387137, "grad_norm": 0.00017197694978676736, "learning_rate": 1.7050911516229437e-05, "loss": 0.009, "step": 18000 }, { "epoch": 0.5734199271344844, "grad_norm": 0.0038332445546984673, "learning_rate": 1.7028679413072477e-05, "loss": 0.0394, "step": 18100 }, { "epoch": 0.576587993030255, "grad_norm": 0.008571327663958073, "learning_rate": 1.700644730991552e-05, "loss": 0.0096, "step": 18200 }, { "epoch": 0.5797560589260257, "grad_norm": 0.09438051283359528, "learning_rate": 1.698421520675856e-05, "loss": 0.0171, "step": 18300 }, { "epoch": 0.5829241248217963, "grad_norm": 0.06961628049612045, "learning_rate": 1.6961983103601604e-05, "loss": 0.0246, "step": 18400 }, { "epoch": 0.5860921907175669, "grad_norm": 0.022431883960962296, "learning_rate": 1.6939751000444644e-05, "loss": 0.0508, "step": 18500 }, { "epoch": 0.5892602566133376, "grad_norm": 0.05012532323598862, "learning_rate": 1.6917518897287684e-05, "loss": 0.0138, "step": 18600 }, { "epoch": 0.5924283225091082, "grad_norm": 0.11273948103189468, "learning_rate": 1.6895286794130724e-05, "loss": 0.0344, "step": 18700 }, { "epoch": 0.5955963884048788, "grad_norm": 2.972710371017456, "learning_rate": 1.6873054690973768e-05, "loss": 0.0345, "step": 18800 }, { "epoch": 0.5987644543006495, "grad_norm": 0.005833790171891451, "learning_rate": 1.6850822587816808e-05, "loss": 0.044, "step": 18900 }, { "epoch": 0.6019325201964201, "grad_norm": 0.002583844820037484, "learning_rate": 1.682859048465985e-05, "loss": 0.0234, "step": 19000 }, { "epoch": 0.6051005860921908, "grad_norm": 0.0027191785629838705, "learning_rate": 1.680635838150289e-05, "loss": 0.0118, "step": 19100 }, { "epoch": 0.6082686519879613, "grad_norm": 0.1673203706741333, "learning_rate": 1.6784126278345935e-05, "loss": 0.0333, "step": 19200 }, { "epoch": 0.611436717883732, "grad_norm": 3.3001066185534e-05, "learning_rate": 1.6761894175188975e-05, "loss": 0.0182, "step": 19300 }, { "epoch": 0.6146047837795026, "grad_norm": 1.063460111618042, "learning_rate": 1.6739662072032015e-05, "loss": 0.0115, "step": 19400 }, { "epoch": 0.6177728496752732, "grad_norm": 0.7981263399124146, "learning_rate": 1.6717429968875055e-05, "loss": 0.0321, "step": 19500 }, { "epoch": 0.6209409155710439, "grad_norm": 1.11823570728302, "learning_rate": 1.66951978657181e-05, "loss": 0.015, "step": 19600 }, { "epoch": 0.6241089814668145, "grad_norm": 43.64628219604492, "learning_rate": 1.667296576256114e-05, "loss": 0.0245, "step": 19700 }, { "epoch": 0.6272770473625852, "grad_norm": 0.006781121250241995, "learning_rate": 1.6650733659404182e-05, "loss": 0.0202, "step": 19800 }, { "epoch": 0.6304451132583557, "grad_norm": 0.030978305265307426, "learning_rate": 1.6628501556247222e-05, "loss": 0.0365, "step": 19900 }, { "epoch": 0.6336131791541264, "grad_norm": 1.2055158615112305, "learning_rate": 1.6606269453090266e-05, "loss": 0.0057, "step": 20000 }, { "epoch": 0.6367812450498971, "grad_norm": 17.737592697143555, "learning_rate": 1.6584037349933306e-05, "loss": 0.1269, "step": 20100 }, { "epoch": 0.6399493109456676, "grad_norm": 0.008062189444899559, "learning_rate": 1.6561805246776346e-05, "loss": 0.0281, "step": 20200 }, { "epoch": 0.6431173768414383, "grad_norm": 0.003408264135941863, "learning_rate": 1.6539573143619386e-05, "loss": 0.0092, "step": 20300 }, { "epoch": 0.6462854427372089, "grad_norm": 0.07715455442667007, "learning_rate": 1.651734104046243e-05, "loss": 0.0486, "step": 20400 }, { "epoch": 0.6494535086329796, "grad_norm": 0.08380332589149475, "learning_rate": 1.649510893730547e-05, "loss": 0.0323, "step": 20500 }, { "epoch": 0.6526215745287502, "grad_norm": 0.0010296939872205257, "learning_rate": 1.6472876834148513e-05, "loss": 0.0175, "step": 20600 }, { "epoch": 0.6557896404245208, "grad_norm": 6.547904922626913e-05, "learning_rate": 1.6450644730991553e-05, "loss": 0.0078, "step": 20700 }, { "epoch": 0.6589577063202915, "grad_norm": 0.002497778506949544, "learning_rate": 1.6428412627834593e-05, "loss": 0.0143, "step": 20800 }, { "epoch": 0.6621257722160621, "grad_norm": 3.8604347705841064, "learning_rate": 1.6406180524677636e-05, "loss": 0.0159, "step": 20900 }, { "epoch": 0.6652938381118327, "grad_norm": 0.0026936056092381477, "learning_rate": 1.6383948421520676e-05, "loss": 0.0261, "step": 21000 }, { "epoch": 0.6684619040076034, "grad_norm": 1.284914255142212, "learning_rate": 1.636171631836372e-05, "loss": 0.0083, "step": 21100 }, { "epoch": 0.671629969903374, "grad_norm": 0.012594709172844887, "learning_rate": 1.633948421520676e-05, "loss": 0.0232, "step": 21200 }, { "epoch": 0.6747980357991447, "grad_norm": 2.2840993404388428, "learning_rate": 1.6317252112049803e-05, "loss": 0.009, "step": 21300 }, { "epoch": 0.6779661016949152, "grad_norm": 0.004042040091007948, "learning_rate": 1.6295020008892843e-05, "loss": 0.0051, "step": 21400 }, { "epoch": 0.6811341675906859, "grad_norm": 0.001544850878417492, "learning_rate": 1.6272787905735884e-05, "loss": 0.0091, "step": 21500 }, { "epoch": 0.6843022334864565, "grad_norm": 0.00791002158075571, "learning_rate": 1.6250555802578924e-05, "loss": 0.0189, "step": 21600 }, { "epoch": 0.6874702993822271, "grad_norm": 0.0008190686348825693, "learning_rate": 1.6228323699421967e-05, "loss": 0.0044, "step": 21700 }, { "epoch": 0.6906383652779978, "grad_norm": 0.012536563910543919, "learning_rate": 1.6206091596265007e-05, "loss": 0.0284, "step": 21800 }, { "epoch": 0.6938064311737684, "grad_norm": 0.010305277071893215, "learning_rate": 1.618385949310805e-05, "loss": 0.0231, "step": 21900 }, { "epoch": 0.6969744970695391, "grad_norm": 0.0036432354245334864, "learning_rate": 1.616162738995109e-05, "loss": 0.0137, "step": 22000 }, { "epoch": 0.7001425629653096, "grad_norm": 0.016494829207658768, "learning_rate": 1.6139395286794134e-05, "loss": 0.0412, "step": 22100 }, { "epoch": 0.7033106288610803, "grad_norm": 0.1826470047235489, "learning_rate": 1.6117163183637174e-05, "loss": 0.0111, "step": 22200 }, { "epoch": 0.706478694756851, "grad_norm": 0.19151180982589722, "learning_rate": 1.6094931080480214e-05, "loss": 0.063, "step": 22300 }, { "epoch": 0.7096467606526216, "grad_norm": 0.00942138209939003, "learning_rate": 1.6072698977323254e-05, "loss": 0.0182, "step": 22400 }, { "epoch": 0.7128148265483922, "grad_norm": 0.007393939886242151, "learning_rate": 1.6050466874166298e-05, "loss": 0.0261, "step": 22500 }, { "epoch": 0.7159828924441628, "grad_norm": 0.00013769048382528126, "learning_rate": 1.6028234771009338e-05, "loss": 0.0221, "step": 22600 }, { "epoch": 0.7191509583399335, "grad_norm": 0.3037724196910858, "learning_rate": 1.600600266785238e-05, "loss": 0.0534, "step": 22700 }, { "epoch": 0.7223190242357042, "grad_norm": 0.03478122875094414, "learning_rate": 1.598377056469542e-05, "loss": 0.0295, "step": 22800 }, { "epoch": 0.7254870901314747, "grad_norm": 1.4521640539169312, "learning_rate": 1.5961538461538465e-05, "loss": 0.0073, "step": 22900 }, { "epoch": 0.7286551560272454, "grad_norm": 1.931846261024475, "learning_rate": 1.5939306358381505e-05, "loss": 0.0075, "step": 23000 }, { "epoch": 0.731823221923016, "grad_norm": 0.15186284482479095, "learning_rate": 1.5917074255224545e-05, "loss": 0.0309, "step": 23100 }, { "epoch": 0.7349912878187866, "grad_norm": 9.98096220428124e-06, "learning_rate": 1.5894842152067585e-05, "loss": 0.0279, "step": 23200 }, { "epoch": 0.7381593537145572, "grad_norm": 0.17749284207820892, "learning_rate": 1.587261004891063e-05, "loss": 0.013, "step": 23300 }, { "epoch": 0.7413274196103279, "grad_norm": 0.0030330857262015343, "learning_rate": 1.585037794575367e-05, "loss": 0.0147, "step": 23400 }, { "epoch": 0.7444954855060986, "grad_norm": 2.5090811252593994, "learning_rate": 1.5828145842596712e-05, "loss": 0.0154, "step": 23500 }, { "epoch": 0.7476635514018691, "grad_norm": 0.030537428334355354, "learning_rate": 1.5805913739439752e-05, "loss": 0.0158, "step": 23600 }, { "epoch": 0.7508316172976398, "grad_norm": 1.8608692698762752e-05, "learning_rate": 1.5783681636282792e-05, "loss": 0.0157, "step": 23700 }, { "epoch": 0.7539996831934104, "grad_norm": 0.40073361992836, "learning_rate": 1.5761449533125836e-05, "loss": 0.0052, "step": 23800 }, { "epoch": 0.757167749089181, "grad_norm": 0.0008514737710356712, "learning_rate": 1.5739217429968876e-05, "loss": 0.0072, "step": 23900 }, { "epoch": 0.7603358149849517, "grad_norm": 0.0016314525855705142, "learning_rate": 1.571698532681192e-05, "loss": 0.0132, "step": 24000 }, { "epoch": 0.7635038808807223, "grad_norm": 0.00047911162255331874, "learning_rate": 1.569475322365496e-05, "loss": 0.0243, "step": 24100 }, { "epoch": 0.766671946776493, "grad_norm": 0.0006689783185720444, "learning_rate": 1.5672521120498003e-05, "loss": 0.0201, "step": 24200 }, { "epoch": 0.7698400126722635, "grad_norm": 0.3778265416622162, "learning_rate": 1.5650289017341043e-05, "loss": 0.0168, "step": 24300 }, { "epoch": 0.7730080785680342, "grad_norm": 0.009497510269284248, "learning_rate": 1.5628056914184083e-05, "loss": 0.0132, "step": 24400 }, { "epoch": 0.7761761444638049, "grad_norm": 0.0015350583707913756, "learning_rate": 1.5605824811027123e-05, "loss": 0.014, "step": 24500 }, { "epoch": 0.7793442103595755, "grad_norm": 0.00025825086049735546, "learning_rate": 1.5583592707870166e-05, "loss": 0.0351, "step": 24600 }, { "epoch": 0.7825122762553461, "grad_norm": 0.005818589590489864, "learning_rate": 1.5561360604713206e-05, "loss": 0.0318, "step": 24700 }, { "epoch": 0.7856803421511167, "grad_norm": 0.0003177553880959749, "learning_rate": 1.553912850155625e-05, "loss": 0.0099, "step": 24800 }, { "epoch": 0.7888484080468874, "grad_norm": 0.029363544657826424, "learning_rate": 1.551689639839929e-05, "loss": 0.0395, "step": 24900 }, { "epoch": 0.7920164739426581, "grad_norm": 0.0023146485909819603, "learning_rate": 1.5494664295242333e-05, "loss": 0.0185, "step": 25000 }, { "epoch": 0.7951845398384286, "grad_norm": 0.7737551927566528, "learning_rate": 1.547243219208537e-05, "loss": 0.0114, "step": 25100 }, { "epoch": 0.7983526057341993, "grad_norm": 0.01664740964770317, "learning_rate": 1.5450200088928414e-05, "loss": 0.0246, "step": 25200 }, { "epoch": 0.8015206716299699, "grad_norm": 0.746607780456543, "learning_rate": 1.5427967985771454e-05, "loss": 0.0392, "step": 25300 }, { "epoch": 0.8046887375257405, "grad_norm": 0.1002197116613388, "learning_rate": 1.5405735882614497e-05, "loss": 0.0042, "step": 25400 }, { "epoch": 0.8078568034215111, "grad_norm": 0.04124956950545311, "learning_rate": 1.5383503779457537e-05, "loss": 0.0188, "step": 25500 }, { "epoch": 0.8110248693172818, "grad_norm": 0.014210743829607964, "learning_rate": 1.536127167630058e-05, "loss": 0.0126, "step": 25600 }, { "epoch": 0.8141929352130525, "grad_norm": 0.9457830786705017, "learning_rate": 1.533903957314362e-05, "loss": 0.0535, "step": 25700 }, { "epoch": 0.817361001108823, "grad_norm": 0.004865521565079689, "learning_rate": 1.531680746998666e-05, "loss": 0.0164, "step": 25800 }, { "epoch": 0.8205290670045937, "grad_norm": 0.050479255616664886, "learning_rate": 1.5294575366829704e-05, "loss": 0.0433, "step": 25900 }, { "epoch": 0.8236971329003643, "grad_norm": 1.1270233392715454, "learning_rate": 1.5272343263672744e-05, "loss": 0.0313, "step": 26000 }, { "epoch": 0.826865198796135, "grad_norm": 0.0021326912101358175, "learning_rate": 1.5250111160515786e-05, "loss": 0.0157, "step": 26100 }, { "epoch": 0.8300332646919056, "grad_norm": 4.062502861022949, "learning_rate": 1.5227879057358828e-05, "loss": 0.0188, "step": 26200 }, { "epoch": 0.8332013305876762, "grad_norm": 0.368022084236145, "learning_rate": 1.520564695420187e-05, "loss": 0.0307, "step": 26300 }, { "epoch": 0.8363693964834469, "grad_norm": 0.003511411137878895, "learning_rate": 1.5183414851044911e-05, "loss": 0.0074, "step": 26400 }, { "epoch": 0.8395374623792174, "grad_norm": 0.24865694344043732, "learning_rate": 1.5161182747887953e-05, "loss": 0.0468, "step": 26500 }, { "epoch": 0.8427055282749881, "grad_norm": 0.001150769297964871, "learning_rate": 1.5138950644730991e-05, "loss": 0.0138, "step": 26600 }, { "epoch": 0.8458735941707588, "grad_norm": 6.395234959200025e-05, "learning_rate": 1.5116718541574033e-05, "loss": 0.0044, "step": 26700 }, { "epoch": 0.8490416600665294, "grad_norm": 2.651227441674564e-05, "learning_rate": 1.5094486438417075e-05, "loss": 0.0366, "step": 26800 }, { "epoch": 0.8522097259623, "grad_norm": 0.0029796832241117954, "learning_rate": 1.5072254335260117e-05, "loss": 0.0343, "step": 26900 }, { "epoch": 0.8553777918580706, "grad_norm": 11.861794471740723, "learning_rate": 1.5050022232103159e-05, "loss": 0.0051, "step": 27000 }, { "epoch": 0.8585458577538413, "grad_norm": 0.002347618341445923, "learning_rate": 1.50277901289462e-05, "loss": 0.0294, "step": 27100 }, { "epoch": 0.8617139236496119, "grad_norm": 1.4324954748153687, "learning_rate": 1.5005558025789242e-05, "loss": 0.0373, "step": 27200 }, { "epoch": 0.8648819895453825, "grad_norm": 23.32402229309082, "learning_rate": 1.4983325922632282e-05, "loss": 0.0097, "step": 27300 }, { "epoch": 0.8680500554411532, "grad_norm": 0.0636548399925232, "learning_rate": 1.4961093819475322e-05, "loss": 0.0177, "step": 27400 }, { "epoch": 0.8712181213369238, "grad_norm": 0.012398977763950825, "learning_rate": 1.4938861716318364e-05, "loss": 0.0124, "step": 27500 }, { "epoch": 0.8743861872326945, "grad_norm": 0.753961980342865, "learning_rate": 1.4916629613161406e-05, "loss": 0.0126, "step": 27600 }, { "epoch": 0.877554253128465, "grad_norm": 0.0002603745961096138, "learning_rate": 1.4894397510004447e-05, "loss": 0.0128, "step": 27700 }, { "epoch": 0.8807223190242357, "grad_norm": 1.4019455909729004, "learning_rate": 1.487216540684749e-05, "loss": 0.01, "step": 27800 }, { "epoch": 0.8838903849200064, "grad_norm": 0.0003745869908016175, "learning_rate": 1.4849933303690531e-05, "loss": 0.0119, "step": 27900 }, { "epoch": 0.8870584508157769, "grad_norm": 0.3194678723812103, "learning_rate": 1.4827701200533571e-05, "loss": 0.0169, "step": 28000 }, { "epoch": 0.8902265167115476, "grad_norm": 0.004225617740303278, "learning_rate": 1.4805469097376613e-05, "loss": 0.0081, "step": 28100 }, { "epoch": 0.8933945826073182, "grad_norm": 11.807644844055176, "learning_rate": 1.4783236994219655e-05, "loss": 0.0075, "step": 28200 }, { "epoch": 0.8965626485030889, "grad_norm": 0.0002042292180703953, "learning_rate": 1.4761004891062696e-05, "loss": 0.0159, "step": 28300 }, { "epoch": 0.8997307143988595, "grad_norm": 0.00018777225341182202, "learning_rate": 1.4738772787905738e-05, "loss": 0.0094, "step": 28400 }, { "epoch": 0.9028987802946301, "grad_norm": 6.943278789520264, "learning_rate": 1.4716540684748778e-05, "loss": 0.0154, "step": 28500 }, { "epoch": 0.9060668461904008, "grad_norm": 9.940530776977539, "learning_rate": 1.469430858159182e-05, "loss": 0.0079, "step": 28600 }, { "epoch": 0.9092349120861714, "grad_norm": 0.0007003561477176845, "learning_rate": 1.467207647843486e-05, "loss": 0.0088, "step": 28700 }, { "epoch": 0.912402977981942, "grad_norm": 0.0008374906610697508, "learning_rate": 1.4649844375277902e-05, "loss": 0.0046, "step": 28800 }, { "epoch": 0.9155710438777127, "grad_norm": 1.0267653465270996, "learning_rate": 1.4627612272120944e-05, "loss": 0.0038, "step": 28900 }, { "epoch": 0.9187391097734833, "grad_norm": 0.0030098266433924437, "learning_rate": 1.4605380168963985e-05, "loss": 0.0132, "step": 29000 }, { "epoch": 0.921907175669254, "grad_norm": 0.00015600144979543984, "learning_rate": 1.4583148065807027e-05, "loss": 0.0128, "step": 29100 }, { "epoch": 0.9250752415650245, "grad_norm": 38.474552154541016, "learning_rate": 1.4560915962650069e-05, "loss": 0.0141, "step": 29200 }, { "epoch": 0.9282433074607952, "grad_norm": 2.2587130069732666, "learning_rate": 1.453868385949311e-05, "loss": 0.0033, "step": 29300 }, { "epoch": 0.9314113733565658, "grad_norm": 0.21571636199951172, "learning_rate": 1.4516451756336149e-05, "loss": 0.0103, "step": 29400 }, { "epoch": 0.9345794392523364, "grad_norm": 0.011366557329893112, "learning_rate": 1.449421965317919e-05, "loss": 0.034, "step": 29500 }, { "epoch": 0.9377475051481071, "grad_norm": 0.13342642784118652, "learning_rate": 1.4471987550022233e-05, "loss": 0.0036, "step": 29600 }, { "epoch": 0.9409155710438777, "grad_norm": 0.008988015353679657, "learning_rate": 1.4449755446865274e-05, "loss": 0.0229, "step": 29700 }, { "epoch": 0.9440836369396484, "grad_norm": 0.0017454049084335566, "learning_rate": 1.4427523343708316e-05, "loss": 0.0113, "step": 29800 }, { "epoch": 0.9472517028354189, "grad_norm": 0.05987678840756416, "learning_rate": 1.4405291240551358e-05, "loss": 0.0136, "step": 29900 }, { "epoch": 0.9504197687311896, "grad_norm": 1.8885809183120728, "learning_rate": 1.43830591373944e-05, "loss": 0.0283, "step": 30000 }, { "epoch": 0.9535878346269603, "grad_norm": 0.00034857913851737976, "learning_rate": 1.436082703423744e-05, "loss": 0.0181, "step": 30100 }, { "epoch": 0.9567559005227309, "grad_norm": 0.009435332380235195, "learning_rate": 1.4338594931080481e-05, "loss": 0.0108, "step": 30200 }, { "epoch": 0.9599239664185015, "grad_norm": 0.06871967762708664, "learning_rate": 1.4316362827923521e-05, "loss": 0.0208, "step": 30300 }, { "epoch": 0.9630920323142721, "grad_norm": 0.018716560676693916, "learning_rate": 1.4294130724766563e-05, "loss": 0.0227, "step": 30400 }, { "epoch": 0.9662600982100428, "grad_norm": 0.01711018569767475, "learning_rate": 1.4271898621609605e-05, "loss": 0.0192, "step": 30500 }, { "epoch": 0.9694281641058135, "grad_norm": 0.002523894887417555, "learning_rate": 1.4249666518452647e-05, "loss": 0.0122, "step": 30600 }, { "epoch": 0.972596230001584, "grad_norm": 0.019669946283102036, "learning_rate": 1.4227434415295689e-05, "loss": 0.0061, "step": 30700 }, { "epoch": 0.9757642958973547, "grad_norm": 0.6150824427604675, "learning_rate": 1.420520231213873e-05, "loss": 0.039, "step": 30800 }, { "epoch": 0.9789323617931253, "grad_norm": 0.0004018946783617139, "learning_rate": 1.418297020898177e-05, "loss": 0.0115, "step": 30900 }, { "epoch": 0.9821004276888959, "grad_norm": 0.0448601096868515, "learning_rate": 1.4160738105824812e-05, "loss": 0.0122, "step": 31000 }, { "epoch": 0.9852684935846666, "grad_norm": 0.0008613928221166134, "learning_rate": 1.4138506002667854e-05, "loss": 0.0085, "step": 31100 }, { "epoch": 0.9884365594804372, "grad_norm": 0.0038725193589925766, "learning_rate": 1.4116273899510896e-05, "loss": 0.035, "step": 31200 }, { "epoch": 0.9916046253762079, "grad_norm": 0.06399986892938614, "learning_rate": 1.4094041796353937e-05, "loss": 0.016, "step": 31300 }, { "epoch": 0.9947726912719784, "grad_norm": 0.0022226041182875633, "learning_rate": 1.4071809693196978e-05, "loss": 0.0337, "step": 31400 }, { "epoch": 0.9979407571677491, "grad_norm": 0.00036096415715292096, "learning_rate": 1.404957759004002e-05, "loss": 0.0056, "step": 31500 }, { "epoch": 1.0011088230635197, "grad_norm": 0.46166709065437317, "learning_rate": 1.402734548688306e-05, "loss": 0.0119, "step": 31600 }, { "epoch": 1.0042768889592903, "grad_norm": 0.0008330343989655375, "learning_rate": 1.4005113383726101e-05, "loss": 0.0046, "step": 31700 }, { "epoch": 1.007444954855061, "grad_norm": 0.011975400149822235, "learning_rate": 1.3982881280569143e-05, "loss": 0.005, "step": 31800 }, { "epoch": 1.0106130207508317, "grad_norm": 0.015499376691877842, "learning_rate": 1.3960649177412185e-05, "loss": 0.0076, "step": 31900 }, { "epoch": 1.0137810866466022, "grad_norm": 8.676903235027567e-05, "learning_rate": 1.3938417074255226e-05, "loss": 0.0067, "step": 32000 }, { "epoch": 1.0169491525423728, "grad_norm": 0.09442619979381561, "learning_rate": 1.3916184971098268e-05, "loss": 0.0047, "step": 32100 }, { "epoch": 1.0201172184381435, "grad_norm": 0.010203162208199501, "learning_rate": 1.389395286794131e-05, "loss": 0.0144, "step": 32200 }, { "epoch": 1.0232852843339142, "grad_norm": 0.006877961102873087, "learning_rate": 1.3871720764784348e-05, "loss": 0.0434, "step": 32300 }, { "epoch": 1.0264533502296849, "grad_norm": 0.00014150477363727987, "learning_rate": 1.384948866162739e-05, "loss": 0.0357, "step": 32400 }, { "epoch": 1.0296214161254553, "grad_norm": 0.0011065603466704488, "learning_rate": 1.3827256558470432e-05, "loss": 0.0062, "step": 32500 }, { "epoch": 1.032789482021226, "grad_norm": 0.2077646106481552, "learning_rate": 1.3805024455313474e-05, "loss": 0.0336, "step": 32600 }, { "epoch": 1.0359575479169967, "grad_norm": 0.0013152187457308173, "learning_rate": 1.3782792352156515e-05, "loss": 0.0352, "step": 32700 }, { "epoch": 1.0391256138127674, "grad_norm": 0.017774872481822968, "learning_rate": 1.3760560248999557e-05, "loss": 0.0043, "step": 32800 }, { "epoch": 1.042293679708538, "grad_norm": 5.051975250244141, "learning_rate": 1.3738328145842599e-05, "loss": 0.0148, "step": 32900 }, { "epoch": 1.0454617456043085, "grad_norm": 3.470256706350483e-05, "learning_rate": 1.3716096042685639e-05, "loss": 0.0042, "step": 33000 }, { "epoch": 1.0486298115000792, "grad_norm": 0.06109689176082611, "learning_rate": 1.3693863939528679e-05, "loss": 0.0044, "step": 33100 }, { "epoch": 1.0517978773958498, "grad_norm": 0.39745813608169556, "learning_rate": 1.367163183637172e-05, "loss": 0.0155, "step": 33200 }, { "epoch": 1.0549659432916205, "grad_norm": 0.00012431037612259388, "learning_rate": 1.3649399733214763e-05, "loss": 0.0251, "step": 33300 }, { "epoch": 1.0581340091873912, "grad_norm": 0.12850160896778107, "learning_rate": 1.3627167630057804e-05, "loss": 0.0092, "step": 33400 }, { "epoch": 1.0613020750831617, "grad_norm": 0.001514806179329753, "learning_rate": 1.3604935526900846e-05, "loss": 0.0207, "step": 33500 }, { "epoch": 1.0644701409789323, "grad_norm": 0.0017514684004709125, "learning_rate": 1.3582703423743888e-05, "loss": 0.0074, "step": 33600 }, { "epoch": 1.067638206874703, "grad_norm": 0.022122355177998543, "learning_rate": 1.3560471320586928e-05, "loss": 0.0352, "step": 33700 }, { "epoch": 1.0708062727704737, "grad_norm": 0.00412428192794323, "learning_rate": 1.353823921742997e-05, "loss": 0.0071, "step": 33800 }, { "epoch": 1.0739743386662441, "grad_norm": 0.15664397180080414, "learning_rate": 1.3516007114273011e-05, "loss": 0.0083, "step": 33900 }, { "epoch": 1.0771424045620148, "grad_norm": 0.18125130236148834, "learning_rate": 1.3493775011116053e-05, "loss": 0.0119, "step": 34000 }, { "epoch": 1.0803104704577855, "grad_norm": 0.07240629196166992, "learning_rate": 1.3471542907959095e-05, "loss": 0.0073, "step": 34100 }, { "epoch": 1.0834785363535562, "grad_norm": 0.006628985516726971, "learning_rate": 1.3449310804802137e-05, "loss": 0.0282, "step": 34200 }, { "epoch": 1.0866466022493269, "grad_norm": 0.8509578704833984, "learning_rate": 1.3427078701645177e-05, "loss": 0.0097, "step": 34300 }, { "epoch": 1.0898146681450975, "grad_norm": 0.044779542833566666, "learning_rate": 1.3404846598488217e-05, "loss": 0.0062, "step": 34400 }, { "epoch": 1.092982734040868, "grad_norm": 4.1824325307970867e-05, "learning_rate": 1.3382614495331259e-05, "loss": 0.0127, "step": 34500 }, { "epoch": 1.0961507999366387, "grad_norm": 0.011673863045871258, "learning_rate": 1.33603823921743e-05, "loss": 0.0117, "step": 34600 }, { "epoch": 1.0993188658324093, "grad_norm": 0.14283762872219086, "learning_rate": 1.3338150289017342e-05, "loss": 0.0163, "step": 34700 }, { "epoch": 1.10248693172818, "grad_norm": 0.009939441457390785, "learning_rate": 1.3315918185860384e-05, "loss": 0.0221, "step": 34800 }, { "epoch": 1.1056549976239505, "grad_norm": 0.09936344623565674, "learning_rate": 1.3293686082703426e-05, "loss": 0.0145, "step": 34900 }, { "epoch": 1.1088230635197212, "grad_norm": 0.0036431141197681427, "learning_rate": 1.3271453979546467e-05, "loss": 0.0073, "step": 35000 }, { "epoch": 1.1119911294154918, "grad_norm": 0.06281910091638565, "learning_rate": 1.324922187638951e-05, "loss": 0.0065, "step": 35100 }, { "epoch": 1.1151591953112625, "grad_norm": 0.0020452861208468676, "learning_rate": 1.3226989773232548e-05, "loss": 0.0333, "step": 35200 }, { "epoch": 1.1183272612070332, "grad_norm": 0.0006033928948454559, "learning_rate": 1.320475767007559e-05, "loss": 0.0048, "step": 35300 }, { "epoch": 1.1214953271028036, "grad_norm": 0.00021041724539827555, "learning_rate": 1.3182525566918631e-05, "loss": 0.0169, "step": 35400 }, { "epoch": 1.1246633929985743, "grad_norm": 0.28758516907691956, "learning_rate": 1.3160293463761673e-05, "loss": 0.0045, "step": 35500 }, { "epoch": 1.127831458894345, "grad_norm": 0.0014314631698653102, "learning_rate": 1.3138061360604715e-05, "loss": 0.0272, "step": 35600 }, { "epoch": 1.1309995247901157, "grad_norm": 0.0035914001055061817, "learning_rate": 1.3115829257447756e-05, "loss": 0.0065, "step": 35700 }, { "epoch": 1.1341675906858864, "grad_norm": 0.0009057251736521721, "learning_rate": 1.3093597154290798e-05, "loss": 0.0026, "step": 35800 }, { "epoch": 1.1373356565816568, "grad_norm": 0.007786700967699289, "learning_rate": 1.3071365051133838e-05, "loss": 0.0139, "step": 35900 }, { "epoch": 1.1405037224774275, "grad_norm": 0.031975653022527695, "learning_rate": 1.3049132947976878e-05, "loss": 0.0219, "step": 36000 }, { "epoch": 1.1436717883731982, "grad_norm": 0.0010592106264084578, "learning_rate": 1.302690084481992e-05, "loss": 0.0132, "step": 36100 }, { "epoch": 1.1468398542689688, "grad_norm": 0.00017434361507184803, "learning_rate": 1.3004668741662962e-05, "loss": 0.0087, "step": 36200 }, { "epoch": 1.1500079201647395, "grad_norm": 0.0018986865179613233, "learning_rate": 1.2982436638506004e-05, "loss": 0.0038, "step": 36300 }, { "epoch": 1.15317598606051, "grad_norm": 0.0015970875974744558, "learning_rate": 1.2960204535349045e-05, "loss": 0.0322, "step": 36400 }, { "epoch": 1.1563440519562806, "grad_norm": 0.2892841696739197, "learning_rate": 1.2937972432192087e-05, "loss": 0.0109, "step": 36500 }, { "epoch": 1.1595121178520513, "grad_norm": 0.0012535974383354187, "learning_rate": 1.2915740329035127e-05, "loss": 0.0059, "step": 36600 }, { "epoch": 1.162680183747822, "grad_norm": 0.01111169159412384, "learning_rate": 1.2893508225878169e-05, "loss": 0.0072, "step": 36700 }, { "epoch": 1.1658482496435927, "grad_norm": 0.00112074613571167, "learning_rate": 1.287127612272121e-05, "loss": 0.0026, "step": 36800 }, { "epoch": 1.1690163155393631, "grad_norm": 0.10183645784854889, "learning_rate": 1.2849044019564253e-05, "loss": 0.0115, "step": 36900 }, { "epoch": 1.1721843814351338, "grad_norm": 0.0021920576691627502, "learning_rate": 1.2826811916407294e-05, "loss": 0.0288, "step": 37000 }, { "epoch": 1.1753524473309045, "grad_norm": 0.013510264456272125, "learning_rate": 1.2804579813250334e-05, "loss": 0.0018, "step": 37100 }, { "epoch": 1.1785205132266752, "grad_norm": 0.003552510868757963, "learning_rate": 1.2782347710093376e-05, "loss": 0.0091, "step": 37200 }, { "epoch": 1.1816885791224458, "grad_norm": 5.933229476795532e-05, "learning_rate": 1.2760115606936416e-05, "loss": 0.0095, "step": 37300 }, { "epoch": 1.1848566450182163, "grad_norm": 0.16859188675880432, "learning_rate": 1.2737883503779458e-05, "loss": 0.0066, "step": 37400 }, { "epoch": 1.188024710913987, "grad_norm": 0.000806248455774039, "learning_rate": 1.27156514006225e-05, "loss": 0.001, "step": 37500 }, { "epoch": 1.1911927768097577, "grad_norm": 0.001987775322049856, "learning_rate": 1.2693419297465541e-05, "loss": 0.0195, "step": 37600 }, { "epoch": 1.1943608427055283, "grad_norm": 0.40488603711128235, "learning_rate": 1.2671187194308583e-05, "loss": 0.0222, "step": 37700 }, { "epoch": 1.1975289086012988, "grad_norm": 0.001327143982052803, "learning_rate": 1.2648955091151625e-05, "loss": 0.0063, "step": 37800 }, { "epoch": 1.2006969744970695, "grad_norm": 0.03306195139884949, "learning_rate": 1.2626722987994667e-05, "loss": 0.0139, "step": 37900 }, { "epoch": 1.2038650403928401, "grad_norm": 0.00011816229380201548, "learning_rate": 1.2604490884837705e-05, "loss": 0.005, "step": 38000 }, { "epoch": 1.2070331062886108, "grad_norm": 0.00011477043153718114, "learning_rate": 1.2582258781680747e-05, "loss": 0.0075, "step": 38100 }, { "epoch": 1.2102011721843815, "grad_norm": 0.020396051928400993, "learning_rate": 1.2560026678523789e-05, "loss": 0.0095, "step": 38200 }, { "epoch": 1.2133692380801522, "grad_norm": 0.002560865134000778, "learning_rate": 1.253779457536683e-05, "loss": 0.0083, "step": 38300 }, { "epoch": 1.2165373039759226, "grad_norm": 0.01047099381685257, "learning_rate": 1.2515562472209872e-05, "loss": 0.0031, "step": 38400 }, { "epoch": 1.2197053698716933, "grad_norm": 2.947988748550415, "learning_rate": 1.2493330369052914e-05, "loss": 0.0026, "step": 38500 }, { "epoch": 1.222873435767464, "grad_norm": 0.005444441922008991, "learning_rate": 1.2471098265895956e-05, "loss": 0.0018, "step": 38600 }, { "epoch": 1.2260415016632347, "grad_norm": 0.05086371675133705, "learning_rate": 1.2448866162738996e-05, "loss": 0.0116, "step": 38700 }, { "epoch": 1.2292095675590051, "grad_norm": 0.0008550622151233256, "learning_rate": 1.2426634059582036e-05, "loss": 0.0037, "step": 38800 }, { "epoch": 1.2323776334547758, "grad_norm": 0.467833548784256, "learning_rate": 1.2404401956425078e-05, "loss": 0.0146, "step": 38900 }, { "epoch": 1.2355456993505465, "grad_norm": 0.002233183477073908, "learning_rate": 1.238216985326812e-05, "loss": 0.0118, "step": 39000 }, { "epoch": 1.2387137652463172, "grad_norm": 0.743407130241394, "learning_rate": 1.2359937750111161e-05, "loss": 0.009, "step": 39100 }, { "epoch": 1.2418818311420878, "grad_norm": 0.03420891985297203, "learning_rate": 1.2337705646954203e-05, "loss": 0.0078, "step": 39200 }, { "epoch": 1.2450498970378585, "grad_norm": 15.531525611877441, "learning_rate": 1.2315473543797245e-05, "loss": 0.0118, "step": 39300 }, { "epoch": 1.248217962933629, "grad_norm": 0.08798499405384064, "learning_rate": 1.2293241440640286e-05, "loss": 0.0061, "step": 39400 }, { "epoch": 1.2513860288293996, "grad_norm": 0.0004404394421726465, "learning_rate": 1.2271009337483327e-05, "loss": 0.0154, "step": 39500 }, { "epoch": 1.2545540947251703, "grad_norm": 0.006552988663315773, "learning_rate": 1.2248777234326368e-05, "loss": 0.0161, "step": 39600 }, { "epoch": 1.257722160620941, "grad_norm": 0.11566559225320816, "learning_rate": 1.222654513116941e-05, "loss": 0.0051, "step": 39700 }, { "epoch": 1.2608902265167115, "grad_norm": 0.0004462143697310239, "learning_rate": 1.2204313028012452e-05, "loss": 0.0113, "step": 39800 }, { "epoch": 1.2640582924124821, "grad_norm": 0.0001944590185303241, "learning_rate": 1.2182080924855494e-05, "loss": 0.0047, "step": 39900 }, { "epoch": 1.2672263583082528, "grad_norm": 0.10759201645851135, "learning_rate": 1.2159848821698534e-05, "loss": 0.0051, "step": 40000 }, { "epoch": 1.2703944242040235, "grad_norm": 0.00022843007172923535, "learning_rate": 1.2137616718541575e-05, "loss": 0.0054, "step": 40100 }, { "epoch": 1.2735624900997942, "grad_norm": 0.00036529748467728496, "learning_rate": 1.2115384615384615e-05, "loss": 0.0085, "step": 40200 }, { "epoch": 1.2767305559955648, "grad_norm": 0.011095472611486912, "learning_rate": 1.2093152512227657e-05, "loss": 0.0097, "step": 40300 }, { "epoch": 1.2798986218913353, "grad_norm": 0.0005841770325787365, "learning_rate": 1.2070920409070699e-05, "loss": 0.009, "step": 40400 }, { "epoch": 1.283066687787106, "grad_norm": 0.041353631764650345, "learning_rate": 1.204868830591374e-05, "loss": 0.0081, "step": 40500 }, { "epoch": 1.2862347536828767, "grad_norm": 0.0013312128139659762, "learning_rate": 1.2026456202756783e-05, "loss": 0.0091, "step": 40600 }, { "epoch": 1.289402819578647, "grad_norm": 0.024310261011123657, "learning_rate": 1.2004224099599824e-05, "loss": 0.0204, "step": 40700 }, { "epoch": 1.2925708854744178, "grad_norm": 1.1424692869186401, "learning_rate": 1.1981991996442866e-05, "loss": 0.0102, "step": 40800 }, { "epoch": 1.2957389513701885, "grad_norm": 0.003675231710076332, "learning_rate": 1.1959759893285904e-05, "loss": 0.0124, "step": 40900 }, { "epoch": 1.2989070172659591, "grad_norm": 0.19512909650802612, "learning_rate": 1.1937527790128946e-05, "loss": 0.0051, "step": 41000 }, { "epoch": 1.3020750831617298, "grad_norm": 0.00013532934826798737, "learning_rate": 1.1915295686971988e-05, "loss": 0.0081, "step": 41100 }, { "epoch": 1.3052431490575005, "grad_norm": 7.001680205576122e-05, "learning_rate": 1.189306358381503e-05, "loss": 0.0011, "step": 41200 }, { "epoch": 1.308411214953271, "grad_norm": 0.056945838034152985, "learning_rate": 1.1870831480658072e-05, "loss": 0.0023, "step": 41300 }, { "epoch": 1.3115792808490416, "grad_norm": 0.30034956336021423, "learning_rate": 1.1848599377501113e-05, "loss": 0.0024, "step": 41400 }, { "epoch": 1.3147473467448123, "grad_norm": 0.0019128896528854966, "learning_rate": 1.1826367274344155e-05, "loss": 0.0155, "step": 41500 }, { "epoch": 1.317915412640583, "grad_norm": 0.001590027124620974, "learning_rate": 1.1804135171187195e-05, "loss": 0.0087, "step": 41600 }, { "epoch": 1.3210834785363534, "grad_norm": 0.019434746354818344, "learning_rate": 1.1781903068030235e-05, "loss": 0.0339, "step": 41700 }, { "epoch": 1.3242515444321241, "grad_norm": 0.0017116613453254104, "learning_rate": 1.1759670964873277e-05, "loss": 0.0044, "step": 41800 }, { "epoch": 1.3274196103278948, "grad_norm": 0.28481578826904297, "learning_rate": 1.1737438861716319e-05, "loss": 0.008, "step": 41900 }, { "epoch": 1.3305876762236655, "grad_norm": 1.551391363143921, "learning_rate": 1.171520675855936e-05, "loss": 0.0261, "step": 42000 }, { "epoch": 1.3337557421194361, "grad_norm": 0.003555042902007699, "learning_rate": 1.1692974655402402e-05, "loss": 0.0026, "step": 42100 }, { "epoch": 1.3369238080152068, "grad_norm": 0.0013489079428836703, "learning_rate": 1.1670742552245444e-05, "loss": 0.0154, "step": 42200 }, { "epoch": 1.3400918739109773, "grad_norm": 0.0012371678603813052, "learning_rate": 1.1648510449088484e-05, "loss": 0.0067, "step": 42300 }, { "epoch": 1.343259939806748, "grad_norm": 0.04170294478535652, "learning_rate": 1.1626278345931526e-05, "loss": 0.0033, "step": 42400 }, { "epoch": 1.3464280057025186, "grad_norm": 0.0005449488526210189, "learning_rate": 1.1604046242774568e-05, "loss": 0.0046, "step": 42500 }, { "epoch": 1.3495960715982893, "grad_norm": 0.4102185368537903, "learning_rate": 1.158181413961761e-05, "loss": 0.011, "step": 42600 }, { "epoch": 1.3527641374940598, "grad_norm": 0.004146751016378403, "learning_rate": 1.1559582036460651e-05, "loss": 0.0054, "step": 42700 }, { "epoch": 1.3559322033898304, "grad_norm": 0.0008668516529724002, "learning_rate": 1.1537349933303691e-05, "loss": 0.0154, "step": 42800 }, { "epoch": 1.3591002692856011, "grad_norm": 0.40073227882385254, "learning_rate": 1.1515117830146733e-05, "loss": 0.0016, "step": 42900 }, { "epoch": 1.3622683351813718, "grad_norm": 0.0008245128556154668, "learning_rate": 1.1492885726989773e-05, "loss": 0.0261, "step": 43000 }, { "epoch": 1.3654364010771425, "grad_norm": 0.0028431194368749857, "learning_rate": 1.1470653623832815e-05, "loss": 0.007, "step": 43100 }, { "epoch": 1.3686044669729132, "grad_norm": 0.0007627083105035126, "learning_rate": 1.1448421520675857e-05, "loss": 0.0011, "step": 43200 }, { "epoch": 1.3717725328686836, "grad_norm": 0.02942449413239956, "learning_rate": 1.1426189417518898e-05, "loss": 0.0058, "step": 43300 }, { "epoch": 1.3749405987644543, "grad_norm": 0.08443868905305862, "learning_rate": 1.140395731436194e-05, "loss": 0.0047, "step": 43400 }, { "epoch": 1.378108664660225, "grad_norm": 1.5906205589999445e-05, "learning_rate": 1.1381725211204982e-05, "loss": 0.0044, "step": 43500 }, { "epoch": 1.3812767305559956, "grad_norm": 0.0010951455915346742, "learning_rate": 1.1359493108048024e-05, "loss": 0.0037, "step": 43600 }, { "epoch": 1.384444796451766, "grad_norm": 0.010305008850991726, "learning_rate": 1.1337261004891062e-05, "loss": 0.0039, "step": 43700 }, { "epoch": 1.3876128623475368, "grad_norm": 0.6063001751899719, "learning_rate": 1.1315028901734104e-05, "loss": 0.028, "step": 43800 }, { "epoch": 1.3907809282433075, "grad_norm": 0.0007881261990405619, "learning_rate": 1.1292796798577146e-05, "loss": 0.0043, "step": 43900 }, { "epoch": 1.3939489941390781, "grad_norm": 0.0031877460423856974, "learning_rate": 1.1270564695420187e-05, "loss": 0.0016, "step": 44000 }, { "epoch": 1.3971170600348488, "grad_norm": 0.008450224064290524, "learning_rate": 1.1248332592263229e-05, "loss": 0.0074, "step": 44100 }, { "epoch": 1.4002851259306195, "grad_norm": 0.01120079681277275, "learning_rate": 1.122610048910627e-05, "loss": 0.0042, "step": 44200 }, { "epoch": 1.40345319182639, "grad_norm": 0.001460696803405881, "learning_rate": 1.1203868385949313e-05, "loss": 0.0025, "step": 44300 }, { "epoch": 1.4066212577221606, "grad_norm": 0.012069291435182095, "learning_rate": 1.1181636282792354e-05, "loss": 0.0257, "step": 44400 }, { "epoch": 1.4097893236179313, "grad_norm": 0.0032543253619223833, "learning_rate": 1.1159404179635394e-05, "loss": 0.0173, "step": 44500 }, { "epoch": 1.4129573895137018, "grad_norm": 0.0007560536032542586, "learning_rate": 1.1137172076478434e-05, "loss": 0.0019, "step": 44600 }, { "epoch": 1.4161254554094724, "grad_norm": 0.0007794055272825062, "learning_rate": 1.1114939973321476e-05, "loss": 0.0024, "step": 44700 }, { "epoch": 1.419293521305243, "grad_norm": 1.3138316717231646e-05, "learning_rate": 1.1092707870164518e-05, "loss": 0.0037, "step": 44800 }, { "epoch": 1.4224615872010138, "grad_norm": 0.000821842928417027, "learning_rate": 1.107047576700756e-05, "loss": 0.004, "step": 44900 }, { "epoch": 1.4256296530967845, "grad_norm": 0.013666361570358276, "learning_rate": 1.1048243663850602e-05, "loss": 0.0015, "step": 45000 }, { "epoch": 1.4287977189925551, "grad_norm": 0.03491951525211334, "learning_rate": 1.1026011560693643e-05, "loss": 0.0062, "step": 45100 }, { "epoch": 1.4319657848883256, "grad_norm": 0.0006090217502787709, "learning_rate": 1.1003779457536683e-05, "loss": 0.0044, "step": 45200 }, { "epoch": 1.4351338507840963, "grad_norm": 0.08735354989767075, "learning_rate": 1.0981547354379725e-05, "loss": 0.0022, "step": 45300 }, { "epoch": 1.438301916679867, "grad_norm": 0.006710866466164589, "learning_rate": 1.0959315251222767e-05, "loss": 0.0013, "step": 45400 }, { "epoch": 1.4414699825756376, "grad_norm": 0.004379246383905411, "learning_rate": 1.0937083148065809e-05, "loss": 0.0038, "step": 45500 }, { "epoch": 1.444638048471408, "grad_norm": 0.001965580740943551, "learning_rate": 1.091485104490885e-05, "loss": 0.0016, "step": 45600 }, { "epoch": 1.4478061143671788, "grad_norm": 0.02562040276825428, "learning_rate": 1.089261894175189e-05, "loss": 0.0129, "step": 45700 }, { "epoch": 1.4509741802629494, "grad_norm": 0.0014989199116826057, "learning_rate": 1.0870386838594932e-05, "loss": 0.0027, "step": 45800 }, { "epoch": 1.4541422461587201, "grad_norm": 0.0005407402641139925, "learning_rate": 1.0848154735437972e-05, "loss": 0.0039, "step": 45900 }, { "epoch": 1.4573103120544908, "grad_norm": 0.019197065383195877, "learning_rate": 1.0825922632281014e-05, "loss": 0.003, "step": 46000 }, { "epoch": 1.4604783779502615, "grad_norm": 5.17616790602915e-05, "learning_rate": 1.0803690529124056e-05, "loss": 0.0015, "step": 46100 }, { "epoch": 1.463646443846032, "grad_norm": 0.0007133186445571482, "learning_rate": 1.0781458425967098e-05, "loss": 0.0005, "step": 46200 }, { "epoch": 1.4668145097418026, "grad_norm": 0.00010192937043029815, "learning_rate": 1.075922632281014e-05, "loss": 0.0031, "step": 46300 }, { "epoch": 1.4699825756375733, "grad_norm": 0.00020953864441253245, "learning_rate": 1.0736994219653181e-05, "loss": 0.0076, "step": 46400 }, { "epoch": 1.473150641533344, "grad_norm": 0.002632291754707694, "learning_rate": 1.0714762116496223e-05, "loss": 0.0008, "step": 46500 }, { "epoch": 1.4763187074291144, "grad_norm": 0.00219705980271101, "learning_rate": 1.0692530013339261e-05, "loss": 0.0019, "step": 46600 }, { "epoch": 1.479486773324885, "grad_norm": 0.6644763946533203, "learning_rate": 1.0670297910182303e-05, "loss": 0.0013, "step": 46700 }, { "epoch": 1.4826548392206558, "grad_norm": 9.365592995891348e-05, "learning_rate": 1.0648065807025345e-05, "loss": 0.0024, "step": 46800 }, { "epoch": 1.4858229051164265, "grad_norm": 0.008505141362547874, "learning_rate": 1.0625833703868387e-05, "loss": 0.0103, "step": 46900 }, { "epoch": 1.4889909710121971, "grad_norm": 0.0001479577913414687, "learning_rate": 1.0603601600711428e-05, "loss": 0.001, "step": 47000 }, { "epoch": 1.4921590369079678, "grad_norm": 8.014648483367637e-05, "learning_rate": 1.058136949755447e-05, "loss": 0.0026, "step": 47100 }, { "epoch": 1.4953271028037383, "grad_norm": 0.0002849689044523984, "learning_rate": 1.0559137394397512e-05, "loss": 0.0019, "step": 47200 }, { "epoch": 1.498495168699509, "grad_norm": 0.19057416915893555, "learning_rate": 1.0536905291240552e-05, "loss": 0.002, "step": 47300 }, { "epoch": 1.5016632345952796, "grad_norm": 9.28570661926642e-05, "learning_rate": 1.0514673188083592e-05, "loss": 0.0023, "step": 47400 }, { "epoch": 1.50483130049105, "grad_norm": 0.0016944918315857649, "learning_rate": 1.0492441084926634e-05, "loss": 0.0019, "step": 47500 }, { "epoch": 1.5079993663868207, "grad_norm": 0.31659239530563354, "learning_rate": 1.0470208981769676e-05, "loss": 0.0032, "step": 47600 }, { "epoch": 1.5111674322825914, "grad_norm": 0.0029370132833719254, "learning_rate": 1.0447976878612717e-05, "loss": 0.0201, "step": 47700 }, { "epoch": 1.514335498178362, "grad_norm": 0.004510208498686552, "learning_rate": 1.0425744775455759e-05, "loss": 0.0023, "step": 47800 }, { "epoch": 1.5175035640741328, "grad_norm": 3.642528099589981e-05, "learning_rate": 1.04035126722988e-05, "loss": 0.0019, "step": 47900 }, { "epoch": 1.5206716299699035, "grad_norm": 0.007735769730061293, "learning_rate": 1.0381280569141841e-05, "loss": 0.0016, "step": 48000 }, { "epoch": 1.5238396958656741, "grad_norm": 0.0006813256768509746, "learning_rate": 1.0359048465984883e-05, "loss": 0.0017, "step": 48100 }, { "epoch": 1.5270077617614446, "grad_norm": 5.778071499662474e-05, "learning_rate": 1.0336816362827924e-05, "loss": 0.0077, "step": 48200 }, { "epoch": 1.5301758276572153, "grad_norm": 0.0040051741525530815, "learning_rate": 1.0314584259670966e-05, "loss": 0.0188, "step": 48300 }, { "epoch": 1.533343893552986, "grad_norm": 0.08614891767501831, "learning_rate": 1.0292352156514008e-05, "loss": 0.0014, "step": 48400 }, { "epoch": 1.5365119594487564, "grad_norm": 12.830080032348633, "learning_rate": 1.0270120053357048e-05, "loss": 0.0057, "step": 48500 }, { "epoch": 1.539680025344527, "grad_norm": 0.0007354695699177682, "learning_rate": 1.024788795020009e-05, "loss": 0.0012, "step": 48600 }, { "epoch": 1.5428480912402978, "grad_norm": 0.000761797244194895, "learning_rate": 1.0225655847043132e-05, "loss": 0.0021, "step": 48700 }, { "epoch": 1.5460161571360684, "grad_norm": 0.33984261751174927, "learning_rate": 1.0203423743886172e-05, "loss": 0.001, "step": 48800 }, { "epoch": 1.5491842230318391, "grad_norm": 0.0008645505295135081, "learning_rate": 1.0181191640729213e-05, "loss": 0.0007, "step": 48900 }, { "epoch": 1.5523522889276098, "grad_norm": 0.0006037073908373713, "learning_rate": 1.0158959537572255e-05, "loss": 0.0032, "step": 49000 }, { "epoch": 1.5555203548233805, "grad_norm": 0.04107392951846123, "learning_rate": 1.0136727434415297e-05, "loss": 0.0015, "step": 49100 }, { "epoch": 1.558688420719151, "grad_norm": 0.0027483664453029633, "learning_rate": 1.0114495331258339e-05, "loss": 0.0006, "step": 49200 }, { "epoch": 1.5618564866149216, "grad_norm": 0.0002005908463615924, "learning_rate": 1.009226322810138e-05, "loss": 0.0234, "step": 49300 }, { "epoch": 1.565024552510692, "grad_norm": 0.004714667797088623, "learning_rate": 1.0070031124944422e-05, "loss": 0.0073, "step": 49400 }, { "epoch": 1.5681926184064627, "grad_norm": 0.022260790690779686, "learning_rate": 1.004779902178746e-05, "loss": 0.002, "step": 49500 }, { "epoch": 1.5713606843022334, "grad_norm": 0.004988121800124645, "learning_rate": 1.0025566918630502e-05, "loss": 0.0009, "step": 49600 }, { "epoch": 1.574528750198004, "grad_norm": 0.05250972509384155, "learning_rate": 1.0003334815473544e-05, "loss": 0.0015, "step": 49700 }, { "epoch": 1.5776968160937748, "grad_norm": 0.0003162017383147031, "learning_rate": 9.981102712316586e-06, "loss": 0.0019, "step": 49800 }, { "epoch": 1.5808648819895454, "grad_norm": 0.0022708792239427567, "learning_rate": 9.958870609159628e-06, "loss": 0.0036, "step": 49900 }, { "epoch": 1.5840329478853161, "grad_norm": 0.07355787605047226, "learning_rate": 9.936638506002668e-06, "loss": 0.0013, "step": 50000 }, { "epoch": 1.5872010137810868, "grad_norm": 0.016092980280518532, "learning_rate": 9.91440640284571e-06, "loss": 0.0036, "step": 50100 }, { "epoch": 1.5903690796768573, "grad_norm": 0.005923141725361347, "learning_rate": 9.892174299688751e-06, "loss": 0.0011, "step": 50200 }, { "epoch": 1.593537145572628, "grad_norm": 0.0008923260029405355, "learning_rate": 9.869942196531793e-06, "loss": 0.0044, "step": 50300 }, { "epoch": 1.5967052114683984, "grad_norm": 0.012484990991652012, "learning_rate": 9.847710093374833e-06, "loss": 0.0013, "step": 50400 }, { "epoch": 1.599873277364169, "grad_norm": 0.006260208319872618, "learning_rate": 9.825477990217875e-06, "loss": 0.0009, "step": 50500 }, { "epoch": 1.6030413432599397, "grad_norm": 0.0003789561160374433, "learning_rate": 9.803245887060917e-06, "loss": 0.0026, "step": 50600 }, { "epoch": 1.6062094091557104, "grad_norm": 0.0023648873902857304, "learning_rate": 9.781013783903958e-06, "loss": 0.0016, "step": 50700 }, { "epoch": 1.609377475051481, "grad_norm": 0.0007971607847139239, "learning_rate": 9.758781680746998e-06, "loss": 0.0179, "step": 50800 }, { "epoch": 1.6125455409472518, "grad_norm": 0.003260316327214241, "learning_rate": 9.73654957759004e-06, "loss": 0.0005, "step": 50900 }, { "epoch": 1.6157136068430225, "grad_norm": 0.009100884199142456, "learning_rate": 9.714317474433082e-06, "loss": 0.0016, "step": 51000 }, { "epoch": 1.6188816727387931, "grad_norm": 0.0007428972166962922, "learning_rate": 9.692085371276124e-06, "loss": 0.0111, "step": 51100 }, { "epoch": 1.6220497386345636, "grad_norm": 0.0015612394781783223, "learning_rate": 9.669853268119166e-06, "loss": 0.0043, "step": 51200 }, { "epoch": 1.6252178045303343, "grad_norm": 0.00023708381922915578, "learning_rate": 9.647621164962207e-06, "loss": 0.0008, "step": 51300 }, { "epoch": 1.6283858704261047, "grad_norm": 0.0002284464571857825, "learning_rate": 9.625389061805247e-06, "loss": 0.0027, "step": 51400 }, { "epoch": 1.6315539363218754, "grad_norm": 0.25235188007354736, "learning_rate": 9.603156958648289e-06, "loss": 0.0037, "step": 51500 }, { "epoch": 1.634722002217646, "grad_norm": 0.04865886643528938, "learning_rate": 9.580924855491331e-06, "loss": 0.0039, "step": 51600 }, { "epoch": 1.6378900681134168, "grad_norm": 0.011536366306245327, "learning_rate": 9.558692752334373e-06, "loss": 0.0295, "step": 51700 }, { "epoch": 1.6410581340091874, "grad_norm": 0.0033942603040486574, "learning_rate": 9.536460649177413e-06, "loss": 0.0041, "step": 51800 }, { "epoch": 1.644226199904958, "grad_norm": 0.0001564373233122751, "learning_rate": 9.514228546020454e-06, "loss": 0.0026, "step": 51900 }, { "epoch": 1.6473942658007288, "grad_norm": 0.004323447588831186, "learning_rate": 9.491996442863496e-06, "loss": 0.001, "step": 52000 }, { "epoch": 1.6505623316964992, "grad_norm": 0.02848782017827034, "learning_rate": 9.469764339706538e-06, "loss": 0.0008, "step": 52100 }, { "epoch": 1.65373039759227, "grad_norm": 0.021708877757191658, "learning_rate": 9.447532236549578e-06, "loss": 0.0016, "step": 52200 }, { "epoch": 1.6568984634880406, "grad_norm": 0.0003262820828240365, "learning_rate": 9.42530013339262e-06, "loss": 0.0009, "step": 52300 }, { "epoch": 1.660066529383811, "grad_norm": 0.001735742436721921, "learning_rate": 9.403068030235662e-06, "loss": 0.0013, "step": 52400 }, { "epoch": 1.6632345952795817, "grad_norm": 4.010135307908058e-05, "learning_rate": 9.380835927078702e-06, "loss": 0.0008, "step": 52500 }, { "epoch": 1.6664026611753524, "grad_norm": 0.0007688233745284379, "learning_rate": 9.358603823921743e-06, "loss": 0.0021, "step": 52600 }, { "epoch": 1.669570727071123, "grad_norm": 0.034028083086013794, "learning_rate": 9.336371720764785e-06, "loss": 0.0004, "step": 52700 }, { "epoch": 1.6727387929668938, "grad_norm": 0.0006701341480948031, "learning_rate": 9.314139617607827e-06, "loss": 0.0027, "step": 52800 }, { "epoch": 1.6759068588626644, "grad_norm": 0.001205560751259327, "learning_rate": 9.291907514450867e-06, "loss": 0.0006, "step": 52900 }, { "epoch": 1.6790749247584351, "grad_norm": 0.0008685667417012155, "learning_rate": 9.269675411293909e-06, "loss": 0.0002, "step": 53000 }, { "epoch": 1.6822429906542056, "grad_norm": 0.008366398513317108, "learning_rate": 9.24744330813695e-06, "loss": 0.0005, "step": 53100 }, { "epoch": 1.6854110565499762, "grad_norm": 0.14362505078315735, "learning_rate": 9.22521120497999e-06, "loss": 0.0054, "step": 53200 }, { "epoch": 1.6885791224457467, "grad_norm": 0.019972218200564384, "learning_rate": 9.202979101823032e-06, "loss": 0.0004, "step": 53300 }, { "epoch": 1.6917471883415174, "grad_norm": 0.0412134975194931, "learning_rate": 9.180746998666074e-06, "loss": 0.0015, "step": 53400 }, { "epoch": 1.694915254237288, "grad_norm": 0.003748115384951234, "learning_rate": 9.158514895509116e-06, "loss": 0.0013, "step": 53500 }, { "epoch": 1.6980833201330587, "grad_norm": 0.052952829748392105, "learning_rate": 9.136282792352158e-06, "loss": 0.0016, "step": 53600 }, { "epoch": 1.7012513860288294, "grad_norm": 0.008989219553768635, "learning_rate": 9.114050689195198e-06, "loss": 0.0072, "step": 53700 }, { "epoch": 1.7044194519246, "grad_norm": 0.047007884830236435, "learning_rate": 9.09181858603824e-06, "loss": 0.0014, "step": 53800 }, { "epoch": 1.7075875178203708, "grad_norm": 0.726980984210968, "learning_rate": 9.069586482881281e-06, "loss": 0.0054, "step": 53900 }, { "epoch": 1.7107555837161414, "grad_norm": 1.7654441762715578e-05, "learning_rate": 9.047354379724323e-06, "loss": 0.0031, "step": 54000 }, { "epoch": 1.713923649611912, "grad_norm": 0.03758937120437622, "learning_rate": 9.025122276567365e-06, "loss": 0.0018, "step": 54100 }, { "epoch": 1.7170917155076826, "grad_norm": 0.0004717516421806067, "learning_rate": 9.002890173410405e-06, "loss": 0.0177, "step": 54200 }, { "epoch": 1.720259781403453, "grad_norm": 0.0008279602043330669, "learning_rate": 8.980658070253447e-06, "loss": 0.0014, "step": 54300 }, { "epoch": 1.7234278472992237, "grad_norm": 0.010010186582803726, "learning_rate": 8.958425967096488e-06, "loss": 0.0019, "step": 54400 }, { "epoch": 1.7265959131949944, "grad_norm": 0.000608959817327559, "learning_rate": 8.93619386393953e-06, "loss": 0.0012, "step": 54500 }, { "epoch": 1.729763979090765, "grad_norm": 0.06244368851184845, "learning_rate": 8.913961760782572e-06, "loss": 0.0005, "step": 54600 }, { "epoch": 1.7329320449865357, "grad_norm": 0.03337325528264046, "learning_rate": 8.891729657625612e-06, "loss": 0.0013, "step": 54700 }, { "epoch": 1.7361001108823064, "grad_norm": 0.23330213129520416, "learning_rate": 8.869497554468654e-06, "loss": 0.0032, "step": 54800 }, { "epoch": 1.739268176778077, "grad_norm": 0.00020378937188070267, "learning_rate": 8.847265451311696e-06, "loss": 0.0028, "step": 54900 }, { "epoch": 1.7424362426738478, "grad_norm": 0.0336778499186039, "learning_rate": 8.825033348154736e-06, "loss": 0.0012, "step": 55000 }, { "epoch": 1.7456043085696182, "grad_norm": 7.928178092697635e-05, "learning_rate": 8.802801244997777e-06, "loss": 0.005, "step": 55100 }, { "epoch": 1.748772374465389, "grad_norm": 0.0005374435568228364, "learning_rate": 8.780569141840819e-06, "loss": 0.0059, "step": 55200 }, { "epoch": 1.7519404403611594, "grad_norm": 8.190758671844378e-05, "learning_rate": 8.758337038683861e-06, "loss": 0.001, "step": 55300 }, { "epoch": 1.75510850625693, "grad_norm": 0.2861631214618683, "learning_rate": 8.736104935526901e-06, "loss": 0.0032, "step": 55400 }, { "epoch": 1.7582765721527007, "grad_norm": 0.32926666736602783, "learning_rate": 8.713872832369943e-06, "loss": 0.0006, "step": 55500 }, { "epoch": 1.7614446380484714, "grad_norm": 0.0030724196694791317, "learning_rate": 8.691640729212984e-06, "loss": 0.01, "step": 55600 }, { "epoch": 1.764612703944242, "grad_norm": 0.032956257462501526, "learning_rate": 8.669408626056025e-06, "loss": 0.0014, "step": 55700 }, { "epoch": 1.7677807698400128, "grad_norm": 5.11103862663731e-05, "learning_rate": 8.647176522899066e-06, "loss": 0.0012, "step": 55800 }, { "epoch": 1.7709488357357834, "grad_norm": 0.12016189843416214, "learning_rate": 8.624944419742108e-06, "loss": 0.002, "step": 55900 }, { "epoch": 1.7741169016315539, "grad_norm": 0.0004587411822285503, "learning_rate": 8.60271231658515e-06, "loss": 0.0024, "step": 56000 }, { "epoch": 1.7772849675273246, "grad_norm": 0.0020646629855036736, "learning_rate": 8.58048021342819e-06, "loss": 0.0006, "step": 56100 }, { "epoch": 1.7804530334230952, "grad_norm": 2.86006870737765e-05, "learning_rate": 8.558248110271232e-06, "loss": 0.002, "step": 56200 }, { "epoch": 1.7836210993188657, "grad_norm": 0.0002436635986668989, "learning_rate": 8.536016007114273e-06, "loss": 0.0019, "step": 56300 }, { "epoch": 1.7867891652146364, "grad_norm": 0.0017174540553241968, "learning_rate": 8.513783903957315e-06, "loss": 0.0004, "step": 56400 }, { "epoch": 1.789957231110407, "grad_norm": 0.0032357415184378624, "learning_rate": 8.491551800800355e-06, "loss": 0.001, "step": 56500 }, { "epoch": 1.7931252970061777, "grad_norm": 0.006789275910705328, "learning_rate": 8.469319697643397e-06, "loss": 0.0032, "step": 56600 }, { "epoch": 1.7962933629019484, "grad_norm": 0.007675839588046074, "learning_rate": 8.447087594486439e-06, "loss": 0.0004, "step": 56700 }, { "epoch": 1.799461428797719, "grad_norm": 0.026306411251425743, "learning_rate": 8.42485549132948e-06, "loss": 0.0015, "step": 56800 }, { "epoch": 1.8026294946934898, "grad_norm": 2.372541348449886e-05, "learning_rate": 8.402623388172522e-06, "loss": 0.0013, "step": 56900 }, { "epoch": 1.8057975605892602, "grad_norm": 5.574413444264792e-05, "learning_rate": 8.380391285015564e-06, "loss": 0.0015, "step": 57000 }, { "epoch": 1.808965626485031, "grad_norm": 0.05931154265999794, "learning_rate": 8.358159181858604e-06, "loss": 0.0024, "step": 57100 }, { "epoch": 1.8121336923808016, "grad_norm": 0.39386412501335144, "learning_rate": 8.335927078701646e-06, "loss": 0.0071, "step": 57200 }, { "epoch": 1.815301758276572, "grad_norm": 0.00321913231164217, "learning_rate": 8.313694975544688e-06, "loss": 0.0096, "step": 57300 }, { "epoch": 1.8184698241723427, "grad_norm": 0.0608585849404335, "learning_rate": 8.29146287238773e-06, "loss": 0.0008, "step": 57400 }, { "epoch": 1.8216378900681134, "grad_norm": 0.0008937619277276099, "learning_rate": 8.26923076923077e-06, "loss": 0.0043, "step": 57500 }, { "epoch": 1.824805955963884, "grad_norm": 0.0015754963969811797, "learning_rate": 8.246998666073811e-06, "loss": 0.0011, "step": 57600 }, { "epoch": 1.8279740218596547, "grad_norm": 0.022448064759373665, "learning_rate": 8.224766562916853e-06, "loss": 0.0009, "step": 57700 }, { "epoch": 1.8311420877554254, "grad_norm": 0.1551535576581955, "learning_rate": 8.202534459759895e-06, "loss": 0.0054, "step": 57800 } ], "logging_steps": 100, "max_steps": 94695, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 100, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }