reward-gpt-b6 / checkpoint-3000 /trainer_state.json
bradmin's picture
Training in progress, step 3000, checkpoint
664566b
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.42774039009923576,
"eval_steps": 500,
"global_step": 3000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 8.999954848325688e-06,
"loss": 0.7277,
"step": 10
},
{
"epoch": 0.0,
"learning_rate": 8.999819394208827e-06,
"loss": 0.6455,
"step": 20
},
{
"epoch": 0.0,
"learning_rate": 8.999593640367633e-06,
"loss": 0.5428,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 8.9992775913324e-06,
"loss": 0.3486,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 8.998871253445414e-06,
"loss": 0.3524,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 8.998374634860822e-06,
"loss": 0.2701,
"step": 60
},
{
"epoch": 0.01,
"learning_rate": 8.997787745544478e-06,
"loss": 0.2106,
"step": 70
},
{
"epoch": 0.01,
"learning_rate": 8.997110597273727e-06,
"loss": 0.2596,
"step": 80
},
{
"epoch": 0.01,
"learning_rate": 8.996343203637181e-06,
"loss": 0.2015,
"step": 90
},
{
"epoch": 0.01,
"learning_rate": 8.995485580034448e-06,
"loss": 0.2462,
"step": 100
},
{
"epoch": 0.02,
"learning_rate": 8.99453774367581e-06,
"loss": 0.2172,
"step": 110
},
{
"epoch": 0.02,
"learning_rate": 8.993499713581887e-06,
"loss": 0.2655,
"step": 120
},
{
"epoch": 0.02,
"learning_rate": 8.992371510583257e-06,
"loss": 0.1823,
"step": 130
},
{
"epoch": 0.02,
"learning_rate": 8.991153157320033e-06,
"loss": 0.227,
"step": 140
},
{
"epoch": 0.02,
"learning_rate": 8.989844678241412e-06,
"loss": 0.2243,
"step": 150
},
{
"epoch": 0.02,
"learning_rate": 8.988446099605179e-06,
"loss": 0.1787,
"step": 160
},
{
"epoch": 0.02,
"learning_rate": 8.986957449477188e-06,
"loss": 0.2211,
"step": 170
},
{
"epoch": 0.03,
"learning_rate": 8.98537875773079e-06,
"loss": 0.1683,
"step": 180
},
{
"epoch": 0.03,
"learning_rate": 8.983710056046243e-06,
"loss": 0.1879,
"step": 190
},
{
"epoch": 0.03,
"learning_rate": 8.98195137791007e-06,
"loss": 0.1809,
"step": 200
},
{
"epoch": 0.03,
"learning_rate": 8.980102758614384e-06,
"loss": 0.2204,
"step": 210
},
{
"epoch": 0.03,
"learning_rate": 8.978164235256191e-06,
"loss": 0.189,
"step": 220
},
{
"epoch": 0.03,
"learning_rate": 8.976135846736634e-06,
"loss": 0.1389,
"step": 230
},
{
"epoch": 0.03,
"learning_rate": 8.97401763376022e-06,
"loss": 0.1231,
"step": 240
},
{
"epoch": 0.04,
"learning_rate": 8.971809638833998e-06,
"loss": 0.2018,
"step": 250
},
{
"epoch": 0.04,
"learning_rate": 8.96951190626671e-06,
"loss": 0.1271,
"step": 260
},
{
"epoch": 0.04,
"learning_rate": 8.967124482167896e-06,
"loss": 0.1364,
"step": 270
},
{
"epoch": 0.04,
"learning_rate": 8.964647414446981e-06,
"loss": 0.1658,
"step": 280
},
{
"epoch": 0.04,
"learning_rate": 8.9620807528123e-06,
"loss": 0.1479,
"step": 290
},
{
"epoch": 0.04,
"learning_rate": 8.959424548770104e-06,
"loss": 0.1353,
"step": 300
},
{
"epoch": 0.04,
"learning_rate": 8.95667885562353e-06,
"loss": 0.1761,
"step": 310
},
{
"epoch": 0.05,
"learning_rate": 8.953843728471533e-06,
"loss": 0.1391,
"step": 320
},
{
"epoch": 0.05,
"learning_rate": 8.950919224207774e-06,
"loss": 0.1776,
"step": 330
},
{
"epoch": 0.05,
"learning_rate": 8.94790540151948e-06,
"loss": 0.2089,
"step": 340
},
{
"epoch": 0.05,
"learning_rate": 8.94480232088627e-06,
"loss": 0.1368,
"step": 350
},
{
"epoch": 0.05,
"learning_rate": 8.941610044578937e-06,
"loss": 0.1501,
"step": 360
},
{
"epoch": 0.05,
"learning_rate": 8.938328636658202e-06,
"loss": 0.1807,
"step": 370
},
{
"epoch": 0.05,
"learning_rate": 8.934958162973425e-06,
"loss": 0.13,
"step": 380
},
{
"epoch": 0.06,
"learning_rate": 8.931498691161287e-06,
"loss": 0.1437,
"step": 390
},
{
"epoch": 0.06,
"learning_rate": 8.92795029064443e-06,
"loss": 0.0828,
"step": 400
},
{
"epoch": 0.06,
"learning_rate": 8.924313032630063e-06,
"loss": 0.1041,
"step": 410
},
{
"epoch": 0.06,
"learning_rate": 8.920586990108539e-06,
"loss": 0.2663,
"step": 420
},
{
"epoch": 0.06,
"learning_rate": 8.916772237851882e-06,
"loss": 0.2055,
"step": 430
},
{
"epoch": 0.06,
"learning_rate": 8.912868852412294e-06,
"loss": 0.151,
"step": 440
},
{
"epoch": 0.06,
"learning_rate": 8.908876912120614e-06,
"loss": 0.1638,
"step": 450
},
{
"epoch": 0.07,
"learning_rate": 8.904796497084747e-06,
"loss": 0.1399,
"step": 460
},
{
"epoch": 0.07,
"learning_rate": 8.900627689188059e-06,
"loss": 0.1615,
"step": 470
},
{
"epoch": 0.07,
"learning_rate": 8.89637057208773e-06,
"loss": 0.1209,
"step": 480
},
{
"epoch": 0.07,
"learning_rate": 8.892025231213077e-06,
"loss": 0.1605,
"step": 490
},
{
"epoch": 0.07,
"learning_rate": 8.887591753763842e-06,
"loss": 0.124,
"step": 500
},
{
"epoch": 0.07,
"eval_accuracy": 0.0,
"eval_loss": 0.13451002538204193,
"eval_runtime": 446.6893,
"eval_samples_per_second": 52.338,
"eval_steps_per_second": 8.724,
"step": 500
},
{
"epoch": 0.07,
"learning_rate": 8.883070228708436e-06,
"loss": 0.1045,
"step": 510
},
{
"epoch": 0.07,
"learning_rate": 8.878460746782163e-06,
"loss": 0.2208,
"step": 520
},
{
"epoch": 0.08,
"learning_rate": 8.873763400485387e-06,
"loss": 0.1483,
"step": 530
},
{
"epoch": 0.08,
"learning_rate": 8.868978284081688e-06,
"loss": 0.1322,
"step": 540
},
{
"epoch": 0.08,
"learning_rate": 8.864105493595962e-06,
"loss": 0.1227,
"step": 550
},
{
"epoch": 0.08,
"learning_rate": 8.859145126812497e-06,
"loss": 0.1371,
"step": 560
},
{
"epoch": 0.08,
"learning_rate": 8.854097283273012e-06,
"loss": 0.1909,
"step": 570
},
{
"epoch": 0.08,
"learning_rate": 8.848962064274655e-06,
"loss": 0.0993,
"step": 580
},
{
"epoch": 0.08,
"learning_rate": 8.843739572867977e-06,
"loss": 0.1053,
"step": 590
},
{
"epoch": 0.09,
"learning_rate": 8.838429913854859e-06,
"loss": 0.1401,
"step": 600
},
{
"epoch": 0.09,
"learning_rate": 8.83303319378641e-06,
"loss": 0.09,
"step": 610
},
{
"epoch": 0.09,
"learning_rate": 8.827549520960827e-06,
"loss": 0.1667,
"step": 620
},
{
"epoch": 0.09,
"learning_rate": 8.821979005421223e-06,
"loss": 0.1069,
"step": 630
},
{
"epoch": 0.09,
"learning_rate": 8.816321758953426e-06,
"loss": 0.1659,
"step": 640
},
{
"epoch": 0.09,
"learning_rate": 8.810577895083726e-06,
"loss": 0.1309,
"step": 650
},
{
"epoch": 0.09,
"learning_rate": 8.804747529076592e-06,
"loss": 0.16,
"step": 660
},
{
"epoch": 0.1,
"learning_rate": 8.79883077793238e-06,
"loss": 0.1041,
"step": 670
},
{
"epoch": 0.1,
"learning_rate": 8.792827760384965e-06,
"loss": 0.1344,
"step": 680
},
{
"epoch": 0.1,
"learning_rate": 8.786738596899362e-06,
"loss": 0.0922,
"step": 690
},
{
"epoch": 0.1,
"learning_rate": 8.780563409669322e-06,
"loss": 0.0813,
"step": 700
},
{
"epoch": 0.1,
"learning_rate": 8.774302322614858e-06,
"loss": 0.1795,
"step": 710
},
{
"epoch": 0.1,
"learning_rate": 8.767955461379783e-06,
"loss": 0.1409,
"step": 720
},
{
"epoch": 0.1,
"learning_rate": 8.761522953329163e-06,
"loss": 0.1362,
"step": 730
},
{
"epoch": 0.11,
"learning_rate": 8.75500492754678e-06,
"loss": 0.1227,
"step": 740
},
{
"epoch": 0.11,
"learning_rate": 8.74840151483254e-06,
"loss": 0.0643,
"step": 750
},
{
"epoch": 0.11,
"learning_rate": 8.741712847699832e-06,
"loss": 0.1895,
"step": 760
},
{
"epoch": 0.11,
"learning_rate": 8.734939060372892e-06,
"loss": 0.0959,
"step": 770
},
{
"epoch": 0.11,
"learning_rate": 8.72808028878409e-06,
"loss": 0.1089,
"step": 780
},
{
"epoch": 0.11,
"learning_rate": 8.721136670571216e-06,
"loss": 0.0959,
"step": 790
},
{
"epoch": 0.11,
"learning_rate": 8.714108345074704e-06,
"loss": 0.1264,
"step": 800
},
{
"epoch": 0.12,
"learning_rate": 8.706995453334855e-06,
"loss": 0.1078,
"step": 810
},
{
"epoch": 0.12,
"learning_rate": 8.699798138088988e-06,
"loss": 0.1278,
"step": 820
},
{
"epoch": 0.12,
"learning_rate": 8.692516543768581e-06,
"loss": 0.1337,
"step": 830
},
{
"epoch": 0.12,
"learning_rate": 8.685150816496384e-06,
"loss": 0.0815,
"step": 840
},
{
"epoch": 0.12,
"learning_rate": 8.677701104083467e-06,
"loss": 0.1033,
"step": 850
},
{
"epoch": 0.12,
"learning_rate": 8.670167556026274e-06,
"loss": 0.1412,
"step": 860
},
{
"epoch": 0.12,
"learning_rate": 8.662550323503606e-06,
"loss": 0.197,
"step": 870
},
{
"epoch": 0.13,
"learning_rate": 8.654849559373595e-06,
"loss": 0.1118,
"step": 880
},
{
"epoch": 0.13,
"learning_rate": 8.647065418170644e-06,
"loss": 0.1132,
"step": 890
},
{
"epoch": 0.13,
"learning_rate": 8.639198056102309e-06,
"loss": 0.0928,
"step": 900
},
{
"epoch": 0.13,
"learning_rate": 8.631247631046176e-06,
"loss": 0.0723,
"step": 910
},
{
"epoch": 0.13,
"learning_rate": 8.623214302546691e-06,
"loss": 0.126,
"step": 920
},
{
"epoch": 0.13,
"learning_rate": 8.615098231811958e-06,
"loss": 0.1363,
"step": 930
},
{
"epoch": 0.13,
"learning_rate": 8.606899581710505e-06,
"loss": 0.1243,
"step": 940
},
{
"epoch": 0.14,
"learning_rate": 8.598618516768005e-06,
"loss": 0.0382,
"step": 950
},
{
"epoch": 0.14,
"learning_rate": 8.590255203163995e-06,
"loss": 0.1295,
"step": 960
},
{
"epoch": 0.14,
"learning_rate": 8.581809808728524e-06,
"loss": 0.1145,
"step": 970
},
{
"epoch": 0.14,
"learning_rate": 8.573282502938791e-06,
"loss": 0.0915,
"step": 980
},
{
"epoch": 0.14,
"learning_rate": 8.564673456915746e-06,
"loss": 0.1918,
"step": 990
},
{
"epoch": 0.14,
"learning_rate": 8.55598284342065e-06,
"loss": 0.1732,
"step": 1000
},
{
"epoch": 0.14,
"eval_accuracy": 0.0,
"eval_loss": 0.12965914607048035,
"eval_runtime": 448.5256,
"eval_samples_per_second": 52.124,
"eval_steps_per_second": 8.688,
"step": 1000
},
{
"epoch": 0.14,
"learning_rate": 8.547210836851614e-06,
"loss": 0.0833,
"step": 1010
},
{
"epoch": 0.15,
"learning_rate": 8.538357613240099e-06,
"loss": 0.1597,
"step": 1020
},
{
"epoch": 0.15,
"learning_rate": 8.52942335024738e-06,
"loss": 0.1041,
"step": 1030
},
{
"epoch": 0.15,
"learning_rate": 8.520408227160983e-06,
"loss": 0.1819,
"step": 1040
},
{
"epoch": 0.15,
"learning_rate": 8.511312424891083e-06,
"loss": 0.1112,
"step": 1050
},
{
"epoch": 0.15,
"learning_rate": 8.502136125966886e-06,
"loss": 0.1372,
"step": 1060
},
{
"epoch": 0.15,
"learning_rate": 8.49287951453295e-06,
"loss": 0.1596,
"step": 1070
},
{
"epoch": 0.15,
"learning_rate": 8.483542776345496e-06,
"loss": 0.1558,
"step": 1080
},
{
"epoch": 0.16,
"learning_rate": 8.47412609876869e-06,
"loss": 0.069,
"step": 1090
},
{
"epoch": 0.16,
"learning_rate": 8.464629670770866e-06,
"loss": 0.1468,
"step": 1100
},
{
"epoch": 0.16,
"learning_rate": 8.455053682920747e-06,
"loss": 0.106,
"step": 1110
},
{
"epoch": 0.16,
"learning_rate": 8.445398327383616e-06,
"loss": 0.1562,
"step": 1120
},
{
"epoch": 0.16,
"learning_rate": 8.435663797917459e-06,
"loss": 0.1404,
"step": 1130
},
{
"epoch": 0.16,
"learning_rate": 8.425850289869074e-06,
"loss": 0.1198,
"step": 1140
},
{
"epoch": 0.16,
"learning_rate": 8.415958000170163e-06,
"loss": 0.107,
"step": 1150
},
{
"epoch": 0.17,
"learning_rate": 8.405987127333367e-06,
"loss": 0.0911,
"step": 1160
},
{
"epoch": 0.17,
"learning_rate": 8.395937871448282e-06,
"loss": 0.066,
"step": 1170
},
{
"epoch": 0.17,
"learning_rate": 8.385810434177462e-06,
"loss": 0.0988,
"step": 1180
},
{
"epoch": 0.17,
"learning_rate": 8.375605018752344e-06,
"loss": 0.1648,
"step": 1190
},
{
"epoch": 0.17,
"learning_rate": 8.365321829969196e-06,
"loss": 0.1124,
"step": 1200
},
{
"epoch": 0.17,
"learning_rate": 8.35496107418499e-06,
"loss": 0.1251,
"step": 1210
},
{
"epoch": 0.17,
"learning_rate": 8.34452295931327e-06,
"loss": 0.1167,
"step": 1220
},
{
"epoch": 0.18,
"learning_rate": 8.334007694819972e-06,
"loss": 0.1238,
"step": 1230
},
{
"epoch": 0.18,
"learning_rate": 8.323415491719233e-06,
"loss": 0.1127,
"step": 1240
},
{
"epoch": 0.18,
"learning_rate": 8.312746562569141e-06,
"loss": 0.0823,
"step": 1250
},
{
"epoch": 0.18,
"learning_rate": 8.302001121467481e-06,
"loss": 0.145,
"step": 1260
},
{
"epoch": 0.18,
"learning_rate": 8.291179384047432e-06,
"loss": 0.1257,
"step": 1270
},
{
"epoch": 0.18,
"learning_rate": 8.280281567473247e-06,
"loss": 0.1014,
"step": 1280
},
{
"epoch": 0.18,
"learning_rate": 8.269307890435886e-06,
"loss": 0.1289,
"step": 1290
},
{
"epoch": 0.19,
"learning_rate": 8.258258573148637e-06,
"loss": 0.104,
"step": 1300
},
{
"epoch": 0.19,
"learning_rate": 8.247133837342685e-06,
"loss": 0.1231,
"step": 1310
},
{
"epoch": 0.19,
"learning_rate": 8.235933906262679e-06,
"loss": 0.1594,
"step": 1320
},
{
"epoch": 0.19,
"learning_rate": 8.22465900466223e-06,
"loss": 0.0999,
"step": 1330
},
{
"epoch": 0.19,
"learning_rate": 8.213309358799427e-06,
"loss": 0.061,
"step": 1340
},
{
"epoch": 0.19,
"learning_rate": 8.201885196432273e-06,
"loss": 0.0618,
"step": 1350
},
{
"epoch": 0.19,
"learning_rate": 8.190386746814125e-06,
"loss": 0.1607,
"step": 1360
},
{
"epoch": 0.2,
"learning_rate": 8.1788142406891e-06,
"loss": 0.0726,
"step": 1370
},
{
"epoch": 0.2,
"learning_rate": 8.167167910287428e-06,
"loss": 0.0594,
"step": 1380
},
{
"epoch": 0.2,
"learning_rate": 8.155447989320808e-06,
"loss": 0.1341,
"step": 1390
},
{
"epoch": 0.2,
"learning_rate": 8.14365471297771e-06,
"loss": 0.1253,
"step": 1400
},
{
"epoch": 0.2,
"learning_rate": 8.131788317918651e-06,
"loss": 0.0865,
"step": 1410
},
{
"epoch": 0.2,
"learning_rate": 8.119849042271458e-06,
"loss": 0.1227,
"step": 1420
},
{
"epoch": 0.2,
"learning_rate": 8.10783712562648e-06,
"loss": 0.1024,
"step": 1430
},
{
"epoch": 0.21,
"learning_rate": 8.095752809031783e-06,
"loss": 0.1495,
"step": 1440
},
{
"epoch": 0.21,
"learning_rate": 8.083596334988313e-06,
"loss": 0.1098,
"step": 1450
},
{
"epoch": 0.21,
"learning_rate": 8.071367947445025e-06,
"loss": 0.1044,
"step": 1460
},
{
"epoch": 0.21,
"learning_rate": 8.059067891794002e-06,
"loss": 0.0757,
"step": 1470
},
{
"epoch": 0.21,
"learning_rate": 8.046696414865507e-06,
"loss": 0.0991,
"step": 1480
},
{
"epoch": 0.21,
"learning_rate": 8.034253764923052e-06,
"loss": 0.141,
"step": 1490
},
{
"epoch": 0.21,
"learning_rate": 8.021740191658405e-06,
"loss": 0.1023,
"step": 1500
},
{
"epoch": 0.21,
"eval_accuracy": 0.0,
"eval_loss": 0.11160164326429367,
"eval_runtime": 448.3846,
"eval_samples_per_second": 52.141,
"eval_steps_per_second": 8.691,
"step": 1500
},
{
"epoch": 0.22,
"learning_rate": 8.009155946186583e-06,
"loss": 0.0782,
"step": 1510
},
{
"epoch": 0.22,
"learning_rate": 7.996501281040809e-06,
"loss": 0.1132,
"step": 1520
},
{
"epoch": 0.22,
"learning_rate": 7.983776450167446e-06,
"loss": 0.042,
"step": 1530
},
{
"epoch": 0.22,
"learning_rate": 7.970981708920902e-06,
"loss": 0.0896,
"step": 1540
},
{
"epoch": 0.22,
"learning_rate": 7.958117314058508e-06,
"loss": 0.1038,
"step": 1550
},
{
"epoch": 0.22,
"learning_rate": 7.945183523735359e-06,
"loss": 0.052,
"step": 1560
},
{
"epoch": 0.22,
"learning_rate": 7.932180597499137e-06,
"loss": 0.1224,
"step": 1570
},
{
"epoch": 0.23,
"learning_rate": 7.919108796284908e-06,
"loss": 0.1473,
"step": 1580
},
{
"epoch": 0.23,
"learning_rate": 7.905968382409874e-06,
"loss": 0.1499,
"step": 1590
},
{
"epoch": 0.23,
"learning_rate": 7.89275961956812e-06,
"loss": 0.0728,
"step": 1600
},
{
"epoch": 0.23,
"learning_rate": 7.879482772825316e-06,
"loss": 0.1159,
"step": 1610
},
{
"epoch": 0.23,
"learning_rate": 7.8661381086134e-06,
"loss": 0.1409,
"step": 1620
},
{
"epoch": 0.23,
"learning_rate": 7.85272589472523e-06,
"loss": 0.1065,
"step": 1630
},
{
"epoch": 0.23,
"learning_rate": 7.839246400309213e-06,
"loss": 0.1386,
"step": 1640
},
{
"epoch": 0.24,
"learning_rate": 7.8256998958639e-06,
"loss": 0.0995,
"step": 1650
},
{
"epoch": 0.24,
"learning_rate": 7.812086653232559e-06,
"loss": 0.0947,
"step": 1660
},
{
"epoch": 0.24,
"learning_rate": 7.798406945597726e-06,
"loss": 0.1507,
"step": 1670
},
{
"epoch": 0.24,
"learning_rate": 7.784661047475712e-06,
"loss": 0.0693,
"step": 1680
},
{
"epoch": 0.24,
"learning_rate": 7.7708492347111e-06,
"loss": 0.1194,
"step": 1690
},
{
"epoch": 0.24,
"learning_rate": 7.756971784471213e-06,
"loss": 0.1297,
"step": 1700
},
{
"epoch": 0.24,
"learning_rate": 7.743028975240545e-06,
"loss": 0.0933,
"step": 1710
},
{
"epoch": 0.25,
"learning_rate": 7.729021086815177e-06,
"loss": 0.1029,
"step": 1720
},
{
"epoch": 0.25,
"learning_rate": 7.71494840029716e-06,
"loss": 0.0572,
"step": 1730
},
{
"epoch": 0.25,
"learning_rate": 7.700811198088876e-06,
"loss": 0.0754,
"step": 1740
},
{
"epoch": 0.25,
"learning_rate": 7.68660976388737e-06,
"loss": 0.0957,
"step": 1750
},
{
"epoch": 0.25,
"learning_rate": 7.672344382678653e-06,
"loss": 0.1321,
"step": 1760
},
{
"epoch": 0.25,
"learning_rate": 7.658015340731994e-06,
"loss": 0.0711,
"step": 1770
},
{
"epoch": 0.25,
"learning_rate": 7.643622925594161e-06,
"loss": 0.1211,
"step": 1780
},
{
"epoch": 0.26,
"learning_rate": 7.629167426083665e-06,
"loss": 0.0726,
"step": 1790
},
{
"epoch": 0.26,
"learning_rate": 7.614649132284948e-06,
"loss": 0.0481,
"step": 1800
},
{
"epoch": 0.26,
"learning_rate": 7.600068335542579e-06,
"loss": 0.1569,
"step": 1810
},
{
"epoch": 0.26,
"learning_rate": 7.585425328455395e-06,
"loss": 0.046,
"step": 1820
},
{
"epoch": 0.26,
"learning_rate": 7.5707204048706365e-06,
"loss": 0.0909,
"step": 1830
},
{
"epoch": 0.26,
"learning_rate": 7.5559538598780434e-06,
"loss": 0.1273,
"step": 1840
},
{
"epoch": 0.26,
"learning_rate": 7.541125989803943e-06,
"loss": 0.0636,
"step": 1850
},
{
"epoch": 0.27,
"learning_rate": 7.526237092205292e-06,
"loss": 0.1124,
"step": 1860
},
{
"epoch": 0.27,
"learning_rate": 7.511287465863719e-06,
"loss": 0.1033,
"step": 1870
},
{
"epoch": 0.27,
"learning_rate": 7.496277410779513e-06,
"loss": 0.1055,
"step": 1880
},
{
"epoch": 0.27,
"learning_rate": 7.481207228165619e-06,
"loss": 0.0778,
"step": 1890
},
{
"epoch": 0.27,
"learning_rate": 7.466077220441581e-06,
"loss": 0.0965,
"step": 1900
},
{
"epoch": 0.27,
"learning_rate": 7.45088769122748e-06,
"loss": 0.0561,
"step": 1910
},
{
"epoch": 0.27,
"learning_rate": 7.435638945337836e-06,
"loss": 0.1318,
"step": 1920
},
{
"epoch": 0.28,
"learning_rate": 7.420331288775502e-06,
"loss": 0.0561,
"step": 1930
},
{
"epoch": 0.28,
"learning_rate": 7.404965028725507e-06,
"loss": 0.0997,
"step": 1940
},
{
"epoch": 0.28,
"learning_rate": 7.3895404735489035e-06,
"loss": 0.135,
"step": 1950
},
{
"epoch": 0.28,
"learning_rate": 7.374057932776579e-06,
"loss": 0.0591,
"step": 1960
},
{
"epoch": 0.28,
"learning_rate": 7.358517717103039e-06,
"loss": 0.0574,
"step": 1970
},
{
"epoch": 0.28,
"learning_rate": 7.342920138380175e-06,
"loss": 0.0589,
"step": 1980
},
{
"epoch": 0.28,
"learning_rate": 7.327265509611007e-06,
"loss": 0.0685,
"step": 1990
},
{
"epoch": 0.29,
"learning_rate": 7.311554144943401e-06,
"loss": 0.0386,
"step": 2000
},
{
"epoch": 0.29,
"eval_accuracy": 0.0,
"eval_loss": 0.14258363842964172,
"eval_runtime": 447.0432,
"eval_samples_per_second": 52.297,
"eval_steps_per_second": 8.717,
"step": 2000
},
{
"epoch": 0.29,
"learning_rate": 7.295786359663766e-06,
"loss": 0.0679,
"step": 2010
},
{
"epoch": 0.29,
"learning_rate": 7.2799624701907275e-06,
"loss": 0.1962,
"step": 2020
},
{
"epoch": 0.29,
"learning_rate": 7.264082794068776e-06,
"loss": 0.0871,
"step": 2030
},
{
"epoch": 0.29,
"learning_rate": 7.248147649961896e-06,
"loss": 0.1062,
"step": 2040
},
{
"epoch": 0.29,
"learning_rate": 7.232157357647171e-06,
"loss": 0.0415,
"step": 2050
},
{
"epoch": 0.29,
"learning_rate": 7.216112238008363e-06,
"loss": 0.1604,
"step": 2060
},
{
"epoch": 0.3,
"learning_rate": 7.200012613029481e-06,
"loss": 0.0401,
"step": 2070
},
{
"epoch": 0.3,
"learning_rate": 7.1838588057883136e-06,
"loss": 0.0683,
"step": 2080
},
{
"epoch": 0.3,
"learning_rate": 7.167651140449947e-06,
"loss": 0.0891,
"step": 2090
},
{
"epoch": 0.3,
"learning_rate": 7.1513899422602585e-06,
"loss": 0.1354,
"step": 2100
},
{
"epoch": 0.3,
"learning_rate": 7.135075537539394e-06,
"loss": 0.0904,
"step": 2110
},
{
"epoch": 0.3,
"learning_rate": 7.118708253675213e-06,
"loss": 0.1138,
"step": 2120
},
{
"epoch": 0.3,
"learning_rate": 7.1022884191167256e-06,
"loss": 0.0917,
"step": 2130
},
{
"epoch": 0.31,
"learning_rate": 7.085816363367499e-06,
"loss": 0.0465,
"step": 2140
},
{
"epoch": 0.31,
"learning_rate": 7.0692924169790395e-06,
"loss": 0.0826,
"step": 2150
},
{
"epoch": 0.31,
"learning_rate": 7.05271691154417e-06,
"loss": 0.0623,
"step": 2160
},
{
"epoch": 0.31,
"learning_rate": 7.036090179690365e-06,
"loss": 0.1098,
"step": 2170
},
{
"epoch": 0.31,
"learning_rate": 7.019412555073087e-06,
"loss": 0.0722,
"step": 2180
},
{
"epoch": 0.31,
"learning_rate": 7.0026843723690755e-06,
"loss": 0.1178,
"step": 2190
},
{
"epoch": 0.31,
"learning_rate": 6.9859059672696475e-06,
"loss": 0.0345,
"step": 2200
},
{
"epoch": 0.32,
"learning_rate": 6.969077676473951e-06,
"loss": 0.0445,
"step": 2210
},
{
"epoch": 0.32,
"learning_rate": 6.95219983768221e-06,
"loss": 0.1169,
"step": 2220
},
{
"epoch": 0.32,
"learning_rate": 6.9352727895889486e-06,
"loss": 0.0563,
"step": 2230
},
{
"epoch": 0.32,
"learning_rate": 6.918296871876196e-06,
"loss": 0.0492,
"step": 2240
},
{
"epoch": 0.32,
"learning_rate": 6.901272425206663e-06,
"loss": 0.1094,
"step": 2250
},
{
"epoch": 0.32,
"learning_rate": 6.884199791216922e-06,
"loss": 0.0515,
"step": 2260
},
{
"epoch": 0.32,
"learning_rate": 6.867079312510525e-06,
"loss": 0.1382,
"step": 2270
},
{
"epoch": 0.33,
"learning_rate": 6.849911332651154e-06,
"loss": 0.1041,
"step": 2280
},
{
"epoch": 0.33,
"learning_rate": 6.832696196155716e-06,
"loss": 0.0913,
"step": 2290
},
{
"epoch": 0.33,
"learning_rate": 6.815434248487425e-06,
"loss": 0.0794,
"step": 2300
},
{
"epoch": 0.33,
"learning_rate": 6.798125836048876e-06,
"loss": 0.0864,
"step": 2310
},
{
"epoch": 0.33,
"learning_rate": 6.780771306175093e-06,
"loss": 0.1276,
"step": 2320
},
{
"epoch": 0.33,
"learning_rate": 6.763371007126555e-06,
"loss": 0.1068,
"step": 2330
},
{
"epoch": 0.33,
"learning_rate": 6.745925288082212e-06,
"loss": 0.1093,
"step": 2340
},
{
"epoch": 0.34,
"learning_rate": 6.728434499132473e-06,
"loss": 0.099,
"step": 2350
},
{
"epoch": 0.34,
"learning_rate": 6.710898991272189e-06,
"loss": 0.0841,
"step": 2360
},
{
"epoch": 0.34,
"learning_rate": 6.693319116393596e-06,
"loss": 0.0863,
"step": 2370
},
{
"epoch": 0.34,
"learning_rate": 6.6756952272792685e-06,
"loss": 0.0726,
"step": 2380
},
{
"epoch": 0.34,
"learning_rate": 6.658027677595027e-06,
"loss": 0.1138,
"step": 2390
},
{
"epoch": 0.34,
"learning_rate": 6.6403168218828494e-06,
"loss": 0.102,
"step": 2400
},
{
"epoch": 0.34,
"learning_rate": 6.622563015553752e-06,
"loss": 0.085,
"step": 2410
},
{
"epoch": 0.35,
"learning_rate": 6.604766614880662e-06,
"loss": 0.0646,
"step": 2420
},
{
"epoch": 0.35,
"learning_rate": 6.586927976991262e-06,
"loss": 0.0487,
"step": 2430
},
{
"epoch": 0.35,
"learning_rate": 6.569047459860825e-06,
"loss": 0.0929,
"step": 2440
},
{
"epoch": 0.35,
"learning_rate": 6.551125422305035e-06,
"loss": 0.097,
"step": 2450
},
{
"epoch": 0.35,
"learning_rate": 6.533162223972784e-06,
"loss": 0.0976,
"step": 2460
},
{
"epoch": 0.35,
"learning_rate": 6.515158225338948e-06,
"loss": 0.1232,
"step": 2470
},
{
"epoch": 0.35,
"learning_rate": 6.497113787697166e-06,
"loss": 0.0646,
"step": 2480
},
{
"epoch": 0.36,
"learning_rate": 6.479029273152582e-06,
"loss": 0.0285,
"step": 2490
},
{
"epoch": 0.36,
"learning_rate": 6.460905044614577e-06,
"loss": 0.0716,
"step": 2500
},
{
"epoch": 0.36,
"eval_accuracy": 0.0,
"eval_loss": 0.11192344129085541,
"eval_runtime": 452.1105,
"eval_samples_per_second": 51.711,
"eval_steps_per_second": 8.62,
"step": 2500
},
{
"epoch": 0.36,
"learning_rate": 6.442741465789492e-06,
"loss": 0.0429,
"step": 2510
},
{
"epoch": 0.36,
"learning_rate": 6.424538901173323e-06,
"loss": 0.0803,
"step": 2520
},
{
"epoch": 0.36,
"learning_rate": 6.406297716044413e-06,
"loss": 0.0858,
"step": 2530
},
{
"epoch": 0.36,
"learning_rate": 6.388018276456118e-06,
"loss": 0.0605,
"step": 2540
},
{
"epoch": 0.36,
"learning_rate": 6.3697009492294605e-06,
"loss": 0.0681,
"step": 2550
},
{
"epoch": 0.37,
"learning_rate": 6.351346101945772e-06,
"loss": 0.0607,
"step": 2560
},
{
"epoch": 0.37,
"learning_rate": 6.332954102939312e-06,
"loss": 0.1236,
"step": 2570
},
{
"epoch": 0.37,
"learning_rate": 6.314525321289882e-06,
"loss": 0.0361,
"step": 2580
},
{
"epoch": 0.37,
"learning_rate": 6.296060126815411e-06,
"loss": 0.0547,
"step": 2590
},
{
"epoch": 0.37,
"learning_rate": 6.277558890064545e-06,
"loss": 0.1043,
"step": 2600
},
{
"epoch": 0.37,
"learning_rate": 6.259021982309201e-06,
"loss": 0.084,
"step": 2610
},
{
"epoch": 0.37,
"learning_rate": 6.240449775537121e-06,
"loss": 0.058,
"step": 2620
},
{
"epoch": 0.37,
"learning_rate": 6.221842642444411e-06,
"loss": 0.0862,
"step": 2630
},
{
"epoch": 0.38,
"learning_rate": 6.20320095642805e-06,
"loss": 0.1117,
"step": 2640
},
{
"epoch": 0.38,
"learning_rate": 6.184525091578414e-06,
"loss": 0.1552,
"step": 2650
},
{
"epoch": 0.38,
"learning_rate": 6.165815422671752e-06,
"loss": 0.0642,
"step": 2660
},
{
"epoch": 0.38,
"learning_rate": 6.147072325162679e-06,
"loss": 0.048,
"step": 2670
},
{
"epoch": 0.38,
"learning_rate": 6.128296175176628e-06,
"loss": 0.0966,
"step": 2680
},
{
"epoch": 0.38,
"learning_rate": 6.109487349502319e-06,
"loss": 0.0853,
"step": 2690
},
{
"epoch": 0.38,
"learning_rate": 6.090646225584179e-06,
"loss": 0.0884,
"step": 2700
},
{
"epoch": 0.39,
"learning_rate": 6.071773181514785e-06,
"loss": 0.0887,
"step": 2710
},
{
"epoch": 0.39,
"learning_rate": 6.052868596027264e-06,
"loss": 0.0736,
"step": 2720
},
{
"epoch": 0.39,
"learning_rate": 6.033932848487701e-06,
"loss": 0.1141,
"step": 2730
},
{
"epoch": 0.39,
"learning_rate": 6.014966318887518e-06,
"loss": 0.0481,
"step": 2740
},
{
"epoch": 0.39,
"learning_rate": 5.995969387835857e-06,
"loss": 0.0891,
"step": 2750
},
{
"epoch": 0.39,
"learning_rate": 5.976942436551939e-06,
"loss": 0.0881,
"step": 2760
},
{
"epoch": 0.39,
"learning_rate": 5.95788584685741e-06,
"loss": 0.0967,
"step": 2770
},
{
"epoch": 0.4,
"learning_rate": 5.938800001168685e-06,
"loss": 0.0753,
"step": 2780
},
{
"epoch": 0.4,
"learning_rate": 5.919685282489269e-06,
"loss": 0.0671,
"step": 2790
},
{
"epoch": 0.4,
"learning_rate": 5.900542074402076e-06,
"loss": 0.1066,
"step": 2800
},
{
"epoch": 0.4,
"learning_rate": 5.881370761061724e-06,
"loss": 0.0888,
"step": 2810
},
{
"epoch": 0.4,
"learning_rate": 5.862171727186837e-06,
"loss": 0.0783,
"step": 2820
},
{
"epoch": 0.4,
"learning_rate": 5.842945358052311e-06,
"loss": 0.0945,
"step": 2830
},
{
"epoch": 0.4,
"learning_rate": 5.823692039481596e-06,
"loss": 0.103,
"step": 2840
},
{
"epoch": 0.41,
"learning_rate": 5.804412157838945e-06,
"loss": 0.0861,
"step": 2850
},
{
"epoch": 0.41,
"learning_rate": 5.785106100021663e-06,
"loss": 0.0627,
"step": 2860
},
{
"epoch": 0.41,
"learning_rate": 5.765774253452345e-06,
"loss": 0.0828,
"step": 2870
},
{
"epoch": 0.41,
"learning_rate": 5.746417006071094e-06,
"loss": 0.0629,
"step": 2880
},
{
"epoch": 0.41,
"learning_rate": 5.727034746327746e-06,
"loss": 0.1149,
"step": 2890
},
{
"epoch": 0.41,
"learning_rate": 5.707627863174073e-06,
"loss": 0.0865,
"step": 2900
},
{
"epoch": 0.41,
"learning_rate": 5.688196746055968e-06,
"loss": 0.0598,
"step": 2910
},
{
"epoch": 0.42,
"learning_rate": 5.668741784905642e-06,
"loss": 0.0564,
"step": 2920
},
{
"epoch": 0.42,
"learning_rate": 5.649263370133794e-06,
"loss": 0.0993,
"step": 2930
},
{
"epoch": 0.42,
"learning_rate": 5.629761892621775e-06,
"loss": 0.0692,
"step": 2940
},
{
"epoch": 0.42,
"learning_rate": 5.610237743713742e-06,
"loss": 0.0787,
"step": 2950
},
{
"epoch": 0.42,
"learning_rate": 5.590691315208815e-06,
"loss": 0.045,
"step": 2960
},
{
"epoch": 0.42,
"learning_rate": 5.571122999353203e-06,
"loss": 0.0221,
"step": 2970
},
{
"epoch": 0.42,
"learning_rate": 5.551533188832339e-06,
"loss": 0.065,
"step": 2980
},
{
"epoch": 0.43,
"learning_rate": 5.531922276763001e-06,
"loss": 0.0351,
"step": 2990
},
{
"epoch": 0.43,
"learning_rate": 5.512290656685415e-06,
"loss": 0.0655,
"step": 3000
},
{
"epoch": 0.43,
"eval_accuracy": 0.0,
"eval_loss": 0.12717388570308685,
"eval_runtime": 447.1731,
"eval_samples_per_second": 52.282,
"eval_steps_per_second": 8.715,
"step": 3000
}
],
"logging_steps": 10,
"max_steps": 7013,
"num_train_epochs": 1,
"save_steps": 500,
"total_flos": 0.0,
"trial_name": null,
"trial_params": null
}