| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 25.0, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.5333333333333334e-05, | |
| "loss": 1.135, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.1333333333333334e-05, | |
| "loss": 0.9975, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 4.7333333333333336e-05, | |
| "loss": 0.8521, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 6.333333333333333e-05, | |
| "loss": 0.6949, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 7.933333333333334e-05, | |
| "loss": 0.5953, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 9.533333333333334e-05, | |
| "loss": 0.5087, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 0.00011133333333333333, | |
| "loss": 0.4767, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 0.00012733333333333336, | |
| "loss": 0.4546, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 0.00014333333333333334, | |
| "loss": 0.4363, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.00015933333333333332, | |
| "loss": 0.4296, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 0.00017533333333333336, | |
| "loss": 0.3956, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 0.00019133333333333334, | |
| "loss": 0.3875, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 0.0001991851851851852, | |
| "loss": 0.3885, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 0.00019740740740740743, | |
| "loss": 0.3827, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 0.00019562962962962964, | |
| "loss": 0.378, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.00019385185185185187, | |
| "loss": 0.3337, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 0.00019207407407407408, | |
| "loss": 0.3448, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 0.00019029629629629632, | |
| "loss": 0.3409, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 0.00018851851851851853, | |
| "loss": 0.3446, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.00018674074074074074, | |
| "loss": 0.3396, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "learning_rate": 0.00018496296296296297, | |
| "loss": 0.2972, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 0.00018318518518518518, | |
| "loss": 0.3032, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 0.00018140740740740742, | |
| "loss": 0.3051, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 0.00017962962962962963, | |
| "loss": 0.307, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.00017785185185185186, | |
| "loss": 0.3093, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 0.00017607407407407407, | |
| "loss": 0.261, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "learning_rate": 0.0001742962962962963, | |
| "loss": 0.2686, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 0.00017251851851851852, | |
| "loss": 0.2753, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 0.00017074074074074075, | |
| "loss": 0.2802, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 0.00016896296296296296, | |
| "loss": 0.2805, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 0.0001671851851851852, | |
| "loss": 0.2335, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 0.0001654074074074074, | |
| "loss": 0.234, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 0.00016362962962962964, | |
| "loss": 0.2434, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 0.00016185185185185185, | |
| "loss": 0.253, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 0.0001600740740740741, | |
| "loss": 0.2481, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 0.0001582962962962963, | |
| "loss": 0.2053, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "learning_rate": 0.00015651851851851854, | |
| "loss": 0.2058, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 0.00015474074074074074, | |
| "loss": 0.213, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 0.00015296296296296298, | |
| "loss": 0.2155, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 0.0001511851851851852, | |
| "loss": 0.2218, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "learning_rate": 0.00014940740740740743, | |
| "loss": 0.1721, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 0.00014762962962962964, | |
| "loss": 0.1812, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 0.00014585185185185187, | |
| "loss": 0.1858, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 0.00014407407407407408, | |
| "loss": 0.187, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 0.00014229629629629632, | |
| "loss": 0.1929, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 0.00014051851851851853, | |
| "loss": 0.1449, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "learning_rate": 0.00013874074074074076, | |
| "loss": 0.1539, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 0.00013696296296296297, | |
| "loss": 0.1556, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 0.00013518518518518518, | |
| "loss": 0.1615, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.00013340740740740742, | |
| "loss": 0.1704, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 0.00013162962962962963, | |
| "loss": 0.1237, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 0.00012985185185185186, | |
| "loss": 0.1282, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "learning_rate": 0.00012807407407407407, | |
| "loss": 0.1335, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 0.0001262962962962963, | |
| "loss": 0.1401, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 0.00012451851851851852, | |
| "loss": 0.14, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 0.00012274074074074075, | |
| "loss": 0.0996, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "learning_rate": 0.00012096296296296296, | |
| "loss": 0.1093, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "learning_rate": 0.0001191851851851852, | |
| "loss": 0.1096, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "learning_rate": 0.00011740740740740741, | |
| "loss": 0.1137, | |
| "step": 1416 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 0.00011562962962962964, | |
| "loss": 0.1187, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "learning_rate": 0.00011385185185185185, | |
| "loss": 0.0852, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "learning_rate": 0.00011207407407407409, | |
| "loss": 0.0901, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "learning_rate": 0.0001102962962962963, | |
| "loss": 0.0909, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 0.00010851851851851853, | |
| "loss": 0.094, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 0.00010674074074074074, | |
| "loss": 0.1011, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 0.00010496296296296298, | |
| "loss": 0.0703, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 13.4, | |
| "learning_rate": 0.00010318518518518519, | |
| "loss": 0.0726, | |
| "step": 1608 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 0.00010140740740740741, | |
| "loss": 0.0769, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 13.8, | |
| "learning_rate": 9.962962962962963e-05, | |
| "loss": 0.0787, | |
| "step": 1656 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 9.785185185185186e-05, | |
| "loss": 0.0848, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 14.2, | |
| "learning_rate": 9.607407407407408e-05, | |
| "loss": 0.0604, | |
| "step": 1704 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 9.42962962962963e-05, | |
| "loss": 0.0586, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 14.6, | |
| "learning_rate": 9.251851851851852e-05, | |
| "loss": 0.0616, | |
| "step": 1752 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 9.074074074074075e-05, | |
| "loss": 0.0663, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 8.896296296296297e-05, | |
| "loss": 0.0651, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "learning_rate": 8.718518518518519e-05, | |
| "loss": 0.0479, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 15.4, | |
| "learning_rate": 8.540740740740742e-05, | |
| "loss": 0.05, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 8.362962962962964e-05, | |
| "loss": 0.0521, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "learning_rate": 8.185185185185186e-05, | |
| "loss": 0.0523, | |
| "step": 1896 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 8.007407407407408e-05, | |
| "loss": 0.0535, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 16.2, | |
| "learning_rate": 7.82962962962963e-05, | |
| "loss": 0.0402, | |
| "step": 1944 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "learning_rate": 7.651851851851853e-05, | |
| "loss": 0.0406, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 16.6, | |
| "learning_rate": 7.474074074074074e-05, | |
| "loss": 0.0426, | |
| "step": 1992 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "learning_rate": 7.296296296296296e-05, | |
| "loss": 0.0414, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "learning_rate": 7.118518518518518e-05, | |
| "loss": 0.0438, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 6.94074074074074e-05, | |
| "loss": 0.0336, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 17.4, | |
| "learning_rate": 6.762962962962963e-05, | |
| "loss": 0.0332, | |
| "step": 2088 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 6.585185185185185e-05, | |
| "loss": 0.0342, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 17.8, | |
| "learning_rate": 6.407407407407407e-05, | |
| "loss": 0.0342, | |
| "step": 2136 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 6.22962962962963e-05, | |
| "loss": 0.0356, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 18.2, | |
| "learning_rate": 6.051851851851852e-05, | |
| "loss": 0.0259, | |
| "step": 2184 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 5.874074074074074e-05, | |
| "loss": 0.0269, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 18.6, | |
| "learning_rate": 5.6962962962962965e-05, | |
| "loss": 0.0282, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "learning_rate": 5.518518518518519e-05, | |
| "loss": 0.0291, | |
| "step": 2256 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "learning_rate": 5.34074074074074e-05, | |
| "loss": 0.0297, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 5.1629629629629626e-05, | |
| "loss": 0.0236, | |
| "step": 2304 | |
| }, | |
| { | |
| "epoch": 19.4, | |
| "learning_rate": 4.9851851851851855e-05, | |
| "loss": 0.0234, | |
| "step": 2328 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 4.807407407407408e-05, | |
| "loss": 0.0233, | |
| "step": 2352 | |
| }, | |
| { | |
| "epoch": 19.8, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.0231, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 4.4518518518518523e-05, | |
| "loss": 0.0231, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 20.2, | |
| "learning_rate": 4.274074074074074e-05, | |
| "loss": 0.0183, | |
| "step": 2424 | |
| }, | |
| { | |
| "epoch": 20.4, | |
| "learning_rate": 4.096296296296296e-05, | |
| "loss": 0.0196, | |
| "step": 2448 | |
| }, | |
| { | |
| "epoch": 20.6, | |
| "learning_rate": 3.9185185185185185e-05, | |
| "loss": 0.0197, | |
| "step": 2472 | |
| }, | |
| { | |
| "epoch": 20.8, | |
| "learning_rate": 3.740740740740741e-05, | |
| "loss": 0.0198, | |
| "step": 2496 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "learning_rate": 3.562962962962963e-05, | |
| "loss": 0.0196, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 21.2, | |
| "learning_rate": 3.385185185185185e-05, | |
| "loss": 0.0162, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 21.4, | |
| "learning_rate": 3.2074074074074075e-05, | |
| "loss": 0.016, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 21.6, | |
| "learning_rate": 3.02962962962963e-05, | |
| "loss": 0.0166, | |
| "step": 2592 | |
| }, | |
| { | |
| "epoch": 21.8, | |
| "learning_rate": 2.851851851851852e-05, | |
| "loss": 0.0168, | |
| "step": 2616 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "learning_rate": 2.6740740740740743e-05, | |
| "loss": 0.0167, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 22.2, | |
| "learning_rate": 2.4962962962962963e-05, | |
| "loss": 0.0145, | |
| "step": 2664 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "learning_rate": 2.318518518518519e-05, | |
| "loss": 0.014, | |
| "step": 2688 | |
| }, | |
| { | |
| "epoch": 22.6, | |
| "learning_rate": 2.1407407407407408e-05, | |
| "loss": 0.0135, | |
| "step": 2712 | |
| }, | |
| { | |
| "epoch": 22.8, | |
| "learning_rate": 1.962962962962963e-05, | |
| "loss": 0.014, | |
| "step": 2736 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "learning_rate": 1.7851851851851853e-05, | |
| "loss": 0.0144, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 23.2, | |
| "learning_rate": 1.6074074074074076e-05, | |
| "loss": 0.0123, | |
| "step": 2784 | |
| }, | |
| { | |
| "epoch": 23.4, | |
| "learning_rate": 1.4296296296296297e-05, | |
| "loss": 0.0128, | |
| "step": 2808 | |
| }, | |
| { | |
| "epoch": 23.6, | |
| "learning_rate": 1.2518518518518518e-05, | |
| "loss": 0.0128, | |
| "step": 2832 | |
| }, | |
| { | |
| "epoch": 23.8, | |
| "learning_rate": 1.074074074074074e-05, | |
| "loss": 0.0124, | |
| "step": 2856 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "learning_rate": 8.962962962962963e-06, | |
| "loss": 0.0125, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 24.2, | |
| "learning_rate": 7.185185185185185e-06, | |
| "loss": 0.0113, | |
| "step": 2904 | |
| }, | |
| { | |
| "epoch": 24.4, | |
| "learning_rate": 5.407407407407407e-06, | |
| "loss": 0.0112, | |
| "step": 2928 | |
| }, | |
| { | |
| "epoch": 24.6, | |
| "learning_rate": 3.6296296296296302e-06, | |
| "loss": 0.0114, | |
| "step": 2952 | |
| }, | |
| { | |
| "epoch": 24.8, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.0118, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 7.407407407407407e-08, | |
| "loss": 0.0116, | |
| "step": 3000 | |
| } | |
| ], | |
| "max_steps": 3000, | |
| "num_train_epochs": 25, | |
| "total_flos": 4.87764705411072e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |