| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 5445, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.018369690011481057, | |
| "grad_norm": 0.8683544993400574, | |
| "learning_rate": 9.5e-05, | |
| "loss": 5.5925, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03673938002296211, | |
| "grad_norm": 0.7329673171043396, | |
| "learning_rate": 9.970832054037457e-05, | |
| "loss": 5.4275, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05510907003444317, | |
| "grad_norm": 0.658450722694397, | |
| "learning_rate": 9.940128953024256e-05, | |
| "loss": 5.3361, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07347876004592423, | |
| "grad_norm": 0.6886904239654541, | |
| "learning_rate": 9.909425852011053e-05, | |
| "loss": 5.2746, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09184845005740529, | |
| "grad_norm": 0.908559262752533, | |
| "learning_rate": 9.878722750997851e-05, | |
| "loss": 5.178, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11021814006888633, | |
| "grad_norm": 0.613741397857666, | |
| "learning_rate": 9.848019649984648e-05, | |
| "loss": 5.3301, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1285878300803674, | |
| "grad_norm": 0.8728771805763245, | |
| "learning_rate": 9.817316548971447e-05, | |
| "loss": 5.2269, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14695752009184845, | |
| "grad_norm": 0.6336873769760132, | |
| "learning_rate": 9.786613447958243e-05, | |
| "loss": 5.2032, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1653272101033295, | |
| "grad_norm": 0.7317622303962708, | |
| "learning_rate": 9.755910346945042e-05, | |
| "loss": 5.1694, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.18369690011481057, | |
| "grad_norm": 0.624251663684845, | |
| "learning_rate": 9.72520724593184e-05, | |
| "loss": 5.15, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2020665901262916, | |
| "grad_norm": 0.6709743142127991, | |
| "learning_rate": 9.694504144918637e-05, | |
| "loss": 5.233, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22043628013777267, | |
| "grad_norm": 0.6651622653007507, | |
| "learning_rate": 9.663801043905434e-05, | |
| "loss": 5.233, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.23880597014925373, | |
| "grad_norm": 0.6254720091819763, | |
| "learning_rate": 9.633097942892233e-05, | |
| "loss": 5.2223, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2571756601607348, | |
| "grad_norm": 0.6242057681083679, | |
| "learning_rate": 9.602394841879029e-05, | |
| "loss": 5.1399, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2755453501722158, | |
| "grad_norm": 0.6652698516845703, | |
| "learning_rate": 9.571691740865828e-05, | |
| "loss": 5.2068, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2939150401836969, | |
| "grad_norm": 0.8067322373390198, | |
| "learning_rate": 9.540988639852626e-05, | |
| "loss": 5.1546, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.31228473019517794, | |
| "grad_norm": 0.8131216764450073, | |
| "learning_rate": 9.510285538839423e-05, | |
| "loss": 5.2022, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.330654420206659, | |
| "grad_norm": 0.7776964902877808, | |
| "learning_rate": 9.47958243782622e-05, | |
| "loss": 5.2079, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.34902411021814006, | |
| "grad_norm": 0.7237767577171326, | |
| "learning_rate": 9.448879336813019e-05, | |
| "loss": 5.1812, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.36739380022962115, | |
| "grad_norm": 0.8634713292121887, | |
| "learning_rate": 9.418176235799817e-05, | |
| "loss": 5.1857, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3857634902411022, | |
| "grad_norm": 0.6349618434906006, | |
| "learning_rate": 9.387473134786614e-05, | |
| "loss": 5.1357, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4041331802525832, | |
| "grad_norm": 0.6394773721694946, | |
| "learning_rate": 9.356770033773412e-05, | |
| "loss": 5.1004, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4225028702640643, | |
| "grad_norm": 0.7400641441345215, | |
| "learning_rate": 9.326066932760209e-05, | |
| "loss": 5.1123, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.44087256027554533, | |
| "grad_norm": 0.6799914836883545, | |
| "learning_rate": 9.295363831747008e-05, | |
| "loss": 5.1645, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4592422502870264, | |
| "grad_norm": 0.7447289824485779, | |
| "learning_rate": 9.264660730733804e-05, | |
| "loss": 5.0634, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.47761194029850745, | |
| "grad_norm": 0.6091820597648621, | |
| "learning_rate": 9.233957629720603e-05, | |
| "loss": 5.1632, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.49598163030998854, | |
| "grad_norm": 0.7650216221809387, | |
| "learning_rate": 9.2032545287074e-05, | |
| "loss": 5.1083, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5143513203214696, | |
| "grad_norm": 0.6420841813087463, | |
| "learning_rate": 9.172551427694198e-05, | |
| "loss": 5.1332, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5327210103329506, | |
| "grad_norm": 0.6499956846237183, | |
| "learning_rate": 9.141848326680995e-05, | |
| "loss": 5.1091, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5510907003444316, | |
| "grad_norm": 0.6708125472068787, | |
| "learning_rate": 9.111145225667794e-05, | |
| "loss": 5.1761, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5694603903559128, | |
| "grad_norm": 0.6679044961929321, | |
| "learning_rate": 9.08044212465459e-05, | |
| "loss": 5.1747, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5878300803673938, | |
| "grad_norm": 0.6638289093971252, | |
| "learning_rate": 9.049739023641389e-05, | |
| "loss": 5.2075, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6061997703788748, | |
| "grad_norm": 0.6478453278541565, | |
| "learning_rate": 9.019035922628186e-05, | |
| "loss": 5.0887, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6245694603903559, | |
| "grad_norm": 0.7284302711486816, | |
| "learning_rate": 8.988332821614984e-05, | |
| "loss": 5.031, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.642939150401837, | |
| "grad_norm": 0.635795533657074, | |
| "learning_rate": 8.957629720601781e-05, | |
| "loss": 5.1412, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.661308840413318, | |
| "grad_norm": 0.6595786213874817, | |
| "learning_rate": 8.92692661958858e-05, | |
| "loss": 5.1112, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6796785304247991, | |
| "grad_norm": 0.8448815941810608, | |
| "learning_rate": 8.896223518575376e-05, | |
| "loss": 5.1198, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6980482204362801, | |
| "grad_norm": 0.7050981521606445, | |
| "learning_rate": 8.865520417562175e-05, | |
| "loss": 5.1194, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7164179104477612, | |
| "grad_norm": 0.618584394454956, | |
| "learning_rate": 8.834817316548972e-05, | |
| "loss": 5.0858, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7347876004592423, | |
| "grad_norm": 0.711953341960907, | |
| "learning_rate": 8.80411421553577e-05, | |
| "loss": 5.0608, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7531572904707233, | |
| "grad_norm": 0.6575554609298706, | |
| "learning_rate": 8.773411114522567e-05, | |
| "loss": 5.122, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7715269804822044, | |
| "grad_norm": 0.6340029835700989, | |
| "learning_rate": 8.742708013509366e-05, | |
| "loss": 5.1082, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7898966704936854, | |
| "grad_norm": 0.6645310521125793, | |
| "learning_rate": 8.712004912496162e-05, | |
| "loss": 5.0621, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8082663605051664, | |
| "grad_norm": 0.7187919020652771, | |
| "learning_rate": 8.681301811482961e-05, | |
| "loss": 5.0188, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.8266360505166476, | |
| "grad_norm": 0.591472327709198, | |
| "learning_rate": 8.650598710469758e-05, | |
| "loss": 5.0586, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8450057405281286, | |
| "grad_norm": 0.7377704381942749, | |
| "learning_rate": 8.619895609456556e-05, | |
| "loss": 5.0832, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8633754305396096, | |
| "grad_norm": 0.7042536735534668, | |
| "learning_rate": 8.589192508443353e-05, | |
| "loss": 5.0497, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8817451205510907, | |
| "grad_norm": 0.630856990814209, | |
| "learning_rate": 8.558489407430152e-05, | |
| "loss": 5.1131, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9001148105625718, | |
| "grad_norm": 0.7756277322769165, | |
| "learning_rate": 8.527786306416948e-05, | |
| "loss": 5.1407, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9184845005740528, | |
| "grad_norm": 0.6865280866622925, | |
| "learning_rate": 8.497083205403747e-05, | |
| "loss": 5.0728, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9368541905855339, | |
| "grad_norm": 0.84028160572052, | |
| "learning_rate": 8.466380104390543e-05, | |
| "loss": 5.0406, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9552238805970149, | |
| "grad_norm": 0.700905442237854, | |
| "learning_rate": 8.435677003377342e-05, | |
| "loss": 5.0948, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.9735935706084959, | |
| "grad_norm": 0.8270923495292664, | |
| "learning_rate": 8.404973902364139e-05, | |
| "loss": 5.1278, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9919632606199771, | |
| "grad_norm": 0.7246171236038208, | |
| "learning_rate": 8.374270801350936e-05, | |
| "loss": 5.177, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 4.98171854019165, | |
| "eval_runtime": 159.7626, | |
| "eval_samples_per_second": 6.059, | |
| "eval_steps_per_second": 3.029, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 1.0101033295063147, | |
| "grad_norm": 0.7068125605583191, | |
| "learning_rate": 8.343567700337734e-05, | |
| "loss": 4.9922, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.0284730195177956, | |
| "grad_norm": 0.6738723516464233, | |
| "learning_rate": 8.312864599324533e-05, | |
| "loss": 4.9379, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.0468427095292767, | |
| "grad_norm": 0.7351597547531128, | |
| "learning_rate": 8.282161498311329e-05, | |
| "loss": 4.8823, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.0652123995407576, | |
| "grad_norm": 0.7661564350128174, | |
| "learning_rate": 8.251458397298128e-05, | |
| "loss": 4.8654, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.0835820895522388, | |
| "grad_norm": 0.7068904638290405, | |
| "learning_rate": 8.220755296284925e-05, | |
| "loss": 4.8782, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.10195177956372, | |
| "grad_norm": 0.8419404029846191, | |
| "learning_rate": 8.190052195271722e-05, | |
| "loss": 4.8636, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.1203214695752008, | |
| "grad_norm": 0.9159519076347351, | |
| "learning_rate": 8.15934909425852e-05, | |
| "loss": 4.8858, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.138691159586682, | |
| "grad_norm": 0.743350088596344, | |
| "learning_rate": 8.128645993245319e-05, | |
| "loss": 4.8809, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.157060849598163, | |
| "grad_norm": 0.737196683883667, | |
| "learning_rate": 8.097942892232115e-05, | |
| "loss": 4.8634, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.175430539609644, | |
| "grad_norm": 0.7130833864212036, | |
| "learning_rate": 8.067239791218914e-05, | |
| "loss": 4.9034, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.1938002296211252, | |
| "grad_norm": 0.8160459995269775, | |
| "learning_rate": 8.036536690205711e-05, | |
| "loss": 4.8548, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.2121699196326061, | |
| "grad_norm": 0.6838780641555786, | |
| "learning_rate": 8.005833589192508e-05, | |
| "loss": 4.9454, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.2305396096440873, | |
| "grad_norm": 0.85508131980896, | |
| "learning_rate": 7.975130488179306e-05, | |
| "loss": 4.9991, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.2489092996555684, | |
| "grad_norm": 0.7988117933273315, | |
| "learning_rate": 7.944427387166105e-05, | |
| "loss": 4.8513, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.2672789896670493, | |
| "grad_norm": 0.7242785096168518, | |
| "learning_rate": 7.913724286152902e-05, | |
| "loss": 4.9535, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.2856486796785305, | |
| "grad_norm": 0.8718686699867249, | |
| "learning_rate": 7.8830211851397e-05, | |
| "loss": 4.8674, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.3040183696900116, | |
| "grad_norm": 0.6738030910491943, | |
| "learning_rate": 7.852318084126497e-05, | |
| "loss": 4.9205, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.3223880597014925, | |
| "grad_norm": 0.8262588381767273, | |
| "learning_rate": 7.821614983113295e-05, | |
| "loss": 4.9733, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.3407577497129735, | |
| "grad_norm": 0.807981014251709, | |
| "learning_rate": 7.790911882100092e-05, | |
| "loss": 4.9438, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.3591274397244546, | |
| "grad_norm": 0.7458767890930176, | |
| "learning_rate": 7.760208781086891e-05, | |
| "loss": 4.8945, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.3774971297359357, | |
| "grad_norm": 0.747597873210907, | |
| "learning_rate": 7.729505680073688e-05, | |
| "loss": 4.8947, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.3958668197474169, | |
| "grad_norm": 0.724322497844696, | |
| "learning_rate": 7.698802579060486e-05, | |
| "loss": 4.9132, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.4142365097588978, | |
| "grad_norm": 0.7647684812545776, | |
| "learning_rate": 7.668099478047283e-05, | |
| "loss": 4.9085, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.432606199770379, | |
| "grad_norm": 0.7980235815048218, | |
| "learning_rate": 7.63739637703408e-05, | |
| "loss": 4.8458, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.4509758897818599, | |
| "grad_norm": 0.9437423944473267, | |
| "learning_rate": 7.606693276020879e-05, | |
| "loss": 4.8771, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.469345579793341, | |
| "grad_norm": 0.7015650272369385, | |
| "learning_rate": 7.575990175007675e-05, | |
| "loss": 4.9099, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.4877152698048222, | |
| "grad_norm": 0.8654283285140991, | |
| "learning_rate": 7.545287073994474e-05, | |
| "loss": 4.8848, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.506084959816303, | |
| "grad_norm": 0.8487702012062073, | |
| "learning_rate": 7.514583972981272e-05, | |
| "loss": 4.9274, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.524454649827784, | |
| "grad_norm": 0.894727349281311, | |
| "learning_rate": 7.483880871968069e-05, | |
| "loss": 4.9121, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.5428243398392651, | |
| "grad_norm": 0.7763797640800476, | |
| "learning_rate": 7.453177770954867e-05, | |
| "loss": 4.9296, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.5611940298507463, | |
| "grad_norm": 0.7502116560935974, | |
| "learning_rate": 7.422474669941665e-05, | |
| "loss": 4.9129, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.5795637198622274, | |
| "grad_norm": 0.9286855459213257, | |
| "learning_rate": 7.391771568928461e-05, | |
| "loss": 4.8938, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.5979334098737084, | |
| "grad_norm": 0.8818989396095276, | |
| "learning_rate": 7.36106846791526e-05, | |
| "loss": 4.8487, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.6163030998851893, | |
| "grad_norm": 0.8271352648735046, | |
| "learning_rate": 7.330365366902058e-05, | |
| "loss": 4.9022, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.6346727898966704, | |
| "grad_norm": 0.7982690334320068, | |
| "learning_rate": 7.299662265888855e-05, | |
| "loss": 4.8769, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.6530424799081516, | |
| "grad_norm": 0.8425403833389282, | |
| "learning_rate": 7.268959164875653e-05, | |
| "loss": 4.9173, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.6714121699196327, | |
| "grad_norm": 0.7963587045669556, | |
| "learning_rate": 7.238256063862451e-05, | |
| "loss": 4.849, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.6897818599311136, | |
| "grad_norm": 0.8965281248092651, | |
| "learning_rate": 7.207552962849247e-05, | |
| "loss": 4.9365, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.7081515499425948, | |
| "grad_norm": 0.7666225433349609, | |
| "learning_rate": 7.176849861836046e-05, | |
| "loss": 4.828, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.7265212399540757, | |
| "grad_norm": 0.752511203289032, | |
| "learning_rate": 7.146146760822844e-05, | |
| "loss": 4.8823, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.7448909299655568, | |
| "grad_norm": 0.9949547052383423, | |
| "learning_rate": 7.115443659809641e-05, | |
| "loss": 4.9664, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.763260619977038, | |
| "grad_norm": 0.8455734848976135, | |
| "learning_rate": 7.084740558796439e-05, | |
| "loss": 4.9238, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.781630309988519, | |
| "grad_norm": 0.8804666996002197, | |
| "learning_rate": 7.054037457783237e-05, | |
| "loss": 4.9111, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.8722001910209656, | |
| "learning_rate": 7.023334356770033e-05, | |
| "loss": 4.8461, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.818369690011481, | |
| "grad_norm": 0.9130426049232483, | |
| "learning_rate": 6.992631255756832e-05, | |
| "loss": 4.8927, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.836739380022962, | |
| "grad_norm": 0.8905944228172302, | |
| "learning_rate": 6.96192815474363e-05, | |
| "loss": 4.9238, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.8551090700344433, | |
| "grad_norm": 0.8010729551315308, | |
| "learning_rate": 6.931225053730427e-05, | |
| "loss": 4.8703, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.8734787600459242, | |
| "grad_norm": 0.7367790937423706, | |
| "learning_rate": 6.900521952717225e-05, | |
| "loss": 4.8673, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.8918484500574053, | |
| "grad_norm": 0.7457343339920044, | |
| "learning_rate": 6.869818851704023e-05, | |
| "loss": 4.8702, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.9102181400688862, | |
| "grad_norm": 0.9334909319877625, | |
| "learning_rate": 6.83911575069082e-05, | |
| "loss": 4.973, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.9285878300803674, | |
| "grad_norm": 0.7645395994186401, | |
| "learning_rate": 6.808412649677618e-05, | |
| "loss": 4.8706, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.9469575200918485, | |
| "grad_norm": 0.8864799737930298, | |
| "learning_rate": 6.777709548664414e-05, | |
| "loss": 4.9114, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.9653272101033297, | |
| "grad_norm": 0.995495617389679, | |
| "learning_rate": 6.747006447651213e-05, | |
| "loss": 4.9208, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.9836969001148106, | |
| "grad_norm": 0.7982215881347656, | |
| "learning_rate": 6.71630334663801e-05, | |
| "loss": 4.8581, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 4.952657699584961, | |
| "eval_runtime": 161.3919, | |
| "eval_samples_per_second": 5.998, | |
| "eval_steps_per_second": 2.999, | |
| "step": 2178 | |
| }, | |
| { | |
| "epoch": 2.001836969001148, | |
| "grad_norm": 0.6282178163528442, | |
| "learning_rate": 6.685600245624808e-05, | |
| "loss": 4.8444, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 2.0202066590126293, | |
| "grad_norm": 0.894764244556427, | |
| "learning_rate": 6.654897144611605e-05, | |
| "loss": 4.6306, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.03857634902411, | |
| "grad_norm": 0.9108575582504272, | |
| "learning_rate": 6.624194043598404e-05, | |
| "loss": 4.6439, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 2.056946039035591, | |
| "grad_norm": 0.8883677124977112, | |
| "learning_rate": 6.5934909425852e-05, | |
| "loss": 4.6667, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 2.0753157290470723, | |
| "grad_norm": 0.8742470741271973, | |
| "learning_rate": 6.562787841571999e-05, | |
| "loss": 4.6403, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 2.0936854190585534, | |
| "grad_norm": 0.883115291595459, | |
| "learning_rate": 6.532084740558797e-05, | |
| "loss": 4.7058, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 2.1120551090700346, | |
| "grad_norm": 0.9962688088417053, | |
| "learning_rate": 6.501381639545594e-05, | |
| "loss": 4.6428, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.1304247990815153, | |
| "grad_norm": 0.8967293500900269, | |
| "learning_rate": 6.470678538532391e-05, | |
| "loss": 4.7033, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 2.1487944890929964, | |
| "grad_norm": 0.8904005289077759, | |
| "learning_rate": 6.43997543751919e-05, | |
| "loss": 4.6883, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 2.1671641791044776, | |
| "grad_norm": 0.9912065267562866, | |
| "learning_rate": 6.409272336505986e-05, | |
| "loss": 4.6034, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 2.1855338691159587, | |
| "grad_norm": 0.7966124415397644, | |
| "learning_rate": 6.378569235492785e-05, | |
| "loss": 4.5939, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.20390355912744, | |
| "grad_norm": 0.8692038059234619, | |
| "learning_rate": 6.347866134479583e-05, | |
| "loss": 4.635, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.2222732491389205, | |
| "grad_norm": 0.9523972868919373, | |
| "learning_rate": 6.31716303346638e-05, | |
| "loss": 4.5906, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.2406429391504017, | |
| "grad_norm": 1.0009050369262695, | |
| "learning_rate": 6.286459932453177e-05, | |
| "loss": 4.622, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.259012629161883, | |
| "grad_norm": 0.9459270238876343, | |
| "learning_rate": 6.255756831439976e-05, | |
| "loss": 4.6509, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.277382319173364, | |
| "grad_norm": 0.8043718934059143, | |
| "learning_rate": 6.225053730426774e-05, | |
| "loss": 4.6445, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.295752009184845, | |
| "grad_norm": 0.916283130645752, | |
| "learning_rate": 6.194350629413571e-05, | |
| "loss": 4.6549, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.314121699196326, | |
| "grad_norm": 0.9356532096862793, | |
| "learning_rate": 6.163647528400369e-05, | |
| "loss": 4.6616, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 2.332491389207807, | |
| "grad_norm": 0.9441368579864502, | |
| "learning_rate": 6.132944427387166e-05, | |
| "loss": 4.6529, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 2.350861079219288, | |
| "grad_norm": 0.8593648672103882, | |
| "learning_rate": 6.102241326373964e-05, | |
| "loss": 4.6266, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 2.3692307692307693, | |
| "grad_norm": 1.0274901390075684, | |
| "learning_rate": 6.071538225360762e-05, | |
| "loss": 4.6506, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 2.3876004592422504, | |
| "grad_norm": 0.8973293900489807, | |
| "learning_rate": 6.040835124347559e-05, | |
| "loss": 4.6335, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.405970149253731, | |
| "grad_norm": 0.7993806600570679, | |
| "learning_rate": 6.010132023334357e-05, | |
| "loss": 4.6048, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 2.4243398392652122, | |
| "grad_norm": 1.0597416162490845, | |
| "learning_rate": 5.979428922321155e-05, | |
| "loss": 4.6689, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.4427095292766934, | |
| "grad_norm": 0.8965485095977783, | |
| "learning_rate": 5.948725821307952e-05, | |
| "loss": 4.7001, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 2.4610792192881745, | |
| "grad_norm": 1.2203742265701294, | |
| "learning_rate": 5.91802272029475e-05, | |
| "loss": 4.6136, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.4794489092996557, | |
| "grad_norm": 0.8789367079734802, | |
| "learning_rate": 5.887319619281547e-05, | |
| "loss": 4.6247, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.497818599311137, | |
| "grad_norm": 0.997754693031311, | |
| "learning_rate": 5.856616518268345e-05, | |
| "loss": 4.6653, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.516188289322618, | |
| "grad_norm": 1.009928822517395, | |
| "learning_rate": 5.825913417255143e-05, | |
| "loss": 4.6168, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.5345579793340987, | |
| "grad_norm": 0.919626772403717, | |
| "learning_rate": 5.79521031624194e-05, | |
| "loss": 4.6677, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.55292766934558, | |
| "grad_norm": 0.8129732012748718, | |
| "learning_rate": 5.764507215228738e-05, | |
| "loss": 4.6061, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.571297359357061, | |
| "grad_norm": 0.8821771144866943, | |
| "learning_rate": 5.733804114215536e-05, | |
| "loss": 4.5725, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.5896670493685416, | |
| "grad_norm": 1.150881052017212, | |
| "learning_rate": 5.703101013202333e-05, | |
| "loss": 4.6755, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.6080367393800232, | |
| "grad_norm": 1.1142388582229614, | |
| "learning_rate": 5.672397912189131e-05, | |
| "loss": 4.6339, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.626406429391504, | |
| "grad_norm": 1.0733916759490967, | |
| "learning_rate": 5.641694811175929e-05, | |
| "loss": 4.6091, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.644776119402985, | |
| "grad_norm": 0.9986577033996582, | |
| "learning_rate": 5.610991710162726e-05, | |
| "loss": 4.6656, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.663145809414466, | |
| "grad_norm": 1.7999745607376099, | |
| "learning_rate": 5.580288609149524e-05, | |
| "loss": 4.7289, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.681515499425947, | |
| "grad_norm": 0.8412841558456421, | |
| "learning_rate": 5.549585508136322e-05, | |
| "loss": 4.6984, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 2.6998851894374285, | |
| "grad_norm": 0.8901891708374023, | |
| "learning_rate": 5.5188824071231197e-05, | |
| "loss": 4.6874, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.718254879448909, | |
| "grad_norm": 0.946198046207428, | |
| "learning_rate": 5.488179306109917e-05, | |
| "loss": 4.669, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 2.7366245694603903, | |
| "grad_norm": 0.9087753891944885, | |
| "learning_rate": 5.457476205096715e-05, | |
| "loss": 4.6541, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 2.7549942594718715, | |
| "grad_norm": 1.0043200254440308, | |
| "learning_rate": 5.426773104083513e-05, | |
| "loss": 4.6556, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.7733639494833526, | |
| "grad_norm": 0.968223512172699, | |
| "learning_rate": 5.39607000307031e-05, | |
| "loss": 4.7655, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 2.7917336394948338, | |
| "grad_norm": 1.2286486625671387, | |
| "learning_rate": 5.365366902057108e-05, | |
| "loss": 4.6863, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 2.8101033295063145, | |
| "grad_norm": 1.153100609779358, | |
| "learning_rate": 5.334663801043906e-05, | |
| "loss": 4.6555, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 2.8284730195177956, | |
| "grad_norm": 1.01992666721344, | |
| "learning_rate": 5.303960700030703e-05, | |
| "loss": 4.6923, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 2.8468427095292768, | |
| "grad_norm": 1.1643304824829102, | |
| "learning_rate": 5.273257599017501e-05, | |
| "loss": 4.6761, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.865212399540758, | |
| "grad_norm": 0.9204754829406738, | |
| "learning_rate": 5.242554498004299e-05, | |
| "loss": 4.6726, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 2.883582089552239, | |
| "grad_norm": 1.0050371885299683, | |
| "learning_rate": 5.211851396991097e-05, | |
| "loss": 4.7501, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 2.9019517795637197, | |
| "grad_norm": 0.9284995794296265, | |
| "learning_rate": 5.181148295977894e-05, | |
| "loss": 4.6609, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 2.920321469575201, | |
| "grad_norm": 0.9417170286178589, | |
| "learning_rate": 5.150445194964692e-05, | |
| "loss": 4.6667, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 2.938691159586682, | |
| "grad_norm": 1.4755141735076904, | |
| "learning_rate": 5.11974209395149e-05, | |
| "loss": 4.6151, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.957060849598163, | |
| "grad_norm": 0.9110747575759888, | |
| "learning_rate": 5.089038992938287e-05, | |
| "loss": 4.6703, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 2.9754305396096443, | |
| "grad_norm": 1.295502781867981, | |
| "learning_rate": 5.058335891925085e-05, | |
| "loss": 4.7312, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 2.993800229621125, | |
| "grad_norm": 1.072573184967041, | |
| "learning_rate": 5.027632790911883e-05, | |
| "loss": 4.6101, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 5.012691974639893, | |
| "eval_runtime": 161.4677, | |
| "eval_samples_per_second": 5.995, | |
| "eval_steps_per_second": 2.998, | |
| "step": 3267 | |
| }, | |
| { | |
| "epoch": 3.011940298507463, | |
| "grad_norm": 0.8994636535644531, | |
| "learning_rate": 4.99692968989868e-05, | |
| "loss": 4.4452, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 3.030309988518944, | |
| "grad_norm": 1.228932499885559, | |
| "learning_rate": 4.966226588885478e-05, | |
| "loss": 4.4106, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.0486796785304247, | |
| "grad_norm": 1.0340349674224854, | |
| "learning_rate": 4.935523487872276e-05, | |
| "loss": 4.4157, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 3.067049368541906, | |
| "grad_norm": 1.091071367263794, | |
| "learning_rate": 4.904820386859073e-05, | |
| "loss": 4.3893, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 3.085419058553387, | |
| "grad_norm": 0.974187433719635, | |
| "learning_rate": 4.874117285845871e-05, | |
| "loss": 4.4317, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 3.103788748564868, | |
| "grad_norm": 1.13187837600708, | |
| "learning_rate": 4.843414184832668e-05, | |
| "loss": 4.4274, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 3.1221584385763492, | |
| "grad_norm": 0.9200457334518433, | |
| "learning_rate": 4.812711083819466e-05, | |
| "loss": 4.3789, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.14052812858783, | |
| "grad_norm": 1.0937250852584839, | |
| "learning_rate": 4.782007982806264e-05, | |
| "loss": 4.3659, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 3.158897818599311, | |
| "grad_norm": 1.0695732831954956, | |
| "learning_rate": 4.751304881793061e-05, | |
| "loss": 4.417, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 3.177267508610792, | |
| "grad_norm": 1.0328212976455688, | |
| "learning_rate": 4.720601780779859e-05, | |
| "loss": 4.3615, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 3.1956371986222734, | |
| "grad_norm": 0.9559247493743896, | |
| "learning_rate": 4.689898679766657e-05, | |
| "loss": 4.4404, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 3.2140068886337545, | |
| "grad_norm": 1.1086647510528564, | |
| "learning_rate": 4.659195578753454e-05, | |
| "loss": 4.3502, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.232376578645235, | |
| "grad_norm": 1.0105242729187012, | |
| "learning_rate": 4.628492477740252e-05, | |
| "loss": 4.3852, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 3.2507462686567163, | |
| "grad_norm": 1.1470004320144653, | |
| "learning_rate": 4.59778937672705e-05, | |
| "loss": 4.4163, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 3.2691159586681975, | |
| "grad_norm": 1.0169470310211182, | |
| "learning_rate": 4.567086275713847e-05, | |
| "loss": 4.4186, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 3.2874856486796786, | |
| "grad_norm": 1.2762373685836792, | |
| "learning_rate": 4.536383174700645e-05, | |
| "loss": 4.4134, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 3.3058553386911598, | |
| "grad_norm": 1.4351933002471924, | |
| "learning_rate": 4.505680073687443e-05, | |
| "loss": 4.3983, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.3242250287026405, | |
| "grad_norm": 1.1447066068649292, | |
| "learning_rate": 4.47497697267424e-05, | |
| "loss": 4.3949, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 3.3425947187141216, | |
| "grad_norm": 1.0704272985458374, | |
| "learning_rate": 4.4442738716610376e-05, | |
| "loss": 4.409, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 3.3609644087256028, | |
| "grad_norm": 1.346151351928711, | |
| "learning_rate": 4.413570770647836e-05, | |
| "loss": 4.4447, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 3.379334098737084, | |
| "grad_norm": 1.2971051931381226, | |
| "learning_rate": 4.382867669634633e-05, | |
| "loss": 4.4367, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 3.397703788748565, | |
| "grad_norm": 0.9847575426101685, | |
| "learning_rate": 4.3521645686214306e-05, | |
| "loss": 4.4296, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.4160734787600457, | |
| "grad_norm": 1.247763752937317, | |
| "learning_rate": 4.321461467608229e-05, | |
| "loss": 4.3508, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 3.434443168771527, | |
| "grad_norm": 1.2263673543930054, | |
| "learning_rate": 4.290758366595026e-05, | |
| "loss": 4.4014, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 3.452812858783008, | |
| "grad_norm": 0.925714910030365, | |
| "learning_rate": 4.2600552655818236e-05, | |
| "loss": 4.3741, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 3.471182548794489, | |
| "grad_norm": 1.2875808477401733, | |
| "learning_rate": 4.229352164568622e-05, | |
| "loss": 4.4081, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 3.4895522388059703, | |
| "grad_norm": 1.1704316139221191, | |
| "learning_rate": 4.198649063555419e-05, | |
| "loss": 4.4046, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 3.507921928817451, | |
| "grad_norm": 1.2848256826400757, | |
| "learning_rate": 4.1679459625422166e-05, | |
| "loss": 4.4377, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 3.526291618828932, | |
| "grad_norm": 1.0369383096694946, | |
| "learning_rate": 4.137242861529015e-05, | |
| "loss": 4.4256, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 3.5446613088404133, | |
| "grad_norm": 1.3157178163528442, | |
| "learning_rate": 4.106539760515812e-05, | |
| "loss": 4.4593, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 3.5630309988518944, | |
| "grad_norm": 1.056178331375122, | |
| "learning_rate": 4.0758366595026096e-05, | |
| "loss": 4.3933, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 3.5814006888633756, | |
| "grad_norm": 1.218313217163086, | |
| "learning_rate": 4.045133558489408e-05, | |
| "loss": 4.3811, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 3.5997703788748563, | |
| "grad_norm": 1.447138786315918, | |
| "learning_rate": 4.014430457476205e-05, | |
| "loss": 4.3999, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 3.6181400688863374, | |
| "grad_norm": 1.0364395380020142, | |
| "learning_rate": 3.9837273564630026e-05, | |
| "loss": 4.3874, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 3.6365097588978186, | |
| "grad_norm": 1.1889121532440186, | |
| "learning_rate": 3.9530242554498e-05, | |
| "loss": 4.4304, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 3.6548794489092997, | |
| "grad_norm": 1.056187629699707, | |
| "learning_rate": 3.922321154436598e-05, | |
| "loss": 4.3609, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 3.673249138920781, | |
| "grad_norm": 1.1961441040039062, | |
| "learning_rate": 3.8916180534233957e-05, | |
| "loss": 4.4391, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.6916188289322616, | |
| "grad_norm": 1.176295518875122, | |
| "learning_rate": 3.860914952410193e-05, | |
| "loss": 4.478, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 3.7099885189437427, | |
| "grad_norm": 1.1515350341796875, | |
| "learning_rate": 3.830211851396991e-05, | |
| "loss": 4.3721, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 3.728358208955224, | |
| "grad_norm": 1.13578200340271, | |
| "learning_rate": 3.7995087503837887e-05, | |
| "loss": 4.4138, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 3.746727898966705, | |
| "grad_norm": 1.1046961545944214, | |
| "learning_rate": 3.768805649370587e-05, | |
| "loss": 4.4361, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 3.765097588978186, | |
| "grad_norm": 1.043782353401184, | |
| "learning_rate": 3.738102548357384e-05, | |
| "loss": 4.4268, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 3.783467278989667, | |
| "grad_norm": 1.1587445735931396, | |
| "learning_rate": 3.707399447344182e-05, | |
| "loss": 4.4048, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 3.801836969001148, | |
| "grad_norm": 1.2178658246994019, | |
| "learning_rate": 3.67669634633098e-05, | |
| "loss": 4.426, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 3.820206659012629, | |
| "grad_norm": 1.1141862869262695, | |
| "learning_rate": 3.645993245317777e-05, | |
| "loss": 4.4415, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 3.8385763490241103, | |
| "grad_norm": 1.062812089920044, | |
| "learning_rate": 3.6152901443045753e-05, | |
| "loss": 4.4025, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 3.8569460390355914, | |
| "grad_norm": 1.1655863523483276, | |
| "learning_rate": 3.584587043291373e-05, | |
| "loss": 4.4325, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.875315729047072, | |
| "grad_norm": 1.1558030843734741, | |
| "learning_rate": 3.55388394227817e-05, | |
| "loss": 4.4379, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 3.8936854190585533, | |
| "grad_norm": 1.0342259407043457, | |
| "learning_rate": 3.5231808412649684e-05, | |
| "loss": 4.4413, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 3.9120551090700344, | |
| "grad_norm": 0.9829803109169006, | |
| "learning_rate": 3.492477740251766e-05, | |
| "loss": 4.4537, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 3.9304247990815155, | |
| "grad_norm": 1.0746711492538452, | |
| "learning_rate": 3.461774639238563e-05, | |
| "loss": 4.3644, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 3.9487944890929967, | |
| "grad_norm": 1.125643253326416, | |
| "learning_rate": 3.4310715382253614e-05, | |
| "loss": 4.3606, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 3.9671641791044774, | |
| "grad_norm": 1.1107122898101807, | |
| "learning_rate": 3.400368437212159e-05, | |
| "loss": 4.3981, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 3.9855338691159585, | |
| "grad_norm": 1.262173056602478, | |
| "learning_rate": 3.369665336198956e-05, | |
| "loss": 4.4326, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 5.145928382873535, | |
| "eval_runtime": 159.4863, | |
| "eval_samples_per_second": 6.069, | |
| "eval_steps_per_second": 3.035, | |
| "step": 4356 | |
| }, | |
| { | |
| "epoch": 4.003673938002296, | |
| "grad_norm": 0.9681905508041382, | |
| "learning_rate": 3.3389622351857544e-05, | |
| "loss": 4.3781, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 4.022043628013777, | |
| "grad_norm": 1.497965693473816, | |
| "learning_rate": 3.308259134172552e-05, | |
| "loss": 4.2113, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 4.040413318025259, | |
| "grad_norm": 0.9628502726554871, | |
| "learning_rate": 3.277556033159349e-05, | |
| "loss": 4.2562, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.058783008036739, | |
| "grad_norm": 1.0885639190673828, | |
| "learning_rate": 3.2468529321461474e-05, | |
| "loss": 4.2, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 4.07715269804822, | |
| "grad_norm": 0.9081327319145203, | |
| "learning_rate": 3.216149831132945e-05, | |
| "loss": 4.1522, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 4.095522388059702, | |
| "grad_norm": 0.95115065574646, | |
| "learning_rate": 3.185446730119742e-05, | |
| "loss": 4.2762, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 4.113892078071182, | |
| "grad_norm": 1.2338329553604126, | |
| "learning_rate": 3.15474362910654e-05, | |
| "loss": 4.2049, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 4.132261768082664, | |
| "grad_norm": 1.256455421447754, | |
| "learning_rate": 3.124040528093338e-05, | |
| "loss": 4.2357, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.150631458094145, | |
| "grad_norm": 1.296912431716919, | |
| "learning_rate": 3.093337427080135e-05, | |
| "loss": 4.2239, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 4.169001148105625, | |
| "grad_norm": 1.1014831066131592, | |
| "learning_rate": 3.062634326066933e-05, | |
| "loss": 4.191, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 4.187370838117107, | |
| "grad_norm": 1.0076754093170166, | |
| "learning_rate": 3.0319312250537308e-05, | |
| "loss": 4.2087, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 4.205740528128588, | |
| "grad_norm": 0.9794378280639648, | |
| "learning_rate": 3.0012281240405283e-05, | |
| "loss": 4.2736, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 4.224110218140069, | |
| "grad_norm": 1.0224289894104004, | |
| "learning_rate": 2.9705250230273257e-05, | |
| "loss": 4.2309, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 4.24247990815155, | |
| "grad_norm": 1.120897889137268, | |
| "learning_rate": 2.939821922014124e-05, | |
| "loss": 4.2337, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 4.2608495981630305, | |
| "grad_norm": 1.2583802938461304, | |
| "learning_rate": 2.9091188210009213e-05, | |
| "loss": 4.1689, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 4.279219288174512, | |
| "grad_norm": 1.421479344367981, | |
| "learning_rate": 2.8784157199877187e-05, | |
| "loss": 4.2721, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 4.297588978185993, | |
| "grad_norm": 1.2407679557800293, | |
| "learning_rate": 2.847712618974517e-05, | |
| "loss": 4.2649, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 4.315958668197474, | |
| "grad_norm": 1.1917908191680908, | |
| "learning_rate": 2.8170095179613143e-05, | |
| "loss": 4.2253, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 4.334328358208955, | |
| "grad_norm": 1.1925767660140991, | |
| "learning_rate": 2.7863064169481117e-05, | |
| "loss": 4.2396, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 4.352698048220436, | |
| "grad_norm": 1.028193712234497, | |
| "learning_rate": 2.75560331593491e-05, | |
| "loss": 4.262, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 4.371067738231917, | |
| "grad_norm": 1.3707988262176514, | |
| "learning_rate": 2.7249002149217073e-05, | |
| "loss": 4.2115, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 4.389437428243398, | |
| "grad_norm": 1.086790919303894, | |
| "learning_rate": 2.6941971139085047e-05, | |
| "loss": 4.1938, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 4.40780711825488, | |
| "grad_norm": 1.0577547550201416, | |
| "learning_rate": 2.6634940128953022e-05, | |
| "loss": 4.2009, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 4.42617680826636, | |
| "grad_norm": 1.1174390316009521, | |
| "learning_rate": 2.6327909118821003e-05, | |
| "loss": 4.22, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 4.444546498277841, | |
| "grad_norm": 1.0326007604599, | |
| "learning_rate": 2.6020878108688977e-05, | |
| "loss": 4.2575, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 4.462916188289323, | |
| "grad_norm": 1.172747254371643, | |
| "learning_rate": 2.5713847098556952e-05, | |
| "loss": 4.1741, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 4.481285878300803, | |
| "grad_norm": 1.2684015035629272, | |
| "learning_rate": 2.5406816088424933e-05, | |
| "loss": 4.2803, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 4.499655568312285, | |
| "grad_norm": 1.3229318857192993, | |
| "learning_rate": 2.5099785078292907e-05, | |
| "loss": 4.2587, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 4.518025258323766, | |
| "grad_norm": 1.0893479585647583, | |
| "learning_rate": 2.4792754068160885e-05, | |
| "loss": 4.1393, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 4.536394948335246, | |
| "grad_norm": 1.18522310256958, | |
| "learning_rate": 2.4485723058028863e-05, | |
| "loss": 4.2688, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 4.554764638346728, | |
| "grad_norm": 1.2020633220672607, | |
| "learning_rate": 2.417869204789684e-05, | |
| "loss": 4.1873, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 4.573134328358209, | |
| "grad_norm": 1.0598649978637695, | |
| "learning_rate": 2.3871661037764815e-05, | |
| "loss": 4.1961, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 4.59150401836969, | |
| "grad_norm": 1.320818305015564, | |
| "learning_rate": 2.3564630027632793e-05, | |
| "loss": 4.2311, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.609873708381171, | |
| "grad_norm": 1.1840718984603882, | |
| "learning_rate": 2.325759901750077e-05, | |
| "loss": 4.1921, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 4.628243398392652, | |
| "grad_norm": 1.0860955715179443, | |
| "learning_rate": 2.2950568007368745e-05, | |
| "loss": 4.1796, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 4.646613088404133, | |
| "grad_norm": 1.1883424520492554, | |
| "learning_rate": 2.2643536997236723e-05, | |
| "loss": 4.2878, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 4.664982778415614, | |
| "grad_norm": 1.2330527305603027, | |
| "learning_rate": 2.23365059871047e-05, | |
| "loss": 4.2017, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 4.6833524684270955, | |
| "grad_norm": 1.153104543685913, | |
| "learning_rate": 2.2029474976972675e-05, | |
| "loss": 4.2254, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 4.701722158438576, | |
| "grad_norm": 1.078261137008667, | |
| "learning_rate": 2.1722443966840653e-05, | |
| "loss": 4.2249, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 4.720091848450057, | |
| "grad_norm": 1.3064528703689575, | |
| "learning_rate": 2.1415412956708628e-05, | |
| "loss": 4.1816, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 4.7384615384615385, | |
| "grad_norm": 1.1915974617004395, | |
| "learning_rate": 2.1108381946576606e-05, | |
| "loss": 4.1787, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 4.756831228473019, | |
| "grad_norm": 1.1211307048797607, | |
| "learning_rate": 2.0801350936444583e-05, | |
| "loss": 4.2094, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 4.775200918484501, | |
| "grad_norm": 1.2985018491744995, | |
| "learning_rate": 2.0494319926312558e-05, | |
| "loss": 4.2306, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 4.7935706084959815, | |
| "grad_norm": 1.0229589939117432, | |
| "learning_rate": 2.0187288916180536e-05, | |
| "loss": 4.2341, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 4.811940298507462, | |
| "grad_norm": 1.3006722927093506, | |
| "learning_rate": 1.988025790604851e-05, | |
| "loss": 4.2555, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 4.830309988518944, | |
| "grad_norm": 1.1146833896636963, | |
| "learning_rate": 1.9573226895916488e-05, | |
| "loss": 4.1993, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 4.8486796785304245, | |
| "grad_norm": 1.0311157703399658, | |
| "learning_rate": 1.9266195885784466e-05, | |
| "loss": 4.2419, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 4.867049368541906, | |
| "grad_norm": 1.187200665473938, | |
| "learning_rate": 1.895916487565244e-05, | |
| "loss": 4.1751, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 4.885419058553387, | |
| "grad_norm": 1.1463189125061035, | |
| "learning_rate": 1.8652133865520418e-05, | |
| "loss": 4.2724, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 4.903788748564868, | |
| "grad_norm": 1.110522747039795, | |
| "learning_rate": 1.8345102855388396e-05, | |
| "loss": 4.2305, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 4.922158438576349, | |
| "grad_norm": 1.139212727546692, | |
| "learning_rate": 1.803807184525637e-05, | |
| "loss": 4.1795, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 4.94052812858783, | |
| "grad_norm": 1.0388927459716797, | |
| "learning_rate": 1.7731040835124348e-05, | |
| "loss": 4.1886, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 4.958897818599311, | |
| "grad_norm": 1.5335136651992798, | |
| "learning_rate": 1.7424009824992322e-05, | |
| "loss": 4.2074, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 4.977267508610792, | |
| "grad_norm": 1.2070115804672241, | |
| "learning_rate": 1.71169788148603e-05, | |
| "loss": 4.2166, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 4.995637198622274, | |
| "grad_norm": 1.0662022829055786, | |
| "learning_rate": 1.6809947804728278e-05, | |
| "loss": 4.2369, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 5.327496528625488, | |
| "eval_runtime": 160.2538, | |
| "eval_samples_per_second": 6.04, | |
| "eval_steps_per_second": 3.02, | |
| "step": 5445 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 6534, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.192414285567118e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |