| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997315436241611, | |
| "eval_steps": 500, | |
| "global_step": 931, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010738255033557046, | |
| "grad_norm": 22.873438223486826, | |
| "learning_rate": 1.0638297872340426e-07, | |
| "loss": 1.3013, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005369127516778523, | |
| "grad_norm": 21.159527745955735, | |
| "learning_rate": 5.319148936170213e-07, | |
| "loss": 1.3281, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010738255033557046, | |
| "grad_norm": 8.15243855646548, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "loss": 1.1934, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.016107382550335572, | |
| "grad_norm": 7.876837115931939, | |
| "learning_rate": 1.595744680851064e-06, | |
| "loss": 1.0403, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.021476510067114093, | |
| "grad_norm": 3.0351073481295128, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 0.9152, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.026845637583892617, | |
| "grad_norm": 2.4261598045634685, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "loss": 0.8749, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.032214765100671144, | |
| "grad_norm": 2.216819042623432, | |
| "learning_rate": 3.191489361702128e-06, | |
| "loss": 0.8366, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03758389261744966, | |
| "grad_norm": 2.282450409349314, | |
| "learning_rate": 3.723404255319149e-06, | |
| "loss": 0.818, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.042953020134228186, | |
| "grad_norm": 2.2660596456862643, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.7985, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04832214765100671, | |
| "grad_norm": 2.367380807091964, | |
| "learning_rate": 4.787234042553192e-06, | |
| "loss": 0.7886, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.053691275167785234, | |
| "grad_norm": 2.385525049214499, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 0.7777, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05906040268456376, | |
| "grad_norm": 2.5856947730866446, | |
| "learning_rate": 5.851063829787235e-06, | |
| "loss": 0.7546, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06442953020134229, | |
| "grad_norm": 2.4085256812489084, | |
| "learning_rate": 6.382978723404256e-06, | |
| "loss": 0.7524, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0697986577181208, | |
| "grad_norm": 2.4401920680003157, | |
| "learning_rate": 6.914893617021278e-06, | |
| "loss": 0.7291, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07516778523489932, | |
| "grad_norm": 2.3253305278745215, | |
| "learning_rate": 7.446808510638298e-06, | |
| "loss": 0.7272, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08053691275167785, | |
| "grad_norm": 2.45450168917715, | |
| "learning_rate": 7.97872340425532e-06, | |
| "loss": 0.7127, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08590604026845637, | |
| "grad_norm": 2.127447895073466, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.7041, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0912751677852349, | |
| "grad_norm": 2.3028444099891705, | |
| "learning_rate": 9.042553191489362e-06, | |
| "loss": 0.7037, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09664429530201342, | |
| "grad_norm": 2.5309118167527616, | |
| "learning_rate": 9.574468085106385e-06, | |
| "loss": 0.6988, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10201342281879194, | |
| "grad_norm": 2.4219930767340743, | |
| "learning_rate": 9.999964780082996e-06, | |
| "loss": 0.6903, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10738255033557047, | |
| "grad_norm": 2.111757405385662, | |
| "learning_rate": 9.998732135085665e-06, | |
| "loss": 0.6801, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11275167785234899, | |
| "grad_norm": 2.2202581015134077, | |
| "learning_rate": 9.995738990383743e-06, | |
| "loss": 0.6907, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11812080536912752, | |
| "grad_norm": 2.626765418959801, | |
| "learning_rate": 9.990986400130607e-06, | |
| "loss": 0.686, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12348993288590604, | |
| "grad_norm": 2.2905312730653424, | |
| "learning_rate": 9.984476038137437e-06, | |
| "loss": 0.6687, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12885906040268458, | |
| "grad_norm": 2.190663676539029, | |
| "learning_rate": 9.97621019728372e-06, | |
| "loss": 0.6776, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1342281879194631, | |
| "grad_norm": 2.2515596629136825, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.6694, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1395973154362416, | |
| "grad_norm": 2.168375563832784, | |
| "learning_rate": 9.954424340791195e-06, | |
| "loss": 0.6752, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14496644295302014, | |
| "grad_norm": 2.088272443654766, | |
| "learning_rate": 9.940911997896774e-06, | |
| "loss": 0.6661, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15033557046979865, | |
| "grad_norm": 2.173874083902431, | |
| "learning_rate": 9.925659518928316e-06, | |
| "loss": 0.6563, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15570469798657718, | |
| "grad_norm": 2.0589658736370757, | |
| "learning_rate": 9.908672275644898e-06, | |
| "loss": 0.6607, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1610738255033557, | |
| "grad_norm": 2.147351094566772, | |
| "learning_rate": 9.889956250770933e-06, | |
| "loss": 0.6575, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16644295302013423, | |
| "grad_norm": 2.3349226051109584, | |
| "learning_rate": 9.86951803588912e-06, | |
| "loss": 0.6527, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.17181208053691274, | |
| "grad_norm": 2.281471579219723, | |
| "learning_rate": 9.847364829118963e-06, | |
| "loss": 0.6544, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17718120805369128, | |
| "grad_norm": 2.1332428982031715, | |
| "learning_rate": 9.82350443258166e-06, | |
| "loss": 0.6479, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1825503355704698, | |
| "grad_norm": 2.0804446588419045, | |
| "learning_rate": 9.797945249652295e-06, | |
| "loss": 0.6439, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.18791946308724833, | |
| "grad_norm": 2.0585417674642636, | |
| "learning_rate": 9.770696282000245e-06, | |
| "loss": 0.6321, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.19328859060402684, | |
| "grad_norm": 2.1666332876300762, | |
| "learning_rate": 9.741767126418898e-06, | |
| "loss": 0.6336, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19865771812080538, | |
| "grad_norm": 2.0205631171866543, | |
| "learning_rate": 9.711167971445766e-06, | |
| "loss": 0.6306, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2040268456375839, | |
| "grad_norm": 2.4514209883366807, | |
| "learning_rate": 9.67890959377418e-06, | |
| "loss": 0.624, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20939597315436242, | |
| "grad_norm": 2.011077236989635, | |
| "learning_rate": 9.645003354457872e-06, | |
| "loss": 0.6176, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.21476510067114093, | |
| "grad_norm": 2.157067702163037, | |
| "learning_rate": 9.60946119490972e-06, | |
| "loss": 0.6231, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22013422818791947, | |
| "grad_norm": 1.851876954712529, | |
| "learning_rate": 9.57229563269612e-06, | |
| "loss": 0.6203, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22550335570469798, | |
| "grad_norm": 1.9636046456639136, | |
| "learning_rate": 9.533519757128426e-06, | |
| "loss": 0.6186, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23087248322147652, | |
| "grad_norm": 2.374396960285375, | |
| "learning_rate": 9.49314722465304e-06, | |
| "loss": 0.6157, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.23624161073825503, | |
| "grad_norm": 2.072084337879624, | |
| "learning_rate": 9.451192254041759e-06, | |
| "loss": 0.6157, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24161073825503357, | |
| "grad_norm": 2.0213326681347605, | |
| "learning_rate": 9.407669621384073e-06, | |
| "loss": 0.619, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.24697986577181208, | |
| "grad_norm": 2.05541918839593, | |
| "learning_rate": 9.362594654883185e-06, | |
| "loss": 0.6043, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2523489932885906, | |
| "grad_norm": 1.9647313060735423, | |
| "learning_rate": 9.31598322945759e-06, | |
| "loss": 0.5877, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.25771812080536916, | |
| "grad_norm": 2.321446787508396, | |
| "learning_rate": 9.267851761150092e-06, | |
| "loss": 0.6038, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26308724832214764, | |
| "grad_norm": 1.954968634006319, | |
| "learning_rate": 9.218217201346251e-06, | |
| "loss": 0.5973, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2684563758389262, | |
| "grad_norm": 1.9588200511779104, | |
| "learning_rate": 9.167097030804289e-06, | |
| "loss": 0.5952, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2738255033557047, | |
| "grad_norm": 1.978200194451611, | |
| "learning_rate": 9.114509253498554e-06, | |
| "loss": 0.5956, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2791946308724832, | |
| "grad_norm": 1.9082868769634187, | |
| "learning_rate": 9.060472390278717e-06, | |
| "loss": 0.5834, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.28456375838926173, | |
| "grad_norm": 1.9947479259010255, | |
| "learning_rate": 9.005005472346923e-06, | |
| "loss": 0.5878, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.28993288590604027, | |
| "grad_norm": 2.0101949394033443, | |
| "learning_rate": 8.948128034555212e-06, | |
| "loss": 0.5815, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2953020134228188, | |
| "grad_norm": 1.855173856877532, | |
| "learning_rate": 8.889860108525544e-06, | |
| "loss": 0.5749, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3006711409395973, | |
| "grad_norm": 1.8947595029633266, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5597, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.30604026845637583, | |
| "grad_norm": 2.0225835192845825, | |
| "learning_rate": 8.76923535958783e-06, | |
| "loss": 0.5683, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.31140939597315437, | |
| "grad_norm": 2.047769165835403, | |
| "learning_rate": 8.706921019419237e-06, | |
| "loss": 0.5693, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3167785234899329, | |
| "grad_norm": 1.996841996415469, | |
| "learning_rate": 8.643301141529619e-06, | |
| "loss": 0.576, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3221476510067114, | |
| "grad_norm": 2.085317513003649, | |
| "learning_rate": 8.578398132155846e-06, | |
| "loss": 0.5586, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3275167785234899, | |
| "grad_norm": 2.2696870964241342, | |
| "learning_rate": 8.512234849439887e-06, | |
| "loss": 0.5487, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.33288590604026846, | |
| "grad_norm": 2.295292164613049, | |
| "learning_rate": 8.444834595378434e-06, | |
| "loss": 0.5657, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.338255033557047, | |
| "grad_norm": 1.9535177447525518, | |
| "learning_rate": 8.376221107616187e-06, | |
| "loss": 0.5492, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3436241610738255, | |
| "grad_norm": 1.935697674348923, | |
| "learning_rate": 8.306418551085707e-06, | |
| "loss": 0.5502, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.348993288590604, | |
| "grad_norm": 1.912637650045647, | |
| "learning_rate": 8.23545150949679e-06, | |
| "loss": 0.5659, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.35436241610738256, | |
| "grad_norm": 2.0813900852367437, | |
| "learning_rate": 8.163344976678342e-06, | |
| "loss": 0.5512, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3597315436241611, | |
| "grad_norm": 1.9178576647529397, | |
| "learning_rate": 8.090124347775837e-06, | |
| "loss": 0.5399, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3651006711409396, | |
| "grad_norm": 1.9921209746990098, | |
| "learning_rate": 8.0158154103074e-06, | |
| "loss": 0.5335, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3704697986577181, | |
| "grad_norm": 2.0251179740048686, | |
| "learning_rate": 7.940444335081733e-06, | |
| "loss": 0.5412, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.37583892617449666, | |
| "grad_norm": 2.0237480393180536, | |
| "learning_rate": 7.864037666981037e-06, | |
| "loss": 0.5318, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3812080536912752, | |
| "grad_norm": 2.062903496053255, | |
| "learning_rate": 7.786622315612182e-06, | |
| "loss": 0.5391, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3865771812080537, | |
| "grad_norm": 1.931828289441635, | |
| "learning_rate": 7.708225545829446e-06, | |
| "loss": 0.525, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3919463087248322, | |
| "grad_norm": 2.004972595205748, | |
| "learning_rate": 7.6288749681321115e-06, | |
| "loss": 0.5173, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.39731543624161075, | |
| "grad_norm": 2.0462367199517937, | |
| "learning_rate": 7.548598528940354e-06, | |
| "loss": 0.5154, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.40268456375838924, | |
| "grad_norm": 2.3727598847724636, | |
| "learning_rate": 7.4674245007528135e-06, | |
| "loss": 0.5279, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4080536912751678, | |
| "grad_norm": 1.965518745162171, | |
| "learning_rate": 7.385381472189321e-06, | |
| "loss": 0.5255, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4134228187919463, | |
| "grad_norm": 2.0728393681212225, | |
| "learning_rate": 7.302498337922293e-06, | |
| "loss": 0.5152, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.41879194630872485, | |
| "grad_norm": 2.0074609174878906, | |
| "learning_rate": 7.218804288500343e-06, | |
| "loss": 0.5224, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.42416107382550333, | |
| "grad_norm": 1.9472600550525063, | |
| "learning_rate": 7.134328800067684e-06, | |
| "loss": 0.5209, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.42953020134228187, | |
| "grad_norm": 1.9949043902713617, | |
| "learning_rate": 7.049101623982938e-06, | |
| "loss": 0.5082, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4348993288590604, | |
| "grad_norm": 1.834654212769176, | |
| "learning_rate": 6.963152776341044e-06, | |
| "loss": 0.5149, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.44026845637583895, | |
| "grad_norm": 2.0502871429630822, | |
| "learning_rate": 6.876512527401897e-06, | |
| "loss": 0.5131, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.44563758389261743, | |
| "grad_norm": 2.000365576779467, | |
| "learning_rate": 6.789211390929497e-06, | |
| "loss": 0.5016, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.45100671140939597, | |
| "grad_norm": 2.0157623562806353, | |
| "learning_rate": 6.701280113445324e-06, | |
| "loss": 0.4958, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4563758389261745, | |
| "grad_norm": 1.970116584111867, | |
| "learning_rate": 6.6127496633997475e-06, | |
| "loss": 0.5003, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.46174496644295304, | |
| "grad_norm": 1.9483867090820592, | |
| "learning_rate": 6.523651220265269e-06, | |
| "loss": 0.4942, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4671140939597315, | |
| "grad_norm": 2.099467823611523, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.4952, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.47248322147651006, | |
| "grad_norm": 1.9558069283541033, | |
| "learning_rate": 6.343876061773385e-06, | |
| "loss": 0.4924, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4778523489932886, | |
| "grad_norm": 1.9847822609941113, | |
| "learning_rate": 6.2532626612936035e-06, | |
| "loss": 0.4823, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.48322147651006714, | |
| "grad_norm": 2.0688837915005993, | |
| "learning_rate": 6.162207875181354e-06, | |
| "loss": 0.4913, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4885906040268456, | |
| "grad_norm": 1.9342878293544008, | |
| "learning_rate": 6.070743771953157e-06, | |
| "loss": 0.4769, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.49395973154362416, | |
| "grad_norm": 2.043215000887027, | |
| "learning_rate": 5.978902564282616e-06, | |
| "loss": 0.4752, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4993288590604027, | |
| "grad_norm": 1.8676181620631764, | |
| "learning_rate": 5.886716597655472e-06, | |
| "loss": 0.4701, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5046979865771812, | |
| "grad_norm": 1.9595541482385936, | |
| "learning_rate": 5.7942183389778536e-06, | |
| "loss": 0.4696, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5100671140939598, | |
| "grad_norm": 1.825167080567174, | |
| "learning_rate": 5.701440365141799e-06, | |
| "loss": 0.4625, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5154362416107383, | |
| "grad_norm": 1.9786129201297613, | |
| "learning_rate": 5.608415351552014e-06, | |
| "loss": 0.4596, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5208053691275167, | |
| "grad_norm": 1.9920386733808053, | |
| "learning_rate": 5.515176060617945e-06, | |
| "loss": 0.4684, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5261744966442953, | |
| "grad_norm": 1.9116799488281817, | |
| "learning_rate": 5.421755330215223e-06, | |
| "loss": 0.4659, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5315436241610738, | |
| "grad_norm": 1.948129234992053, | |
| "learning_rate": 5.328186062120509e-06, | |
| "loss": 0.4436, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5369127516778524, | |
| "grad_norm": 2.078748185289108, | |
| "learning_rate": 5.23450121042383e-06, | |
| "loss": 0.4548, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5422818791946309, | |
| "grad_norm": 1.9444657978297817, | |
| "learning_rate": 5.140733769922525e-06, | |
| "loss": 0.4533, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5476510067114094, | |
| "grad_norm": 1.9186109778492548, | |
| "learning_rate": 5.0469167645008245e-06, | |
| "loss": 0.4484, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.553020134228188, | |
| "grad_norm": 2.115910772434904, | |
| "learning_rate": 4.953083235499177e-06, | |
| "loss": 0.4557, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5583892617449664, | |
| "grad_norm": 1.9469832075906568, | |
| "learning_rate": 4.859266230077474e-06, | |
| "loss": 0.4439, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5637583892617449, | |
| "grad_norm": 1.90470271131317, | |
| "learning_rate": 4.7654987895761705e-06, | |
| "loss": 0.4431, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5691275167785235, | |
| "grad_norm": 1.941709380965833, | |
| "learning_rate": 4.671813937879494e-06, | |
| "loss": 0.4517, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.574496644295302, | |
| "grad_norm": 1.9633282455666656, | |
| "learning_rate": 4.5782446697847775e-06, | |
| "loss": 0.4469, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5798657718120805, | |
| "grad_norm": 1.8845576495595482, | |
| "learning_rate": 4.484823939382056e-06, | |
| "loss": 0.4316, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5852348993288591, | |
| "grad_norm": 1.8811789086409214, | |
| "learning_rate": 4.391584648447989e-06, | |
| "loss": 0.4423, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5906040268456376, | |
| "grad_norm": 1.9555309537231733, | |
| "learning_rate": 4.298559634858202e-06, | |
| "loss": 0.4368, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5959731543624162, | |
| "grad_norm": 2.0555513727023578, | |
| "learning_rate": 4.205781661022146e-06, | |
| "loss": 0.4298, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6013422818791946, | |
| "grad_norm": 1.8777742161685087, | |
| "learning_rate": 4.1132834023445304e-06, | |
| "loss": 0.426, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6067114093959731, | |
| "grad_norm": 1.939266556920773, | |
| "learning_rate": 4.021097435717386e-06, | |
| "loss": 0.4216, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6120805369127517, | |
| "grad_norm": 1.9261147196325419, | |
| "learning_rate": 3.929256228046845e-06, | |
| "loss": 0.4285, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6174496644295302, | |
| "grad_norm": 1.805107172046572, | |
| "learning_rate": 3.837792124818647e-06, | |
| "loss": 0.416, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6228187919463087, | |
| "grad_norm": 1.8704231334410426, | |
| "learning_rate": 3.7467373387063973e-06, | |
| "loss": 0.4191, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6281879194630873, | |
| "grad_norm": 1.9935959039830586, | |
| "learning_rate": 3.656123938226618e-06, | |
| "loss": 0.4183, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6335570469798658, | |
| "grad_norm": 1.9248727434625008, | |
| "learning_rate": 3.5659838364445505e-06, | |
| "loss": 0.4239, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6389261744966444, | |
| "grad_norm": 1.9448194253706668, | |
| "learning_rate": 3.476348779734732e-06, | |
| "loss": 0.4118, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6442953020134228, | |
| "grad_norm": 1.8244567974798083, | |
| "learning_rate": 3.387250336600254e-06, | |
| "loss": 0.4108, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6496644295302013, | |
| "grad_norm": 1.8766359486725541, | |
| "learning_rate": 3.298719886554677e-06, | |
| "loss": 0.4094, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6550335570469799, | |
| "grad_norm": 1.9861783019472632, | |
| "learning_rate": 3.2107886090705035e-06, | |
| "loss": 0.4024, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6604026845637584, | |
| "grad_norm": 1.8918824877700715, | |
| "learning_rate": 3.1234874725981045e-06, | |
| "loss": 0.4086, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6657718120805369, | |
| "grad_norm": 2.0714978550264003, | |
| "learning_rate": 3.036847223658958e-06, | |
| "loss": 0.396, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6711409395973155, | |
| "grad_norm": 1.9560452224763145, | |
| "learning_rate": 2.950898376017064e-06, | |
| "loss": 0.4035, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.676510067114094, | |
| "grad_norm": 2.043363405835207, | |
| "learning_rate": 2.865671199932318e-06, | |
| "loss": 0.4068, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6818791946308724, | |
| "grad_norm": 1.8816775523611597, | |
| "learning_rate": 2.781195711499658e-06, | |
| "loss": 0.393, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.687248322147651, | |
| "grad_norm": 1.8488910820957973, | |
| "learning_rate": 2.697501662077707e-06, | |
| "loss": 0.4013, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6926174496644295, | |
| "grad_norm": 1.862219172494916, | |
| "learning_rate": 2.6146185278106807e-06, | |
| "loss": 0.3909, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.697986577181208, | |
| "grad_norm": 2.012779150472371, | |
| "learning_rate": 2.5325754992471886e-06, | |
| "loss": 0.3933, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7033557046979866, | |
| "grad_norm": 1.925776837174688, | |
| "learning_rate": 2.4514014710596467e-06, | |
| "loss": 0.3794, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7087248322147651, | |
| "grad_norm": 1.8323121691067996, | |
| "learning_rate": 2.3711250318678906e-06, | |
| "loss": 0.3906, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7140939597315437, | |
| "grad_norm": 1.8449642618290978, | |
| "learning_rate": 2.2917744541705544e-06, | |
| "loss": 0.3804, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.7194630872483222, | |
| "grad_norm": 1.8385577849219443, | |
| "learning_rate": 2.2133776843878185e-06, | |
| "loss": 0.3896, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7248322147651006, | |
| "grad_norm": 1.8863509359096609, | |
| "learning_rate": 2.1359623330189655e-06, | |
| "loss": 0.3906, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7302013422818792, | |
| "grad_norm": 1.9412331601540616, | |
| "learning_rate": 2.059555664918268e-06, | |
| "loss": 0.3872, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7355704697986577, | |
| "grad_norm": 1.795387975158009, | |
| "learning_rate": 1.9841845896926022e-06, | |
| "loss": 0.3798, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7409395973154362, | |
| "grad_norm": 1.843173352636606, | |
| "learning_rate": 1.9098756522241634e-06, | |
| "loss": 0.3788, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7463087248322148, | |
| "grad_norm": 1.912174039399666, | |
| "learning_rate": 1.8366550233216584e-06, | |
| "loss": 0.3695, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7516778523489933, | |
| "grad_norm": 1.8237677647091424, | |
| "learning_rate": 1.7645484905032129e-06, | |
| "loss": 0.3933, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7570469798657719, | |
| "grad_norm": 1.7822160523413584, | |
| "learning_rate": 1.6935814489142937e-06, | |
| "loss": 0.3781, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7624161073825504, | |
| "grad_norm": 1.8878271721301074, | |
| "learning_rate": 1.6237788923838149e-06, | |
| "loss": 0.3751, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7677852348993288, | |
| "grad_norm": 1.9082160841545055, | |
| "learning_rate": 1.555165404621567e-06, | |
| "loss": 0.3728, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7731543624161074, | |
| "grad_norm": 1.816281209572928, | |
| "learning_rate": 1.487765150560116e-06, | |
| "loss": 0.3777, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7785234899328859, | |
| "grad_norm": 1.91354765663735, | |
| "learning_rate": 1.4216018678441558e-06, | |
| "loss": 0.3699, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7838926174496644, | |
| "grad_norm": 1.9355520287879238, | |
| "learning_rate": 1.3566988584703817e-06, | |
| "loss": 0.371, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.789261744966443, | |
| "grad_norm": 1.7280419120289419, | |
| "learning_rate": 1.293078980580766e-06, | |
| "loss": 0.3764, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7946308724832215, | |
| "grad_norm": 1.9044555736805637, | |
| "learning_rate": 1.2307646404121692e-06, | |
| "loss": 0.3664, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.891432724737264, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.3664, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8053691275167785, | |
| "grad_norm": 1.6814136447042585, | |
| "learning_rate": 1.1101398914744565e-06, | |
| "loss": 0.3574, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.810738255033557, | |
| "grad_norm": 1.8225153893838018, | |
| "learning_rate": 1.0518719654447896e-06, | |
| "loss": 0.3684, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.8161073825503355, | |
| "grad_norm": 1.8275415575673986, | |
| "learning_rate": 9.949945276530782e-07, | |
| "loss": 0.3603, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8214765100671141, | |
| "grad_norm": 1.853253611406343, | |
| "learning_rate": 9.395276097212841e-07, | |
| "loss": 0.3651, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8268456375838926, | |
| "grad_norm": 1.835073747374209, | |
| "learning_rate": 8.854907465014479e-07, | |
| "loss": 0.3677, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8322147651006712, | |
| "grad_norm": 1.7695339228605491, | |
| "learning_rate": 8.329029691957124e-07, | |
| "loss": 0.3606, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8375838926174497, | |
| "grad_norm": 1.749975183558492, | |
| "learning_rate": 7.817827986537508e-07, | |
| "loss": 0.3511, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8429530201342282, | |
| "grad_norm": 1.9426041133261787, | |
| "learning_rate": 7.321482388499096e-07, | |
| "loss": 0.358, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8483221476510067, | |
| "grad_norm": 1.8171481735775137, | |
| "learning_rate": 6.840167705424106e-07, | |
| "loss": 0.3544, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8536912751677852, | |
| "grad_norm": 1.8412832647944588, | |
| "learning_rate": 6.374053451168166e-07, | |
| "loss": 0.3661, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8590604026845637, | |
| "grad_norm": 1.8350930759774366, | |
| "learning_rate": 5.92330378615929e-07, | |
| "loss": 0.3537, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8644295302013423, | |
| "grad_norm": 1.723224003576655, | |
| "learning_rate": 5.488077459582425e-07, | |
| "loss": 0.3553, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8697986577181208, | |
| "grad_norm": 1.8473917985800616, | |
| "learning_rate": 5.068527753469604e-07, | |
| "loss": 0.3603, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8751677852348994, | |
| "grad_norm": 1.8325217665657294, | |
| "learning_rate": 4.664802428715753e-07, | |
| "loss": 0.348, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8805369127516779, | |
| "grad_norm": 1.7346304200820148, | |
| "learning_rate": 4.2770436730388166e-07, | |
| "loss": 0.3498, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8859060402684564, | |
| "grad_norm": 1.7575151506469913, | |
| "learning_rate": 3.9053880509028086e-07, | |
| "loss": 0.3616, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8912751677852349, | |
| "grad_norm": 1.800873655206166, | |
| "learning_rate": 3.549966455421305e-07, | |
| "loss": 0.3507, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8966442953020134, | |
| "grad_norm": 1.710054226395714, | |
| "learning_rate": 3.2109040622582186e-07, | |
| "loss": 0.3527, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.9020134228187919, | |
| "grad_norm": 1.6705083298206793, | |
| "learning_rate": 2.8883202855423676e-07, | |
| "loss": 0.3516, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.9073825503355705, | |
| "grad_norm": 1.7474504216215188, | |
| "learning_rate": 2.582328735811029e-07, | |
| "loss": 0.3452, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.912751677852349, | |
| "grad_norm": 1.7463913866112468, | |
| "learning_rate": 2.2930371799975593e-07, | |
| "loss": 0.3471, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9181208053691275, | |
| "grad_norm": 1.7297820385087201, | |
| "learning_rate": 2.0205475034770606e-07, | |
| "loss": 0.3468, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9234899328859061, | |
| "grad_norm": 1.7482101229826381, | |
| "learning_rate": 1.7649556741833995e-07, | |
| "loss": 0.3478, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9288590604026845, | |
| "grad_norm": 1.8150110385873126, | |
| "learning_rate": 1.5263517088103862e-07, | |
| "loss": 0.3586, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.934228187919463, | |
| "grad_norm": 1.8153438746846808, | |
| "learning_rate": 1.304819641108801e-07, | |
| "loss": 0.3555, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9395973154362416, | |
| "grad_norm": 1.8087296470250676, | |
| "learning_rate": 1.1004374922906846e-07, | |
| "loss": 0.3457, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9449664429530201, | |
| "grad_norm": 1.7195355098862968, | |
| "learning_rate": 9.132772435510362e-08, | |
| "loss": 0.349, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9503355704697987, | |
| "grad_norm": 1.7647364179507874, | |
| "learning_rate": 7.434048107168523e-08, | |
| "loss": 0.3466, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9557046979865772, | |
| "grad_norm": 1.9088484806569315, | |
| "learning_rate": 5.908800210322696e-08, | |
| "loss": 0.3411, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9610738255033557, | |
| "grad_norm": 1.7271817479075642, | |
| "learning_rate": 4.55756592088058e-08, | |
| "loss": 0.3484, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.9664429530201343, | |
| "grad_norm": 1.7372584442728163, | |
| "learning_rate": 3.3808211290284886e-08, | |
| "loss": 0.3433, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9718120805369127, | |
| "grad_norm": 1.808380468788094, | |
| "learning_rate": 2.378980271628195e-08, | |
| "loss": 0.3528, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9771812080536912, | |
| "grad_norm": 1.7605143851131846, | |
| "learning_rate": 1.552396186256411e-08, | |
| "loss": 0.3445, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9825503355704698, | |
| "grad_norm": 1.8047618073453124, | |
| "learning_rate": 9.013599869394096e-09, | |
| "loss": 0.3497, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9879194630872483, | |
| "grad_norm": 1.7837820069864043, | |
| "learning_rate": 4.261009616257638e-09, | |
| "loss": 0.3452, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9932885906040269, | |
| "grad_norm": 1.7898194618346237, | |
| "learning_rate": 1.2678649143349485e-09, | |
| "loss": 0.3497, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9986577181208054, | |
| "grad_norm": 1.8036764007244224, | |
| "learning_rate": 3.5219917003948003e-11, | |
| "loss": 0.346, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9997315436241611, | |
| "eval_loss": 0.2977491617202759, | |
| "eval_runtime": 1.3637, | |
| "eval_samples_per_second": 1.467, | |
| "eval_steps_per_second": 0.733, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.9997315436241611, | |
| "step": 931, | |
| "total_flos": 194880114524160.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 0.0085, | |
| "train_samples_per_second": 3497894.361, | |
| "train_steps_per_second": 109279.854 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 931, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 194880114524160.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |