{ "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "global_step": 113705, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02, "learning_rate": 1.991240490743591e-05, "loss": 0.4731, "step": 500 }, { "epoch": 0.04, "learning_rate": 1.982445802735148e-05, "loss": 0.4151, "step": 1000 }, { "epoch": 0.07, "learning_rate": 1.973668704102722e-05, "loss": 0.3915, "step": 1500 }, { "epoch": 0.09, "learning_rate": 1.9648740160942792e-05, "loss": 0.3795, "step": 2000 }, { "epoch": 0.11, "learning_rate": 1.9560793280858363e-05, "loss": 0.3621, "step": 2500 }, { "epoch": 0.13, "learning_rate": 1.9472846400773934e-05, "loss": 0.3572, "step": 3000 }, { "epoch": 0.15, "learning_rate": 1.9384899520689505e-05, "loss": 0.3508, "step": 3500 }, { "epoch": 0.18, "learning_rate": 1.9296952640605076e-05, "loss": 0.3451, "step": 4000 }, { "epoch": 0.2, "learning_rate": 1.9209357548040985e-05, "loss": 0.3363, "step": 4500 }, { "epoch": 0.22, "learning_rate": 1.9121410667956555e-05, "loss": 0.3358, "step": 5000 }, { "epoch": 0.24, "learning_rate": 1.9033463787872126e-05, "loss": 0.3361, "step": 5500 }, { "epoch": 0.26, "learning_rate": 1.8945516907787697e-05, "loss": 0.3357, "step": 6000 }, { "epoch": 0.29, "learning_rate": 1.8857570027703268e-05, "loss": 0.3401, "step": 6500 }, { "epoch": 0.31, "learning_rate": 1.8769623147618842e-05, "loss": 0.3219, "step": 7000 }, { "epoch": 0.33, "learning_rate": 1.868167626753441e-05, "loss": 0.3215, "step": 7500 }, { "epoch": 0.35, "learning_rate": 1.859372938744998e-05, "loss": 0.3214, "step": 8000 }, { "epoch": 0.37, "learning_rate": 1.850595840112572e-05, "loss": 0.3059, "step": 8500 }, { "epoch": 0.4, "learning_rate": 1.841801152104129e-05, "loss": 0.3091, "step": 9000 }, { "epoch": 0.42, "learning_rate": 1.8330064640956862e-05, "loss": 0.3176, "step": 9500 }, { "epoch": 0.44, "learning_rate": 1.8242117760872437e-05, "loss": 0.3061, "step": 10000 }, { "epoch": 0.46, "learning_rate": 1.8154170880788004e-05, "loss": 0.3113, "step": 10500 }, { "epoch": 0.48, "learning_rate": 1.8066224000703575e-05, "loss": 0.2971, "step": 11000 }, { "epoch": 0.51, "learning_rate": 1.797827712061915e-05, "loss": 0.2939, "step": 11500 }, { "epoch": 0.53, "learning_rate": 1.789033024053472e-05, "loss": 0.3074, "step": 12000 }, { "epoch": 0.55, "learning_rate": 1.780255925421046e-05, "loss": 0.3049, "step": 12500 }, { "epoch": 0.57, "learning_rate": 1.771461237412603e-05, "loss": 0.3016, "step": 13000 }, { "epoch": 0.59, "learning_rate": 1.76266654940416e-05, "loss": 0.3004, "step": 13500 }, { "epoch": 0.62, "learning_rate": 1.753889450771734e-05, "loss": 0.3107, "step": 14000 }, { "epoch": 0.64, "learning_rate": 1.745094762763291e-05, "loss": 0.3007, "step": 14500 }, { "epoch": 0.66, "learning_rate": 1.7363000747548483e-05, "loss": 0.2941, "step": 15000 }, { "epoch": 0.68, "learning_rate": 1.7275053867464054e-05, "loss": 0.2782, "step": 15500 }, { "epoch": 0.7, "learning_rate": 1.7187106987379625e-05, "loss": 0.2985, "step": 16000 }, { "epoch": 0.73, "learning_rate": 1.7099160107295193e-05, "loss": 0.2921, "step": 16500 }, { "epoch": 0.75, "learning_rate": 1.7011213227210767e-05, "loss": 0.2866, "step": 17000 }, { "epoch": 0.77, "learning_rate": 1.6923266347126338e-05, "loss": 0.2932, "step": 17500 }, { "epoch": 0.79, "learning_rate": 1.6835495360802078e-05, "loss": 0.2872, "step": 18000 }, { "epoch": 0.81, "learning_rate": 1.674754848071765e-05, "loss": 0.2884, "step": 18500 }, { "epoch": 0.84, "learning_rate": 1.665960160063322e-05, "loss": 0.2825, "step": 19000 }, { "epoch": 0.86, "learning_rate": 1.657165472054879e-05, "loss": 0.297, "step": 19500 }, { "epoch": 0.88, "learning_rate": 1.648370784046436e-05, "loss": 0.2862, "step": 20000 }, { "epoch": 0.9, "learning_rate": 1.63959368541401e-05, "loss": 0.2925, "step": 20500 }, { "epoch": 0.92, "learning_rate": 1.630816586781584e-05, "loss": 0.2723, "step": 21000 }, { "epoch": 0.95, "learning_rate": 1.6220218987731412e-05, "loss": 0.2803, "step": 21500 }, { "epoch": 0.97, "learning_rate": 1.6132272107646983e-05, "loss": 0.2857, "step": 22000 }, { "epoch": 0.99, "learning_rate": 1.6044325227562554e-05, "loss": 0.2856, "step": 22500 }, { "epoch": 1.01, "learning_rate": 1.5956554241238293e-05, "loss": 0.2384, "step": 23000 }, { "epoch": 1.03, "learning_rate": 1.5868607361153864e-05, "loss": 0.2204, "step": 23500 }, { "epoch": 1.06, "learning_rate": 1.5780660481069435e-05, "loss": 0.2099, "step": 24000 }, { "epoch": 1.08, "learning_rate": 1.5692713600985006e-05, "loss": 0.2198, "step": 24500 }, { "epoch": 1.1, "learning_rate": 1.5604942614660746e-05, "loss": 0.2214, "step": 25000 }, { "epoch": 1.12, "learning_rate": 1.5517347522096655e-05, "loss": 0.2198, "step": 25500 }, { "epoch": 1.14, "learning_rate": 1.5429400642012226e-05, "loss": 0.2174, "step": 26000 }, { "epoch": 1.17, "learning_rate": 1.5341453761927797e-05, "loss": 0.2113, "step": 26500 }, { "epoch": 1.19, "learning_rate": 1.5253506881843367e-05, "loss": 0.2223, "step": 27000 }, { "epoch": 1.21, "learning_rate": 1.5165560001758938e-05, "loss": 0.2237, "step": 27500 }, { "epoch": 1.23, "learning_rate": 1.5077613121674511e-05, "loss": 0.2238, "step": 28000 }, { "epoch": 1.25, "learning_rate": 1.4989666241590082e-05, "loss": 0.2201, "step": 28500 }, { "epoch": 1.28, "learning_rate": 1.4901719361505651e-05, "loss": 0.2153, "step": 29000 }, { "epoch": 1.3, "learning_rate": 1.4813772481421222e-05, "loss": 0.2161, "step": 29500 }, { "epoch": 1.32, "learning_rate": 1.4725825601336794e-05, "loss": 0.2205, "step": 30000 }, { "epoch": 1.34, "learning_rate": 1.4637878721252365e-05, "loss": 0.2156, "step": 30500 }, { "epoch": 1.36, "learning_rate": 1.4549931841167934e-05, "loss": 0.2263, "step": 31000 }, { "epoch": 1.39, "learning_rate": 1.4462160854843676e-05, "loss": 0.2209, "step": 31500 }, { "epoch": 1.41, "learning_rate": 1.4374213974759245e-05, "loss": 0.2295, "step": 32000 }, { "epoch": 1.43, "learning_rate": 1.4286267094674818e-05, "loss": 0.2179, "step": 32500 }, { "epoch": 1.45, "learning_rate": 1.4198320214590389e-05, "loss": 0.2189, "step": 33000 }, { "epoch": 1.47, "learning_rate": 1.411037333450596e-05, "loss": 0.215, "step": 33500 }, { "epoch": 1.5, "learning_rate": 1.4022426454421532e-05, "loss": 0.2234, "step": 34000 }, { "epoch": 1.52, "learning_rate": 1.3934655468097272e-05, "loss": 0.2161, "step": 34500 }, { "epoch": 1.54, "learning_rate": 1.3846708588012841e-05, "loss": 0.2109, "step": 35000 }, { "epoch": 1.56, "learning_rate": 1.3758761707928412e-05, "loss": 0.2167, "step": 35500 }, { "epoch": 1.58, "learning_rate": 1.3670814827843983e-05, "loss": 0.2169, "step": 36000 }, { "epoch": 1.61, "learning_rate": 1.3582867947759556e-05, "loss": 0.223, "step": 36500 }, { "epoch": 1.63, "learning_rate": 1.3495272855195463e-05, "loss": 0.2128, "step": 37000 }, { "epoch": 1.65, "learning_rate": 1.3407325975111034e-05, "loss": 0.2084, "step": 37500 }, { "epoch": 1.67, "learning_rate": 1.3319379095026606e-05, "loss": 0.2194, "step": 38000 }, { "epoch": 1.69, "learning_rate": 1.3231432214942177e-05, "loss": 0.2148, "step": 38500 }, { "epoch": 1.71, "learning_rate": 1.3143485334857746e-05, "loss": 0.2103, "step": 39000 }, { "epoch": 1.74, "learning_rate": 1.3055538454773317e-05, "loss": 0.2205, "step": 39500 }, { "epoch": 1.76, "learning_rate": 1.296759157468889e-05, "loss": 0.2114, "step": 40000 }, { "epoch": 1.78, "learning_rate": 1.287964469460446e-05, "loss": 0.2168, "step": 40500 }, { "epoch": 1.8, "learning_rate": 1.279169781452003e-05, "loss": 0.2204, "step": 41000 }, { "epoch": 1.82, "learning_rate": 1.27037509344356e-05, "loss": 0.2137, "step": 41500 }, { "epoch": 1.85, "learning_rate": 1.2615804054351173e-05, "loss": 0.2147, "step": 42000 }, { "epoch": 1.87, "learning_rate": 1.2527857174266744e-05, "loss": 0.2211, "step": 42500 }, { "epoch": 1.89, "learning_rate": 1.2439910294182317e-05, "loss": 0.2124, "step": 43000 }, { "epoch": 1.91, "learning_rate": 1.2351963414097886e-05, "loss": 0.2146, "step": 43500 }, { "epoch": 1.93, "learning_rate": 1.2264192427773627e-05, "loss": 0.2235, "step": 44000 }, { "epoch": 1.96, "learning_rate": 1.2176245547689197e-05, "loss": 0.208, "step": 44500 }, { "epoch": 1.98, "learning_rate": 1.2088298667604767e-05, "loss": 0.2143, "step": 45000 }, { "epoch": 2.0, "learning_rate": 1.2000351787520338e-05, "loss": 0.2148, "step": 45500 }, { "epoch": 2.02, "learning_rate": 1.1912404907435911e-05, "loss": 0.157, "step": 46000 }, { "epoch": 2.04, "learning_rate": 1.182463392111165e-05, "loss": 0.1545, "step": 46500 }, { "epoch": 2.07, "learning_rate": 1.1736687041027222e-05, "loss": 0.1594, "step": 47000 }, { "epoch": 2.09, "learning_rate": 1.1648916054702962e-05, "loss": 0.1603, "step": 47500 }, { "epoch": 2.11, "learning_rate": 1.156096917461853e-05, "loss": 0.1643, "step": 48000 }, { "epoch": 2.13, "learning_rate": 1.1473022294534102e-05, "loss": 0.1595, "step": 48500 }, { "epoch": 2.15, "learning_rate": 1.1385075414449674e-05, "loss": 0.1603, "step": 49000 }, { "epoch": 2.18, "learning_rate": 1.1297128534365245e-05, "loss": 0.1581, "step": 49500 }, { "epoch": 2.2, "learning_rate": 1.1209357548040985e-05, "loss": 0.1706, "step": 50000 }, { "epoch": 2.22, "learning_rate": 1.1121586561716725e-05, "loss": 0.1682, "step": 50500 }, { "epoch": 2.24, "learning_rate": 1.1033639681632296e-05, "loss": 0.159, "step": 51000 }, { "epoch": 2.26, "learning_rate": 1.0945692801547867e-05, "loss": 0.163, "step": 51500 }, { "epoch": 2.29, "learning_rate": 1.0857745921463436e-05, "loss": 0.184, "step": 52000 }, { "epoch": 2.31, "learning_rate": 1.0769799041379008e-05, "loss": 0.1671, "step": 52500 }, { "epoch": 2.33, "learning_rate": 1.068185216129458e-05, "loss": 0.1633, "step": 53000 }, { "epoch": 2.35, "learning_rate": 1.059390528121015e-05, "loss": 0.1679, "step": 53500 }, { "epoch": 2.37, "learning_rate": 1.050595840112572e-05, "loss": 0.1689, "step": 54000 }, { "epoch": 2.4, "learning_rate": 1.0418187414801462e-05, "loss": 0.168, "step": 54500 }, { "epoch": 2.42, "learning_rate": 1.0330240534717032e-05, "loss": 0.1716, "step": 55000 }, { "epoch": 2.44, "learning_rate": 1.0242293654632603e-05, "loss": 0.1701, "step": 55500 }, { "epoch": 2.46, "learning_rate": 1.0154346774548173e-05, "loss": 0.1673, "step": 56000 }, { "epoch": 2.48, "learning_rate": 1.0066575788223913e-05, "loss": 0.1527, "step": 56500 }, { "epoch": 2.51, "learning_rate": 9.978628908139484e-06, "loss": 0.1674, "step": 57000 }, { "epoch": 2.53, "learning_rate": 9.890682028055055e-06, "loss": 0.1725, "step": 57500 }, { "epoch": 2.55, "learning_rate": 9.802735147970626e-06, "loss": 0.1669, "step": 58000 }, { "epoch": 2.57, "learning_rate": 9.714788267886197e-06, "loss": 0.1722, "step": 58500 }, { "epoch": 2.59, "learning_rate": 9.627017281561938e-06, "loss": 0.168, "step": 59000 }, { "epoch": 2.62, "learning_rate": 9.539070401477508e-06, "loss": 0.1751, "step": 59500 }, { "epoch": 2.64, "learning_rate": 9.451299415153247e-06, "loss": 0.1734, "step": 60000 }, { "epoch": 2.66, "learning_rate": 9.36335253506882e-06, "loss": 0.1658, "step": 60500 }, { "epoch": 2.68, "learning_rate": 9.275405654984391e-06, "loss": 0.1616, "step": 61000 }, { "epoch": 2.7, "learning_rate": 9.187458774899962e-06, "loss": 0.177, "step": 61500 }, { "epoch": 2.73, "learning_rate": 9.099511894815533e-06, "loss": 0.1615, "step": 62000 }, { "epoch": 2.75, "learning_rate": 9.011565014731104e-06, "loss": 0.1619, "step": 62500 }, { "epoch": 2.77, "learning_rate": 8.923618134646674e-06, "loss": 0.1707, "step": 63000 }, { "epoch": 2.79, "learning_rate": 8.835671254562245e-06, "loss": 0.1658, "step": 63500 }, { "epoch": 2.81, "learning_rate": 8.747724374477816e-06, "loss": 0.174, "step": 64000 }, { "epoch": 2.84, "learning_rate": 8.659777494393387e-06, "loss": 0.159, "step": 64500 }, { "epoch": 2.86, "learning_rate": 8.572006508069127e-06, "loss": 0.1691, "step": 65000 }, { "epoch": 2.88, "learning_rate": 8.484235521744867e-06, "loss": 0.1687, "step": 65500 }, { "epoch": 2.9, "learning_rate": 8.396288641660438e-06, "loss": 0.1695, "step": 66000 }, { "epoch": 2.92, "learning_rate": 8.308341761576009e-06, "loss": 0.1718, "step": 66500 }, { "epoch": 2.95, "learning_rate": 8.22039488149158e-06, "loss": 0.1808, "step": 67000 }, { "epoch": 2.97, "learning_rate": 8.13244800140715e-06, "loss": 0.1618, "step": 67500 }, { "epoch": 2.99, "learning_rate": 8.044501121322721e-06, "loss": 0.1715, "step": 68000 }, { "epoch": 3.01, "learning_rate": 7.956730134998461e-06, "loss": 0.1222, "step": 68500 }, { "epoch": 3.03, "learning_rate": 7.868783254914034e-06, "loss": 0.0954, "step": 69000 }, { "epoch": 3.06, "learning_rate": 7.780836374829603e-06, "loss": 0.1102, "step": 69500 }, { "epoch": 3.08, "learning_rate": 7.692889494745175e-06, "loss": 0.1102, "step": 70000 }, { "epoch": 3.1, "learning_rate": 7.6049426146607454e-06, "loss": 0.1179, "step": 70500 }, { "epoch": 3.12, "learning_rate": 7.516995734576316e-06, "loss": 0.1137, "step": 71000 }, { "epoch": 3.14, "learning_rate": 7.429048854491887e-06, "loss": 0.1092, "step": 71500 }, { "epoch": 3.17, "learning_rate": 7.341101974407459e-06, "loss": 0.1074, "step": 72000 }, { "epoch": 3.19, "learning_rate": 7.253330988083198e-06, "loss": 0.1096, "step": 72500 }, { "epoch": 3.21, "learning_rate": 7.16538410799877e-06, "loss": 0.1087, "step": 73000 }, { "epoch": 3.23, "learning_rate": 7.07743722791434e-06, "loss": 0.1194, "step": 73500 }, { "epoch": 3.25, "learning_rate": 6.9894903478299114e-06, "loss": 0.1085, "step": 74000 }, { "epoch": 3.28, "learning_rate": 6.9015434677454815e-06, "loss": 0.1198, "step": 74500 }, { "epoch": 3.3, "learning_rate": 6.813772481421222e-06, "loss": 0.1187, "step": 75000 }, { "epoch": 3.32, "learning_rate": 6.725825601336793e-06, "loss": 0.1093, "step": 75500 }, { "epoch": 3.34, "learning_rate": 6.637878721252364e-06, "loss": 0.1142, "step": 76000 }, { "epoch": 3.36, "learning_rate": 6.549931841167935e-06, "loss": 0.1131, "step": 76500 }, { "epoch": 3.39, "learning_rate": 6.4619849610835065e-06, "loss": 0.1236, "step": 77000 }, { "epoch": 3.41, "learning_rate": 6.374038080999077e-06, "loss": 0.1164, "step": 77500 }, { "epoch": 3.43, "learning_rate": 6.286091200914648e-06, "loss": 0.1046, "step": 78000 }, { "epoch": 3.45, "learning_rate": 6.198320214590387e-06, "loss": 0.1181, "step": 78500 }, { "epoch": 3.47, "learning_rate": 6.110373334505959e-06, "loss": 0.1083, "step": 79000 }, { "epoch": 3.5, "learning_rate": 6.022426454421529e-06, "loss": 0.1121, "step": 79500 }, { "epoch": 3.52, "learning_rate": 5.934479574337101e-06, "loss": 0.1156, "step": 80000 }, { "epoch": 3.54, "learning_rate": 5.8465326942526725e-06, "loss": 0.1282, "step": 80500 }, { "epoch": 3.56, "learning_rate": 5.7585858141682426e-06, "loss": 0.112, "step": 81000 }, { "epoch": 3.58, "learning_rate": 5.670638934083814e-06, "loss": 0.1263, "step": 81500 }, { "epoch": 3.61, "learning_rate": 5.582867947759554e-06, "loss": 0.1021, "step": 82000 }, { "epoch": 3.63, "learning_rate": 5.494921067675124e-06, "loss": 0.1184, "step": 82500 }, { "epoch": 3.65, "learning_rate": 5.406974187590696e-06, "loss": 0.1229, "step": 83000 }, { "epoch": 3.67, "learning_rate": 5.319027307506267e-06, "loss": 0.112, "step": 83500 }, { "epoch": 3.69, "learning_rate": 5.231080427421838e-06, "loss": 0.1189, "step": 84000 }, { "epoch": 3.72, "learning_rate": 5.143309441097577e-06, "loss": 0.1077, "step": 84500 }, { "epoch": 3.74, "learning_rate": 5.0553625610131484e-06, "loss": 0.1129, "step": 85000 }, { "epoch": 3.76, "learning_rate": 4.967415680928719e-06, "loss": 0.1139, "step": 85500 }, { "epoch": 3.78, "learning_rate": 4.879644694604459e-06, "loss": 0.1057, "step": 86000 }, { "epoch": 3.8, "learning_rate": 4.79169781452003e-06, "loss": 0.0967, "step": 86500 }, { "epoch": 3.83, "learning_rate": 4.703750934435602e-06, "loss": 0.1182, "step": 87000 }, { "epoch": 3.85, "learning_rate": 4.615804054351173e-06, "loss": 0.1097, "step": 87500 }, { "epoch": 3.87, "learning_rate": 4.5278571742667436e-06, "loss": 0.1101, "step": 88000 }, { "epoch": 3.89, "learning_rate": 4.4399102941823144e-06, "loss": 0.1192, "step": 88500 }, { "epoch": 3.91, "learning_rate": 4.352139307858054e-06, "loss": 0.1148, "step": 89000 }, { "epoch": 3.94, "learning_rate": 4.264192427773625e-06, "loss": 0.1178, "step": 89500 }, { "epoch": 3.96, "learning_rate": 4.176245547689196e-06, "loss": 0.1114, "step": 90000 }, { "epoch": 3.98, "learning_rate": 4.088298667604767e-06, "loss": 0.1051, "step": 90500 }, { "epoch": 4.0, "learning_rate": 4.000351787520338e-06, "loss": 0.1081, "step": 91000 }, { "epoch": 4.02, "learning_rate": 3.912404907435909e-06, "loss": 0.0648, "step": 91500 }, { "epoch": 4.05, "learning_rate": 3.8244580273514796e-06, "loss": 0.0627, "step": 92000 }, { "epoch": 4.07, "learning_rate": 3.73668704102722e-06, "loss": 0.0557, "step": 92500 }, { "epoch": 4.09, "learning_rate": 3.6487401609427908e-06, "loss": 0.0549, "step": 93000 }, { "epoch": 4.11, "learning_rate": 3.5607932808583616e-06, "loss": 0.0482, "step": 93500 }, { "epoch": 4.13, "learning_rate": 3.4728464007739325e-06, "loss": 0.0658, "step": 94000 }, { "epoch": 4.16, "learning_rate": 3.3848995206895034e-06, "loss": 0.0647, "step": 94500 }, { "epoch": 4.18, "learning_rate": 3.296952640605075e-06, "loss": 0.0655, "step": 95000 }, { "epoch": 4.2, "learning_rate": 3.209005760520646e-06, "loss": 0.0612, "step": 95500 }, { "epoch": 4.22, "learning_rate": 3.121058880436217e-06, "loss": 0.0652, "step": 96000 }, { "epoch": 4.24, "learning_rate": 3.0332878941119563e-06, "loss": 0.0529, "step": 96500 }, { "epoch": 4.27, "learning_rate": 2.945341014027527e-06, "loss": 0.0602, "step": 97000 }, { "epoch": 4.29, "learning_rate": 2.857394133943099e-06, "loss": 0.065, "step": 97500 }, { "epoch": 4.31, "learning_rate": 2.76944725385867e-06, "loss": 0.0581, "step": 98000 }, { "epoch": 4.33, "learning_rate": 2.6815003737742407e-06, "loss": 0.0565, "step": 98500 }, { "epoch": 4.35, "learning_rate": 2.59372938744998e-06, "loss": 0.0687, "step": 99000 }, { "epoch": 4.38, "learning_rate": 2.5059584011257204e-06, "loss": 0.0625, "step": 99500 }, { "epoch": 4.4, "learning_rate": 2.4180115210412913e-06, "loss": 0.0529, "step": 100000 }, { "epoch": 4.42, "learning_rate": 2.330064640956862e-06, "loss": 0.0527, "step": 100500 }, { "epoch": 4.44, "learning_rate": 2.242117760872433e-06, "loss": 0.0501, "step": 101000 }, { "epoch": 4.46, "learning_rate": 2.1541708807880044e-06, "loss": 0.0672, "step": 101500 }, { "epoch": 4.49, "learning_rate": 2.0662240007035753e-06, "loss": 0.0591, "step": 102000 }, { "epoch": 4.51, "learning_rate": 1.978277120619146e-06, "loss": 0.0591, "step": 102500 }, { "epoch": 4.53, "learning_rate": 1.890330240534717e-06, "loss": 0.0558, "step": 103000 }, { "epoch": 4.55, "learning_rate": 1.8023833604502883e-06, "loss": 0.0589, "step": 103500 }, { "epoch": 4.57, "learning_rate": 1.7147882678861968e-06, "loss": 0.0606, "step": 104000 }, { "epoch": 4.6, "learning_rate": 1.6268413878017679e-06, "loss": 0.058, "step": 104500 }, { "epoch": 4.62, "learning_rate": 1.538894507717339e-06, "loss": 0.0613, "step": 105000 }, { "epoch": 4.64, "learning_rate": 1.4509476276329098e-06, "loss": 0.0623, "step": 105500 }, { "epoch": 4.66, "learning_rate": 1.3631766413086497e-06, "loss": 0.0622, "step": 106000 }, { "epoch": 4.68, "learning_rate": 1.2752297612242206e-06, "loss": 0.0519, "step": 106500 }, { "epoch": 4.71, "learning_rate": 1.1872828811397917e-06, "loss": 0.0562, "step": 107000 }, { "epoch": 4.73, "learning_rate": 1.0993360010553625e-06, "loss": 0.0524, "step": 107500 }, { "epoch": 4.75, "learning_rate": 1.0113891209709336e-06, "loss": 0.0707, "step": 108000 }, { "epoch": 4.77, "learning_rate": 9.234422408865046e-07, "loss": 0.0564, "step": 108500 }, { "epoch": 4.79, "learning_rate": 8.354953608020756e-07, "loss": 0.0508, "step": 109000 }, { "epoch": 4.82, "learning_rate": 7.475484807176466e-07, "loss": 0.0644, "step": 109500 }, { "epoch": 4.84, "learning_rate": 6.596016006332176e-07, "loss": 0.0553, "step": 110000 }, { "epoch": 4.86, "learning_rate": 5.718306143089574e-07, "loss": 0.0438, "step": 110500 }, { "epoch": 4.88, "learning_rate": 4.838837342245284e-07, "loss": 0.0442, "step": 111000 }, { "epoch": 4.9, "learning_rate": 3.959368541400994e-07, "loss": 0.0497, "step": 111500 }, { "epoch": 4.93, "learning_rate": 3.0816586781583924e-07, "loss": 0.0648, "step": 112000 }, { "epoch": 4.95, "learning_rate": 2.2021898773141022e-07, "loss": 0.0603, "step": 112500 }, { "epoch": 4.97, "learning_rate": 1.3227210764698123e-07, "loss": 0.0533, "step": 113000 }, { "epoch": 4.99, "learning_rate": 4.4325227562552225e-08, "loss": 0.0525, "step": 113500 }, { "epoch": 5.0, "step": 113705, "total_flos": 1.1883972189487104e+17, "train_loss": 0.17438626093734647, "train_runtime": 5196.6919, "train_samples_per_second": 350.075, "train_steps_per_second": 21.88 } ], "max_steps": 113705, "num_train_epochs": 5, "total_flos": 1.1883972189487104e+17, "trial_name": null, "trial_params": null }