| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.9951749095295535, | |
| "eval_steps": 500, | |
| "global_step": 2484, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.606425702811245e-05, | |
| "loss": 1.1216, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.132530120481928e-05, | |
| "loss": 0.9435, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.738955823293173e-05, | |
| "loss": 0.7337, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 6.345381526104418e-05, | |
| "loss": 0.4553, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 7.951807228915663e-05, | |
| "loss": 0.2854, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 9.558232931726909e-05, | |
| "loss": 0.2014, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 0.00011164658634538152, | |
| "loss": 0.1777, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 0.00012771084337349396, | |
| "loss": 0.164, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.00014377510040160642, | |
| "loss": 0.1561, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 0.00015983935742971888, | |
| "loss": 0.1513, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 0.00017590361445783134, | |
| "loss": 0.1481, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 0.00019196787148594377, | |
| "loss": 0.1468, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 0.00019910514541387027, | |
| "loss": 0.1437, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 0.00019731543624161075, | |
| "loss": 0.1408, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 0.00019552572706935123, | |
| "loss": 0.1414, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.00019373601789709173, | |
| "loss": 0.137, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 0.0001919463087248322, | |
| "loss": 0.1371, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 0.0001901565995525727, | |
| "loss": 0.1367, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 0.0001883668903803132, | |
| "loss": 0.1358, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.0001865771812080537, | |
| "loss": 0.1362, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 0.0001847874720357942, | |
| "loss": 0.1339, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 0.0001829977628635347, | |
| "loss": 0.1345, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 0.00018120805369127517, | |
| "loss": 0.1309, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 0.00017941834451901567, | |
| "loss": 0.1315, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 0.00017762863534675615, | |
| "loss": 0.1353, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 0.00017583892617449665, | |
| "loss": 0.1339, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 0.00017404921700223716, | |
| "loss": 0.1319, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 0.00017225950782997763, | |
| "loss": 0.1328, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00017046979865771814, | |
| "loss": 0.1332, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 0.0001686800894854586, | |
| "loss": 0.131, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.00016689038031319912, | |
| "loss": 0.1278, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 0.00016510067114093962, | |
| "loss": 0.1331, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 0.0001633109619686801, | |
| "loss": 0.1286, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 0.0001615212527964206, | |
| "loss": 0.1276, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 0.00015973154362416107, | |
| "loss": 0.1287, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 0.00015794183445190158, | |
| "loss": 0.1312, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 0.00015615212527964208, | |
| "loss": 0.1298, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 0.00015436241610738256, | |
| "loss": 0.1288, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 0.00015257270693512303, | |
| "loss": 0.1297, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 0.00015078299776286354, | |
| "loss": 0.129, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 0.00014899328859060404, | |
| "loss": 0.1286, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 0.00014720357941834454, | |
| "loss": 0.1289, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 0.00014541387024608502, | |
| "loss": 0.1285, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 0.0001436241610738255, | |
| "loss": 0.1283, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 0.00014183445190156602, | |
| "loss": 0.1247, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 0.0001400447427293065, | |
| "loss": 0.1279, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 0.00013825503355704698, | |
| "loss": 0.1249, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 0.00013646532438478748, | |
| "loss": 0.1244, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 0.00013467561521252796, | |
| "loss": 0.1259, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 0.00013288590604026846, | |
| "loss": 0.1257, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 0.00013109619686800896, | |
| "loss": 0.1249, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 0.00012930648769574944, | |
| "loss": 0.1244, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 0.00012751677852348994, | |
| "loss": 0.125, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 0.00012572706935123044, | |
| "loss": 0.1213, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 0.00012393736017897092, | |
| "loss": 0.1235, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 0.00012214765100671142, | |
| "loss": 0.1265, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 0.0001203579418344519, | |
| "loss": 0.1245, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 0.00011856823266219239, | |
| "loss": 0.1245, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 0.0001167785234899329, | |
| "loss": 0.1249, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 0.00011498881431767338, | |
| "loss": 0.1252, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 0.00011319910514541387, | |
| "loss": 0.1236, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 0.00011140939597315436, | |
| "loss": 0.1213, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 0.00010961968680089485, | |
| "loss": 0.1254, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 0.00010782997762863535, | |
| "loss": 0.1221, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 0.00010604026845637584, | |
| "loss": 0.1207, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 0.00010425055928411633, | |
| "loss": 0.1218, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 0.00010246085011185682, | |
| "loss": 0.1244, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 0.00010067114093959733, | |
| "loss": 0.1218, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 9.888143176733782e-05, | |
| "loss": 0.1245, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 9.70917225950783e-05, | |
| "loss": 0.1229, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 9.53020134228188e-05, | |
| "loss": 0.123, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 9.351230425055928e-05, | |
| "loss": 0.1214, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 9.172259507829977e-05, | |
| "loss": 0.1231, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 8.993288590604028e-05, | |
| "loss": 0.123, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 8.814317673378077e-05, | |
| "loss": 0.123, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 8.635346756152126e-05, | |
| "loss": 0.1179, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 8.456375838926175e-05, | |
| "loss": 0.1233, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 8.277404921700224e-05, | |
| "loss": 0.1195, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 8.098434004474274e-05, | |
| "loss": 0.1195, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 7.919463087248322e-05, | |
| "loss": 0.1204, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 7.740492170022372e-05, | |
| "loss": 0.1197, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 7.561521252796421e-05, | |
| "loss": 0.1195, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 7.382550335570471e-05, | |
| "loss": 0.1205, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 7.203579418344519e-05, | |
| "loss": 0.1205, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 7.024608501118568e-05, | |
| "loss": 0.116, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 6.845637583892618e-05, | |
| "loss": 0.1184, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.1225, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 6.487695749440716e-05, | |
| "loss": 0.1195, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 6.308724832214765e-05, | |
| "loss": 0.1189, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 6.129753914988815e-05, | |
| "loss": 0.1206, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 5.9507829977628635e-05, | |
| "loss": 0.1199, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 5.771812080536914e-05, | |
| "loss": 0.1189, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 5.592841163310962e-05, | |
| "loss": 0.1174, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 5.413870246085011e-05, | |
| "loss": 0.1197, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 5.234899328859061e-05, | |
| "loss": 0.119, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 5.05592841163311e-05, | |
| "loss": 0.1161, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 4.8769574944071586e-05, | |
| "loss": 0.1171, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "learning_rate": 4.697986577181208e-05, | |
| "loss": 0.1201, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 4.519015659955257e-05, | |
| "loss": 0.1191, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 4.340044742729307e-05, | |
| "loss": 0.1193, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 4.161073825503356e-05, | |
| "loss": 0.119, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 3.9821029082774055e-05, | |
| "loss": 0.1188, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 3.8031319910514545e-05, | |
| "loss": 0.1165, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 3.6241610738255034e-05, | |
| "loss": 0.1204, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "learning_rate": 3.4451901565995524e-05, | |
| "loss": 0.1196, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 3.266219239373602e-05, | |
| "loss": 0.1186, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 3.087248322147651e-05, | |
| "loss": 0.1135, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "learning_rate": 2.9082774049217003e-05, | |
| "loss": 0.1196, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 2.7293064876957496e-05, | |
| "loss": 0.1161, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 2.550335570469799e-05, | |
| "loss": 0.1162, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 2.371364653243848e-05, | |
| "loss": 0.1165, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 2.192393736017897e-05, | |
| "loss": 0.1164, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 2.013422818791946e-05, | |
| "loss": 0.116, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 1.8344519015659954e-05, | |
| "loss": 0.1172, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 1.6554809843400447e-05, | |
| "loss": 0.117, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 1.4765100671140942e-05, | |
| "loss": 0.1127, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 1.2975391498881432e-05, | |
| "loss": 0.1157, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 1.1185682326621925e-05, | |
| "loss": 0.1193, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 9.395973154362418e-06, | |
| "loss": 0.1153, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 7.606263982102908e-06, | |
| "loss": 0.1164, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 5.8165548098434e-06, | |
| "loss": 0.1181, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 4.026845637583892e-06, | |
| "loss": 0.1166, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "learning_rate": 2.237136465324385e-06, | |
| "loss": 0.1156, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 4.4742729306487696e-07, | |
| "loss": 0.1148, | |
| "step": 2480 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 2484, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 1.2566003090639028e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |