| { | |
| "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": 23.23349641925915, | |
| "learning_rate": 1.0638297872340426e-07, | |
| "loss": 1.3224, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005369127516778523, | |
| "grad_norm": 20.864176480583073, | |
| "learning_rate": 5.319148936170213e-07, | |
| "loss": 1.3084, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010738255033557046, | |
| "grad_norm": 8.533873275523032, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "loss": 1.1987, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.016107382550335572, | |
| "grad_norm": 7.665296900301775, | |
| "learning_rate": 1.595744680851064e-06, | |
| "loss": 1.0267, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.021476510067114093, | |
| "grad_norm": 2.9761601522815546, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 0.9131, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.026845637583892617, | |
| "grad_norm": 2.433875752280275, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "loss": 0.8629, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.032214765100671144, | |
| "grad_norm": 2.309359589452698, | |
| "learning_rate": 3.191489361702128e-06, | |
| "loss": 0.8432, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03758389261744966, | |
| "grad_norm": 2.3739666896282, | |
| "learning_rate": 3.723404255319149e-06, | |
| "loss": 0.8237, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.042953020134228186, | |
| "grad_norm": 2.411414400789539, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.8004, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04832214765100671, | |
| "grad_norm": 2.2779093841935176, | |
| "learning_rate": 4.787234042553192e-06, | |
| "loss": 0.7798, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.053691275167785234, | |
| "grad_norm": 2.390464966558602, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 0.777, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05906040268456376, | |
| "grad_norm": 2.30705290045996, | |
| "learning_rate": 5.851063829787235e-06, | |
| "loss": 0.7534, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06442953020134229, | |
| "grad_norm": 2.432088997553532, | |
| "learning_rate": 6.382978723404256e-06, | |
| "loss": 0.7363, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0697986577181208, | |
| "grad_norm": 2.231418575279909, | |
| "learning_rate": 6.914893617021278e-06, | |
| "loss": 0.7258, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07516778523489932, | |
| "grad_norm": 2.3360484244386015, | |
| "learning_rate": 7.446808510638298e-06, | |
| "loss": 0.7239, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08053691275167785, | |
| "grad_norm": 2.2583495209303805, | |
| "learning_rate": 7.97872340425532e-06, | |
| "loss": 0.7158, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08590604026845637, | |
| "grad_norm": 2.166632602977575, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.698, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0912751677852349, | |
| "grad_norm": 2.261852982800451, | |
| "learning_rate": 9.042553191489362e-06, | |
| "loss": 0.6978, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09664429530201342, | |
| "grad_norm": 2.2449184892395593, | |
| "learning_rate": 9.574468085106385e-06, | |
| "loss": 0.6914, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10201342281879194, | |
| "grad_norm": 2.6796428545722906, | |
| "learning_rate": 9.999964780082996e-06, | |
| "loss": 0.6925, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10738255033557047, | |
| "grad_norm": 2.445223053748917, | |
| "learning_rate": 9.998732135085665e-06, | |
| "loss": 0.6894, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11275167785234899, | |
| "grad_norm": 2.101448463042792, | |
| "learning_rate": 9.995738990383743e-06, | |
| "loss": 0.6842, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11812080536912752, | |
| "grad_norm": 2.375756679579466, | |
| "learning_rate": 9.990986400130607e-06, | |
| "loss": 0.6881, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12348993288590604, | |
| "grad_norm": 2.0584289730843675, | |
| "learning_rate": 9.984476038137437e-06, | |
| "loss": 0.6789, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12885906040268458, | |
| "grad_norm": 2.1097466944401946, | |
| "learning_rate": 9.97621019728372e-06, | |
| "loss": 0.6794, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1342281879194631, | |
| "grad_norm": 2.1000573026325813, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.6622, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1395973154362416, | |
| "grad_norm": 2.21126026156604, | |
| "learning_rate": 9.954424340791195e-06, | |
| "loss": 0.6659, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14496644295302014, | |
| "grad_norm": 2.233783388990834, | |
| "learning_rate": 9.940911997896774e-06, | |
| "loss": 0.6708, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15033557046979865, | |
| "grad_norm": 2.3220703167361068, | |
| "learning_rate": 9.925659518928316e-06, | |
| "loss": 0.6546, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15570469798657718, | |
| "grad_norm": 2.1406068535723626, | |
| "learning_rate": 9.908672275644898e-06, | |
| "loss": 0.6606, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1610738255033557, | |
| "grad_norm": 1.9796338120208212, | |
| "learning_rate": 9.889956250770933e-06, | |
| "loss": 0.6543, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16644295302013423, | |
| "grad_norm": 2.038004062758435, | |
| "learning_rate": 9.86951803588912e-06, | |
| "loss": 0.639, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.17181208053691274, | |
| "grad_norm": 1.9023679569925318, | |
| "learning_rate": 9.847364829118963e-06, | |
| "loss": 0.6394, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17718120805369128, | |
| "grad_norm": 1.908984775733616, | |
| "learning_rate": 9.82350443258166e-06, | |
| "loss": 0.6449, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1825503355704698, | |
| "grad_norm": 2.010057009992793, | |
| "learning_rate": 9.797945249652295e-06, | |
| "loss": 0.6417, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.18791946308724833, | |
| "grad_norm": 2.2539350542085006, | |
| "learning_rate": 9.770696282000245e-06, | |
| "loss": 0.6333, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.19328859060402684, | |
| "grad_norm": 1.9065312726735626, | |
| "learning_rate": 9.741767126418898e-06, | |
| "loss": 0.6441, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19865771812080538, | |
| "grad_norm": 2.0928470931750023, | |
| "learning_rate": 9.711167971445766e-06, | |
| "loss": 0.6213, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2040268456375839, | |
| "grad_norm": 2.006004562606479, | |
| "learning_rate": 9.67890959377418e-06, | |
| "loss": 0.6278, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20939597315436242, | |
| "grad_norm": 1.9301686303505856, | |
| "learning_rate": 9.645003354457872e-06, | |
| "loss": 0.6325, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.21476510067114093, | |
| "grad_norm": 2.038658207178946, | |
| "learning_rate": 9.60946119490972e-06, | |
| "loss": 0.6361, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22013422818791947, | |
| "grad_norm": 2.2159213381686604, | |
| "learning_rate": 9.57229563269612e-06, | |
| "loss": 0.619, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22550335570469798, | |
| "grad_norm": 2.2179012384166685, | |
| "learning_rate": 9.533519757128426e-06, | |
| "loss": 0.6296, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23087248322147652, | |
| "grad_norm": 1.9118609526745118, | |
| "learning_rate": 9.49314722465304e-06, | |
| "loss": 0.6172, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.23624161073825503, | |
| "grad_norm": 2.0780299928973753, | |
| "learning_rate": 9.451192254041759e-06, | |
| "loss": 0.6238, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24161073825503357, | |
| "grad_norm": 1.880745326989776, | |
| "learning_rate": 9.407669621384073e-06, | |
| "loss": 0.603, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.24697986577181208, | |
| "grad_norm": 2.0325599687685325, | |
| "learning_rate": 9.362594654883185e-06, | |
| "loss": 0.6098, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2523489932885906, | |
| "grad_norm": 2.1114795284546006, | |
| "learning_rate": 9.31598322945759e-06, | |
| "loss": 0.5996, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.25771812080536916, | |
| "grad_norm": 2.147915062806659, | |
| "learning_rate": 9.267851761150092e-06, | |
| "loss": 0.5922, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26308724832214764, | |
| "grad_norm": 1.8963761596238329, | |
| "learning_rate": 9.218217201346251e-06, | |
| "loss": 0.5977, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2684563758389262, | |
| "grad_norm": 2.0596529484621406, | |
| "learning_rate": 9.167097030804289e-06, | |
| "loss": 0.6019, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2738255033557047, | |
| "grad_norm": 1.9613750119136377, | |
| "learning_rate": 9.114509253498554e-06, | |
| "loss": 0.5823, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2791946308724832, | |
| "grad_norm": 1.9493095780956693, | |
| "learning_rate": 9.060472390278717e-06, | |
| "loss": 0.5895, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.28456375838926173, | |
| "grad_norm": 1.9649255622722053, | |
| "learning_rate": 9.005005472346923e-06, | |
| "loss": 0.5861, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.28993288590604027, | |
| "grad_norm": 1.916731628798239, | |
| "learning_rate": 8.948128034555212e-06, | |
| "loss": 0.5741, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2953020134228188, | |
| "grad_norm": 1.9882471136788202, | |
| "learning_rate": 8.889860108525544e-06, | |
| "loss": 0.5891, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3006711409395973, | |
| "grad_norm": 1.8657052588450402, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5662, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.30604026845637583, | |
| "grad_norm": 1.8876272358030999, | |
| "learning_rate": 8.76923535958783e-06, | |
| "loss": 0.5611, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.31140939597315437, | |
| "grad_norm": 1.9352681945969277, | |
| "learning_rate": 8.706921019419237e-06, | |
| "loss": 0.5703, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3167785234899329, | |
| "grad_norm": 2.061655573531995, | |
| "learning_rate": 8.643301141529619e-06, | |
| "loss": 0.5704, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3221476510067114, | |
| "grad_norm": 1.9955648734094908, | |
| "learning_rate": 8.578398132155846e-06, | |
| "loss": 0.5646, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3275167785234899, | |
| "grad_norm": 1.9172018423696509, | |
| "learning_rate": 8.512234849439887e-06, | |
| "loss": 0.5603, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.33288590604026846, | |
| "grad_norm": 2.748943489124992, | |
| "learning_rate": 8.444834595378434e-06, | |
| "loss": 0.5635, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.338255033557047, | |
| "grad_norm": 2.031805396620389, | |
| "learning_rate": 8.376221107616187e-06, | |
| "loss": 0.56, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3436241610738255, | |
| "grad_norm": 2.310478182840465, | |
| "learning_rate": 8.306418551085707e-06, | |
| "loss": 0.5518, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.348993288590604, | |
| "grad_norm": 2.143453110103673, | |
| "learning_rate": 8.23545150949679e-06, | |
| "loss": 0.545, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.35436241610738256, | |
| "grad_norm": 2.074828415385107, | |
| "learning_rate": 8.163344976678342e-06, | |
| "loss": 0.5417, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3597315436241611, | |
| "grad_norm": 1.95080697885721, | |
| "learning_rate": 8.090124347775837e-06, | |
| "loss": 0.5461, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3651006711409396, | |
| "grad_norm": 1.8940424474498196, | |
| "learning_rate": 8.0158154103074e-06, | |
| "loss": 0.5415, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3704697986577181, | |
| "grad_norm": 1.9899368306488876, | |
| "learning_rate": 7.940444335081733e-06, | |
| "loss": 0.5334, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.37583892617449666, | |
| "grad_norm": 3.3190577946485074, | |
| "learning_rate": 7.864037666981037e-06, | |
| "loss": 0.5274, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3812080536912752, | |
| "grad_norm": 1.9132824928499255, | |
| "learning_rate": 7.786622315612182e-06, | |
| "loss": 0.5395, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3865771812080537, | |
| "grad_norm": 1.9703232913555617, | |
| "learning_rate": 7.708225545829446e-06, | |
| "loss": 0.5213, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3919463087248322, | |
| "grad_norm": 2.076133929106472, | |
| "learning_rate": 7.6288749681321115e-06, | |
| "loss": 0.5356, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.39731543624161075, | |
| "grad_norm": 2.0858297493245637, | |
| "learning_rate": 7.548598528940354e-06, | |
| "loss": 0.5286, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.40268456375838924, | |
| "grad_norm": 1.9498254037168414, | |
| "learning_rate": 7.4674245007528135e-06, | |
| "loss": 0.5217, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4080536912751678, | |
| "grad_norm": 1.9718043172163682, | |
| "learning_rate": 7.385381472189321e-06, | |
| "loss": 0.5185, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4134228187919463, | |
| "grad_norm": 2.0714251046483136, | |
| "learning_rate": 7.302498337922293e-06, | |
| "loss": 0.5212, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.41879194630872485, | |
| "grad_norm": 2.059741936941689, | |
| "learning_rate": 7.218804288500343e-06, | |
| "loss": 0.5068, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.42416107382550333, | |
| "grad_norm": 2.0677503384082656, | |
| "learning_rate": 7.134328800067684e-06, | |
| "loss": 0.5128, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.42953020134228187, | |
| "grad_norm": 2.1270598611660008, | |
| "learning_rate": 7.049101623982938e-06, | |
| "loss": 0.5056, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4348993288590604, | |
| "grad_norm": 1.7730274826372387, | |
| "learning_rate": 6.963152776341044e-06, | |
| "loss": 0.5007, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.44026845637583895, | |
| "grad_norm": 2.092171850544149, | |
| "learning_rate": 6.876512527401897e-06, | |
| "loss": 0.4969, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.44563758389261743, | |
| "grad_norm": 1.8511960894357482, | |
| "learning_rate": 6.789211390929497e-06, | |
| "loss": 0.5069, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.45100671140939597, | |
| "grad_norm": 2.003279508110089, | |
| "learning_rate": 6.701280113445324e-06, | |
| "loss": 0.4868, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4563758389261745, | |
| "grad_norm": 1.9910852122088196, | |
| "learning_rate": 6.6127496633997475e-06, | |
| "loss": 0.4856, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.46174496644295304, | |
| "grad_norm": 2.0142055270514425, | |
| "learning_rate": 6.523651220265269e-06, | |
| "loss": 0.4942, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4671140939597315, | |
| "grad_norm": 2.102485865135929, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.4848, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.47248322147651006, | |
| "grad_norm": 2.1181467567869623, | |
| "learning_rate": 6.343876061773385e-06, | |
| "loss": 0.4886, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4778523489932886, | |
| "grad_norm": 2.294633696659754, | |
| "learning_rate": 6.2532626612936035e-06, | |
| "loss": 0.4862, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.48322147651006714, | |
| "grad_norm": 1.9401139366264295, | |
| "learning_rate": 6.162207875181354e-06, | |
| "loss": 0.4834, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4885906040268456, | |
| "grad_norm": 1.8938897786171476, | |
| "learning_rate": 6.070743771953157e-06, | |
| "loss": 0.4769, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.49395973154362416, | |
| "grad_norm": 2.035803586489251, | |
| "learning_rate": 5.978902564282616e-06, | |
| "loss": 0.4819, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4993288590604027, | |
| "grad_norm": 1.8943417586526035, | |
| "learning_rate": 5.886716597655472e-06, | |
| "loss": 0.4759, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5046979865771812, | |
| "grad_norm": 1.8388956694087875, | |
| "learning_rate": 5.7942183389778536e-06, | |
| "loss": 0.4683, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5100671140939598, | |
| "grad_norm": 1.9348760408299013, | |
| "learning_rate": 5.701440365141799e-06, | |
| "loss": 0.4644, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5154362416107383, | |
| "grad_norm": 2.0489951854475796, | |
| "learning_rate": 5.608415351552014e-06, | |
| "loss": 0.4662, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5208053691275167, | |
| "grad_norm": 1.9191064778476148, | |
| "learning_rate": 5.515176060617945e-06, | |
| "loss": 0.4629, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5261744966442953, | |
| "grad_norm": 2.1725377637259924, | |
| "learning_rate": 5.421755330215223e-06, | |
| "loss": 0.4562, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5315436241610738, | |
| "grad_norm": 1.9581424804999248, | |
| "learning_rate": 5.328186062120509e-06, | |
| "loss": 0.4635, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5369127516778524, | |
| "grad_norm": 1.8929231268679836, | |
| "learning_rate": 5.23450121042383e-06, | |
| "loss": 0.454, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5422818791946309, | |
| "grad_norm": 1.9802801188021257, | |
| "learning_rate": 5.140733769922525e-06, | |
| "loss": 0.4641, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5476510067114094, | |
| "grad_norm": 2.008695527313478, | |
| "learning_rate": 5.0469167645008245e-06, | |
| "loss": 0.4482, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.553020134228188, | |
| "grad_norm": 2.010797136402869, | |
| "learning_rate": 4.953083235499177e-06, | |
| "loss": 0.4407, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5583892617449664, | |
| "grad_norm": 1.9094101117542013, | |
| "learning_rate": 4.859266230077474e-06, | |
| "loss": 0.4479, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5637583892617449, | |
| "grad_norm": 1.9310237131735821, | |
| "learning_rate": 4.7654987895761705e-06, | |
| "loss": 0.4421, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5691275167785235, | |
| "grad_norm": 1.9040107482286928, | |
| "learning_rate": 4.671813937879494e-06, | |
| "loss": 0.446, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.574496644295302, | |
| "grad_norm": 2.0403382751342387, | |
| "learning_rate": 4.5782446697847775e-06, | |
| "loss": 0.4469, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5798657718120805, | |
| "grad_norm": 1.827855233692127, | |
| "learning_rate": 4.484823939382056e-06, | |
| "loss": 0.4349, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5852348993288591, | |
| "grad_norm": 1.8653574780339666, | |
| "learning_rate": 4.391584648447989e-06, | |
| "loss": 0.4288, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5906040268456376, | |
| "grad_norm": 1.986995981845508, | |
| "learning_rate": 4.298559634858202e-06, | |
| "loss": 0.4372, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5959731543624162, | |
| "grad_norm": 1.9833632848096547, | |
| "learning_rate": 4.205781661022146e-06, | |
| "loss": 0.4332, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6013422818791946, | |
| "grad_norm": 1.9751140496991637, | |
| "learning_rate": 4.1132834023445304e-06, | |
| "loss": 0.4478, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6067114093959731, | |
| "grad_norm": 2.0032559169326456, | |
| "learning_rate": 4.021097435717386e-06, | |
| "loss": 0.4221, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6120805369127517, | |
| "grad_norm": 1.9079295664632885, | |
| "learning_rate": 3.929256228046845e-06, | |
| "loss": 0.4157, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6174496644295302, | |
| "grad_norm": 1.8781628903110001, | |
| "learning_rate": 3.837792124818647e-06, | |
| "loss": 0.4277, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6228187919463087, | |
| "grad_norm": 1.7915826382801951, | |
| "learning_rate": 3.7467373387063973e-06, | |
| "loss": 0.4241, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6281879194630873, | |
| "grad_norm": 1.8613980836008768, | |
| "learning_rate": 3.656123938226618e-06, | |
| "loss": 0.4153, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6335570469798658, | |
| "grad_norm": 1.8036271371193608, | |
| "learning_rate": 3.5659838364445505e-06, | |
| "loss": 0.4087, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6389261744966444, | |
| "grad_norm": 1.876894262488646, | |
| "learning_rate": 3.476348779734732e-06, | |
| "loss": 0.4171, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6442953020134228, | |
| "grad_norm": 2.005365914493537, | |
| "learning_rate": 3.387250336600254e-06, | |
| "loss": 0.4117, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6496644295302013, | |
| "grad_norm": 1.847827057920518, | |
| "learning_rate": 3.298719886554677e-06, | |
| "loss": 0.4129, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6550335570469799, | |
| "grad_norm": 1.8912137540309444, | |
| "learning_rate": 3.2107886090705035e-06, | |
| "loss": 0.4006, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6604026845637584, | |
| "grad_norm": 1.8594474550600166, | |
| "learning_rate": 3.1234874725981045e-06, | |
| "loss": 0.4043, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6657718120805369, | |
| "grad_norm": 1.8497041114366601, | |
| "learning_rate": 3.036847223658958e-06, | |
| "loss": 0.4033, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6711409395973155, | |
| "grad_norm": 1.960254189224025, | |
| "learning_rate": 2.950898376017064e-06, | |
| "loss": 0.4036, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.676510067114094, | |
| "grad_norm": 2.10629670479599, | |
| "learning_rate": 2.865671199932318e-06, | |
| "loss": 0.4044, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6818791946308724, | |
| "grad_norm": 1.8824907032670186, | |
| "learning_rate": 2.781195711499658e-06, | |
| "loss": 0.4095, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.687248322147651, | |
| "grad_norm": 1.9655088963405234, | |
| "learning_rate": 2.697501662077707e-06, | |
| "loss": 0.4021, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6926174496644295, | |
| "grad_norm": 1.889034517852432, | |
| "learning_rate": 2.6146185278106807e-06, | |
| "loss": 0.3825, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.697986577181208, | |
| "grad_norm": 1.781359751443117, | |
| "learning_rate": 2.5325754992471886e-06, | |
| "loss": 0.403, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7033557046979866, | |
| "grad_norm": 2.124217994937085, | |
| "learning_rate": 2.4514014710596467e-06, | |
| "loss": 0.3938, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7087248322147651, | |
| "grad_norm": 2.0049604807831636, | |
| "learning_rate": 2.3711250318678906e-06, | |
| "loss": 0.3943, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7140939597315437, | |
| "grad_norm": 1.8963570644553214, | |
| "learning_rate": 2.2917744541705544e-06, | |
| "loss": 0.3886, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.7194630872483222, | |
| "grad_norm": 1.9231403525555266, | |
| "learning_rate": 2.2133776843878185e-06, | |
| "loss": 0.3918, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7248322147651006, | |
| "grad_norm": 1.9140311429947394, | |
| "learning_rate": 2.1359623330189655e-06, | |
| "loss": 0.3924, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7302013422818792, | |
| "grad_norm": 1.790471011384921, | |
| "learning_rate": 2.059555664918268e-06, | |
| "loss": 0.3861, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7355704697986577, | |
| "grad_norm": 1.8245278498502604, | |
| "learning_rate": 1.9841845896926022e-06, | |
| "loss": 0.3837, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7409395973154362, | |
| "grad_norm": 1.835597641802119, | |
| "learning_rate": 1.9098756522241634e-06, | |
| "loss": 0.3833, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7463087248322148, | |
| "grad_norm": 1.7733798731081618, | |
| "learning_rate": 1.8366550233216584e-06, | |
| "loss": 0.3859, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7516778523489933, | |
| "grad_norm": 1.8909772302753176, | |
| "learning_rate": 1.7645484905032129e-06, | |
| "loss": 0.3725, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7570469798657719, | |
| "grad_norm": 1.8395377457755753, | |
| "learning_rate": 1.6935814489142937e-06, | |
| "loss": 0.375, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7624161073825504, | |
| "grad_norm": 2.004189882277981, | |
| "learning_rate": 1.6237788923838149e-06, | |
| "loss": 0.3745, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7677852348993288, | |
| "grad_norm": 1.8493982881721263, | |
| "learning_rate": 1.555165404621567e-06, | |
| "loss": 0.3851, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7731543624161074, | |
| "grad_norm": 1.8731392963530693, | |
| "learning_rate": 1.487765150560116e-06, | |
| "loss": 0.3719, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7785234899328859, | |
| "grad_norm": 1.800997454509494, | |
| "learning_rate": 1.4216018678441558e-06, | |
| "loss": 0.3735, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7838926174496644, | |
| "grad_norm": 1.858224940166772, | |
| "learning_rate": 1.3566988584703817e-06, | |
| "loss": 0.3674, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.789261744966443, | |
| "grad_norm": 1.8306556119241966, | |
| "learning_rate": 1.293078980580766e-06, | |
| "loss": 0.3731, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7946308724832215, | |
| "grad_norm": 1.7780036225720621, | |
| "learning_rate": 1.2307646404121692e-06, | |
| "loss": 0.362, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.9057965494467726, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.3711, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8053691275167785, | |
| "grad_norm": 1.7752487152814125, | |
| "learning_rate": 1.1101398914744565e-06, | |
| "loss": 0.3644, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.810738255033557, | |
| "grad_norm": 1.8233485934046587, | |
| "learning_rate": 1.0518719654447896e-06, | |
| "loss": 0.3635, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.8161073825503355, | |
| "grad_norm": 1.7956083349387602, | |
| "learning_rate": 9.949945276530782e-07, | |
| "loss": 0.361, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8214765100671141, | |
| "grad_norm": 1.7050898287128047, | |
| "learning_rate": 9.395276097212841e-07, | |
| "loss": 0.3632, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8268456375838926, | |
| "grad_norm": 1.8682808768340036, | |
| "learning_rate": 8.854907465014479e-07, | |
| "loss": 0.3679, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8322147651006712, | |
| "grad_norm": 1.7567008107111888, | |
| "learning_rate": 8.329029691957124e-07, | |
| "loss": 0.3663, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8375838926174497, | |
| "grad_norm": 1.8090009300620273, | |
| "learning_rate": 7.817827986537508e-07, | |
| "loss": 0.3552, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8429530201342282, | |
| "grad_norm": 1.8320510025717955, | |
| "learning_rate": 7.321482388499096e-07, | |
| "loss": 0.3595, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8483221476510067, | |
| "grad_norm": 1.80583578265842, | |
| "learning_rate": 6.840167705424106e-07, | |
| "loss": 0.3652, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8536912751677852, | |
| "grad_norm": 1.8327102347329405, | |
| "learning_rate": 6.374053451168166e-07, | |
| "loss": 0.3561, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8590604026845637, | |
| "grad_norm": 1.8311020050096911, | |
| "learning_rate": 5.92330378615929e-07, | |
| "loss": 0.3585, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8644295302013423, | |
| "grad_norm": 1.8055339358917153, | |
| "learning_rate": 5.488077459582425e-07, | |
| "loss": 0.3593, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8697986577181208, | |
| "grad_norm": 1.8142078683309888, | |
| "learning_rate": 5.068527753469604e-07, | |
| "loss": 0.3556, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8751677852348994, | |
| "grad_norm": 1.8006022957772634, | |
| "learning_rate": 4.664802428715753e-07, | |
| "loss": 0.3501, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8805369127516779, | |
| "grad_norm": 1.8115511806620141, | |
| "learning_rate": 4.2770436730388166e-07, | |
| "loss": 0.3555, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8859060402684564, | |
| "grad_norm": 1.8145406688101844, | |
| "learning_rate": 3.9053880509028086e-07, | |
| "loss": 0.345, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8912751677852349, | |
| "grad_norm": 1.7552883220570528, | |
| "learning_rate": 3.549966455421305e-07, | |
| "loss": 0.3484, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8966442953020134, | |
| "grad_norm": 1.7075911431534658, | |
| "learning_rate": 3.2109040622582186e-07, | |
| "loss": 0.3588, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.9020134228187919, | |
| "grad_norm": 1.6934879805875171, | |
| "learning_rate": 2.8883202855423676e-07, | |
| "loss": 0.3583, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.9073825503355705, | |
| "grad_norm": 1.77775983103448, | |
| "learning_rate": 2.582328735811029e-07, | |
| "loss": 0.351, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.912751677852349, | |
| "grad_norm": 1.8618113993543914, | |
| "learning_rate": 2.2930371799975593e-07, | |
| "loss": 0.3491, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9181208053691275, | |
| "grad_norm": 1.858661836026022, | |
| "learning_rate": 2.0205475034770606e-07, | |
| "loss": 0.3458, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9234899328859061, | |
| "grad_norm": 1.7009249535049955, | |
| "learning_rate": 1.7649556741833995e-07, | |
| "loss": 0.3456, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9288590604026845, | |
| "grad_norm": 1.792030192988469, | |
| "learning_rate": 1.5263517088103862e-07, | |
| "loss": 0.3575, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.934228187919463, | |
| "grad_norm": 1.9256857866336536, | |
| "learning_rate": 1.304819641108801e-07, | |
| "loss": 0.3571, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9395973154362416, | |
| "grad_norm": 1.84021155443979, | |
| "learning_rate": 1.1004374922906846e-07, | |
| "loss": 0.3478, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9449664429530201, | |
| "grad_norm": 1.706774577248585, | |
| "learning_rate": 9.132772435510362e-08, | |
| "loss": 0.3459, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9503355704697987, | |
| "grad_norm": 1.737896821789404, | |
| "learning_rate": 7.434048107168523e-08, | |
| "loss": 0.3487, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9557046979865772, | |
| "grad_norm": 1.754018893680602, | |
| "learning_rate": 5.908800210322696e-08, | |
| "loss": 0.3533, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9610738255033557, | |
| "grad_norm": 1.6854396215038576, | |
| "learning_rate": 4.55756592088058e-08, | |
| "loss": 0.3493, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.9664429530201343, | |
| "grad_norm": 1.7570339706613218, | |
| "learning_rate": 3.3808211290284886e-08, | |
| "loss": 0.3389, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9718120805369127, | |
| "grad_norm": 1.83552844908017, | |
| "learning_rate": 2.378980271628195e-08, | |
| "loss": 0.3455, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9771812080536912, | |
| "grad_norm": 1.6940562777678623, | |
| "learning_rate": 1.552396186256411e-08, | |
| "loss": 0.3478, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9825503355704698, | |
| "grad_norm": 1.7576783825666717, | |
| "learning_rate": 9.013599869394096e-09, | |
| "loss": 0.3393, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9879194630872483, | |
| "grad_norm": 1.7480069996022527, | |
| "learning_rate": 4.261009616257638e-09, | |
| "loss": 0.3499, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9932885906040269, | |
| "grad_norm": 1.6906969685299382, | |
| "learning_rate": 1.2678649143349485e-09, | |
| "loss": 0.3528, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9986577181208054, | |
| "grad_norm": 1.77167724060171, | |
| "learning_rate": 3.5219917003948003e-11, | |
| "loss": 0.3509, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9997315436241611, | |
| "eval_loss": 0.3463418781757355, | |
| "eval_runtime": 1.1928, | |
| "eval_samples_per_second": 1.677, | |
| "eval_steps_per_second": 0.838, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.9997315436241611, | |
| "step": 931, | |
| "total_flos": 194880114524160.0, | |
| "train_loss": 0.5085661182086014, | |
| "train_runtime": 21123.8288, | |
| "train_samples_per_second": 1.41, | |
| "train_steps_per_second": 0.044 | |
| } | |
| ], | |
| "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 | |
| } | |