{ "best_metric": null, "best_model_checkpoint": null, "epoch": 50.0, "eval_steps": 500, "global_step": 161050, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.1552312946289972, "grad_norm": 2.5311179161071777, "learning_rate": 4.9844768705371e-05, "loss": 7.3069, "step": 500 }, { "epoch": 0.3104625892579944, "grad_norm": 1.969574213027954, "learning_rate": 4.968953741074201e-05, "loss": 6.5393, "step": 1000 }, { "epoch": 0.4656938838869916, "grad_norm": 2.073662757873535, "learning_rate": 4.953430611611301e-05, "loss": 6.1794, "step": 1500 }, { "epoch": 0.6209251785159888, "grad_norm": 1.9679251909255981, "learning_rate": 4.937907482148401e-05, "loss": 5.8754, "step": 2000 }, { "epoch": 0.776156473144986, "grad_norm": 2.1763973236083984, "learning_rate": 4.922384352685502e-05, "loss": 5.6265, "step": 2500 }, { "epoch": 0.9313877677739832, "grad_norm": 2.12394642829895, "learning_rate": 4.906861223222602e-05, "loss": 5.4302, "step": 3000 }, { "epoch": 1.0866190624029803, "grad_norm": 2.004790782928467, "learning_rate": 4.891338093759702e-05, "loss": 5.2491, "step": 3500 }, { "epoch": 1.2418503570319777, "grad_norm": 1.9632400274276733, "learning_rate": 4.875814964296802e-05, "loss": 5.1163, "step": 4000 }, { "epoch": 1.397081651660975, "grad_norm": 2.08528733253479, "learning_rate": 4.8602918348339024e-05, "loss": 5.0118, "step": 4500 }, { "epoch": 1.552312946289972, "grad_norm": 2.016941547393799, "learning_rate": 4.8447687053710025e-05, "loss": 4.9387, "step": 5000 }, { "epoch": 1.7075442409189692, "grad_norm": 2.050320863723755, "learning_rate": 4.829245575908103e-05, "loss": 4.8555, "step": 5500 }, { "epoch": 1.8627755355479665, "grad_norm": 2.028923273086548, "learning_rate": 4.8137224464452034e-05, "loss": 4.801, "step": 6000 }, { "epoch": 2.018006830176964, "grad_norm": 2.3016574382781982, "learning_rate": 4.7981993169823035e-05, "loss": 4.7415, "step": 6500 }, { "epoch": 2.1732381248059607, "grad_norm": 2.362550735473633, "learning_rate": 4.782676187519404e-05, "loss": 4.6673, "step": 7000 }, { "epoch": 2.328469419434958, "grad_norm": 2.2543234825134277, "learning_rate": 4.7671530580565043e-05, "loss": 4.6305, "step": 7500 }, { "epoch": 2.4837007140639553, "grad_norm": 2.4154460430145264, "learning_rate": 4.7516299285936044e-05, "loss": 4.5969, "step": 8000 }, { "epoch": 2.6389320086929526, "grad_norm": 2.489636182785034, "learning_rate": 4.736106799130705e-05, "loss": 4.5431, "step": 8500 }, { "epoch": 2.79416330332195, "grad_norm": 2.3382210731506348, "learning_rate": 4.720583669667805e-05, "loss": 4.5128, "step": 9000 }, { "epoch": 2.949394597950947, "grad_norm": 2.371870756149292, "learning_rate": 4.7050605402049054e-05, "loss": 4.481, "step": 9500 }, { "epoch": 3.104625892579944, "grad_norm": 2.4025254249572754, "learning_rate": 4.689537410742006e-05, "loss": 4.4316, "step": 10000 }, { "epoch": 3.2598571872089415, "grad_norm": 2.3801567554473877, "learning_rate": 4.674014281279106e-05, "loss": 4.392, "step": 10500 }, { "epoch": 3.4150884818379383, "grad_norm": 2.7669832706451416, "learning_rate": 4.6584911518162063e-05, "loss": 4.3718, "step": 11000 }, { "epoch": 3.5703197764669357, "grad_norm": 2.3554067611694336, "learning_rate": 4.6429680223533064e-05, "loss": 4.3374, "step": 11500 }, { "epoch": 3.725551071095933, "grad_norm": 2.672555923461914, "learning_rate": 4.6274448928904065e-05, "loss": 4.2903, "step": 12000 }, { "epoch": 3.8807823657249303, "grad_norm": 3.017108678817749, "learning_rate": 4.611921763427507e-05, "loss": 4.2753, "step": 12500 }, { "epoch": 4.036013660353928, "grad_norm": 2.9149186611175537, "learning_rate": 4.5963986339646074e-05, "loss": 4.2174, "step": 13000 }, { "epoch": 4.191244954982925, "grad_norm": 2.5483713150024414, "learning_rate": 4.5808755045017075e-05, "loss": 4.1771, "step": 13500 }, { "epoch": 4.346476249611921, "grad_norm": 2.8469910621643066, "learning_rate": 4.565352375038808e-05, "loss": 4.1374, "step": 14000 }, { "epoch": 4.501707544240919, "grad_norm": 2.5643463134765625, "learning_rate": 4.5498292455759083e-05, "loss": 4.1179, "step": 14500 }, { "epoch": 4.656938838869916, "grad_norm": 2.949036121368408, "learning_rate": 4.5343061161130084e-05, "loss": 4.0832, "step": 15000 }, { "epoch": 4.812170133498913, "grad_norm": 2.8270998001098633, "learning_rate": 4.518782986650109e-05, "loss": 4.066, "step": 15500 }, { "epoch": 4.967401428127911, "grad_norm": 2.839745283126831, "learning_rate": 4.503259857187209e-05, "loss": 4.0297, "step": 16000 }, { "epoch": 5.122632722756908, "grad_norm": 2.8617799282073975, "learning_rate": 4.4877367277243094e-05, "loss": 3.9769, "step": 16500 }, { "epoch": 5.277864017385905, "grad_norm": 2.988461494445801, "learning_rate": 4.47221359826141e-05, "loss": 3.9605, "step": 17000 }, { "epoch": 5.433095312014903, "grad_norm": 2.9263391494750977, "learning_rate": 4.45669046879851e-05, "loss": 3.9326, "step": 17500 }, { "epoch": 5.5883266066439, "grad_norm": 3.013662099838257, "learning_rate": 4.4411673393356104e-05, "loss": 3.9117, "step": 18000 }, { "epoch": 5.743557901272896, "grad_norm": 3.1931121349334717, "learning_rate": 4.4256442098727104e-05, "loss": 3.8977, "step": 18500 }, { "epoch": 5.898789195901894, "grad_norm": 3.0486295223236084, "learning_rate": 4.4101210804098105e-05, "loss": 3.8594, "step": 19000 }, { "epoch": 6.054020490530891, "grad_norm": 3.090029001235962, "learning_rate": 4.3945979509469106e-05, "loss": 3.8299, "step": 19500 }, { "epoch": 6.209251785159888, "grad_norm": 2.7931933403015137, "learning_rate": 4.3790748214840114e-05, "loss": 3.7919, "step": 20000 }, { "epoch": 6.364483079788886, "grad_norm": 3.0710182189941406, "learning_rate": 4.3635516920211115e-05, "loss": 3.7688, "step": 20500 }, { "epoch": 6.519714374417883, "grad_norm": 3.5505247116088867, "learning_rate": 4.3480285625582116e-05, "loss": 3.7545, "step": 21000 }, { "epoch": 6.674945669046879, "grad_norm": 3.376587390899658, "learning_rate": 4.3325054330953124e-05, "loss": 3.7285, "step": 21500 }, { "epoch": 6.830176963675877, "grad_norm": 4.988298416137695, "learning_rate": 4.3169823036324124e-05, "loss": 3.7192, "step": 22000 }, { "epoch": 6.985408258304874, "grad_norm": 3.1776342391967773, "learning_rate": 4.3014591741695125e-05, "loss": 3.6964, "step": 22500 }, { "epoch": 7.140639552933871, "grad_norm": 3.3717660903930664, "learning_rate": 4.285936044706613e-05, "loss": 3.6509, "step": 23000 }, { "epoch": 7.295870847562869, "grad_norm": 3.224043607711792, "learning_rate": 4.2704129152437134e-05, "loss": 3.6125, "step": 23500 }, { "epoch": 7.451102142191866, "grad_norm": 3.1459500789642334, "learning_rate": 4.2548897857808135e-05, "loss": 3.6209, "step": 24000 }, { "epoch": 7.606333436820863, "grad_norm": 3.296095132827759, "learning_rate": 4.239366656317914e-05, "loss": 3.5956, "step": 24500 }, { "epoch": 7.761564731449861, "grad_norm": 3.7389063835144043, "learning_rate": 4.2238435268550144e-05, "loss": 3.5743, "step": 25000 }, { "epoch": 7.916796026078858, "grad_norm": 3.22495174407959, "learning_rate": 4.2083203973921145e-05, "loss": 3.5633, "step": 25500 }, { "epoch": 8.072027320707855, "grad_norm": 3.2648046016693115, "learning_rate": 4.192797267929215e-05, "loss": 3.5174, "step": 26000 }, { "epoch": 8.227258615336853, "grad_norm": 3.516303539276123, "learning_rate": 4.177274138466315e-05, "loss": 3.4799, "step": 26500 }, { "epoch": 8.38248990996585, "grad_norm": 3.50242280960083, "learning_rate": 4.1617510090034154e-05, "loss": 3.4678, "step": 27000 }, { "epoch": 8.537721204594845, "grad_norm": 3.312835454940796, "learning_rate": 4.1462278795405155e-05, "loss": 3.4551, "step": 27500 }, { "epoch": 8.692952499223843, "grad_norm": 3.0370001792907715, "learning_rate": 4.1307047500776156e-05, "loss": 3.4393, "step": 28000 }, { "epoch": 8.84818379385284, "grad_norm": 3.387784242630005, "learning_rate": 4.115181620614716e-05, "loss": 3.415, "step": 28500 }, { "epoch": 9.003415088481837, "grad_norm": 3.6201112270355225, "learning_rate": 4.0996584911518165e-05, "loss": 3.418, "step": 29000 }, { "epoch": 9.158646383110835, "grad_norm": 3.2895004749298096, "learning_rate": 4.0841353616889166e-05, "loss": 3.3479, "step": 29500 }, { "epoch": 9.313877677739832, "grad_norm": 4.205784320831299, "learning_rate": 4.0686122322260166e-05, "loss": 3.3433, "step": 30000 }, { "epoch": 9.46910897236883, "grad_norm": 4.602181434631348, "learning_rate": 4.0530891027631174e-05, "loss": 3.3223, "step": 30500 }, { "epoch": 9.624340266997827, "grad_norm": 3.7277743816375732, "learning_rate": 4.0375659733002175e-05, "loss": 3.3039, "step": 31000 }, { "epoch": 9.779571561626824, "grad_norm": 3.491147994995117, "learning_rate": 4.0220428438373176e-05, "loss": 3.2895, "step": 31500 }, { "epoch": 9.934802856255821, "grad_norm": 3.2562263011932373, "learning_rate": 4.0065197143744184e-05, "loss": 3.2754, "step": 32000 }, { "epoch": 10.090034150884819, "grad_norm": 3.8227834701538086, "learning_rate": 3.9909965849115185e-05, "loss": 3.239, "step": 32500 }, { "epoch": 10.245265445513816, "grad_norm": 3.5297625064849854, "learning_rate": 3.9754734554486186e-05, "loss": 3.2036, "step": 33000 }, { "epoch": 10.400496740142813, "grad_norm": 3.621588945388794, "learning_rate": 3.959950325985719e-05, "loss": 3.1944, "step": 33500 }, { "epoch": 10.55572803477181, "grad_norm": 3.2394556999206543, "learning_rate": 3.9444271965228194e-05, "loss": 3.1961, "step": 34000 }, { "epoch": 10.710959329400808, "grad_norm": 3.6555140018463135, "learning_rate": 3.9289040670599195e-05, "loss": 3.1687, "step": 34500 }, { "epoch": 10.866190624029805, "grad_norm": 3.675229787826538, "learning_rate": 3.9133809375970196e-05, "loss": 3.1664, "step": 35000 }, { "epoch": 11.021421918658802, "grad_norm": 3.4863462448120117, "learning_rate": 3.89785780813412e-05, "loss": 3.1411, "step": 35500 }, { "epoch": 11.176653213287798, "grad_norm": 3.564634084701538, "learning_rate": 3.88233467867122e-05, "loss": 3.0942, "step": 36000 }, { "epoch": 11.331884507916795, "grad_norm": 3.6480257511138916, "learning_rate": 3.8668115492083206e-05, "loss": 3.0866, "step": 36500 }, { "epoch": 11.487115802545793, "grad_norm": 3.6294379234313965, "learning_rate": 3.8512884197454207e-05, "loss": 3.075, "step": 37000 }, { "epoch": 11.64234709717479, "grad_norm": 3.706338882446289, "learning_rate": 3.8357652902825214e-05, "loss": 3.0711, "step": 37500 }, { "epoch": 11.797578391803787, "grad_norm": 3.5022192001342773, "learning_rate": 3.8202421608196215e-05, "loss": 3.0543, "step": 38000 }, { "epoch": 11.952809686432785, "grad_norm": 3.718973159790039, "learning_rate": 3.8047190313567216e-05, "loss": 3.0499, "step": 38500 }, { "epoch": 12.108040981061782, "grad_norm": 3.5916123390197754, "learning_rate": 3.7891959018938224e-05, "loss": 3.0056, "step": 39000 }, { "epoch": 12.26327227569078, "grad_norm": 3.694257974624634, "learning_rate": 3.7736727724309225e-05, "loss": 2.9916, "step": 39500 }, { "epoch": 12.418503570319777, "grad_norm": 3.5255913734436035, "learning_rate": 3.7581496429680226e-05, "loss": 2.9823, "step": 40000 }, { "epoch": 12.573734864948774, "grad_norm": 3.4169323444366455, "learning_rate": 3.742626513505123e-05, "loss": 2.9726, "step": 40500 }, { "epoch": 12.728966159577771, "grad_norm": 3.2919702529907227, "learning_rate": 3.7271033840422234e-05, "loss": 2.9586, "step": 41000 }, { "epoch": 12.884197454206769, "grad_norm": 3.52095890045166, "learning_rate": 3.7115802545793235e-05, "loss": 2.9576, "step": 41500 }, { "epoch": 13.039428748835766, "grad_norm": 4.317255020141602, "learning_rate": 3.6960571251164236e-05, "loss": 2.936, "step": 42000 }, { "epoch": 13.194660043464763, "grad_norm": 3.2749974727630615, "learning_rate": 3.680533995653524e-05, "loss": 2.894, "step": 42500 }, { "epoch": 13.34989133809376, "grad_norm": 4.149952411651611, "learning_rate": 3.665010866190624e-05, "loss": 2.8892, "step": 43000 }, { "epoch": 13.505122632722756, "grad_norm": 3.8253960609436035, "learning_rate": 3.6494877367277246e-05, "loss": 2.8957, "step": 43500 }, { "epoch": 13.660353927351753, "grad_norm": 3.3494088649749756, "learning_rate": 3.6339646072648247e-05, "loss": 2.8821, "step": 44000 }, { "epoch": 13.81558522198075, "grad_norm": 3.426492691040039, "learning_rate": 3.618441477801925e-05, "loss": 2.8655, "step": 44500 }, { "epoch": 13.970816516609748, "grad_norm": 3.3416435718536377, "learning_rate": 3.6029183483390255e-05, "loss": 2.8685, "step": 45000 }, { "epoch": 14.126047811238745, "grad_norm": 3.6365139484405518, "learning_rate": 3.5873952188761256e-05, "loss": 2.8184, "step": 45500 }, { "epoch": 14.281279105867743, "grad_norm": 3.4732120037078857, "learning_rate": 3.571872089413226e-05, "loss": 2.8102, "step": 46000 }, { "epoch": 14.43651040049674, "grad_norm": 3.4674530029296875, "learning_rate": 3.5563489599503265e-05, "loss": 2.8077, "step": 46500 }, { "epoch": 14.591741695125737, "grad_norm": 3.3378286361694336, "learning_rate": 3.5408258304874266e-05, "loss": 2.808, "step": 47000 }, { "epoch": 14.746972989754735, "grad_norm": 3.5950191020965576, "learning_rate": 3.525302701024527e-05, "loss": 2.7938, "step": 47500 }, { "epoch": 14.902204284383732, "grad_norm": 3.4225220680236816, "learning_rate": 3.5097795715616274e-05, "loss": 2.7817, "step": 48000 }, { "epoch": 15.05743557901273, "grad_norm": 3.353314161300659, "learning_rate": 3.4942564420987275e-05, "loss": 2.7783, "step": 48500 }, { "epoch": 15.212666873641727, "grad_norm": 3.483828067779541, "learning_rate": 3.4787333126358276e-05, "loss": 2.7388, "step": 49000 }, { "epoch": 15.367898168270724, "grad_norm": 3.295490026473999, "learning_rate": 3.463210183172928e-05, "loss": 2.7404, "step": 49500 }, { "epoch": 15.523129462899721, "grad_norm": 3.7673990726470947, "learning_rate": 3.447687053710028e-05, "loss": 2.7192, "step": 50000 }, { "epoch": 15.678360757528718, "grad_norm": 3.3685121536254883, "learning_rate": 3.432163924247128e-05, "loss": 2.7341, "step": 50500 }, { "epoch": 15.833592052157716, "grad_norm": 3.8389620780944824, "learning_rate": 3.416640794784229e-05, "loss": 2.7239, "step": 51000 }, { "epoch": 15.988823346786713, "grad_norm": 3.8301634788513184, "learning_rate": 3.401117665321329e-05, "loss": 2.716, "step": 51500 }, { "epoch": 16.14405464141571, "grad_norm": 3.7196969985961914, "learning_rate": 3.385594535858429e-05, "loss": 2.6732, "step": 52000 }, { "epoch": 16.299285936044708, "grad_norm": 3.847682476043701, "learning_rate": 3.3700714063955296e-05, "loss": 2.6561, "step": 52500 }, { "epoch": 16.454517230673705, "grad_norm": 3.079871416091919, "learning_rate": 3.35454827693263e-05, "loss": 2.6765, "step": 53000 }, { "epoch": 16.609748525302702, "grad_norm": 3.7256991863250732, "learning_rate": 3.33902514746973e-05, "loss": 2.6754, "step": 53500 }, { "epoch": 16.7649798199317, "grad_norm": 3.556029796600342, "learning_rate": 3.3235020180068306e-05, "loss": 2.6675, "step": 54000 }, { "epoch": 16.920211114560697, "grad_norm": 3.9556143283843994, "learning_rate": 3.307978888543931e-05, "loss": 2.6507, "step": 54500 }, { "epoch": 17.075442409189694, "grad_norm": 3.482832431793213, "learning_rate": 3.292455759081031e-05, "loss": 2.6343, "step": 55000 }, { "epoch": 17.230673703818688, "grad_norm": 4.063329696655273, "learning_rate": 3.2769326296181315e-05, "loss": 2.6071, "step": 55500 }, { "epoch": 17.385904998447685, "grad_norm": 3.9433131217956543, "learning_rate": 3.2614095001552316e-05, "loss": 2.6064, "step": 56000 }, { "epoch": 17.541136293076683, "grad_norm": 3.5680246353149414, "learning_rate": 3.245886370692332e-05, "loss": 2.6091, "step": 56500 }, { "epoch": 17.69636758770568, "grad_norm": 3.5213463306427, "learning_rate": 3.230363241229432e-05, "loss": 2.6058, "step": 57000 }, { "epoch": 17.851598882334677, "grad_norm": 3.5061028003692627, "learning_rate": 3.214840111766532e-05, "loss": 2.5981, "step": 57500 }, { "epoch": 18.006830176963675, "grad_norm": 3.4912548065185547, "learning_rate": 3.199316982303632e-05, "loss": 2.5998, "step": 58000 }, { "epoch": 18.162061471592672, "grad_norm": 3.2917845249176025, "learning_rate": 3.183793852840733e-05, "loss": 2.5569, "step": 58500 }, { "epoch": 18.31729276622167, "grad_norm": 3.707597017288208, "learning_rate": 3.168270723377833e-05, "loss": 2.555, "step": 59000 }, { "epoch": 18.472524060850667, "grad_norm": 3.4082422256469727, "learning_rate": 3.152747593914933e-05, "loss": 2.5539, "step": 59500 }, { "epoch": 18.627755355479664, "grad_norm": 3.6042168140411377, "learning_rate": 3.137224464452034e-05, "loss": 2.5485, "step": 60000 }, { "epoch": 18.78298665010866, "grad_norm": 3.665288209915161, "learning_rate": 3.121701334989134e-05, "loss": 2.5499, "step": 60500 }, { "epoch": 18.93821794473766, "grad_norm": 3.590956211090088, "learning_rate": 3.1061782055262346e-05, "loss": 2.5479, "step": 61000 }, { "epoch": 19.093449239366656, "grad_norm": 3.468674898147583, "learning_rate": 3.090655076063335e-05, "loss": 2.5156, "step": 61500 }, { "epoch": 19.248680533995653, "grad_norm": 3.4389026165008545, "learning_rate": 3.075131946600435e-05, "loss": 2.5037, "step": 62000 }, { "epoch": 19.40391182862465, "grad_norm": 3.568042039871216, "learning_rate": 3.0596088171375355e-05, "loss": 2.514, "step": 62500 }, { "epoch": 19.559143123253648, "grad_norm": 3.448721408843994, "learning_rate": 3.0440856876746353e-05, "loss": 2.5004, "step": 63000 }, { "epoch": 19.714374417882645, "grad_norm": 3.3219761848449707, "learning_rate": 3.0285625582117354e-05, "loss": 2.5072, "step": 63500 }, { "epoch": 19.869605712511643, "grad_norm": 3.5406880378723145, "learning_rate": 3.013039428748836e-05, "loss": 2.5006, "step": 64000 }, { "epoch": 20.02483700714064, "grad_norm": 3.606174945831299, "learning_rate": 2.9975162992859363e-05, "loss": 2.5015, "step": 64500 }, { "epoch": 20.180068301769637, "grad_norm": 3.5096230506896973, "learning_rate": 2.9819931698230363e-05, "loss": 2.456, "step": 65000 }, { "epoch": 20.335299596398634, "grad_norm": 3.478529691696167, "learning_rate": 2.966470040360137e-05, "loss": 2.4515, "step": 65500 }, { "epoch": 20.490530891027632, "grad_norm": 3.3920540809631348, "learning_rate": 2.9509469108972372e-05, "loss": 2.4625, "step": 66000 }, { "epoch": 20.64576218565663, "grad_norm": 3.6235618591308594, "learning_rate": 2.9354237814343373e-05, "loss": 2.4665, "step": 66500 }, { "epoch": 20.800993480285626, "grad_norm": 3.580286979675293, "learning_rate": 2.9199006519714377e-05, "loss": 2.4577, "step": 67000 }, { "epoch": 20.956224774914624, "grad_norm": 3.5689404010772705, "learning_rate": 2.9043775225085378e-05, "loss": 2.4619, "step": 67500 }, { "epoch": 21.11145606954362, "grad_norm": 3.5909037590026855, "learning_rate": 2.888854393045638e-05, "loss": 2.4148, "step": 68000 }, { "epoch": 21.26668736417262, "grad_norm": 3.4171316623687744, "learning_rate": 2.8733312635827387e-05, "loss": 2.4071, "step": 68500 }, { "epoch": 21.421918658801616, "grad_norm": 3.457381248474121, "learning_rate": 2.8578081341198388e-05, "loss": 2.4169, "step": 69000 }, { "epoch": 21.577149953430613, "grad_norm": 3.420381784439087, "learning_rate": 2.842285004656939e-05, "loss": 2.42, "step": 69500 }, { "epoch": 21.73238124805961, "grad_norm": 3.2285988330841064, "learning_rate": 2.8267618751940393e-05, "loss": 2.4262, "step": 70000 }, { "epoch": 21.887612542688608, "grad_norm": 3.473289728164673, "learning_rate": 2.8112387457311394e-05, "loss": 2.413, "step": 70500 }, { "epoch": 22.042843837317605, "grad_norm": 3.272094488143921, "learning_rate": 2.7957156162682395e-05, "loss": 2.4166, "step": 71000 }, { "epoch": 22.1980751319466, "grad_norm": 3.903505563735962, "learning_rate": 2.7801924868053403e-05, "loss": 2.3766, "step": 71500 }, { "epoch": 22.353306426575596, "grad_norm": 3.58394193649292, "learning_rate": 2.7646693573424404e-05, "loss": 2.3844, "step": 72000 }, { "epoch": 22.508537721204593, "grad_norm": 3.6447036266326904, "learning_rate": 2.7491462278795404e-05, "loss": 2.3683, "step": 72500 }, { "epoch": 22.66376901583359, "grad_norm": 3.740295648574829, "learning_rate": 2.7336230984166412e-05, "loss": 2.3822, "step": 73000 }, { "epoch": 22.819000310462588, "grad_norm": 3.439695358276367, "learning_rate": 2.7180999689537413e-05, "loss": 2.3893, "step": 73500 }, { "epoch": 22.974231605091585, "grad_norm": 3.6966145038604736, "learning_rate": 2.7025768394908414e-05, "loss": 2.3803, "step": 74000 }, { "epoch": 23.129462899720583, "grad_norm": 3.1649584770202637, "learning_rate": 2.687053710027942e-05, "loss": 2.343, "step": 74500 }, { "epoch": 23.28469419434958, "grad_norm": 3.4965484142303467, "learning_rate": 2.671530580565042e-05, "loss": 2.3351, "step": 75000 }, { "epoch": 23.439925488978577, "grad_norm": 3.3004109859466553, "learning_rate": 2.656007451102142e-05, "loss": 2.335, "step": 75500 }, { "epoch": 23.595156783607575, "grad_norm": 3.5990171432495117, "learning_rate": 2.6404843216392428e-05, "loss": 2.3474, "step": 76000 }, { "epoch": 23.750388078236572, "grad_norm": 3.2564170360565186, "learning_rate": 2.624961192176343e-05, "loss": 2.3447, "step": 76500 }, { "epoch": 23.90561937286557, "grad_norm": 3.103309392929077, "learning_rate": 2.609438062713443e-05, "loss": 2.3536, "step": 77000 }, { "epoch": 24.060850667494567, "grad_norm": 3.434351921081543, "learning_rate": 2.5939149332505437e-05, "loss": 2.3314, "step": 77500 }, { "epoch": 24.216081962123564, "grad_norm": 3.0751280784606934, "learning_rate": 2.5783918037876435e-05, "loss": 2.3113, "step": 78000 }, { "epoch": 24.37131325675256, "grad_norm": 3.3030903339385986, "learning_rate": 2.5628686743247436e-05, "loss": 2.3074, "step": 78500 }, { "epoch": 24.52654455138156, "grad_norm": 3.4424233436584473, "learning_rate": 2.5473455448618444e-05, "loss": 2.3074, "step": 79000 }, { "epoch": 24.681775846010556, "grad_norm": 3.4242780208587646, "learning_rate": 2.5318224153989445e-05, "loss": 2.3127, "step": 79500 }, { "epoch": 24.837007140639553, "grad_norm": 3.478363275527954, "learning_rate": 2.5162992859360445e-05, "loss": 2.3068, "step": 80000 }, { "epoch": 24.99223843526855, "grad_norm": 3.381448745727539, "learning_rate": 2.5007761564731453e-05, "loss": 2.3128, "step": 80500 }, { "epoch": 25.147469729897548, "grad_norm": 3.3895316123962402, "learning_rate": 2.4852530270102454e-05, "loss": 2.2783, "step": 81000 }, { "epoch": 25.302701024526545, "grad_norm": 3.522529363632202, "learning_rate": 2.469729897547346e-05, "loss": 2.2744, "step": 81500 }, { "epoch": 25.457932319155542, "grad_norm": 3.2338192462921143, "learning_rate": 2.454206768084446e-05, "loss": 2.2724, "step": 82000 }, { "epoch": 25.61316361378454, "grad_norm": 3.4074482917785645, "learning_rate": 2.438683638621546e-05, "loss": 2.2792, "step": 82500 }, { "epoch": 25.768394908413537, "grad_norm": 4.039905548095703, "learning_rate": 2.4231605091586465e-05, "loss": 2.2841, "step": 83000 }, { "epoch": 25.923626203042534, "grad_norm": 3.547403335571289, "learning_rate": 2.407637379695747e-05, "loss": 2.2819, "step": 83500 }, { "epoch": 26.07885749767153, "grad_norm": 3.3868236541748047, "learning_rate": 2.392114250232847e-05, "loss": 2.2798, "step": 84000 }, { "epoch": 26.23408879230053, "grad_norm": 3.4890501499176025, "learning_rate": 2.3765911207699474e-05, "loss": 2.2533, "step": 84500 }, { "epoch": 26.389320086929526, "grad_norm": 3.4324049949645996, "learning_rate": 2.361067991307048e-05, "loss": 2.2503, "step": 85000 }, { "epoch": 26.544551381558524, "grad_norm": 3.444091558456421, "learning_rate": 2.345544861844148e-05, "loss": 2.2475, "step": 85500 }, { "epoch": 26.69978267618752, "grad_norm": 3.4090046882629395, "learning_rate": 2.330021732381248e-05, "loss": 2.2484, "step": 86000 }, { "epoch": 26.85501397081652, "grad_norm": 3.447425365447998, "learning_rate": 2.3144986029183485e-05, "loss": 2.2488, "step": 86500 }, { "epoch": 27.010245265445512, "grad_norm": 3.323572874069214, "learning_rate": 2.2989754734554486e-05, "loss": 2.2459, "step": 87000 }, { "epoch": 27.16547656007451, "grad_norm": 3.2703440189361572, "learning_rate": 2.283452343992549e-05, "loss": 2.2174, "step": 87500 }, { "epoch": 27.320707854703507, "grad_norm": 3.5882294178009033, "learning_rate": 2.2679292145296494e-05, "loss": 2.2148, "step": 88000 }, { "epoch": 27.475939149332504, "grad_norm": 3.4588534832000732, "learning_rate": 2.2524060850667495e-05, "loss": 2.2225, "step": 88500 }, { "epoch": 27.6311704439615, "grad_norm": 3.370229482650757, "learning_rate": 2.23688295560385e-05, "loss": 2.2274, "step": 89000 }, { "epoch": 27.7864017385905, "grad_norm": 3.271768093109131, "learning_rate": 2.22135982614095e-05, "loss": 2.2272, "step": 89500 }, { "epoch": 27.941633033219496, "grad_norm": 3.4835493564605713, "learning_rate": 2.20583669667805e-05, "loss": 2.2324, "step": 90000 }, { "epoch": 28.096864327848493, "grad_norm": 3.6498279571533203, "learning_rate": 2.1903135672151506e-05, "loss": 2.1981, "step": 90500 }, { "epoch": 28.25209562247749, "grad_norm": 3.356560468673706, "learning_rate": 2.174790437752251e-05, "loss": 2.2064, "step": 91000 }, { "epoch": 28.407326917106488, "grad_norm": 3.2081451416015625, "learning_rate": 2.1592673082893514e-05, "loss": 2.1893, "step": 91500 }, { "epoch": 28.562558211735485, "grad_norm": 3.4001569747924805, "learning_rate": 2.1437441788264515e-05, "loss": 2.2014, "step": 92000 }, { "epoch": 28.717789506364483, "grad_norm": 3.529869794845581, "learning_rate": 2.128221049363552e-05, "loss": 2.2016, "step": 92500 }, { "epoch": 28.87302080099348, "grad_norm": 3.313579797744751, "learning_rate": 2.112697919900652e-05, "loss": 2.1937, "step": 93000 }, { "epoch": 29.028252095622477, "grad_norm": 3.336879253387451, "learning_rate": 2.097174790437752e-05, "loss": 2.1945, "step": 93500 }, { "epoch": 29.183483390251475, "grad_norm": 3.567126512527466, "learning_rate": 2.0816516609748526e-05, "loss": 2.1729, "step": 94000 }, { "epoch": 29.338714684880472, "grad_norm": 3.5103540420532227, "learning_rate": 2.066128531511953e-05, "loss": 2.1772, "step": 94500 }, { "epoch": 29.49394597950947, "grad_norm": 3.479119300842285, "learning_rate": 2.050605402049053e-05, "loss": 2.1786, "step": 95000 }, { "epoch": 29.649177274138466, "grad_norm": 3.516080856323242, "learning_rate": 2.0350822725861535e-05, "loss": 2.1714, "step": 95500 }, { "epoch": 29.804408568767464, "grad_norm": 3.3175048828125, "learning_rate": 2.019559143123254e-05, "loss": 2.176, "step": 96000 }, { "epoch": 29.95963986339646, "grad_norm": 3.5741403102874756, "learning_rate": 2.004036013660354e-05, "loss": 2.1747, "step": 96500 }, { "epoch": 30.11487115802546, "grad_norm": 3.3326079845428467, "learning_rate": 1.988512884197454e-05, "loss": 2.1465, "step": 97000 }, { "epoch": 30.270102452654456, "grad_norm": 3.591451406478882, "learning_rate": 1.9729897547345546e-05, "loss": 2.1452, "step": 97500 }, { "epoch": 30.425333747283453, "grad_norm": 3.386875629425049, "learning_rate": 1.9574666252716547e-05, "loss": 2.1527, "step": 98000 }, { "epoch": 30.58056504191245, "grad_norm": 3.5038528442382812, "learning_rate": 1.941943495808755e-05, "loss": 2.1562, "step": 98500 }, { "epoch": 30.735796336541448, "grad_norm": 3.5655734539031982, "learning_rate": 1.9264203663458555e-05, "loss": 2.1579, "step": 99000 }, { "epoch": 30.891027631170445, "grad_norm": 3.556865692138672, "learning_rate": 1.9108972368829556e-05, "loss": 2.1654, "step": 99500 }, { "epoch": 31.046258925799442, "grad_norm": 3.171966791152954, "learning_rate": 1.895374107420056e-05, "loss": 2.1375, "step": 100000 }, { "epoch": 31.20149022042844, "grad_norm": 3.2655224800109863, "learning_rate": 1.8798509779571565e-05, "loss": 2.1242, "step": 100500 }, { "epoch": 31.356721515057437, "grad_norm": 3.5983667373657227, "learning_rate": 1.8643278484942562e-05, "loss": 2.1241, "step": 101000 }, { "epoch": 31.511952809686434, "grad_norm": 3.4082624912261963, "learning_rate": 1.8488047190313567e-05, "loss": 2.1286, "step": 101500 }, { "epoch": 31.66718410431543, "grad_norm": 3.6842904090881348, "learning_rate": 1.833281589568457e-05, "loss": 2.1307, "step": 102000 }, { "epoch": 31.82241539894443, "grad_norm": 3.506561040878296, "learning_rate": 1.8177584601055575e-05, "loss": 2.128, "step": 102500 }, { "epoch": 31.977646693573423, "grad_norm": 3.1754329204559326, "learning_rate": 1.8022353306426576e-05, "loss": 2.1546, "step": 103000 }, { "epoch": 32.132877988202424, "grad_norm": 3.4814705848693848, "learning_rate": 1.786712201179758e-05, "loss": 2.1155, "step": 103500 }, { "epoch": 32.28810928283142, "grad_norm": 3.7139673233032227, "learning_rate": 1.7711890717168585e-05, "loss": 2.1056, "step": 104000 }, { "epoch": 32.44334057746042, "grad_norm": 3.498924732208252, "learning_rate": 1.7556659422539586e-05, "loss": 2.1142, "step": 104500 }, { "epoch": 32.598571872089416, "grad_norm": 3.50618052482605, "learning_rate": 1.7401428127910587e-05, "loss": 2.1133, "step": 105000 }, { "epoch": 32.75380316671841, "grad_norm": 3.3597817420959473, "learning_rate": 1.724619683328159e-05, "loss": 2.1023, "step": 105500 }, { "epoch": 32.90903446134741, "grad_norm": 3.1257002353668213, "learning_rate": 1.7090965538652592e-05, "loss": 2.1164, "step": 106000 }, { "epoch": 33.06426575597641, "grad_norm": 3.2539827823638916, "learning_rate": 1.6935734244023596e-05, "loss": 2.094, "step": 106500 }, { "epoch": 33.219497050605405, "grad_norm": 3.5922114849090576, "learning_rate": 1.67805029493946e-05, "loss": 2.0797, "step": 107000 }, { "epoch": 33.3747283452344, "grad_norm": 3.4829044342041016, "learning_rate": 1.66252716547656e-05, "loss": 2.0991, "step": 107500 }, { "epoch": 33.5299596398634, "grad_norm": 3.402447462081909, "learning_rate": 1.6470040360136606e-05, "loss": 2.095, "step": 108000 }, { "epoch": 33.6851909344924, "grad_norm": 3.4598045349121094, "learning_rate": 1.6314809065507607e-05, "loss": 2.0866, "step": 108500 }, { "epoch": 33.840422229121394, "grad_norm": 3.806218147277832, "learning_rate": 1.6159577770878608e-05, "loss": 2.1014, "step": 109000 }, { "epoch": 33.99565352375039, "grad_norm": 3.7168009281158447, "learning_rate": 1.6004346476249612e-05, "loss": 2.103, "step": 109500 }, { "epoch": 34.15088481837939, "grad_norm": 3.703636884689331, "learning_rate": 1.5849115181620616e-05, "loss": 2.0691, "step": 110000 }, { "epoch": 34.306116113008386, "grad_norm": 3.4325764179229736, "learning_rate": 1.5693883886991617e-05, "loss": 2.077, "step": 110500 }, { "epoch": 34.461347407637376, "grad_norm": 3.5435197353363037, "learning_rate": 1.553865259236262e-05, "loss": 2.0839, "step": 111000 }, { "epoch": 34.616578702266374, "grad_norm": 3.507366418838501, "learning_rate": 1.5383421297733626e-05, "loss": 2.0709, "step": 111500 }, { "epoch": 34.77180999689537, "grad_norm": 3.401642084121704, "learning_rate": 1.5228190003104625e-05, "loss": 2.0656, "step": 112000 }, { "epoch": 34.92704129152437, "grad_norm": 3.4845223426818848, "learning_rate": 1.507295870847563e-05, "loss": 2.0848, "step": 112500 }, { "epoch": 35.082272586153366, "grad_norm": 3.477248191833496, "learning_rate": 1.4917727413846632e-05, "loss": 2.0595, "step": 113000 }, { "epoch": 35.23750388078236, "grad_norm": 3.3301658630371094, "learning_rate": 1.4762496119217633e-05, "loss": 2.056, "step": 113500 }, { "epoch": 35.39273517541136, "grad_norm": 3.5794308185577393, "learning_rate": 1.4607264824588637e-05, "loss": 2.0579, "step": 114000 }, { "epoch": 35.54796647004036, "grad_norm": 3.680616617202759, "learning_rate": 1.4452033529959642e-05, "loss": 2.0607, "step": 114500 }, { "epoch": 35.703197764669355, "grad_norm": 3.581709384918213, "learning_rate": 1.4296802235330644e-05, "loss": 2.0583, "step": 115000 }, { "epoch": 35.85842905929835, "grad_norm": 3.3622961044311523, "learning_rate": 1.4141570940701645e-05, "loss": 2.0646, "step": 115500 }, { "epoch": 36.01366035392735, "grad_norm": 3.7327592372894287, "learning_rate": 1.398633964607265e-05, "loss": 2.0665, "step": 116000 }, { "epoch": 36.16889164855635, "grad_norm": 3.1711039543151855, "learning_rate": 1.3831108351443654e-05, "loss": 2.0485, "step": 116500 }, { "epoch": 36.324122943185344, "grad_norm": 3.367382764816284, "learning_rate": 1.3675877056814653e-05, "loss": 2.0385, "step": 117000 }, { "epoch": 36.47935423781434, "grad_norm": 3.5484557151794434, "learning_rate": 1.3520645762185657e-05, "loss": 2.0345, "step": 117500 }, { "epoch": 36.63458553244334, "grad_norm": 3.6032321453094482, "learning_rate": 1.3365414467556662e-05, "loss": 2.0468, "step": 118000 }, { "epoch": 36.789816827072336, "grad_norm": 4.026392936706543, "learning_rate": 1.3210183172927663e-05, "loss": 2.0481, "step": 118500 }, { "epoch": 36.94504812170133, "grad_norm": 3.644474983215332, "learning_rate": 1.3054951878298665e-05, "loss": 2.0447, "step": 119000 }, { "epoch": 37.10027941633033, "grad_norm": 3.28544545173645, "learning_rate": 1.289972058366967e-05, "loss": 2.0358, "step": 119500 }, { "epoch": 37.25551071095933, "grad_norm": 3.2941246032714844, "learning_rate": 1.274448928904067e-05, "loss": 2.0084, "step": 120000 }, { "epoch": 37.410742005588325, "grad_norm": 3.7883927822113037, "learning_rate": 1.2589257994411675e-05, "loss": 2.0329, "step": 120500 }, { "epoch": 37.56597330021732, "grad_norm": 3.23514986038208, "learning_rate": 1.2434026699782676e-05, "loss": 2.0297, "step": 121000 }, { "epoch": 37.72120459484632, "grad_norm": 4.0385894775390625, "learning_rate": 1.227879540515368e-05, "loss": 2.0347, "step": 121500 }, { "epoch": 37.87643588947532, "grad_norm": 3.6746459007263184, "learning_rate": 1.2123564110524683e-05, "loss": 2.0386, "step": 122000 }, { "epoch": 38.031667184104315, "grad_norm": 3.5813486576080322, "learning_rate": 1.1968332815895685e-05, "loss": 2.0275, "step": 122500 }, { "epoch": 38.18689847873331, "grad_norm": 3.3303112983703613, "learning_rate": 1.1813101521266688e-05, "loss": 2.0086, "step": 123000 }, { "epoch": 38.34212977336231, "grad_norm": 3.5184788703918457, "learning_rate": 1.165787022663769e-05, "loss": 2.016, "step": 123500 }, { "epoch": 38.49736106799131, "grad_norm": 3.505143642425537, "learning_rate": 1.1502638932008695e-05, "loss": 2.0179, "step": 124000 }, { "epoch": 38.652592362620304, "grad_norm": 3.432009696960449, "learning_rate": 1.1347407637379696e-05, "loss": 2.02, "step": 124500 }, { "epoch": 38.8078236572493, "grad_norm": 3.268419027328491, "learning_rate": 1.1192176342750698e-05, "loss": 2.0185, "step": 125000 }, { "epoch": 38.9630549518783, "grad_norm": 3.895332098007202, "learning_rate": 1.1036945048121703e-05, "loss": 2.0213, "step": 125500 }, { "epoch": 39.118286246507296, "grad_norm": 3.481630325317383, "learning_rate": 1.0881713753492705e-05, "loss": 1.9974, "step": 126000 }, { "epoch": 39.27351754113629, "grad_norm": 3.3944613933563232, "learning_rate": 1.0726482458863706e-05, "loss": 2.0012, "step": 126500 }, { "epoch": 39.42874883576529, "grad_norm": 4.5043182373046875, "learning_rate": 1.057125116423471e-05, "loss": 2.0044, "step": 127000 }, { "epoch": 39.58398013039429, "grad_norm": 3.6557388305664062, "learning_rate": 1.0416019869605713e-05, "loss": 2.0059, "step": 127500 }, { "epoch": 39.739211425023285, "grad_norm": 3.3339288234710693, "learning_rate": 1.0260788574976716e-05, "loss": 2.0018, "step": 128000 }, { "epoch": 39.89444271965228, "grad_norm": 3.376677989959717, "learning_rate": 1.0105557280347718e-05, "loss": 2.0102, "step": 128500 }, { "epoch": 40.04967401428128, "grad_norm": 3.601083278656006, "learning_rate": 9.950325985718721e-06, "loss": 2.0019, "step": 129000 }, { "epoch": 40.20490530891028, "grad_norm": 3.458583116531372, "learning_rate": 9.795094691089725e-06, "loss": 1.9888, "step": 129500 }, { "epoch": 40.360136603539274, "grad_norm": 3.3313772678375244, "learning_rate": 9.639863396460728e-06, "loss": 1.9947, "step": 130000 }, { "epoch": 40.51536789816827, "grad_norm": 3.6328794956207275, "learning_rate": 9.484632101831729e-06, "loss": 1.985, "step": 130500 }, { "epoch": 40.67059919279727, "grad_norm": 3.165858030319214, "learning_rate": 9.329400807202733e-06, "loss": 1.9962, "step": 131000 }, { "epoch": 40.825830487426266, "grad_norm": 3.330787420272827, "learning_rate": 9.174169512573736e-06, "loss": 1.9982, "step": 131500 }, { "epoch": 40.981061782055264, "grad_norm": 3.3106303215026855, "learning_rate": 9.018938217944738e-06, "loss": 1.9974, "step": 132000 }, { "epoch": 41.13629307668426, "grad_norm": 3.49682354927063, "learning_rate": 8.863706923315741e-06, "loss": 1.9706, "step": 132500 }, { "epoch": 41.29152437131326, "grad_norm": 3.306004047393799, "learning_rate": 8.708475628686744e-06, "loss": 1.9795, "step": 133000 }, { "epoch": 41.446755665942256, "grad_norm": 3.715866804122925, "learning_rate": 8.553244334057746e-06, "loss": 1.9736, "step": 133500 }, { "epoch": 41.60198696057125, "grad_norm": 3.6723625659942627, "learning_rate": 8.398013039428749e-06, "loss": 1.9844, "step": 134000 }, { "epoch": 41.75721825520025, "grad_norm": 3.4786245822906494, "learning_rate": 8.242781744799752e-06, "loss": 1.9902, "step": 134500 }, { "epoch": 41.91244954982925, "grad_norm": 3.588338851928711, "learning_rate": 8.087550450170754e-06, "loss": 1.9829, "step": 135000 }, { "epoch": 42.067680844458245, "grad_norm": 3.7299747467041016, "learning_rate": 7.932319155541758e-06, "loss": 1.9803, "step": 135500 }, { "epoch": 42.22291213908724, "grad_norm": 3.1764190196990967, "learning_rate": 7.77708786091276e-06, "loss": 1.968, "step": 136000 }, { "epoch": 42.37814343371624, "grad_norm": 3.7980525493621826, "learning_rate": 7.621856566283764e-06, "loss": 1.9643, "step": 136500 }, { "epoch": 42.53337472834524, "grad_norm": 3.3517909049987793, "learning_rate": 7.466625271654766e-06, "loss": 1.9786, "step": 137000 }, { "epoch": 42.688606022974234, "grad_norm": 3.4372501373291016, "learning_rate": 7.311393977025768e-06, "loss": 1.9649, "step": 137500 }, { "epoch": 42.84383731760323, "grad_norm": 3.6716599464416504, "learning_rate": 7.156162682396772e-06, "loss": 1.9805, "step": 138000 }, { "epoch": 42.99906861223223, "grad_norm": 3.4524483680725098, "learning_rate": 7.000931387767774e-06, "loss": 1.9822, "step": 138500 }, { "epoch": 43.154299906861226, "grad_norm": 3.7548680305480957, "learning_rate": 6.845700093138777e-06, "loss": 1.9648, "step": 139000 }, { "epoch": 43.30953120149022, "grad_norm": 3.3365092277526855, "learning_rate": 6.69046879850978e-06, "loss": 1.9602, "step": 139500 }, { "epoch": 43.46476249611922, "grad_norm": 3.7342677116394043, "learning_rate": 6.535237503880783e-06, "loss": 1.9615, "step": 140000 }, { "epoch": 43.61999379074822, "grad_norm": 3.863659381866455, "learning_rate": 6.380006209251785e-06, "loss": 1.9672, "step": 140500 }, { "epoch": 43.775225085377215, "grad_norm": 3.3871288299560547, "learning_rate": 6.224774914622788e-06, "loss": 1.9663, "step": 141000 }, { "epoch": 43.93045638000621, "grad_norm": 3.8932695388793945, "learning_rate": 6.069543619993791e-06, "loss": 1.9654, "step": 141500 }, { "epoch": 44.08568767463521, "grad_norm": 3.8590352535247803, "learning_rate": 5.914312325364793e-06, "loss": 1.9528, "step": 142000 }, { "epoch": 44.2409189692642, "grad_norm": 3.2460689544677734, "learning_rate": 5.759081030735797e-06, "loss": 1.9557, "step": 142500 }, { "epoch": 44.3961502638932, "grad_norm": 3.6797304153442383, "learning_rate": 5.6038497361067995e-06, "loss": 1.9552, "step": 143000 }, { "epoch": 44.551381558522195, "grad_norm": 3.47025203704834, "learning_rate": 5.448618441477802e-06, "loss": 1.9548, "step": 143500 }, { "epoch": 44.70661285315119, "grad_norm": 3.376739978790283, "learning_rate": 5.293387146848805e-06, "loss": 1.9584, "step": 144000 }, { "epoch": 44.86184414778019, "grad_norm": 3.447455644607544, "learning_rate": 5.138155852219808e-06, "loss": 1.9524, "step": 144500 }, { "epoch": 45.01707544240919, "grad_norm": 3.5013909339904785, "learning_rate": 4.98292455759081e-06, "loss": 1.9537, "step": 145000 }, { "epoch": 45.172306737038184, "grad_norm": 3.857957124710083, "learning_rate": 4.827693262961813e-06, "loss": 1.9446, "step": 145500 }, { "epoch": 45.32753803166718, "grad_norm": 3.5706942081451416, "learning_rate": 4.672461968332816e-06, "loss": 1.9512, "step": 146000 }, { "epoch": 45.48276932629618, "grad_norm": 3.5100932121276855, "learning_rate": 4.5172306737038195e-06, "loss": 1.9573, "step": 146500 }, { "epoch": 45.638000620925176, "grad_norm": 3.7259585857391357, "learning_rate": 4.361999379074821e-06, "loss": 1.9461, "step": 147000 }, { "epoch": 45.79323191555417, "grad_norm": 3.6146180629730225, "learning_rate": 4.206768084445825e-06, "loss": 1.946, "step": 147500 }, { "epoch": 45.94846321018317, "grad_norm": 3.5233962535858154, "learning_rate": 4.051536789816827e-06, "loss": 1.9454, "step": 148000 }, { "epoch": 46.10369450481217, "grad_norm": 3.628469944000244, "learning_rate": 3.89630549518783e-06, "loss": 1.9394, "step": 148500 }, { "epoch": 46.258925799441165, "grad_norm": 3.5431318283081055, "learning_rate": 3.7410742005588326e-06, "loss": 1.9319, "step": 149000 }, { "epoch": 46.41415709407016, "grad_norm": 3.5288734436035156, "learning_rate": 3.5858429059298356e-06, "loss": 1.9436, "step": 149500 }, { "epoch": 46.56938838869916, "grad_norm": 3.5967633724212646, "learning_rate": 3.4306116113008387e-06, "loss": 1.9393, "step": 150000 }, { "epoch": 46.72461968332816, "grad_norm": 3.546693801879883, "learning_rate": 3.275380316671841e-06, "loss": 1.9502, "step": 150500 }, { "epoch": 46.879850977957155, "grad_norm": 3.5913896560668945, "learning_rate": 3.120149022042844e-06, "loss": 1.9514, "step": 151000 }, { "epoch": 47.03508227258615, "grad_norm": 3.8581769466400146, "learning_rate": 2.964917727413847e-06, "loss": 1.9383, "step": 151500 }, { "epoch": 47.19031356721515, "grad_norm": 3.577950954437256, "learning_rate": 2.8096864327848496e-06, "loss": 1.9267, "step": 152000 }, { "epoch": 47.34554486184415, "grad_norm": 3.6761834621429443, "learning_rate": 2.6544551381558526e-06, "loss": 1.9418, "step": 152500 }, { "epoch": 47.500776156473144, "grad_norm": 3.6187174320220947, "learning_rate": 2.4992238435268553e-06, "loss": 1.9362, "step": 153000 }, { "epoch": 47.65600745110214, "grad_norm": 3.6399784088134766, "learning_rate": 2.343992548897858e-06, "loss": 1.9375, "step": 153500 }, { "epoch": 47.81123874573114, "grad_norm": 3.451267957687378, "learning_rate": 2.188761254268861e-06, "loss": 1.9349, "step": 154000 }, { "epoch": 47.966470040360136, "grad_norm": 3.334773540496826, "learning_rate": 2.0335299596398635e-06, "loss": 1.9418, "step": 154500 }, { "epoch": 48.12170133498913, "grad_norm": 3.2175698280334473, "learning_rate": 1.8782986650108664e-06, "loss": 1.9339, "step": 155000 }, { "epoch": 48.27693262961813, "grad_norm": 3.638575315475464, "learning_rate": 1.723067370381869e-06, "loss": 1.9272, "step": 155500 }, { "epoch": 48.43216392424713, "grad_norm": 3.2722606658935547, "learning_rate": 1.567836075752872e-06, "loss": 1.9311, "step": 156000 }, { "epoch": 48.587395218876125, "grad_norm": 3.3455374240875244, "learning_rate": 1.4126047811238747e-06, "loss": 1.9318, "step": 156500 }, { "epoch": 48.74262651350512, "grad_norm": 3.634216785430908, "learning_rate": 1.2573734864948775e-06, "loss": 1.9335, "step": 157000 }, { "epoch": 48.89785780813412, "grad_norm": 3.6753644943237305, "learning_rate": 1.1021421918658803e-06, "loss": 1.9397, "step": 157500 }, { "epoch": 49.05308910276312, "grad_norm": 3.3982911109924316, "learning_rate": 9.469108972368831e-07, "loss": 1.9312, "step": 158000 }, { "epoch": 49.208320397392114, "grad_norm": 3.403062343597412, "learning_rate": 7.916796026078859e-07, "loss": 1.9299, "step": 158500 }, { "epoch": 49.36355169202111, "grad_norm": 3.539356231689453, "learning_rate": 6.364483079788885e-07, "loss": 1.9314, "step": 159000 }, { "epoch": 49.51878298665011, "grad_norm": 3.549177885055542, "learning_rate": 4.812170133498913e-07, "loss": 1.9192, "step": 159500 }, { "epoch": 49.674014281279106, "grad_norm": 3.6044740676879883, "learning_rate": 3.259857187208941e-07, "loss": 1.9283, "step": 160000 }, { "epoch": 49.829245575908104, "grad_norm": 3.24859619140625, "learning_rate": 1.7075442409189693e-07, "loss": 1.9268, "step": 160500 }, { "epoch": 49.9844768705371, "grad_norm": 3.433534860610962, "learning_rate": 1.552312946289972e-08, "loss": 1.9389, "step": 161000 }, { "epoch": 50.0, "step": 161050, "total_flos": 2.8277448966144e+17, "train_loss": 2.6664713644307803, "train_runtime": 39272.3209, "train_samples_per_second": 262.414, "train_steps_per_second": 4.101 } ], "logging_steps": 500, "max_steps": 161050, "num_input_tokens_seen": 0, "num_train_epochs": 50, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.8277448966144e+17, "train_batch_size": 64, "trial_name": null, "trial_params": null }