{ "best_global_step": 416, "best_metric": 0.10842076689004898, "best_model_checkpoint": "saves_bts_preliminary/base/llama-3.2-1b-instruct/train_mrpc_42_1776331557/checkpoint-416", "epoch": 5.0, "eval_steps": 104, "global_step": 2065, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.012106537530266344, "grad_norm": 320.0541076660156, "learning_rate": 9.661835748792271e-08, "loss": 0.81, "num_input_tokens_seen": 4352, "step": 5 }, { "epoch": 0.024213075060532687, "grad_norm": 293.6126403808594, "learning_rate": 2.173913043478261e-07, "loss": 0.8, "num_input_tokens_seen": 8768, "step": 10 }, { "epoch": 0.03631961259079903, "grad_norm": 176.21604919433594, "learning_rate": 3.3816425120772945e-07, "loss": 0.6335, "num_input_tokens_seen": 12992, "step": 15 }, { "epoch": 0.048426150121065374, "grad_norm": 67.18666076660156, "learning_rate": 4.5893719806763294e-07, "loss": 0.3717, "num_input_tokens_seen": 17344, "step": 20 }, { "epoch": 0.06053268765133172, "grad_norm": 41.770694732666016, "learning_rate": 5.797101449275363e-07, "loss": 0.2438, "num_input_tokens_seen": 21696, "step": 25 }, { "epoch": 0.07263922518159806, "grad_norm": 54.93859100341797, "learning_rate": 7.004830917874397e-07, "loss": 0.2523, "num_input_tokens_seen": 26112, "step": 30 }, { "epoch": 0.0847457627118644, "grad_norm": 14.744933128356934, "learning_rate": 8.212560386473431e-07, "loss": 0.2223, "num_input_tokens_seen": 30208, "step": 35 }, { "epoch": 0.09685230024213075, "grad_norm": 6.022367477416992, "learning_rate": 9.420289855072465e-07, "loss": 0.2184, "num_input_tokens_seen": 34688, "step": 40 }, { "epoch": 0.1089588377723971, "grad_norm": 39.68272399902344, "learning_rate": 1.0628019323671499e-06, "loss": 0.2163, "num_input_tokens_seen": 38784, "step": 45 }, { "epoch": 0.12106537530266344, "grad_norm": 9.175082206726074, "learning_rate": 1.1835748792270531e-06, "loss": 0.2198, "num_input_tokens_seen": 43200, "step": 50 }, { "epoch": 0.13317191283292978, "grad_norm": 42.212303161621094, "learning_rate": 1.3043478260869566e-06, "loss": 0.2224, "num_input_tokens_seen": 47296, "step": 55 }, { "epoch": 0.14527845036319612, "grad_norm": 5.82745885848999, "learning_rate": 1.42512077294686e-06, "loss": 0.2272, "num_input_tokens_seen": 51712, "step": 60 }, { "epoch": 0.15738498789346247, "grad_norm": 18.959779739379883, "learning_rate": 1.5458937198067634e-06, "loss": 0.1665, "num_input_tokens_seen": 55872, "step": 65 }, { "epoch": 0.1694915254237288, "grad_norm": 54.883968353271484, "learning_rate": 1.6666666666666667e-06, "loss": 0.1835, "num_input_tokens_seen": 59840, "step": 70 }, { "epoch": 0.18159806295399517, "grad_norm": 87.76690673828125, "learning_rate": 1.7874396135265702e-06, "loss": 0.1898, "num_input_tokens_seen": 64000, "step": 75 }, { "epoch": 0.1937046004842615, "grad_norm": 22.004396438598633, "learning_rate": 1.9082125603864736e-06, "loss": 0.2149, "num_input_tokens_seen": 68352, "step": 80 }, { "epoch": 0.20581113801452786, "grad_norm": 13.515002250671387, "learning_rate": 2.028985507246377e-06, "loss": 0.1519, "num_input_tokens_seen": 72768, "step": 85 }, { "epoch": 0.2179176755447942, "grad_norm": 22.01325798034668, "learning_rate": 2.1497584541062806e-06, "loss": 0.1468, "num_input_tokens_seen": 77120, "step": 90 }, { "epoch": 0.23002421307506055, "grad_norm": 24.60677146911621, "learning_rate": 2.270531400966184e-06, "loss": 0.2277, "num_input_tokens_seen": 81664, "step": 95 }, { "epoch": 0.24213075060532688, "grad_norm": 19.255220413208008, "learning_rate": 2.391304347826087e-06, "loss": 0.1552, "num_input_tokens_seen": 86080, "step": 100 }, { "epoch": 0.25181598062953997, "eval_loss": 0.1484687179327011, "eval_runtime": 0.6174, "eval_samples_per_second": 594.395, "eval_steps_per_second": 74.502, "num_input_tokens_seen": 89600, "step": 104 }, { "epoch": 0.2542372881355932, "grad_norm": 51.306358337402344, "learning_rate": 2.5120772946859904e-06, "loss": 0.1673, "num_input_tokens_seen": 90432, "step": 105 }, { "epoch": 0.26634382566585957, "grad_norm": 20.150880813598633, "learning_rate": 2.632850241545894e-06, "loss": 0.1694, "num_input_tokens_seen": 94528, "step": 110 }, { "epoch": 0.2784503631961259, "grad_norm": 16.875028610229492, "learning_rate": 2.7536231884057974e-06, "loss": 0.1627, "num_input_tokens_seen": 98816, "step": 115 }, { "epoch": 0.29055690072639223, "grad_norm": 17.16132164001465, "learning_rate": 2.8743961352657007e-06, "loss": 0.2205, "num_input_tokens_seen": 103104, "step": 120 }, { "epoch": 0.3026634382566586, "grad_norm": 19.57474708557129, "learning_rate": 2.995169082125604e-06, "loss": 0.1841, "num_input_tokens_seen": 107328, "step": 125 }, { "epoch": 0.31476997578692495, "grad_norm": 17.67765998840332, "learning_rate": 3.1159420289855073e-06, "loss": 0.1779, "num_input_tokens_seen": 111488, "step": 130 }, { "epoch": 0.3268765133171913, "grad_norm": 21.58500099182129, "learning_rate": 3.236714975845411e-06, "loss": 0.158, "num_input_tokens_seen": 115968, "step": 135 }, { "epoch": 0.3389830508474576, "grad_norm": 41.02679443359375, "learning_rate": 3.3574879227053142e-06, "loss": 0.2089, "num_input_tokens_seen": 120192, "step": 140 }, { "epoch": 0.35108958837772397, "grad_norm": 13.272953987121582, "learning_rate": 3.4782608695652175e-06, "loss": 0.1007, "num_input_tokens_seen": 124416, "step": 145 }, { "epoch": 0.36319612590799033, "grad_norm": 58.32763671875, "learning_rate": 3.5990338164251208e-06, "loss": 0.2718, "num_input_tokens_seen": 128832, "step": 150 }, { "epoch": 0.37530266343825663, "grad_norm": 46.48350524902344, "learning_rate": 3.7198067632850245e-06, "loss": 0.3704, "num_input_tokens_seen": 132992, "step": 155 }, { "epoch": 0.387409200968523, "grad_norm": 13.923331260681152, "learning_rate": 3.840579710144928e-06, "loss": 0.1945, "num_input_tokens_seen": 137280, "step": 160 }, { "epoch": 0.39951573849878935, "grad_norm": 76.48035430908203, "learning_rate": 3.961352657004831e-06, "loss": 0.253, "num_input_tokens_seen": 141568, "step": 165 }, { "epoch": 0.4116222760290557, "grad_norm": 8.948328971862793, "learning_rate": 4.082125603864734e-06, "loss": 0.1622, "num_input_tokens_seen": 145984, "step": 170 }, { "epoch": 0.423728813559322, "grad_norm": 44.44621276855469, "learning_rate": 4.202898550724638e-06, "loss": 0.1974, "num_input_tokens_seen": 150144, "step": 175 }, { "epoch": 0.4358353510895884, "grad_norm": 12.341259956359863, "learning_rate": 4.323671497584541e-06, "loss": 0.2191, "num_input_tokens_seen": 154624, "step": 180 }, { "epoch": 0.44794188861985473, "grad_norm": 13.725370407104492, "learning_rate": 4.444444444444444e-06, "loss": 0.2192, "num_input_tokens_seen": 158784, "step": 185 }, { "epoch": 0.4600484261501211, "grad_norm": 9.8717622756958, "learning_rate": 4.565217391304348e-06, "loss": 0.1887, "num_input_tokens_seen": 163072, "step": 190 }, { "epoch": 0.4721549636803874, "grad_norm": 15.11948299407959, "learning_rate": 4.6859903381642516e-06, "loss": 0.1951, "num_input_tokens_seen": 167104, "step": 195 }, { "epoch": 0.48426150121065376, "grad_norm": 10.298319816589355, "learning_rate": 4.806763285024155e-06, "loss": 0.1486, "num_input_tokens_seen": 171456, "step": 200 }, { "epoch": 0.4963680387409201, "grad_norm": 22.512561798095703, "learning_rate": 4.927536231884059e-06, "loss": 0.2178, "num_input_tokens_seen": 175808, "step": 205 }, { "epoch": 0.5036319612590799, "eval_loss": 0.1319892704486847, "eval_runtime": 0.6163, "eval_samples_per_second": 595.512, "eval_steps_per_second": 74.642, "num_input_tokens_seen": 178688, "step": 208 }, { "epoch": 0.5084745762711864, "grad_norm": 11.861852645874023, "learning_rate": 4.999985705205496e-06, "loss": 0.1052, "num_input_tokens_seen": 180224, "step": 210 }, { "epoch": 0.5205811138014528, "grad_norm": 15.08198070526123, "learning_rate": 4.999824890644693e-06, "loss": 0.1655, "num_input_tokens_seen": 184704, "step": 215 }, { "epoch": 0.5326876513317191, "grad_norm": 9.965168952941895, "learning_rate": 4.999485404562269e-06, "loss": 0.3684, "num_input_tokens_seen": 189184, "step": 220 }, { "epoch": 0.5447941888619855, "grad_norm": 7.275655269622803, "learning_rate": 4.998967271222521e-06, "loss": 0.1527, "num_input_tokens_seen": 193536, "step": 225 }, { "epoch": 0.5569007263922519, "grad_norm": 7.0880584716796875, "learning_rate": 4.998270527658311e-06, "loss": 0.1238, "num_input_tokens_seen": 197888, "step": 230 }, { "epoch": 0.5690072639225182, "grad_norm": 27.60887908935547, "learning_rate": 4.997395223668422e-06, "loss": 0.2147, "num_input_tokens_seen": 202112, "step": 235 }, { "epoch": 0.5811138014527845, "grad_norm": 43.02740478515625, "learning_rate": 4.996341421813993e-06, "loss": 0.1162, "num_input_tokens_seen": 206528, "step": 240 }, { "epoch": 0.5932203389830508, "grad_norm": 30.406055450439453, "learning_rate": 4.995109197414051e-06, "loss": 0.1311, "num_input_tokens_seen": 210944, "step": 245 }, { "epoch": 0.6053268765133172, "grad_norm": 14.91820240020752, "learning_rate": 4.9936986385401305e-06, "loss": 0.1437, "num_input_tokens_seen": 215104, "step": 250 }, { "epoch": 0.6174334140435835, "grad_norm": 20.09491729736328, "learning_rate": 4.992109846009972e-06, "loss": 0.1597, "num_input_tokens_seen": 219328, "step": 255 }, { "epoch": 0.6295399515738499, "grad_norm": 6.193624973297119, "learning_rate": 4.990342933380321e-06, "loss": 0.1878, "num_input_tokens_seen": 223680, "step": 260 }, { "epoch": 0.6416464891041163, "grad_norm": 6.540223121643066, "learning_rate": 4.988398026938811e-06, "loss": 0.1445, "num_input_tokens_seen": 227904, "step": 265 }, { "epoch": 0.6537530266343826, "grad_norm": 17.89214515686035, "learning_rate": 4.986275265694935e-06, "loss": 0.0992, "num_input_tokens_seen": 231936, "step": 270 }, { "epoch": 0.6658595641646489, "grad_norm": 0.7999329566955566, "learning_rate": 4.983974801370115e-06, "loss": 0.0608, "num_input_tokens_seen": 236160, "step": 275 }, { "epoch": 0.6779661016949152, "grad_norm": 36.78638458251953, "learning_rate": 4.981496798386849e-06, "loss": 0.2262, "num_input_tokens_seen": 240320, "step": 280 }, { "epoch": 0.6900726392251816, "grad_norm": 18.634634017944336, "learning_rate": 4.9788414338569715e-06, "loss": 0.1165, "num_input_tokens_seen": 244800, "step": 285 }, { "epoch": 0.7021791767554479, "grad_norm": 35.7069091796875, "learning_rate": 4.9760088975689815e-06, "loss": 0.2377, "num_input_tokens_seen": 249152, "step": 290 }, { "epoch": 0.7142857142857143, "grad_norm": 10.031001091003418, "learning_rate": 4.972999391974488e-06, "loss": 0.1377, "num_input_tokens_seen": 253376, "step": 295 }, { "epoch": 0.7263922518159807, "grad_norm": 7.3871612548828125, "learning_rate": 4.969813132173735e-06, "loss": 0.19, "num_input_tokens_seen": 257664, "step": 300 }, { "epoch": 0.738498789346247, "grad_norm": 16.09194564819336, "learning_rate": 4.966450345900229e-06, "loss": 0.1146, "num_input_tokens_seen": 262016, "step": 305 }, { "epoch": 0.7506053268765133, "grad_norm": 7.399415969848633, "learning_rate": 4.962911273504461e-06, "loss": 0.1165, "num_input_tokens_seen": 266432, "step": 310 }, { "epoch": 0.7554479418886199, "eval_loss": 0.11303775012493134, "eval_runtime": 1.7273, "eval_samples_per_second": 212.47, "eval_steps_per_second": 26.631, "num_input_tokens_seen": 267968, "step": 312 }, { "epoch": 0.7627118644067796, "grad_norm": 11.618612289428711, "learning_rate": 4.959196167936729e-06, "loss": 0.181, "num_input_tokens_seen": 270464, "step": 315 }, { "epoch": 0.774818401937046, "grad_norm": 11.343527793884277, "learning_rate": 4.955305294729056e-06, "loss": 0.0946, "num_input_tokens_seen": 274688, "step": 320 }, { "epoch": 0.7869249394673123, "grad_norm": 9.781023025512695, "learning_rate": 4.9512389319762165e-06, "loss": 0.1293, "num_input_tokens_seen": 278848, "step": 325 }, { "epoch": 0.7990314769975787, "grad_norm": 23.389354705810547, "learning_rate": 4.946997370315857e-06, "loss": 0.124, "num_input_tokens_seen": 283136, "step": 330 }, { "epoch": 0.8111380145278451, "grad_norm": 21.524974822998047, "learning_rate": 4.9425809129077204e-06, "loss": 0.1767, "num_input_tokens_seen": 287680, "step": 335 }, { "epoch": 0.8232445520581114, "grad_norm": 12.489716529846191, "learning_rate": 4.937989875411986e-06, "loss": 0.0811, "num_input_tokens_seen": 292224, "step": 340 }, { "epoch": 0.8353510895883777, "grad_norm": 11.355611801147461, "learning_rate": 4.933224585966696e-06, "loss": 0.1567, "num_input_tokens_seen": 296448, "step": 345 }, { "epoch": 0.847457627118644, "grad_norm": 24.681901931762695, "learning_rate": 4.928285385164316e-06, "loss": 0.1363, "num_input_tokens_seen": 300736, "step": 350 }, { "epoch": 0.8595641646489104, "grad_norm": 8.876485824584961, "learning_rate": 4.92317262602738e-06, "loss": 0.1348, "num_input_tokens_seen": 304960, "step": 355 }, { "epoch": 0.8716707021791767, "grad_norm": 20.233808517456055, "learning_rate": 4.917886673983267e-06, "loss": 0.1694, "num_input_tokens_seen": 309184, "step": 360 }, { "epoch": 0.8837772397094431, "grad_norm": 20.60609245300293, "learning_rate": 4.912427906838079e-06, "loss": 0.1352, "num_input_tokens_seen": 313408, "step": 365 }, { "epoch": 0.8958837772397095, "grad_norm": 6.115711688995361, "learning_rate": 4.906796714749635e-06, "loss": 0.0933, "num_input_tokens_seen": 317888, "step": 370 }, { "epoch": 0.9079903147699758, "grad_norm": 18.18195152282715, "learning_rate": 4.900993500199591e-06, "loss": 0.1488, "num_input_tokens_seen": 322048, "step": 375 }, { "epoch": 0.9200968523002422, "grad_norm": 12.584397315979004, "learning_rate": 4.895018677964669e-06, "loss": 0.087, "num_input_tokens_seen": 326592, "step": 380 }, { "epoch": 0.9322033898305084, "grad_norm": 7.245577812194824, "learning_rate": 4.888872675087012e-06, "loss": 0.1017, "num_input_tokens_seen": 330880, "step": 385 }, { "epoch": 0.9443099273607748, "grad_norm": 51.712425231933594, "learning_rate": 4.882555930843664e-06, "loss": 0.1105, "num_input_tokens_seen": 335104, "step": 390 }, { "epoch": 0.9564164648910412, "grad_norm": 25.56722068786621, "learning_rate": 4.876068896715171e-06, "loss": 0.1437, "num_input_tokens_seen": 339392, "step": 395 }, { "epoch": 0.9685230024213075, "grad_norm": 20.534461975097656, "learning_rate": 4.8694120363533105e-06, "loss": 0.146, "num_input_tokens_seen": 343744, "step": 400 }, { "epoch": 0.9806295399515739, "grad_norm": 8.779641151428223, "learning_rate": 4.862585825547957e-06, "loss": 0.0985, "num_input_tokens_seen": 348160, "step": 405 }, { "epoch": 0.9927360774818402, "grad_norm": 17.242847442626953, "learning_rate": 4.855590752193075e-06, "loss": 0.116, "num_input_tokens_seen": 352448, "step": 410 }, { "epoch": 1.0048426150121066, "grad_norm": 13.249277114868164, "learning_rate": 4.848427316251843e-06, "loss": 0.1193, "num_input_tokens_seen": 356656, "step": 415 }, { "epoch": 1.0072639225181599, "eval_loss": 0.10842076689004898, "eval_runtime": 0.63, "eval_samples_per_second": 582.523, "eval_steps_per_second": 73.014, "num_input_tokens_seen": 357488, "step": 416 }, { "epoch": 1.0169491525423728, "grad_norm": 1.1179414987564087, "learning_rate": 4.841096029720921e-06, "loss": 0.073, "num_input_tokens_seen": 360880, "step": 420 }, { "epoch": 1.0290556900726393, "grad_norm": 31.719369888305664, "learning_rate": 4.833597416593861e-06, "loss": 0.0535, "num_input_tokens_seen": 365104, "step": 425 }, { "epoch": 1.0411622276029056, "grad_norm": 48.02503204345703, "learning_rate": 4.825932012823652e-06, "loss": 0.1458, "num_input_tokens_seen": 369776, "step": 430 }, { "epoch": 1.053268765133172, "grad_norm": 97.04767608642578, "learning_rate": 4.818100366284408e-06, "loss": 0.1602, "num_input_tokens_seen": 374000, "step": 435 }, { "epoch": 1.0653753026634383, "grad_norm": 37.753238677978516, "learning_rate": 4.81010303673222e-06, "loss": 0.2577, "num_input_tokens_seen": 378096, "step": 440 }, { "epoch": 1.0774818401937045, "grad_norm": 0.17760241031646729, "learning_rate": 4.80194059576514e-06, "loss": 0.0566, "num_input_tokens_seen": 382256, "step": 445 }, { "epoch": 1.089588377723971, "grad_norm": 34.652015686035156, "learning_rate": 4.793613626782331e-06, "loss": 0.1761, "num_input_tokens_seen": 386672, "step": 450 }, { "epoch": 1.1016949152542372, "grad_norm": 28.04759407043457, "learning_rate": 4.785122724942367e-06, "loss": 0.0591, "num_input_tokens_seen": 390960, "step": 455 }, { "epoch": 1.1138014527845037, "grad_norm": 2.957566976547241, "learning_rate": 4.7764684971206974e-06, "loss": 0.0952, "num_input_tokens_seen": 395440, "step": 460 }, { "epoch": 1.12590799031477, "grad_norm": 44.540069580078125, "learning_rate": 4.767651561866269e-06, "loss": 0.0664, "num_input_tokens_seen": 399600, "step": 465 }, { "epoch": 1.1380145278450362, "grad_norm": 24.51837730407715, "learning_rate": 4.758672549357316e-06, "loss": 0.1001, "num_input_tokens_seen": 403888, "step": 470 }, { "epoch": 1.1501210653753027, "grad_norm": 40.098114013671875, "learning_rate": 4.7495321013563225e-06, "loss": 0.2506, "num_input_tokens_seen": 408176, "step": 475 }, { "epoch": 1.162227602905569, "grad_norm": 2.672497510910034, "learning_rate": 4.740230871164148e-06, "loss": 0.044, "num_input_tokens_seen": 412208, "step": 480 }, { "epoch": 1.1743341404358354, "grad_norm": 0.2532678544521332, "learning_rate": 4.730769523573337e-06, "loss": 0.1472, "num_input_tokens_seen": 416624, "step": 485 }, { "epoch": 1.1864406779661016, "grad_norm": 4.472592830657959, "learning_rate": 4.721148734820605e-06, "loss": 0.1661, "num_input_tokens_seen": 421040, "step": 490 }, { "epoch": 1.1985472154963681, "grad_norm": 37.826171875, "learning_rate": 4.711369192538503e-06, "loss": 0.094, "num_input_tokens_seen": 425136, "step": 495 }, { "epoch": 1.2106537530266344, "grad_norm": 14.548481941223145, "learning_rate": 4.701431595706269e-06, "loss": 0.1282, "num_input_tokens_seen": 429680, "step": 500 }, { "epoch": 1.2227602905569008, "grad_norm": 24.612041473388672, "learning_rate": 4.691336654599873e-06, "loss": 0.0874, "num_input_tokens_seen": 434224, "step": 505 }, { "epoch": 1.234866828087167, "grad_norm": 11.3072509765625, "learning_rate": 4.6810850907412486e-06, "loss": 0.0403, "num_input_tokens_seen": 438320, "step": 510 }, { "epoch": 1.2469733656174333, "grad_norm": 9.126791000366211, "learning_rate": 4.6706776368467236e-06, "loss": 0.0227, "num_input_tokens_seen": 442672, "step": 515 }, { "epoch": 1.2590799031476998, "grad_norm": 23.92775535583496, "learning_rate": 4.6601150367746485e-06, "loss": 0.0685, "num_input_tokens_seen": 446896, "step": 520 }, { "epoch": 1.2590799031476998, "eval_loss": 0.19028596580028534, "eval_runtime": 0.7167, "eval_samples_per_second": 512.081, "eval_steps_per_second": 64.185, "num_input_tokens_seen": 446896, "step": 520 }, { "epoch": 1.271186440677966, "grad_norm": 43.593448638916016, "learning_rate": 4.649398045472235e-06, "loss": 0.1008, "num_input_tokens_seen": 451312, "step": 525 }, { "epoch": 1.2832929782082325, "grad_norm": 8.737812042236328, "learning_rate": 4.638527428921592e-06, "loss": 0.3076, "num_input_tokens_seen": 455408, "step": 530 }, { "epoch": 1.2953995157384988, "grad_norm": 2.4155948162078857, "learning_rate": 4.627503964084981e-06, "loss": 0.0462, "num_input_tokens_seen": 460080, "step": 535 }, { "epoch": 1.307506053268765, "grad_norm": 2.872014284133911, "learning_rate": 4.616328438849284e-06, "loss": 0.0124, "num_input_tokens_seen": 464496, "step": 540 }, { "epoch": 1.3196125907990315, "grad_norm": 34.14030456542969, "learning_rate": 4.605001651969686e-06, "loss": 0.1408, "num_input_tokens_seen": 468720, "step": 545 }, { "epoch": 1.331719128329298, "grad_norm": 59.56473922729492, "learning_rate": 4.5935244130125925e-06, "loss": 0.115, "num_input_tokens_seen": 473264, "step": 550 }, { "epoch": 1.3438256658595642, "grad_norm": 1.2305806875228882, "learning_rate": 4.581897542297761e-06, "loss": 0.0061, "num_input_tokens_seen": 477552, "step": 555 }, { "epoch": 1.3559322033898304, "grad_norm": 52.619632720947266, "learning_rate": 4.570121870839671e-06, "loss": 0.0843, "num_input_tokens_seen": 482032, "step": 560 }, { "epoch": 1.368038740920097, "grad_norm": 97.8170394897461, "learning_rate": 4.558198240288131e-06, "loss": 0.0764, "num_input_tokens_seen": 486384, "step": 565 }, { "epoch": 1.3801452784503632, "grad_norm": 35.3338737487793, "learning_rate": 4.5461275028681186e-06, "loss": 0.1836, "num_input_tokens_seen": 490672, "step": 570 }, { "epoch": 1.3922518159806296, "grad_norm": 27.42061996459961, "learning_rate": 4.533910521318872e-06, "loss": 0.1097, "num_input_tokens_seen": 494960, "step": 575 }, { "epoch": 1.4043583535108959, "grad_norm": 7.799497604370117, "learning_rate": 4.521548168832227e-06, "loss": 0.1144, "num_input_tokens_seen": 499120, "step": 580 }, { "epoch": 1.4164648910411621, "grad_norm": 11.070670127868652, "learning_rate": 4.509041328990204e-06, "loss": 0.0169, "num_input_tokens_seen": 503408, "step": 585 }, { "epoch": 1.4285714285714286, "grad_norm": 5.323161602020264, "learning_rate": 4.496390895701858e-06, "loss": 0.0424, "num_input_tokens_seen": 507312, "step": 590 }, { "epoch": 1.4406779661016949, "grad_norm": 26.43401527404785, "learning_rate": 4.483597773139387e-06, "loss": 0.053, "num_input_tokens_seen": 511600, "step": 595 }, { "epoch": 1.4527845036319613, "grad_norm": 0.05975044146180153, "learning_rate": 4.470662875673506e-06, "loss": 0.0615, "num_input_tokens_seen": 515888, "step": 600 }, { "epoch": 1.4648910411622276, "grad_norm": 52.62843322753906, "learning_rate": 4.4575871278080964e-06, "loss": 0.2071, "num_input_tokens_seen": 519920, "step": 605 }, { "epoch": 1.4769975786924938, "grad_norm": 10.014227867126465, "learning_rate": 4.444371464114126e-06, "loss": 0.0688, "num_input_tokens_seen": 524336, "step": 610 }, { "epoch": 1.4891041162227603, "grad_norm": 0.5611757636070251, "learning_rate": 4.431016829162851e-06, "loss": 0.071, "num_input_tokens_seen": 528496, "step": 615 }, { "epoch": 1.5012106537530268, "grad_norm": 0.10402300208806992, "learning_rate": 4.417524177458309e-06, "loss": 0.0801, "num_input_tokens_seen": 532784, "step": 620 }, { "epoch": 1.5108958837772397, "eval_loss": 0.1981746405363083, "eval_runtime": 0.6398, "eval_samples_per_second": 573.626, "eval_steps_per_second": 71.899, "num_input_tokens_seen": 536176, "step": 624 }, { "epoch": 1.513317191283293, "grad_norm": 24.07758331298828, "learning_rate": 4.403894473369092e-06, "loss": 0.0258, "num_input_tokens_seen": 537136, "step": 625 }, { "epoch": 1.5254237288135593, "grad_norm": 48.458106994628906, "learning_rate": 4.390128691059423e-06, "loss": 0.199, "num_input_tokens_seen": 541552, "step": 630 }, { "epoch": 1.5375302663438255, "grad_norm": 8.228415489196777, "learning_rate": 4.376227814419524e-06, "loss": 0.1964, "num_input_tokens_seen": 545648, "step": 635 }, { "epoch": 1.549636803874092, "grad_norm": 12.28972339630127, "learning_rate": 4.3621928369952995e-06, "loss": 0.06, "num_input_tokens_seen": 550256, "step": 640 }, { "epoch": 1.5617433414043584, "grad_norm": 15.129716873168945, "learning_rate": 4.348024761917321e-06, "loss": 0.1114, "num_input_tokens_seen": 554928, "step": 645 }, { "epoch": 1.5738498789346247, "grad_norm": 0.27470338344573975, "learning_rate": 4.333724601829132e-06, "loss": 0.0725, "num_input_tokens_seen": 559344, "step": 650 }, { "epoch": 1.585956416464891, "grad_norm": 0.1773267388343811, "learning_rate": 4.319293378814868e-06, "loss": 0.1308, "num_input_tokens_seen": 563760, "step": 655 }, { "epoch": 1.5980629539951574, "grad_norm": 14.22355842590332, "learning_rate": 4.3047321243262065e-06, "loss": 0.0653, "num_input_tokens_seen": 568112, "step": 660 }, { "epoch": 1.6101694915254239, "grad_norm": 0.1500890851020813, "learning_rate": 4.290041879108641e-06, "loss": 0.006, "num_input_tokens_seen": 572464, "step": 665 }, { "epoch": 1.6222760290556901, "grad_norm": 30.997364044189453, "learning_rate": 4.275223693127103e-06, "loss": 0.0771, "num_input_tokens_seen": 576752, "step": 670 }, { "epoch": 1.6343825665859564, "grad_norm": 44.13563919067383, "learning_rate": 4.260278625490911e-06, "loss": 0.034, "num_input_tokens_seen": 580976, "step": 675 }, { "epoch": 1.6464891041162226, "grad_norm": 1.0693022012710571, "learning_rate": 4.245207744378075e-06, "loss": 0.1429, "num_input_tokens_seen": 585264, "step": 680 }, { "epoch": 1.658595641646489, "grad_norm": 32.29472351074219, "learning_rate": 4.2300121269589475e-06, "loss": 0.0664, "num_input_tokens_seen": 589744, "step": 685 }, { "epoch": 1.6707021791767556, "grad_norm": 40.11149597167969, "learning_rate": 4.2146928593192375e-06, "loss": 0.0792, "num_input_tokens_seen": 593968, "step": 690 }, { "epoch": 1.6828087167070218, "grad_norm": 43.619258880615234, "learning_rate": 4.19925103638238e-06, "loss": 0.1061, "num_input_tokens_seen": 598256, "step": 695 }, { "epoch": 1.694915254237288, "grad_norm": 0.6466928720474243, "learning_rate": 4.183687761831282e-06, "loss": 0.0958, "num_input_tokens_seen": 602608, "step": 700 }, { "epoch": 1.7070217917675545, "grad_norm": 27.355270385742188, "learning_rate": 4.168004148029435e-06, "loss": 0.1234, "num_input_tokens_seen": 607088, "step": 705 }, { "epoch": 1.7191283292978208, "grad_norm": 17.83440399169922, "learning_rate": 4.152201315941414e-06, "loss": 0.1094, "num_input_tokens_seen": 611248, "step": 710 }, { "epoch": 1.7312348668280872, "grad_norm": 1.9912621974945068, "learning_rate": 4.136280395052754e-06, "loss": 0.1047, "num_input_tokens_seen": 615536, "step": 715 }, { "epoch": 1.7433414043583535, "grad_norm": 7.247110843658447, "learning_rate": 4.120242523289223e-06, "loss": 0.0341, "num_input_tokens_seen": 619952, "step": 720 }, { "epoch": 1.7554479418886197, "grad_norm": 42.79171371459961, "learning_rate": 4.104088846935493e-06, "loss": 0.2066, "num_input_tokens_seen": 624368, "step": 725 }, { "epoch": 1.7627118644067796, "eval_loss": 0.14485575258731842, "eval_runtime": 0.6316, "eval_samples_per_second": 581.029, "eval_steps_per_second": 72.826, "num_input_tokens_seen": 626992, "step": 728 }, { "epoch": 1.7675544794188862, "grad_norm": 15.863067626953125, "learning_rate": 4.087820520553205e-06, "loss": 0.0104, "num_input_tokens_seen": 628720, "step": 730 }, { "epoch": 1.7796610169491527, "grad_norm": 4.8767571449279785, "learning_rate": 4.071438706898457e-06, "loss": 0.0572, "num_input_tokens_seen": 633008, "step": 735 }, { "epoch": 1.791767554479419, "grad_norm": 8.256195068359375, "learning_rate": 4.0549445768386895e-06, "loss": 0.1222, "num_input_tokens_seen": 637360, "step": 740 }, { "epoch": 1.8038740920096852, "grad_norm": 40.003360748291016, "learning_rate": 4.038339309269002e-06, "loss": 0.1171, "num_input_tokens_seen": 641648, "step": 745 }, { "epoch": 1.8159806295399514, "grad_norm": 1.1402125358581543, "learning_rate": 4.021624091027895e-06, "loss": 0.1638, "num_input_tokens_seen": 645552, "step": 750 }, { "epoch": 1.828087167070218, "grad_norm": 19.136348724365234, "learning_rate": 4.00480011681244e-06, "loss": 0.1092, "num_input_tokens_seen": 649904, "step": 755 }, { "epoch": 1.8401937046004844, "grad_norm": 16.968360900878906, "learning_rate": 3.987868589092894e-06, "loss": 0.1118, "num_input_tokens_seen": 654128, "step": 760 }, { "epoch": 1.8523002421307506, "grad_norm": 15.944784164428711, "learning_rate": 3.970830718026746e-06, "loss": 0.1015, "num_input_tokens_seen": 658672, "step": 765 }, { "epoch": 1.8644067796610169, "grad_norm": 20.015836715698242, "learning_rate": 3.9536877213722335e-06, "loss": 0.1207, "num_input_tokens_seen": 663088, "step": 770 }, { "epoch": 1.8765133171912833, "grad_norm": 30.892627716064453, "learning_rate": 3.936440824401299e-06, "loss": 0.083, "num_input_tokens_seen": 667440, "step": 775 }, { "epoch": 1.8886198547215496, "grad_norm": 19.33950424194336, "learning_rate": 3.919091259812013e-06, "loss": 0.0249, "num_input_tokens_seen": 671792, "step": 780 }, { "epoch": 1.900726392251816, "grad_norm": 24.891815185546875, "learning_rate": 3.901640267640475e-06, "loss": 0.0425, "num_input_tokens_seen": 676336, "step": 785 }, { "epoch": 1.9128329297820823, "grad_norm": 28.49574089050293, "learning_rate": 3.884089095172181e-06, "loss": 0.0402, "num_input_tokens_seen": 680624, "step": 790 }, { "epoch": 1.9249394673123486, "grad_norm": 0.016377810388803482, "learning_rate": 3.866438996852873e-06, "loss": 0.0155, "num_input_tokens_seen": 685040, "step": 795 }, { "epoch": 1.937046004842615, "grad_norm": 0.16267594695091248, "learning_rate": 3.848691234198879e-06, "loss": 0.0372, "num_input_tokens_seen": 689392, "step": 800 }, { "epoch": 1.9491525423728815, "grad_norm": 1.1569898128509521, "learning_rate": 3.830847075706957e-06, "loss": 0.1257, "num_input_tokens_seen": 693552, "step": 805 }, { "epoch": 1.9612590799031477, "grad_norm": 1.0089043378829956, "learning_rate": 3.812907796763616e-06, "loss": 0.0454, "num_input_tokens_seen": 698032, "step": 810 }, { "epoch": 1.973365617433414, "grad_norm": 0.9869695901870728, "learning_rate": 3.794874679553975e-06, "loss": 0.2136, "num_input_tokens_seen": 702000, "step": 815 }, { "epoch": 1.9854721549636802, "grad_norm": 23.745838165283203, "learning_rate": 3.7767490129701057e-06, "loss": 0.1643, "num_input_tokens_seen": 706160, "step": 820 }, { "epoch": 1.9975786924939467, "grad_norm": 4.922607898712158, "learning_rate": 3.7585320925189246e-06, "loss": 0.0475, "num_input_tokens_seen": 710768, "step": 825 }, { "epoch": 2.009685230024213, "grad_norm": 0.3702712655067444, "learning_rate": 3.7402252202295876e-06, "loss": 0.0011, "num_input_tokens_seen": 714744, "step": 830 }, { "epoch": 2.0145278450363198, "eval_loss": 0.2067757099866867, "eval_runtime": 0.634, "eval_samples_per_second": 578.871, "eval_steps_per_second": 72.556, "num_input_tokens_seen": 716344, "step": 832 }, { "epoch": 2.0217917675544794, "grad_norm": 66.86534118652344, "learning_rate": 3.7218297045604362e-06, "loss": 0.0057, "num_input_tokens_seen": 718776, "step": 835 }, { "epoch": 2.0338983050847457, "grad_norm": 0.06272957473993301, "learning_rate": 3.703346860305473e-06, "loss": 0.0114, "num_input_tokens_seen": 722744, "step": 840 }, { "epoch": 2.046004842615012, "grad_norm": 0.00682666152715683, "learning_rate": 3.6847780085003908e-06, "loss": 0.0047, "num_input_tokens_seen": 727160, "step": 845 }, { "epoch": 2.0581113801452786, "grad_norm": 60.80389404296875, "learning_rate": 3.666124476328155e-06, "loss": 0.0867, "num_input_tokens_seen": 731576, "step": 850 }, { "epoch": 2.070217917675545, "grad_norm": 15.216704368591309, "learning_rate": 3.647387597024139e-06, "loss": 0.0084, "num_input_tokens_seen": 736184, "step": 855 }, { "epoch": 2.082324455205811, "grad_norm": 0.1580037623643875, "learning_rate": 3.6285687097808396e-06, "loss": 0.0011, "num_input_tokens_seen": 740472, "step": 860 }, { "epoch": 2.0944309927360774, "grad_norm": 56.17728805541992, "learning_rate": 3.609669159652158e-06, "loss": 0.0528, "num_input_tokens_seen": 744760, "step": 865 }, { "epoch": 2.106537530266344, "grad_norm": 0.03734096884727478, "learning_rate": 3.5906902974572623e-06, "loss": 0.0003, "num_input_tokens_seen": 749176, "step": 870 }, { "epoch": 2.1186440677966103, "grad_norm": 0.013891610316932201, "learning_rate": 3.5716334796840403e-06, "loss": 0.0329, "num_input_tokens_seen": 753528, "step": 875 }, { "epoch": 2.1307506053268765, "grad_norm": 0.11342939734458923, "learning_rate": 3.5525000683921467e-06, "loss": 0.0022, "num_input_tokens_seen": 757688, "step": 880 }, { "epoch": 2.142857142857143, "grad_norm": 0.012256304733455181, "learning_rate": 3.533291431115653e-06, "loss": 0.0268, "num_input_tokens_seen": 762040, "step": 885 }, { "epoch": 2.154963680387409, "grad_norm": 55.32374572753906, "learning_rate": 3.514008940765304e-06, "loss": 0.0746, "num_input_tokens_seen": 766200, "step": 890 }, { "epoch": 2.1670702179176757, "grad_norm": 0.006180486176162958, "learning_rate": 3.494653975530388e-06, "loss": 0.0202, "num_input_tokens_seen": 770680, "step": 895 }, { "epoch": 2.179176755447942, "grad_norm": 0.8855127096176147, "learning_rate": 3.475227918780239e-06, "loss": 0.0023, "num_input_tokens_seen": 774840, "step": 900 }, { "epoch": 2.1912832929782082, "grad_norm": 0.0062964423559606075, "learning_rate": 3.455732158965356e-06, "loss": 0.0001, "num_input_tokens_seen": 779192, "step": 905 }, { "epoch": 2.2033898305084745, "grad_norm": 0.005455203354358673, "learning_rate": 3.436168089518168e-06, "loss": 0.0001, "num_input_tokens_seen": 783608, "step": 910 }, { "epoch": 2.2154963680387407, "grad_norm": 40.38529968261719, "learning_rate": 3.4165371087534428e-06, "loss": 0.0365, "num_input_tokens_seen": 788088, "step": 915 }, { "epoch": 2.2276029055690074, "grad_norm": 0.01571381278336048, "learning_rate": 3.396840619768338e-06, "loss": 0.0, "num_input_tokens_seen": 792568, "step": 920 }, { "epoch": 2.2397094430992737, "grad_norm": 0.002371912356466055, "learning_rate": 3.377080030342125e-06, "loss": 0.0001, "num_input_tokens_seen": 797176, "step": 925 }, { "epoch": 2.25181598062954, "grad_norm": 31.288312911987305, "learning_rate": 3.3572567528355614e-06, "loss": 0.0038, "num_input_tokens_seen": 801400, "step": 930 }, { "epoch": 2.263922518159806, "grad_norm": 0.009536500088870525, "learning_rate": 3.3373722040899515e-06, "loss": 0.0059, "num_input_tokens_seen": 805944, "step": 935 }, { "epoch": 2.2663438256658597, "eval_loss": 0.26913806796073914, "eval_runtime": 0.697, "eval_samples_per_second": 526.575, "eval_steps_per_second": 66.001, "num_input_tokens_seen": 806712, "step": 936 }, { "epoch": 2.2760290556900724, "grad_norm": 0.0007474619778804481, "learning_rate": 3.3174278053258753e-06, "loss": 0.0006, "num_input_tokens_seen": 810040, "step": 940 }, { "epoch": 2.288135593220339, "grad_norm": 42.942989349365234, "learning_rate": 3.2974249820416094e-06, "loss": 0.0482, "num_input_tokens_seen": 814392, "step": 945 }, { "epoch": 2.3002421307506054, "grad_norm": 62.229331970214844, "learning_rate": 3.2773651639112432e-06, "loss": 0.0175, "num_input_tokens_seen": 818872, "step": 950 }, { "epoch": 2.3123486682808716, "grad_norm": 0.0007935139001347125, "learning_rate": 3.2572497846824922e-06, "loss": 0.0039, "num_input_tokens_seen": 823096, "step": 955 }, { "epoch": 2.324455205811138, "grad_norm": 2.6152658462524414, "learning_rate": 3.2370802820742273e-06, "loss": 0.0549, "num_input_tokens_seen": 827128, "step": 960 }, { "epoch": 2.3365617433414045, "grad_norm": 0.9258336424827576, "learning_rate": 3.2168580976737105e-06, "loss": 0.0011, "num_input_tokens_seen": 831288, "step": 965 }, { "epoch": 2.348668280871671, "grad_norm": 0.28666022419929504, "learning_rate": 3.1965846768335625e-06, "loss": 0.0202, "num_input_tokens_seen": 835640, "step": 970 }, { "epoch": 2.360774818401937, "grad_norm": 0.0008573018712922931, "learning_rate": 3.176261468568457e-06, "loss": 0.0019, "num_input_tokens_seen": 839736, "step": 975 }, { "epoch": 2.3728813559322033, "grad_norm": 55.564476013183594, "learning_rate": 3.155889925451557e-06, "loss": 0.0363, "num_input_tokens_seen": 844024, "step": 980 }, { "epoch": 2.38498789346247, "grad_norm": 0.025824718177318573, "learning_rate": 3.1354715035106892e-06, "loss": 0.0001, "num_input_tokens_seen": 848248, "step": 985 }, { "epoch": 2.3970944309927362, "grad_norm": 0.0016786637715995312, "learning_rate": 3.115007662124282e-06, "loss": 0.0, "num_input_tokens_seen": 852472, "step": 990 }, { "epoch": 2.4092009685230025, "grad_norm": 0.014527814462780952, "learning_rate": 3.0944998639170544e-06, "loss": 0.0006, "num_input_tokens_seen": 856824, "step": 995 }, { "epoch": 2.4213075060532687, "grad_norm": 4.990849018096924, "learning_rate": 3.0739495746554785e-06, "loss": 0.0018, "num_input_tokens_seen": 860984, "step": 1000 }, { "epoch": 2.433414043583535, "grad_norm": 0.5683162808418274, "learning_rate": 3.0533582631430153e-06, "loss": 0.068, "num_input_tokens_seen": 865272, "step": 1005 }, { "epoch": 2.4455205811138017, "grad_norm": 5.13648796081543, "learning_rate": 3.0327274011151355e-06, "loss": 0.0395, "num_input_tokens_seen": 869560, "step": 1010 }, { "epoch": 2.457627118644068, "grad_norm": 0.0019914067815989256, "learning_rate": 3.012058463134126e-06, "loss": 0.0, "num_input_tokens_seen": 873976, "step": 1015 }, { "epoch": 2.469733656174334, "grad_norm": 0.0011213916586712003, "learning_rate": 2.991352926483702e-06, "loss": 0.0, "num_input_tokens_seen": 878200, "step": 1020 }, { "epoch": 2.4818401937046004, "grad_norm": 0.0030807037837803364, "learning_rate": 2.9706122710634166e-06, "loss": 0.0008, "num_input_tokens_seen": 882872, "step": 1025 }, { "epoch": 2.4939467312348667, "grad_norm": 0.0031619234941899776, "learning_rate": 2.949837979282889e-06, "loss": 0.0, "num_input_tokens_seen": 887096, "step": 1030 }, { "epoch": 2.5060532687651333, "grad_norm": 29.275707244873047, "learning_rate": 2.9290315359558504e-06, "loss": 0.0032, "num_input_tokens_seen": 891576, "step": 1035 }, { "epoch": 2.5181598062953996, "grad_norm": 0.013175534084439278, "learning_rate": 2.908194428194019e-06, "loss": 0.0756, "num_input_tokens_seen": 895736, "step": 1040 }, { "epoch": 2.5181598062953996, "eval_loss": 0.28947436809539795, "eval_runtime": 0.636, "eval_samples_per_second": 577.029, "eval_steps_per_second": 72.325, "num_input_tokens_seen": 895736, "step": 1040 }, { "epoch": 2.530266343825666, "grad_norm": 0.002881130203604698, "learning_rate": 2.88732814530081e-06, "loss": 0.0001, "num_input_tokens_seen": 900024, "step": 1045 }, { "epoch": 2.542372881355932, "grad_norm": 122.53758239746094, "learning_rate": 2.8664341786648932e-06, "loss": 0.0128, "num_input_tokens_seen": 904440, "step": 1050 }, { "epoch": 2.5544794188861983, "grad_norm": 0.0015194405568763614, "learning_rate": 2.845514021653595e-06, "loss": 0.0001, "num_input_tokens_seen": 908728, "step": 1055 }, { "epoch": 2.566585956416465, "grad_norm": 0.13878583908081055, "learning_rate": 2.8245691695061605e-06, "loss": 0.0443, "num_input_tokens_seen": 913016, "step": 1060 }, { "epoch": 2.5786924939467313, "grad_norm": 26.11018180847168, "learning_rate": 2.8036011192268863e-06, "loss": 0.0032, "num_input_tokens_seen": 917304, "step": 1065 }, { "epoch": 2.5907990314769975, "grad_norm": 0.003500137245282531, "learning_rate": 2.7826113694781254e-06, "loss": 0.0001, "num_input_tokens_seen": 921528, "step": 1070 }, { "epoch": 2.6029055690072638, "grad_norm": 0.00933838915079832, "learning_rate": 2.7616014204731683e-06, "loss": 0.0, "num_input_tokens_seen": 925944, "step": 1075 }, { "epoch": 2.61501210653753, "grad_norm": 0.0023421638179570436, "learning_rate": 2.7405727738690193e-06, "loss": 0.0001, "num_input_tokens_seen": 930744, "step": 1080 }, { "epoch": 2.6271186440677967, "grad_norm": 0.009054594673216343, "learning_rate": 2.7195269326590685e-06, "loss": 0.0725, "num_input_tokens_seen": 935352, "step": 1085 }, { "epoch": 2.639225181598063, "grad_norm": 0.02030082233250141, "learning_rate": 2.698465401065667e-06, "loss": 0.0295, "num_input_tokens_seen": 939640, "step": 1090 }, { "epoch": 2.651331719128329, "grad_norm": 0.27530986070632935, "learning_rate": 2.6773896844326126e-06, "loss": 0.0001, "num_input_tokens_seen": 943672, "step": 1095 }, { "epoch": 2.663438256658596, "grad_norm": 0.03071773052215576, "learning_rate": 2.656301289117561e-06, "loss": 0.0001, "num_input_tokens_seen": 947704, "step": 1100 }, { "epoch": 2.6755447941888617, "grad_norm": 0.014886329881846905, "learning_rate": 2.6352017223843584e-06, "loss": 0.0196, "num_input_tokens_seen": 951928, "step": 1105 }, { "epoch": 2.6876513317191284, "grad_norm": 0.06265253573656082, "learning_rate": 2.6140924922953125e-06, "loss": 0.0294, "num_input_tokens_seen": 956216, "step": 1110 }, { "epoch": 2.6997578692493946, "grad_norm": 0.04291946068406105, "learning_rate": 2.592975107603406e-06, "loss": 0.0001, "num_input_tokens_seen": 960504, "step": 1115 }, { "epoch": 2.711864406779661, "grad_norm": 0.01645965874195099, "learning_rate": 2.571851077644461e-06, "loss": 0.0135, "num_input_tokens_seen": 965048, "step": 1120 }, { "epoch": 2.7239709443099276, "grad_norm": 0.007254968397319317, "learning_rate": 2.55072191222926e-06, "loss": 0.0001, "num_input_tokens_seen": 969208, "step": 1125 }, { "epoch": 2.736077481840194, "grad_norm": 0.114701047539711, "learning_rate": 2.5295891215356362e-06, "loss": 0.0991, "num_input_tokens_seen": 973624, "step": 1130 }, { "epoch": 2.74818401937046, "grad_norm": 0.0054718079045414925, "learning_rate": 2.5084542160005338e-06, "loss": 0.0064, "num_input_tokens_seen": 977976, "step": 1135 }, { "epoch": 2.7602905569007263, "grad_norm": 0.28143173456192017, "learning_rate": 2.4873187062120515e-06, "loss": 0.0001, "num_input_tokens_seen": 982200, "step": 1140 }, { "epoch": 2.7699757869249395, "eval_loss": 0.22601255774497986, "eval_runtime": 0.6803, "eval_samples_per_second": 539.434, "eval_steps_per_second": 67.613, "num_input_tokens_seen": 985592, "step": 1144 }, { "epoch": 2.7723970944309926, "grad_norm": 0.02263481356203556, "learning_rate": 2.4661841028014786e-06, "loss": 0.0002, "num_input_tokens_seen": 986488, "step": 1145 }, { "epoch": 2.7845036319612593, "grad_norm": 0.00884742010384798, "learning_rate": 2.445051916335321e-06, "loss": 0.0002, "num_input_tokens_seen": 990456, "step": 1150 }, { "epoch": 2.7966101694915255, "grad_norm": 0.019946428015828133, "learning_rate": 2.4239236572073354e-06, "loss": 0.0766, "num_input_tokens_seen": 994744, "step": 1155 }, { "epoch": 2.8087167070217918, "grad_norm": 0.010249280370771885, "learning_rate": 2.4028008355305817e-06, "loss": 0.0501, "num_input_tokens_seen": 999160, "step": 1160 }, { "epoch": 2.820823244552058, "grad_norm": 56.294708251953125, "learning_rate": 2.3816849610294784e-06, "loss": 0.0289, "num_input_tokens_seen": 1003256, "step": 1165 }, { "epoch": 2.8329297820823243, "grad_norm": 0.007834532298147678, "learning_rate": 2.3605775429319115e-06, "loss": 0.0884, "num_input_tokens_seen": 1007480, "step": 1170 }, { "epoch": 2.845036319612591, "grad_norm": 0.02283744513988495, "learning_rate": 2.3394800898613536e-06, "loss": 0.0004, "num_input_tokens_seen": 1011896, "step": 1175 }, { "epoch": 2.857142857142857, "grad_norm": 0.05437995120882988, "learning_rate": 2.318394109729041e-06, "loss": 0.0004, "num_input_tokens_seen": 1015992, "step": 1180 }, { "epoch": 2.8692493946731235, "grad_norm": 0.41270506381988525, "learning_rate": 2.297321109626198e-06, "loss": 0.003, "num_input_tokens_seen": 1020408, "step": 1185 }, { "epoch": 2.8813559322033897, "grad_norm": 0.04989304393529892, "learning_rate": 2.27626259571632e-06, "loss": 0.0003, "num_input_tokens_seen": 1025016, "step": 1190 }, { "epoch": 2.893462469733656, "grad_norm": 0.09522224217653275, "learning_rate": 2.2552200731275215e-06, "loss": 0.0571, "num_input_tokens_seen": 1029368, "step": 1195 }, { "epoch": 2.9055690072639226, "grad_norm": 3.600926160812378, "learning_rate": 2.2341950458449576e-06, "loss": 0.0007, "num_input_tokens_seen": 1033592, "step": 1200 }, { "epoch": 2.917675544794189, "grad_norm": 0.005034204572439194, "learning_rate": 2.2131890166033333e-06, "loss": 0.0001, "num_input_tokens_seen": 1037688, "step": 1205 }, { "epoch": 2.929782082324455, "grad_norm": 0.01115792989730835, "learning_rate": 2.1922034867794923e-06, "loss": 0.0136, "num_input_tokens_seen": 1041912, "step": 1210 }, { "epoch": 2.9418886198547214, "grad_norm": 0.01456889882683754, "learning_rate": 2.171239956285115e-06, "loss": 0.0003, "num_input_tokens_seen": 1046392, "step": 1215 }, { "epoch": 2.9539951573849876, "grad_norm": 0.009599031880497932, "learning_rate": 2.150299923459505e-06, "loss": 0.0001, "num_input_tokens_seen": 1050616, "step": 1220 }, { "epoch": 2.9661016949152543, "grad_norm": 0.04260379076004028, "learning_rate": 2.1293848849625065e-06, "loss": 0.0001, "num_input_tokens_seen": 1054840, "step": 1225 }, { "epoch": 2.9782082324455206, "grad_norm": 0.006233742460608482, "learning_rate": 2.108496335667527e-06, "loss": 0.0001, "num_input_tokens_seen": 1058936, "step": 1230 }, { "epoch": 2.990314769975787, "grad_norm": 0.005781834479421377, "learning_rate": 2.0876357685546942e-06, "loss": 0.0001, "num_input_tokens_seen": 1063288, "step": 1235 }, { "epoch": 3.002421307506053, "grad_norm": 0.004296495113521814, "learning_rate": 2.0668046746041497e-06, "loss": 0.0, "num_input_tokens_seen": 1067392, "step": 1240 }, { "epoch": 3.0145278450363198, "grad_norm": 0.05696773901581764, "learning_rate": 2.0460045426894816e-06, "loss": 0.0, "num_input_tokens_seen": 1071872, "step": 1245 }, { "epoch": 3.0217917675544794, "eval_loss": 0.22526989877223969, "eval_runtime": 0.6509, "eval_samples_per_second": 563.818, "eval_steps_per_second": 70.669, "num_input_tokens_seen": 1074624, "step": 1248 }, { "epoch": 3.026634382566586, "grad_norm": 0.033577512949705124, "learning_rate": 2.0252368594713083e-06, "loss": 0.0, "num_input_tokens_seen": 1076416, "step": 1250 }, { "epoch": 3.0387409200968523, "grad_norm": 8.000839233398438, "learning_rate": 2.004503109291023e-06, "loss": 0.0024, "num_input_tokens_seen": 1080512, "step": 1255 }, { "epoch": 3.0508474576271185, "grad_norm": 0.0029002963565289974, "learning_rate": 1.9838047740647024e-06, "loss": 0.0, "num_input_tokens_seen": 1084608, "step": 1260 }, { "epoch": 3.062953995157385, "grad_norm": 0.0019197538495063782, "learning_rate": 1.9631433331771886e-06, "loss": 0.0, "num_input_tokens_seen": 1089024, "step": 1265 }, { "epoch": 3.0750605326876514, "grad_norm": 0.1197233721613884, "learning_rate": 1.942520263376351e-06, "loss": 0.0, "num_input_tokens_seen": 1093376, "step": 1270 }, { "epoch": 3.0871670702179177, "grad_norm": 0.005993293132632971, "learning_rate": 1.921937038667539e-06, "loss": 0.0, "num_input_tokens_seen": 1097728, "step": 1275 }, { "epoch": 3.099273607748184, "grad_norm": 8.694519996643066, "learning_rate": 1.901395130208229e-06, "loss": 0.0831, "num_input_tokens_seen": 1101888, "step": 1280 }, { "epoch": 3.11138014527845, "grad_norm": 0.007167529780417681, "learning_rate": 1.880896006202876e-06, "loss": 0.0, "num_input_tokens_seen": 1106176, "step": 1285 }, { "epoch": 3.123486682808717, "grad_norm": 0.014599725604057312, "learning_rate": 1.860441131797977e-06, "loss": 0.0, "num_input_tokens_seen": 1110272, "step": 1290 }, { "epoch": 3.135593220338983, "grad_norm": 0.016075119376182556, "learning_rate": 1.8400319689773474e-06, "loss": 0.0001, "num_input_tokens_seen": 1114496, "step": 1295 }, { "epoch": 3.1476997578692494, "grad_norm": 0.04632039740681648, "learning_rate": 1.8196699764576316e-06, "loss": 0.0001, "num_input_tokens_seen": 1118784, "step": 1300 }, { "epoch": 3.1598062953995156, "grad_norm": 0.014973443932831287, "learning_rate": 1.7993566095840442e-06, "loss": 0.0001, "num_input_tokens_seen": 1123008, "step": 1305 }, { "epoch": 3.171912832929782, "grad_norm": 0.017391176894307137, "learning_rate": 1.7790933202263437e-06, "loss": 0.0001, "num_input_tokens_seen": 1127424, "step": 1310 }, { "epoch": 3.1840193704600486, "grad_norm": 0.0033686573151499033, "learning_rate": 1.7588815566750728e-06, "loss": 0.0001, "num_input_tokens_seen": 1131840, "step": 1315 }, { "epoch": 3.196125907990315, "grad_norm": 0.012609965167939663, "learning_rate": 1.7387227635380362e-06, "loss": 0.0001, "num_input_tokens_seen": 1136192, "step": 1320 }, { "epoch": 3.208232445520581, "grad_norm": 0.12501056492328644, "learning_rate": 1.7186183816370522e-06, "loss": 0.0001, "num_input_tokens_seen": 1140544, "step": 1325 }, { "epoch": 3.2203389830508473, "grad_norm": 0.0041845571249723434, "learning_rate": 1.6985698479049703e-06, "loss": 0.0001, "num_input_tokens_seen": 1145280, "step": 1330 }, { "epoch": 3.232445520581114, "grad_norm": 0.0181956198066473, "learning_rate": 1.6785785952829718e-06, "loss": 0.0039, "num_input_tokens_seen": 1149888, "step": 1335 }, { "epoch": 3.2445520581113803, "grad_norm": 0.005230342503637075, "learning_rate": 1.6586460526181476e-06, "loss": 0.0, "num_input_tokens_seen": 1153920, "step": 1340 }, { "epoch": 3.2566585956416465, "grad_norm": 0.0063026342540979385, "learning_rate": 1.6387736445613772e-06, "loss": 0.0, "num_input_tokens_seen": 1158592, "step": 1345 }, { "epoch": 3.2687651331719128, "grad_norm": 0.004648419097065926, "learning_rate": 1.618962791465501e-06, "loss": 0.0, "num_input_tokens_seen": 1162816, "step": 1350 }, { "epoch": 3.2736077481840193, "eval_loss": 0.25782492756843567, "eval_runtime": 0.6532, "eval_samples_per_second": 561.843, "eval_steps_per_second": 70.422, "num_input_tokens_seen": 1164544, "step": 1352 }, { "epoch": 3.280871670702179, "grad_norm": 0.003744626184925437, "learning_rate": 1.599214909283805e-06, "loss": 0.0002, "num_input_tokens_seen": 1167232, "step": 1355 }, { "epoch": 3.2929782082324457, "grad_norm": 0.002472149208188057, "learning_rate": 1.579531409468815e-06, "loss": 0.0, "num_input_tokens_seen": 1171648, "step": 1360 }, { "epoch": 3.305084745762712, "grad_norm": 0.00250844843685627, "learning_rate": 1.5599136988714186e-06, "loss": 0.0, "num_input_tokens_seen": 1175808, "step": 1365 }, { "epoch": 3.317191283292978, "grad_norm": 0.0014152796939015388, "learning_rate": 1.5403631796403085e-06, "loss": 0.0, "num_input_tokens_seen": 1180224, "step": 1370 }, { "epoch": 3.3292978208232444, "grad_norm": 0.00422420259565115, "learning_rate": 1.5208812491217669e-06, "loss": 0.0, "num_input_tokens_seen": 1184704, "step": 1375 }, { "epoch": 3.341404358353511, "grad_norm": 0.004841359332203865, "learning_rate": 1.5014692997597962e-06, "loss": 0.053, "num_input_tokens_seen": 1188992, "step": 1380 }, { "epoch": 3.3535108958837774, "grad_norm": 0.0019620037637650967, "learning_rate": 1.4821287189965865e-06, "loss": 0.0, "num_input_tokens_seen": 1193408, "step": 1385 }, { "epoch": 3.3656174334140436, "grad_norm": 0.0014666810166090727, "learning_rate": 1.4628608891733626e-06, "loss": 0.0002, "num_input_tokens_seen": 1197760, "step": 1390 }, { "epoch": 3.37772397094431, "grad_norm": 0.0015612264396622777, "learning_rate": 1.443667187431572e-06, "loss": 0.0, "num_input_tokens_seen": 1201792, "step": 1395 }, { "epoch": 3.389830508474576, "grad_norm": 0.03145941346883774, "learning_rate": 1.4245489856144633e-06, "loss": 0.0, "num_input_tokens_seen": 1205824, "step": 1400 }, { "epoch": 3.401937046004843, "grad_norm": 0.0026610263157635927, "learning_rate": 1.4055076501690313e-06, "loss": 0.0, "num_input_tokens_seen": 1210240, "step": 1405 }, { "epoch": 3.414043583535109, "grad_norm": 0.002824948402121663, "learning_rate": 1.3865445420483524e-06, "loss": 0.0, "num_input_tokens_seen": 1214464, "step": 1410 }, { "epoch": 3.4261501210653753, "grad_norm": 3.925107002258301, "learning_rate": 1.367661016614315e-06, "loss": 0.0005, "num_input_tokens_seen": 1218752, "step": 1415 }, { "epoch": 3.4382566585956416, "grad_norm": 0.0029422210063785315, "learning_rate": 1.348858423540744e-06, "loss": 0.0, "num_input_tokens_seen": 1223168, "step": 1420 }, { "epoch": 3.450363196125908, "grad_norm": 0.00631983857601881, "learning_rate": 1.3301381067169367e-06, "loss": 0.0, "num_input_tokens_seen": 1227328, "step": 1425 }, { "epoch": 3.4624697336561745, "grad_norm": 0.0015673706075176597, "learning_rate": 1.3115014041516088e-06, "loss": 0.0, "num_input_tokens_seen": 1231360, "step": 1430 }, { "epoch": 3.4745762711864407, "grad_norm": 0.001485335873439908, "learning_rate": 1.2929496478772635e-06, "loss": 0.0, "num_input_tokens_seen": 1235456, "step": 1435 }, { "epoch": 3.486682808716707, "grad_norm": 0.001463928259909153, "learning_rate": 1.2744841638549843e-06, "loss": 0.0, "num_input_tokens_seen": 1239616, "step": 1440 }, { "epoch": 3.4987893462469732, "grad_norm": 0.013075617142021656, "learning_rate": 1.2561062718796663e-06, "loss": 0.0, "num_input_tokens_seen": 1243968, "step": 1445 }, { "epoch": 3.5108958837772395, "grad_norm": 0.00566933723166585, "learning_rate": 1.2378172854856831e-06, "loss": 0.0001, "num_input_tokens_seen": 1248128, "step": 1450 }, { "epoch": 3.523002421307506, "grad_norm": 0.004368732683360577, "learning_rate": 1.2196185118530063e-06, "loss": 0.0, "num_input_tokens_seen": 1252288, "step": 1455 }, { "epoch": 3.5254237288135593, "eval_loss": 0.2580437958240509, "eval_runtime": 0.6456, "eval_samples_per_second": 568.425, "eval_steps_per_second": 71.247, "num_input_tokens_seen": 1253248, "step": 1456 }, { "epoch": 3.5351089588377724, "grad_norm": 0.0046086618676781654, "learning_rate": 1.2015112517137744e-06, "loss": 0.0, "num_input_tokens_seen": 1256640, "step": 1460 }, { "epoch": 3.5472154963680387, "grad_norm": 0.0021092540118843317, "learning_rate": 1.183496799259326e-06, "loss": 0.0, "num_input_tokens_seen": 1261440, "step": 1465 }, { "epoch": 3.559322033898305, "grad_norm": 0.0011951872147619724, "learning_rate": 1.165576442047699e-06, "loss": 0.0, "num_input_tokens_seen": 1265664, "step": 1470 }, { "epoch": 3.571428571428571, "grad_norm": 0.003473962191492319, "learning_rate": 1.147751460911604e-06, "loss": 0.0, "num_input_tokens_seen": 1270016, "step": 1475 }, { "epoch": 3.583535108958838, "grad_norm": 0.0040196748450398445, "learning_rate": 1.1300231298668786e-06, "loss": 0.0, "num_input_tokens_seen": 1274560, "step": 1480 }, { "epoch": 3.595641646489104, "grad_norm": 0.0007644465658813715, "learning_rate": 1.112392716021429e-06, "loss": 0.0, "num_input_tokens_seen": 1278976, "step": 1485 }, { "epoch": 3.6077481840193704, "grad_norm": 0.00214450154453516, "learning_rate": 1.0948614794846668e-06, "loss": 0.0, "num_input_tokens_seen": 1283200, "step": 1490 }, { "epoch": 3.619854721549637, "grad_norm": 0.004525753669440746, "learning_rate": 1.0774306732774414e-06, "loss": 0.0, "num_input_tokens_seen": 1287296, "step": 1495 }, { "epoch": 3.6319612590799033, "grad_norm": 0.0008026693249121308, "learning_rate": 1.0601015432424818e-06, "loss": 0.0, "num_input_tokens_seen": 1291712, "step": 1500 }, { "epoch": 3.6440677966101696, "grad_norm": 0.0020796298049390316, "learning_rate": 1.0428753279553561e-06, "loss": 0.0328, "num_input_tokens_seen": 1295936, "step": 1505 }, { "epoch": 3.656174334140436, "grad_norm": 16.17505645751953, "learning_rate": 1.0257532586359422e-06, "loss": 0.0527, "num_input_tokens_seen": 1300608, "step": 1510 }, { "epoch": 3.668280871670702, "grad_norm": 0.005263039376586676, "learning_rate": 1.008736559060429e-06, "loss": 0.0, "num_input_tokens_seen": 1305024, "step": 1515 }, { "epoch": 3.6803874092009687, "grad_norm": 0.015520356595516205, "learning_rate": 9.918264454738504e-07, "loss": 0.0001, "num_input_tokens_seen": 1309376, "step": 1520 }, { "epoch": 3.692493946731235, "grad_norm": 0.01658783107995987, "learning_rate": 9.750241265031529e-07, "loss": 0.0001, "num_input_tokens_seen": 1313664, "step": 1525 }, { "epoch": 3.7046004842615012, "grad_norm": 0.0006883519235998392, "learning_rate": 9.583308030708135e-07, "loss": 0.0001, "num_input_tokens_seen": 1318080, "step": 1530 }, { "epoch": 3.7167070217917675, "grad_norm": 0.0012563606724143028, "learning_rate": 9.417476683090007e-07, "loss": 0.0001, "num_input_tokens_seen": 1322432, "step": 1535 }, { "epoch": 3.7288135593220337, "grad_norm": 0.0474010594189167, "learning_rate": 9.252759074743034e-07, "loss": 0.0003, "num_input_tokens_seen": 1326848, "step": 1540 }, { "epoch": 3.7409200968523004, "grad_norm": 0.012124676257371902, "learning_rate": 9.08916697863014e-07, "loss": 0.0, "num_input_tokens_seen": 1331328, "step": 1545 }, { "epoch": 3.7530266343825667, "grad_norm": 0.0076986453495919704, "learning_rate": 8.926712087269801e-07, "loss": 0.0, "num_input_tokens_seen": 1335424, "step": 1550 }, { "epoch": 3.765133171912833, "grad_norm": 0.004324712324887514, "learning_rate": 8.765406011900368e-07, "loss": 0.0, "num_input_tokens_seen": 1339712, "step": 1555 }, { "epoch": 3.777239709443099, "grad_norm": 0.007034031208604574, "learning_rate": 8.605260281650152e-07, "loss": 0.0, "num_input_tokens_seen": 1344000, "step": 1560 }, { "epoch": 3.777239709443099, "eval_loss": 0.2703007757663727, "eval_runtime": 0.6408, "eval_samples_per_second": 572.737, "eval_steps_per_second": 71.787, "num_input_tokens_seen": 1344000, "step": 1560 }, { "epoch": 3.7893462469733654, "grad_norm": 0.00895662046968937, "learning_rate": 8.44628634271342e-07, "loss": 0.0, "num_input_tokens_seen": 1348224, "step": 1565 }, { "epoch": 3.801452784503632, "grad_norm": 0.002478554379194975, "learning_rate": 8.288495557532241e-07, "loss": 0.0017, "num_input_tokens_seen": 1352576, "step": 1570 }, { "epoch": 3.8135593220338984, "grad_norm": 11.973031997680664, "learning_rate": 8.131899203984464e-07, "loss": 0.0616, "num_input_tokens_seen": 1356864, "step": 1575 }, { "epoch": 3.8256658595641646, "grad_norm": 0.0017213321989402175, "learning_rate": 7.976508474577549e-07, "loss": 0.0, "num_input_tokens_seen": 1361152, "step": 1580 }, { "epoch": 3.837772397094431, "grad_norm": 0.0008722307975403965, "learning_rate": 7.822334475648655e-07, "loss": 0.0, "num_input_tokens_seen": 1365376, "step": 1585 }, { "epoch": 3.849878934624697, "grad_norm": 0.0017557705286890268, "learning_rate": 7.66938822657081e-07, "loss": 0.0, "num_input_tokens_seen": 1369728, "step": 1590 }, { "epoch": 3.861985472154964, "grad_norm": 0.0016538921045139432, "learning_rate": 7.517680658965328e-07, "loss": 0.0, "num_input_tokens_seen": 1374144, "step": 1595 }, { "epoch": 3.87409200968523, "grad_norm": 0.0021663156803697348, "learning_rate": 7.367222615920477e-07, "loss": 0.0, "num_input_tokens_seen": 1378368, "step": 1600 }, { "epoch": 3.8861985472154963, "grad_norm": 0.001977034378796816, "learning_rate": 7.21802485121649e-07, "loss": 0.0, "num_input_tokens_seen": 1382464, "step": 1605 }, { "epoch": 3.898305084745763, "grad_norm": 0.012750508263707161, "learning_rate": 7.070098028556949e-07, "loss": 0.0, "num_input_tokens_seen": 1386880, "step": 1610 }, { "epoch": 3.910411622276029, "grad_norm": 0.004628063179552555, "learning_rate": 6.923452720806612e-07, "loss": 0.0, "num_input_tokens_seen": 1391296, "step": 1615 }, { "epoch": 3.9225181598062955, "grad_norm": 0.0018564671045169234, "learning_rate": 6.778099409235739e-07, "loss": 0.0, "num_input_tokens_seen": 1395456, "step": 1620 }, { "epoch": 3.9346246973365617, "grad_norm": 0.0023303565103560686, "learning_rate": 6.634048482770946e-07, "loss": 0.0, "num_input_tokens_seen": 1399616, "step": 1625 }, { "epoch": 3.946731234866828, "grad_norm": 0.0011153841624036431, "learning_rate": 6.491310237252679e-07, "loss": 0.0, "num_input_tokens_seen": 1403712, "step": 1630 }, { "epoch": 3.9588377723970947, "grad_norm": 0.0012228480773046613, "learning_rate": 6.349894874699345e-07, "loss": 0.0, "num_input_tokens_seen": 1408128, "step": 1635 }, { "epoch": 3.970944309927361, "grad_norm": 0.0019976331386715174, "learning_rate": 6.209812502578113e-07, "loss": 0.0, "num_input_tokens_seen": 1412480, "step": 1640 }, { "epoch": 3.983050847457627, "grad_norm": 0.0029911224264651537, "learning_rate": 6.071073133082492e-07, "loss": 0.0, "num_input_tokens_seen": 1416704, "step": 1645 }, { "epoch": 3.9951573849878934, "grad_norm": 0.0036039084661751986, "learning_rate": 5.933686682416759e-07, "loss": 0.0, "num_input_tokens_seen": 1421120, "step": 1650 }, { "epoch": 4.00726392251816, "grad_norm": 0.00185173109639436, "learning_rate": 5.797662970087184e-07, "loss": 0.0, "num_input_tokens_seen": 1424944, "step": 1655 }, { "epoch": 4.019370460048426, "grad_norm": 0.002781663788482547, "learning_rate": 5.663011718200201e-07, "loss": 0.0, "num_input_tokens_seen": 1429296, "step": 1660 }, { "epoch": 4.0290556900726395, "eval_loss": 0.2501881718635559, "eval_runtime": 0.6431, "eval_samples_per_second": 570.66, "eval_steps_per_second": 71.527, "num_input_tokens_seen": 1432880, "step": 1664 }, { "epoch": 4.031476997578692, "grad_norm": 0.0033902269788086414, "learning_rate": 5.529742550767545e-07, "loss": 0.0, "num_input_tokens_seen": 1433776, "step": 1665 }, { "epoch": 4.043583535108959, "grad_norm": 0.005406382493674755, "learning_rate": 5.397864993018367e-07, "loss": 0.0, "num_input_tokens_seen": 1438000, "step": 1670 }, { "epoch": 4.0556900726392255, "grad_norm": 0.0016813945258036256, "learning_rate": 5.267388470718449e-07, "loss": 0.0, "num_input_tokens_seen": 1442352, "step": 1675 }, { "epoch": 4.067796610169491, "grad_norm": 0.0014291841071099043, "learning_rate": 5.138322309496504e-07, "loss": 0.0, "num_input_tokens_seen": 1446704, "step": 1680 }, { "epoch": 4.079903147699758, "grad_norm": 0.0015732025494799018, "learning_rate": 5.010675734177631e-07, "loss": 0.0, "num_input_tokens_seen": 1450864, "step": 1685 }, { "epoch": 4.092009685230024, "grad_norm": 0.0017645714106038213, "learning_rate": 4.884457868124001e-07, "loss": 0.0, "num_input_tokens_seen": 1455088, "step": 1690 }, { "epoch": 4.1041162227602905, "grad_norm": 0.0008963189902715385, "learning_rate": 4.759677732582782e-07, "loss": 0.0051, "num_input_tokens_seen": 1459376, "step": 1695 }, { "epoch": 4.116222760290557, "grad_norm": 0.005383517127484083, "learning_rate": 4.6363442460413215e-07, "loss": 0.0, "num_input_tokens_seen": 1463600, "step": 1700 }, { "epoch": 4.128329297820823, "grad_norm": 0.001712340977974236, "learning_rate": 4.514466223589753e-07, "loss": 0.0, "num_input_tokens_seen": 1468080, "step": 1705 }, { "epoch": 4.14043583535109, "grad_norm": 0.0034400331787765026, "learning_rate": 4.394052376290914e-07, "loss": 0.0253, "num_input_tokens_seen": 1472624, "step": 1710 }, { "epoch": 4.1525423728813555, "grad_norm": 0.0054069641046226025, "learning_rate": 4.2751113105577587e-07, "loss": 0.0, "num_input_tokens_seen": 1477040, "step": 1715 }, { "epoch": 4.164648910411622, "grad_norm": 0.01111331395804882, "learning_rate": 4.157651527538223e-07, "loss": 0.0, "num_input_tokens_seen": 1481328, "step": 1720 }, { "epoch": 4.176755447941889, "grad_norm": 0.0013017026940360665, "learning_rate": 4.041681422507604e-07, "loss": 0.0, "num_input_tokens_seen": 1485808, "step": 1725 }, { "epoch": 4.188861985472155, "grad_norm": 0.002490431070327759, "learning_rate": 3.927209284268535e-07, "loss": 0.0, "num_input_tokens_seen": 1490160, "step": 1730 }, { "epoch": 4.200968523002421, "grad_norm": 0.008939598686993122, "learning_rate": 3.8142432945585425e-07, "loss": 0.0, "num_input_tokens_seen": 1494512, "step": 1735 }, { "epoch": 4.213075060532688, "grad_norm": 0.002455121139064431, "learning_rate": 3.702791527465274e-07, "loss": 0.0, "num_input_tokens_seen": 1498480, "step": 1740 }, { "epoch": 4.225181598062954, "grad_norm": 0.0009606365929357708, "learning_rate": 3.592861948849416e-07, "loss": 0.0, "num_input_tokens_seen": 1502768, "step": 1745 }, { "epoch": 4.237288135593221, "grad_norm": 0.0018587886588647962, "learning_rate": 3.484462415775333e-07, "loss": 0.0, "num_input_tokens_seen": 1506992, "step": 1750 }, { "epoch": 4.249394673123486, "grad_norm": 0.003451196476817131, "learning_rate": 3.377600675949527e-07, "loss": 0.0, "num_input_tokens_seen": 1511472, "step": 1755 }, { "epoch": 4.261501210653753, "grad_norm": 0.002900092862546444, "learning_rate": 3.272284367166825e-07, "loss": 0.0, "num_input_tokens_seen": 1515824, "step": 1760 }, { "epoch": 4.27360774818402, "grad_norm": 0.005786838009953499, "learning_rate": 3.1685210167645336e-07, "loss": 0.0001, "num_input_tokens_seen": 1520176, "step": 1765 }, { "epoch": 4.280871670702179, "eval_loss": 0.25040701031684875, "eval_runtime": 2.3855, "eval_samples_per_second": 153.845, "eval_steps_per_second": 19.283, "num_input_tokens_seen": 1522544, "step": 1768 }, { "epoch": 4.285714285714286, "grad_norm": 0.0017267963849008083, "learning_rate": 3.066318041084398e-07, "loss": 0.0018, "num_input_tokens_seen": 1524336, "step": 1770 }, { "epoch": 4.297820823244552, "grad_norm": 0.011876060627400875, "learning_rate": 2.9656827449425495e-07, "loss": 0.0, "num_input_tokens_seen": 1528560, "step": 1775 }, { "epoch": 4.309927360774818, "grad_norm": 0.005988541524857283, "learning_rate": 2.86662232110739e-07, "loss": 0.026, "num_input_tokens_seen": 1532720, "step": 1780 }, { "epoch": 4.322033898305085, "grad_norm": 0.003772641299292445, "learning_rate": 2.769143849785513e-07, "loss": 0.0, "num_input_tokens_seen": 1536944, "step": 1785 }, { "epoch": 4.3341404358353515, "grad_norm": 0.0012762263650074601, "learning_rate": 2.673254298115646e-07, "loss": 0.0, "num_input_tokens_seen": 1541168, "step": 1790 }, { "epoch": 4.346246973365617, "grad_norm": 0.002060825005173683, "learning_rate": 2.5789605196706675e-07, "loss": 0.0, "num_input_tokens_seen": 1545456, "step": 1795 }, { "epoch": 4.358353510895884, "grad_norm": 0.002298081526532769, "learning_rate": 2.4862692539677907e-07, "loss": 0.0, "num_input_tokens_seen": 1549872, "step": 1800 }, { "epoch": 4.37046004842615, "grad_norm": 0.0016389107331633568, "learning_rate": 2.39518712598685e-07, "loss": 0.0, "num_input_tokens_seen": 1554288, "step": 1805 }, { "epoch": 4.3825665859564165, "grad_norm": 0.0016518625197932124, "learning_rate": 2.3057206456967908e-07, "loss": 0.0, "num_input_tokens_seen": 1558384, "step": 1810 }, { "epoch": 4.394673123486683, "grad_norm": 0.003522375365719199, "learning_rate": 2.2178762075903747e-07, "loss": 0.0, "num_input_tokens_seen": 1562544, "step": 1815 }, { "epoch": 4.406779661016949, "grad_norm": 0.001208233181387186, "learning_rate": 2.131660090227139e-07, "loss": 0.0, "num_input_tokens_seen": 1567216, "step": 1820 }, { "epoch": 4.418886198547216, "grad_norm": 0.0012071267701685429, "learning_rate": 2.0470784557846652e-07, "loss": 0.0, "num_input_tokens_seen": 1571568, "step": 1825 }, { "epoch": 4.4309927360774815, "grad_norm": 0.0016566209960728884, "learning_rate": 1.9641373496181143e-07, "loss": 0.0, "num_input_tokens_seen": 1575792, "step": 1830 }, { "epoch": 4.443099273607748, "grad_norm": 0.002269514137879014, "learning_rate": 1.882842699828169e-07, "loss": 0.0, "num_input_tokens_seen": 1580080, "step": 1835 }, { "epoch": 4.455205811138015, "grad_norm": 0.0023223496973514557, "learning_rate": 1.8032003168373306e-07, "loss": 0.0, "num_input_tokens_seen": 1584112, "step": 1840 }, { "epoch": 4.467312348668281, "grad_norm": 0.001878439332358539, "learning_rate": 1.7252158929746133e-07, "loss": 0.0, "num_input_tokens_seen": 1588400, "step": 1845 }, { "epoch": 4.479418886198547, "grad_norm": 0.009905189275741577, "learning_rate": 1.6488950020686956e-07, "loss": 0.0, "num_input_tokens_seen": 1592816, "step": 1850 }, { "epoch": 4.491525423728813, "grad_norm": 0.0024043757002800703, "learning_rate": 1.5742430990495465e-07, "loss": 0.0, "num_input_tokens_seen": 1597296, "step": 1855 }, { "epoch": 4.50363196125908, "grad_norm": 0.0012564613716676831, "learning_rate": 1.501265519558537e-07, "loss": 0.0184, "num_input_tokens_seen": 1601648, "step": 1860 }, { "epoch": 4.5157384987893465, "grad_norm": 0.014631015248596668, "learning_rate": 1.4299674795670765e-07, "loss": 0.0, "num_input_tokens_seen": 1605936, "step": 1865 }, { "epoch": 4.527845036319612, "grad_norm": 0.002841898240149021, "learning_rate": 1.360354075003828e-07, "loss": 0.0, "num_input_tokens_seen": 1610096, "step": 1870 }, { "epoch": 4.532687651331719, "eval_loss": 0.2488991767168045, "eval_runtime": 0.6331, "eval_samples_per_second": 579.694, "eval_steps_per_second": 72.659, "num_input_tokens_seen": 1611760, "step": 1872 }, { "epoch": 4.539951573849879, "grad_norm": 0.008398376405239105, "learning_rate": 1.2924302813904582e-07, "loss": 0.0, "num_input_tokens_seen": 1614384, "step": 1875 }, { "epoch": 4.552058111380145, "grad_norm": 0.0019534530583769083, "learning_rate": 1.2262009534860368e-07, "loss": 0.0, "num_input_tokens_seen": 1618800, "step": 1880 }, { "epoch": 4.5641646489104115, "grad_norm": 0.0013132602907717228, "learning_rate": 1.161670824940045e-07, "loss": 0.0, "num_input_tokens_seen": 1622960, "step": 1885 }, { "epoch": 4.576271186440678, "grad_norm": 0.0015446435427293181, "learning_rate": 1.0988445079540389e-07, "loss": 0.0, "num_input_tokens_seen": 1627056, "step": 1890 }, { "epoch": 4.588377723970944, "grad_norm": 0.0025299135595560074, "learning_rate": 1.0377264929520126e-07, "loss": 0.0002, "num_input_tokens_seen": 1631408, "step": 1895 }, { "epoch": 4.600484261501211, "grad_norm": 0.0017427592538297176, "learning_rate": 9.783211482594285e-08, "loss": 0.0, "num_input_tokens_seen": 1635888, "step": 1900 }, { "epoch": 4.6125907990314765, "grad_norm": 0.002218902576714754, "learning_rate": 9.206327197910203e-08, "loss": 0.0, "num_input_tokens_seen": 1640176, "step": 1905 }, { "epoch": 4.624697336561743, "grad_norm": 0.004275280050933361, "learning_rate": 8.64665330747308e-08, "loss": 0.0, "num_input_tokens_seen": 1644528, "step": 1910 }, { "epoch": 4.63680387409201, "grad_norm": 0.00501580024138093, "learning_rate": 8.104229813199111e-08, "loss": 0.0, "num_input_tokens_seen": 1649264, "step": 1915 }, { "epoch": 4.648910411622276, "grad_norm": 0.004407305270433426, "learning_rate": 7.579095484056193e-08, "loss": 0.0, "num_input_tokens_seen": 1653808, "step": 1920 }, { "epoch": 4.661016949152542, "grad_norm": 0.001284220372326672, "learning_rate": 7.071287853293141e-08, "loss": 0.0, "num_input_tokens_seen": 1658288, "step": 1925 }, { "epoch": 4.673123486682809, "grad_norm": 0.0011293090647086501, "learning_rate": 6.580843215757082e-08, "loss": 0.0, "num_input_tokens_seen": 1662576, "step": 1930 }, { "epoch": 4.685230024213075, "grad_norm": 0.007353964261710644, "learning_rate": 6.107796625299117e-08, "loss": 0.0, "num_input_tokens_seen": 1667056, "step": 1935 }, { "epoch": 4.697336561743342, "grad_norm": 0.0041248914785683155, "learning_rate": 5.652181892269182e-08, "loss": 0.0, "num_input_tokens_seen": 1671536, "step": 1940 }, { "epoch": 4.709443099273607, "grad_norm": 0.0036760459188371897, "learning_rate": 5.214031581099149e-08, "loss": 0.0, "num_input_tokens_seen": 1675888, "step": 1945 }, { "epoch": 4.721549636803874, "grad_norm": 0.0022840022575110197, "learning_rate": 4.793377007975719e-08, "loss": 0.0, "num_input_tokens_seen": 1680176, "step": 1950 }, { "epoch": 4.733656174334141, "grad_norm": 0.004901141859591007, "learning_rate": 4.3902482386018186e-08, "loss": 0.0, "num_input_tokens_seen": 1684400, "step": 1955 }, { "epoch": 4.745762711864407, "grad_norm": 0.004255094565451145, "learning_rate": 4.004674086047905e-08, "loss": 0.0357, "num_input_tokens_seen": 1688816, "step": 1960 }, { "epoch": 4.757869249394673, "grad_norm": 0.0026743041817098856, "learning_rate": 3.636682108692502e-08, "loss": 0.0, "num_input_tokens_seen": 1693360, "step": 1965 }, { "epoch": 4.76997578692494, "grad_norm": 0.0023194768000394106, "learning_rate": 3.286298608252442e-08, "loss": 0.0, "num_input_tokens_seen": 1697584, "step": 1970 }, { "epoch": 4.782082324455206, "grad_norm": 0.003945828415453434, "learning_rate": 2.953548627903202e-08, "loss": 0.0, "num_input_tokens_seen": 1702000, "step": 1975 }, { "epoch": 4.784503631961259, "eval_loss": 0.2507624924182892, "eval_runtime": 0.6499, "eval_samples_per_second": 564.699, "eval_steps_per_second": 70.78, "num_input_tokens_seen": 1702832, "step": 1976 }, { "epoch": 4.7941888619854724, "grad_norm": 0.0016053339932113886, "learning_rate": 2.6384559504886164e-08, "loss": 0.0, "num_input_tokens_seen": 1706416, "step": 1980 }, { "epoch": 4.806295399515738, "grad_norm": 0.0013666612794622779, "learning_rate": 2.3410430968214825e-08, "loss": 0.0, "num_input_tokens_seen": 1710960, "step": 1985 }, { "epoch": 4.818401937046005, "grad_norm": 0.0037024938501417637, "learning_rate": 2.0613313240735457e-08, "loss": 0.0, "num_input_tokens_seen": 1715440, "step": 1990 }, { "epoch": 4.830508474576272, "grad_norm": 0.0014027768047526479, "learning_rate": 1.7993406242563238e-08, "loss": 0.0, "num_input_tokens_seen": 1719728, "step": 1995 }, { "epoch": 4.842615012106537, "grad_norm": 0.0023848332930356264, "learning_rate": 1.5550897227922522e-08, "loss": 0.0, "num_input_tokens_seen": 1724272, "step": 2000 }, { "epoch": 4.854721549636804, "grad_norm": 0.001933931838721037, "learning_rate": 1.3285960771761696e-08, "loss": 0.0, "num_input_tokens_seen": 1728560, "step": 2005 }, { "epoch": 4.86682808716707, "grad_norm": 0.0011232432443648577, "learning_rate": 1.119875875727705e-08, "loss": 0.0, "num_input_tokens_seen": 1733104, "step": 2010 }, { "epoch": 4.878934624697337, "grad_norm": 0.0012837464455515146, "learning_rate": 9.289440364341484e-09, "loss": 0.0, "num_input_tokens_seen": 1737264, "step": 2015 }, { "epoch": 4.891041162227603, "grad_norm": 0.003382457885891199, "learning_rate": 7.558142058842755e-09, "loss": 0.0, "num_input_tokens_seen": 1741424, "step": 2020 }, { "epoch": 4.903147699757869, "grad_norm": 0.002113591879606247, "learning_rate": 6.004987582929056e-09, "loss": 0.0, "num_input_tokens_seen": 1745648, "step": 2025 }, { "epoch": 4.915254237288136, "grad_norm": 0.0015381674747914076, "learning_rate": 4.6300879461655404e-09, "loss": 0.0, "num_input_tokens_seen": 1749872, "step": 2030 }, { "epoch": 4.927360774818402, "grad_norm": 0.0024200750049203634, "learning_rate": 3.4335414175995506e-09, "loss": 0.0, "num_input_tokens_seen": 1754288, "step": 2035 }, { "epoch": 4.939467312348668, "grad_norm": 0.0022603883408010006, "learning_rate": 2.4154335187365207e-09, "loss": 0.0, "num_input_tokens_seen": 1758640, "step": 2040 }, { "epoch": 4.951573849878935, "grad_norm": 0.001337168039754033, "learning_rate": 1.575837017428472e-09, "loss": 0.0003, "num_input_tokens_seen": 1762928, "step": 2045 }, { "epoch": 4.963680387409201, "grad_norm": 0.002117044758051634, "learning_rate": 9.14811922672898e-10, "loss": 0.0, "num_input_tokens_seen": 1767344, "step": 2050 }, { "epoch": 4.9757869249394675, "grad_norm": 0.000865236681420356, "learning_rate": 4.3240548032230657e-10, "loss": 0.0, "num_input_tokens_seen": 1771632, "step": 2055 }, { "epoch": 4.987893462469733, "grad_norm": 0.001379348454065621, "learning_rate": 1.2865216970914253e-10, "loss": 0.0, "num_input_tokens_seen": 1775728, "step": 2060 }, { "epoch": 5.0, "grad_norm": 0.0014177365228533745, "learning_rate": 3.573701180537015e-12, "loss": 0.0, "num_input_tokens_seen": 1780000, "step": 2065 }, { "epoch": 5.0, "num_input_tokens_seen": 1780000, "step": 2065, "total_flos": 1.039320047616e+16, "train_loss": 0.06261659951716346, "train_runtime": 1141.6604, "train_samples_per_second": 14.457, "train_steps_per_second": 1.809 } ], "logging_steps": 5, "max_steps": 2065, "num_input_tokens_seen": 1780000, "num_train_epochs": 5, "save_steps": 104, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.039320047616e+16, "train_batch_size": 8, "trial_name": null, "trial_params": null }