| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 468, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008547008547008548, | |
| "grad_norm": 0.17424613237380981, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7889, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017094017094017096, | |
| "grad_norm": 0.1736629158258438, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3589, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02564102564102564, | |
| "grad_norm": 0.1898377388715744, | |
| "learning_rate": 6e-05, | |
| "loss": 1.8064, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03418803418803419, | |
| "grad_norm": 0.203221395611763, | |
| "learning_rate": 8e-05, | |
| "loss": 2.0415, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.042735042735042736, | |
| "grad_norm": 0.1799948662519455, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6291, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05128205128205128, | |
| "grad_norm": 0.17798663675785065, | |
| "learning_rate": 0.00012, | |
| "loss": 2.0392, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05982905982905983, | |
| "grad_norm": 0.18702112138271332, | |
| "learning_rate": 0.00014, | |
| "loss": 1.6122, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06837606837606838, | |
| "grad_norm": 0.15779322385787964, | |
| "learning_rate": 0.00016, | |
| "loss": 1.6366, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 0.16097557544708252, | |
| "learning_rate": 0.00018, | |
| "loss": 1.4243, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08547008547008547, | |
| "grad_norm": 0.24233727157115936, | |
| "learning_rate": 0.0002, | |
| "loss": 1.8261, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09401709401709402, | |
| "grad_norm": 0.21837319433689117, | |
| "learning_rate": 0.0001999976474595967, | |
| "loss": 1.8375, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10256410256410256, | |
| "grad_norm": 0.1537744700908661, | |
| "learning_rate": 0.00019999058994907564, | |
| "loss": 1.4731, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.170364648103714, | |
| "learning_rate": 0.00019997882780049847, | |
| "loss": 1.5884, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.11965811965811966, | |
| "grad_norm": 0.15152305364608765, | |
| "learning_rate": 0.0001999623615672837, | |
| "loss": 1.8465, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1282051282051282, | |
| "grad_norm": 0.15080629289150238, | |
| "learning_rate": 0.00019994119202418098, | |
| "loss": 1.7975, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.13675213675213677, | |
| "grad_norm": 0.14586149156093597, | |
| "learning_rate": 0.00019991532016723439, | |
| "loss": 1.4316, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1452991452991453, | |
| "grad_norm": 0.143412247300148, | |
| "learning_rate": 0.00019988474721373568, | |
| "loss": 1.9093, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.1399279683828354, | |
| "learning_rate": 0.00019984947460216707, | |
| "loss": 1.2154, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1623931623931624, | |
| "grad_norm": 0.17332254350185394, | |
| "learning_rate": 0.00019980950399213344, | |
| "loss": 1.5147, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17094017094017094, | |
| "grad_norm": 0.15907926857471466, | |
| "learning_rate": 0.00019976483726428422, | |
| "loss": 1.22, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1794871794871795, | |
| "grad_norm": 0.17602041363716125, | |
| "learning_rate": 0.0001997154765202251, | |
| "loss": 1.2712, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.18803418803418803, | |
| "grad_norm": 0.1826002299785614, | |
| "learning_rate": 0.00019966142408241901, | |
| "loss": 1.5025, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.19658119658119658, | |
| "grad_norm": 0.19422835111618042, | |
| "learning_rate": 0.00019960268249407675, | |
| "loss": 1.5527, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.20512820512820512, | |
| "grad_norm": 0.18412941694259644, | |
| "learning_rate": 0.00019953925451903756, | |
| "loss": 1.4429, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.21367521367521367, | |
| "grad_norm": 0.17428062856197357, | |
| "learning_rate": 0.0001994711431416389, | |
| "loss": 2.1797, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.17244970798492432, | |
| "learning_rate": 0.00019939835156657616, | |
| "loss": 1.9712, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.14775343239307404, | |
| "learning_rate": 0.00019932088321875172, | |
| "loss": 1.4019, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.23931623931623933, | |
| "grad_norm": 0.17615348100662231, | |
| "learning_rate": 0.00019923874174311394, | |
| "loss": 1.1836, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.24786324786324787, | |
| "grad_norm": 0.14396396279335022, | |
| "learning_rate": 0.0001991519310044857, | |
| "loss": 1.7511, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2564102564102564, | |
| "grad_norm": 0.1678997278213501, | |
| "learning_rate": 0.00019906045508738228, | |
| "loss": 1.6097, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.26495726495726496, | |
| "grad_norm": 0.140168696641922, | |
| "learning_rate": 0.0001989643182958196, | |
| "loss": 1.2644, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.27350427350427353, | |
| "grad_norm": 0.1790284663438797, | |
| "learning_rate": 0.00019886352515311134, | |
| "loss": 1.3387, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.28205128205128205, | |
| "grad_norm": 0.160356342792511, | |
| "learning_rate": 0.0001987580804016563, | |
| "loss": 1.6599, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2905982905982906, | |
| "grad_norm": 0.16769911348819733, | |
| "learning_rate": 0.00019864798900271532, | |
| "loss": 1.5747, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.29914529914529914, | |
| "grad_norm": 0.1620282530784607, | |
| "learning_rate": 0.0001985332561361776, | |
| "loss": 1.3, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.14948143064975739, | |
| "learning_rate": 0.00019841388720031727, | |
| "loss": 1.1657, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3162393162393162, | |
| "grad_norm": 0.1771283745765686, | |
| "learning_rate": 0.00019828988781153917, | |
| "loss": 1.4809, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3247863247863248, | |
| "grad_norm": 0.16619449853897095, | |
| "learning_rate": 0.00019816126380411476, | |
| "loss": 1.3213, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.15086504817008972, | |
| "learning_rate": 0.00019802802122990758, | |
| "loss": 1.3772, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3418803418803419, | |
| "grad_norm": 0.18030764162540436, | |
| "learning_rate": 0.00019789016635808837, | |
| "loss": 1.4383, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3504273504273504, | |
| "grad_norm": 0.1466120481491089, | |
| "learning_rate": 0.00019774770567484022, | |
| "loss": 1.4808, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.358974358974359, | |
| "grad_norm": 0.14574064314365387, | |
| "learning_rate": 0.00019760064588305345, | |
| "loss": 1.2758, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.36752136752136755, | |
| "grad_norm": 0.1460503488779068, | |
| "learning_rate": 0.00019744899390201006, | |
| "loss": 1.2585, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.37606837606837606, | |
| "grad_norm": 0.15577222406864166, | |
| "learning_rate": 0.0001972927568670583, | |
| "loss": 1.7296, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.16191129386425018, | |
| "learning_rate": 0.00019713194212927696, | |
| "loss": 1.4425, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.39316239316239315, | |
| "grad_norm": 0.180201917886734, | |
| "learning_rate": 0.00019696655725512933, | |
| "loss": 1.3318, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4017094017094017, | |
| "grad_norm": 0.17621096968650818, | |
| "learning_rate": 0.00019679661002610743, | |
| "loss": 1.3787, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.41025641025641024, | |
| "grad_norm": 0.15179012715816498, | |
| "learning_rate": 0.00019662210843836574, | |
| "loss": 1.2288, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4188034188034188, | |
| "grad_norm": 0.17625261843204498, | |
| "learning_rate": 0.0001964430607023449, | |
| "loss": 1.362, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.42735042735042733, | |
| "grad_norm": 0.2081981599330902, | |
| "learning_rate": 0.00019625947524238563, | |
| "loss": 1.7481, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4358974358974359, | |
| "grad_norm": 0.16698336601257324, | |
| "learning_rate": 0.00019607136069633212, | |
| "loss": 1.3257, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.15349215269088745, | |
| "learning_rate": 0.0001958787259151258, | |
| "loss": 1.2898, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.452991452991453, | |
| "grad_norm": 0.1789768785238266, | |
| "learning_rate": 0.00019568157996238884, | |
| "loss": 1.3996, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.14934459328651428, | |
| "learning_rate": 0.0001954799321139975, | |
| "loss": 1.7246, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4700854700854701, | |
| "grad_norm": 0.15673181414604187, | |
| "learning_rate": 0.00019527379185764612, | |
| "loss": 1.2129, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.47863247863247865, | |
| "grad_norm": 0.19993078708648682, | |
| "learning_rate": 0.00019506316889240027, | |
| "loss": 1.7753, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.48717948717948717, | |
| "grad_norm": 0.1767326444387436, | |
| "learning_rate": 0.00019484807312824067, | |
| "loss": 1.1891, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.49572649572649574, | |
| "grad_norm": 0.2193121314048767, | |
| "learning_rate": 0.0001946285146855968, | |
| "loss": 1.4018, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5042735042735043, | |
| "grad_norm": 0.18395313620567322, | |
| "learning_rate": 0.0001944045038948709, | |
| "loss": 1.2613, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 0.19356532394886017, | |
| "learning_rate": 0.00019417605129595157, | |
| "loss": 1.1606, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5213675213675214, | |
| "grad_norm": 0.20749132335186005, | |
| "learning_rate": 0.0001939431676377183, | |
| "loss": 1.2299, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5299145299145299, | |
| "grad_norm": 0.1726475954055786, | |
| "learning_rate": 0.0001937058638775353, | |
| "loss": 1.0828, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.1452275812625885, | |
| "learning_rate": 0.00019346415118073632, | |
| "loss": 1.9403, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5470085470085471, | |
| "grad_norm": 0.20886297523975372, | |
| "learning_rate": 0.00019321804092009906, | |
| "loss": 1.3598, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.19738954305648804, | |
| "learning_rate": 0.00019296754467531014, | |
| "loss": 1.2469, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5641025641025641, | |
| "grad_norm": 0.23291267454624176, | |
| "learning_rate": 0.00019271267423242024, | |
| "loss": 1.1946, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5726495726495726, | |
| "grad_norm": 0.2172214835882187, | |
| "learning_rate": 0.00019245344158328972, | |
| "loss": 1.5041, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5811965811965812, | |
| "grad_norm": 0.2076561003923416, | |
| "learning_rate": 0.0001921898589250242, | |
| "loss": 1.2608, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5897435897435898, | |
| "grad_norm": 0.187509685754776, | |
| "learning_rate": 0.0001919219386594007, | |
| "loss": 1.2722, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5982905982905983, | |
| "grad_norm": 0.19337813556194305, | |
| "learning_rate": 0.00019164969339228422, | |
| "loss": 1.3573, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6068376068376068, | |
| "grad_norm": 0.2286800742149353, | |
| "learning_rate": 0.00019137313593303463, | |
| "loss": 1.3611, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.17622244358062744, | |
| "learning_rate": 0.00019109227929390378, | |
| "loss": 2.1658, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6239316239316239, | |
| "grad_norm": 0.2051881104707718, | |
| "learning_rate": 0.00019080713668942356, | |
| "loss": 1.2844, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6324786324786325, | |
| "grad_norm": 0.20667484402656555, | |
| "learning_rate": 0.00019051772153578389, | |
| "loss": 1.1692, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6410256410256411, | |
| "grad_norm": 0.21888047456741333, | |
| "learning_rate": 0.00019022404745020163, | |
| "loss": 1.1527, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6495726495726496, | |
| "grad_norm": 0.1887664496898651, | |
| "learning_rate": 0.00018992612825027976, | |
| "loss": 1.2714, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6581196581196581, | |
| "grad_norm": 0.25063708424568176, | |
| "learning_rate": 0.0001896239779533575, | |
| "loss": 1.3904, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.2010890245437622, | |
| "learning_rate": 0.00018931761077585035, | |
| "loss": 1.1843, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6752136752136753, | |
| "grad_norm": 0.20284424722194672, | |
| "learning_rate": 0.00018900704113258165, | |
| "loss": 1.0914, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6837606837606838, | |
| "grad_norm": 0.20292750000953674, | |
| "learning_rate": 0.00018869228363610404, | |
| "loss": 1.1945, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.2656649649143219, | |
| "learning_rate": 0.00018837335309601213, | |
| "loss": 1.3174, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7008547008547008, | |
| "grad_norm": 0.19038134813308716, | |
| "learning_rate": 0.00018805026451824546, | |
| "loss": 1.8758, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7094017094017094, | |
| "grad_norm": 0.19635434448719025, | |
| "learning_rate": 0.00018772303310438275, | |
| "loss": 1.7643, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.717948717948718, | |
| "grad_norm": 0.20333418250083923, | |
| "learning_rate": 0.00018739167425092644, | |
| "loss": 1.7937, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7264957264957265, | |
| "grad_norm": 0.2025371640920639, | |
| "learning_rate": 0.00018705620354857833, | |
| "loss": 1.1183, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7350427350427351, | |
| "grad_norm": 0.22109034657478333, | |
| "learning_rate": 0.00018671663678150607, | |
| "loss": 1.0984, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7435897435897436, | |
| "grad_norm": 0.207436203956604, | |
| "learning_rate": 0.0001863729899266004, | |
| "loss": 1.1577, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7521367521367521, | |
| "grad_norm": 0.23750126361846924, | |
| "learning_rate": 0.0001860252791527236, | |
| "loss": 1.635, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7606837606837606, | |
| "grad_norm": 0.18621255457401276, | |
| "learning_rate": 0.00018567352081994852, | |
| "loss": 1.7687, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.18724265694618225, | |
| "learning_rate": 0.00018531773147878895, | |
| "loss": 1.7867, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.22778168320655823, | |
| "learning_rate": 0.0001849579278694209, | |
| "loss": 1.2273, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.7863247863247863, | |
| "grad_norm": 0.22308748960494995, | |
| "learning_rate": 0.00018459412692089494, | |
| "loss": 1.2479, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7948717948717948, | |
| "grad_norm": 0.21641641855239868, | |
| "learning_rate": 0.0001842263457503397, | |
| "loss": 1.1156, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8034188034188035, | |
| "grad_norm": 0.24323439598083496, | |
| "learning_rate": 0.00018385460166215638, | |
| "loss": 1.3997, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.811965811965812, | |
| "grad_norm": 0.2553081810474396, | |
| "learning_rate": 0.00018347891214720477, | |
| "loss": 1.1956, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8205128205128205, | |
| "grad_norm": 0.192823588848114, | |
| "learning_rate": 0.00018309929488198012, | |
| "loss": 1.813, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8290598290598291, | |
| "grad_norm": 0.24415500462055206, | |
| "learning_rate": 0.00018271576772778154, | |
| "loss": 1.2135, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8376068376068376, | |
| "grad_norm": 0.22568519413471222, | |
| "learning_rate": 0.00018232834872987147, | |
| "loss": 1.8256, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.2517865002155304, | |
| "learning_rate": 0.00018193705611662696, | |
| "loss": 1.2788, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "grad_norm": 0.2798250913619995, | |
| "learning_rate": 0.0001815419082986815, | |
| "loss": 1.4225, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8632478632478633, | |
| "grad_norm": 0.20875456929206848, | |
| "learning_rate": 0.00018114292386805936, | |
| "loss": 1.6024, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8717948717948718, | |
| "grad_norm": 0.1969788372516632, | |
| "learning_rate": 0.00018074012159730032, | |
| "loss": 1.8428, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.8803418803418803, | |
| "grad_norm": 0.20729199051856995, | |
| "learning_rate": 0.00018033352043857675, | |
| "loss": 1.4009, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.21610009670257568, | |
| "learning_rate": 0.00017992313952280172, | |
| "loss": 1.769, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8974358974358975, | |
| "grad_norm": 0.21818678081035614, | |
| "learning_rate": 0.00017950899815872892, | |
| "loss": 1.919, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.905982905982906, | |
| "grad_norm": 0.3048407733440399, | |
| "learning_rate": 0.00017909111583204422, | |
| "loss": 1.7712, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9145299145299145, | |
| "grad_norm": 0.3431422710418701, | |
| "learning_rate": 0.0001786695122044487, | |
| "loss": 1.3565, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.2217400223016739, | |
| "learning_rate": 0.0001782442071127338, | |
| "loss": 1.7571, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9316239316239316, | |
| "grad_norm": 0.29589998722076416, | |
| "learning_rate": 0.0001778152205678477, | |
| "loss": 1.7932, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9401709401709402, | |
| "grad_norm": 0.2844977080821991, | |
| "learning_rate": 0.00017738257275395404, | |
| "loss": 1.3841, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9487179487179487, | |
| "grad_norm": 0.24854885041713715, | |
| "learning_rate": 0.00017694628402748202, | |
| "loss": 1.1902, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9572649572649573, | |
| "grad_norm": 0.2537485957145691, | |
| "learning_rate": 0.0001765063749161688, | |
| "loss": 1.2673, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9658119658119658, | |
| "grad_norm": 0.2731517255306244, | |
| "learning_rate": 0.00017606286611809353, | |
| "loss": 1.2555, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.9743589743589743, | |
| "grad_norm": 0.24051187932491302, | |
| "learning_rate": 0.00017561577850070355, | |
| "loss": 1.3397, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.9829059829059829, | |
| "grad_norm": 0.2200351357460022, | |
| "learning_rate": 0.00017516513309983253, | |
| "loss": 1.1119, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9914529914529915, | |
| "grad_norm": 0.29839056730270386, | |
| "learning_rate": 0.00017471095111871074, | |
| "loss": 1.3709, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.23434622585773468, | |
| "learning_rate": 0.0001742532539269674, | |
| "loss": 1.1741, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0085470085470085, | |
| "grad_norm": 0.23401354253292084, | |
| "learning_rate": 0.00017379206305962526, | |
| "loss": 1.0751, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.017094017094017, | |
| "grad_norm": 0.24140281975269318, | |
| "learning_rate": 0.00017332740021608722, | |
| "loss": 1.077, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0256410256410255, | |
| "grad_norm": 0.27418169379234314, | |
| "learning_rate": 0.00017285928725911562, | |
| "loss": 1.1409, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0341880341880343, | |
| "grad_norm": 0.3128631114959717, | |
| "learning_rate": 0.00017238774621380337, | |
| "loss": 1.1578, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0427350427350428, | |
| "grad_norm": 0.23959437012672424, | |
| "learning_rate": 0.00017191279926653761, | |
| "loss": 1.988, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0512820512820513, | |
| "grad_norm": 0.259440541267395, | |
| "learning_rate": 0.00017143446876395602, | |
| "loss": 1.7755, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.0598290598290598, | |
| "grad_norm": 0.22774603962898254, | |
| "learning_rate": 0.00017095277721189528, | |
| "loss": 1.7069, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.0683760683760684, | |
| "grad_norm": 0.2853987216949463, | |
| "learning_rate": 0.00017046774727433222, | |
| "loss": 1.1868, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.2648456394672394, | |
| "learning_rate": 0.00016997940177231722, | |
| "loss": 1.0607, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.0854700854700854, | |
| "grad_norm": 0.21033689379692078, | |
| "learning_rate": 0.00016948776368290084, | |
| "loss": 1.7483, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.0940170940170941, | |
| "grad_norm": 0.2851806581020355, | |
| "learning_rate": 0.00016899285613805246, | |
| "loss": 0.9906, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.1025641025641026, | |
| "grad_norm": 0.29302412271499634, | |
| "learning_rate": 0.00016849470242357196, | |
| "loss": 1.0733, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.2884501516819, | |
| "learning_rate": 0.00016799332597799413, | |
| "loss": 1.1676, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1196581196581197, | |
| "grad_norm": 0.26243993639945984, | |
| "learning_rate": 0.00016748875039148593, | |
| "loss": 1.5788, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1282051282051282, | |
| "grad_norm": 0.2301926463842392, | |
| "learning_rate": 0.0001669809994047364, | |
| "loss": 1.6793, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.1367521367521367, | |
| "grad_norm": 0.2946681082248688, | |
| "learning_rate": 0.0001664700969078398, | |
| "loss": 1.2275, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1452991452991452, | |
| "grad_norm": 0.2194235473871231, | |
| "learning_rate": 0.00016595606693917142, | |
| "loss": 1.6844, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.24949957430362701, | |
| "learning_rate": 0.00016543893368425666, | |
| "loss": 1.0142, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1623931623931625, | |
| "grad_norm": 0.29410868883132935, | |
| "learning_rate": 0.00016491872147463306, | |
| "loss": 1.0962, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.170940170940171, | |
| "grad_norm": 0.21689262986183167, | |
| "learning_rate": 0.00016439545478670543, | |
| "loss": 2.3927, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.1794871794871795, | |
| "grad_norm": 0.28670018911361694, | |
| "learning_rate": 0.00016386915824059427, | |
| "loss": 1.0522, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.188034188034188, | |
| "grad_norm": 0.25839945673942566, | |
| "learning_rate": 0.00016333985659897735, | |
| "loss": 0.9783, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.1965811965811965, | |
| "grad_norm": 0.3567582964897156, | |
| "learning_rate": 0.00016280757476592466, | |
| "loss": 1.1115, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.205128205128205, | |
| "grad_norm": 0.2851192355155945, | |
| "learning_rate": 0.0001622723377857265, | |
| "loss": 1.1181, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2136752136752136, | |
| "grad_norm": 0.23090043663978577, | |
| "learning_rate": 0.00016173417084171536, | |
| "loss": 1.7546, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.2993018329143524, | |
| "learning_rate": 0.00016119309925508078, | |
| "loss": 1.12, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.3502201437950134, | |
| "learning_rate": 0.0001606491484836782, | |
| "loss": 1.1383, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.2393162393162394, | |
| "grad_norm": 0.3279533386230469, | |
| "learning_rate": 0.00016010234412083086, | |
| "loss": 1.116, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2478632478632479, | |
| "grad_norm": 0.30735889077186584, | |
| "learning_rate": 0.00015955271189412598, | |
| "loss": 1.2025, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2564102564102564, | |
| "grad_norm": 0.3305642008781433, | |
| "learning_rate": 0.00015900027766420393, | |
| "loss": 1.1144, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.264957264957265, | |
| "grad_norm": 0.3071320354938507, | |
| "learning_rate": 0.00015844506742354164, | |
| "loss": 1.1524, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.2735042735042734, | |
| "grad_norm": 0.20446474850177765, | |
| "learning_rate": 0.00015788710729522953, | |
| "loss": 1.5036, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.282051282051282, | |
| "grad_norm": 0.3168157637119293, | |
| "learning_rate": 0.00015732642353174259, | |
| "loss": 1.6971, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2905982905982907, | |
| "grad_norm": 0.2681102156639099, | |
| "learning_rate": 0.0001567630425137049, | |
| "loss": 0.898, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.2991452991452992, | |
| "grad_norm": 0.3173544406890869, | |
| "learning_rate": 0.00015619699074864864, | |
| "loss": 0.9953, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.355228990316391, | |
| "learning_rate": 0.00015562829486976673, | |
| "loss": 1.2261, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.3162393162393162, | |
| "grad_norm": 0.31110596656799316, | |
| "learning_rate": 0.00015505698163465986, | |
| "loss": 1.4145, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.3247863247863247, | |
| "grad_norm": 0.34020379185676575, | |
| "learning_rate": 0.00015448307792407734, | |
| "loss": 1.0687, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.29489684104919434, | |
| "learning_rate": 0.00015390661074065256, | |
| "loss": 1.0736, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.341880341880342, | |
| "grad_norm": 0.26536640524864197, | |
| "learning_rate": 0.00015332760720763232, | |
| "loss": 1.7907, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.3504273504273505, | |
| "grad_norm": 0.37698546051979065, | |
| "learning_rate": 0.00015274609456760073, | |
| "loss": 1.1021, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.358974358974359, | |
| "grad_norm": 0.3642312288284302, | |
| "learning_rate": 0.00015216210018119733, | |
| "loss": 1.3004, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.3675213675213675, | |
| "grad_norm": 0.3472050726413727, | |
| "learning_rate": 0.00015157565152583002, | |
| "loss": 0.9691, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.376068376068376, | |
| "grad_norm": 0.33574604988098145, | |
| "learning_rate": 0.0001509867761943818, | |
| "loss": 1.2569, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.36424562335014343, | |
| "learning_rate": 0.00015039550189391298, | |
| "loss": 1.0695, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.393162393162393, | |
| "grad_norm": 0.2944134771823883, | |
| "learning_rate": 0.0001498018564443571, | |
| "loss": 1.488, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4017094017094016, | |
| "grad_norm": 0.3835750222206116, | |
| "learning_rate": 0.0001492058677772123, | |
| "loss": 1.4408, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.4102564102564101, | |
| "grad_norm": 0.3394653797149658, | |
| "learning_rate": 0.000148607563934227, | |
| "loss": 1.1648, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4188034188034189, | |
| "grad_norm": 0.2613736391067505, | |
| "learning_rate": 0.00014800697306608044, | |
| "loss": 0.8816, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.4273504273504274, | |
| "grad_norm": 0.31841540336608887, | |
| "learning_rate": 0.00014740412343105828, | |
| "loss": 1.3399, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.435897435897436, | |
| "grad_norm": 0.33013221621513367, | |
| "learning_rate": 0.00014679904339372302, | |
| "loss": 1.0821, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.35095587372779846, | |
| "learning_rate": 0.00014619176142357935, | |
| "loss": 1.0055, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.452991452991453, | |
| "grad_norm": 0.3409000635147095, | |
| "learning_rate": 0.0001455823060937347, | |
| "loss": 1.1507, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.3400250971317291, | |
| "learning_rate": 0.00014497070607955476, | |
| "loss": 1.0127, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.4700854700854702, | |
| "grad_norm": 0.3428599536418915, | |
| "learning_rate": 0.00014435699015731448, | |
| "loss": 1.0493, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.4786324786324787, | |
| "grad_norm": 0.3085605502128601, | |
| "learning_rate": 0.00014374118720284388, | |
| "loss": 1.2153, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.4871794871794872, | |
| "grad_norm": 0.26580706238746643, | |
| "learning_rate": 0.00014312332619016965, | |
| "loss": 1.3625, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.4957264957264957, | |
| "grad_norm": 0.34680411219596863, | |
| "learning_rate": 0.0001425034361901516, | |
| "loss": 1.1775, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.5042735042735043, | |
| "grad_norm": 0.317070335149765, | |
| "learning_rate": 0.00014188154636911524, | |
| "loss": 1.7189, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5128205128205128, | |
| "grad_norm": 0.39725956320762634, | |
| "learning_rate": 0.0001412576859874791, | |
| "loss": 1.0802, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.5213675213675213, | |
| "grad_norm": 0.3366641402244568, | |
| "learning_rate": 0.00014063188439837832, | |
| "loss": 1.0153, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.5299145299145298, | |
| "grad_norm": 0.3365512490272522, | |
| "learning_rate": 0.0001400041710462833, | |
| "loss": 1.0023, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.3443971872329712, | |
| "learning_rate": 0.0001393745754656146, | |
| "loss": 1.0784, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.547008547008547, | |
| "grad_norm": 0.3158676028251648, | |
| "learning_rate": 0.00013874312727935292, | |
| "loss": 1.0342, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.3261699676513672, | |
| "learning_rate": 0.00013810985619764572, | |
| "loss": 1.7786, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.564102564102564, | |
| "grad_norm": 0.3323323428630829, | |
| "learning_rate": 0.00013747479201640914, | |
| "loss": 0.984, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.5726495726495726, | |
| "grad_norm": 0.39274442195892334, | |
| "learning_rate": 0.00013683796461592604, | |
| "loss": 1.104, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.5811965811965814, | |
| "grad_norm": 0.3277168273925781, | |
| "learning_rate": 0.00013619940395944027, | |
| "loss": 1.1087, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5897435897435899, | |
| "grad_norm": 0.3574274480342865, | |
| "learning_rate": 0.00013555914009174663, | |
| "loss": 1.0967, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.5982905982905984, | |
| "grad_norm": 0.3351602852344513, | |
| "learning_rate": 0.00013491720313777756, | |
| "loss": 1.5144, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.606837606837607, | |
| "grad_norm": 0.35201290249824524, | |
| "learning_rate": 0.00013427362330118543, | |
| "loss": 1.0471, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.3592745065689087, | |
| "learning_rate": 0.0001336284308629216, | |
| "loss": 1.315, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.623931623931624, | |
| "grad_norm": 0.3243856728076935, | |
| "learning_rate": 0.00013298165617981172, | |
| "loss": 1.345, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6324786324786325, | |
| "grad_norm": 0.3796270787715912, | |
| "learning_rate": 0.00013233332968312715, | |
| "loss": 1.4821, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.641025641025641, | |
| "grad_norm": 0.36001330614089966, | |
| "learning_rate": 0.0001316834818771535, | |
| "loss": 1.1569, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.6495726495726495, | |
| "grad_norm": 0.39287564158439636, | |
| "learning_rate": 0.00013103214333775521, | |
| "loss": 1.1965, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.658119658119658, | |
| "grad_norm": 0.38017645478248596, | |
| "learning_rate": 0.00013037934471093682, | |
| "loss": 1.0135, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.3751702606678009, | |
| "learning_rate": 0.00012972511671140125, | |
| "loss": 1.0617, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.6752136752136753, | |
| "grad_norm": 0.3148857355117798, | |
| "learning_rate": 0.00012906949012110456, | |
| "loss": 1.3559, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.6837606837606838, | |
| "grad_norm": 0.44921553134918213, | |
| "learning_rate": 0.00012841249578780757, | |
| "loss": 1.3055, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.2940453588962555, | |
| "learning_rate": 0.00012775416462362457, | |
| "loss": 1.1186, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7008547008547008, | |
| "grad_norm": 0.31202247738838196, | |
| "learning_rate": 0.00012709452760356884, | |
| "loss": 1.6689, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.7094017094017095, | |
| "grad_norm": 0.34548264741897583, | |
| "learning_rate": 0.00012643361576409516, | |
| "loss": 1.0929, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.717948717948718, | |
| "grad_norm": 0.36168211698532104, | |
| "learning_rate": 0.00012577146020163968, | |
| "loss": 1.1417, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.7264957264957266, | |
| "grad_norm": 0.40510398149490356, | |
| "learning_rate": 0.00012510809207115666, | |
| "loss": 1.0695, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.735042735042735, | |
| "grad_norm": 0.3778018653392792, | |
| "learning_rate": 0.00012444354258465268, | |
| "loss": 1.0432, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.7435897435897436, | |
| "grad_norm": 0.3535318672657013, | |
| "learning_rate": 0.00012377784300971807, | |
| "loss": 1.5086, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.7521367521367521, | |
| "grad_norm": 0.33403706550598145, | |
| "learning_rate": 0.0001231110246680558, | |
| "loss": 1.8465, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.7606837606837606, | |
| "grad_norm": 0.41274493932724, | |
| "learning_rate": 0.00012244311893400763, | |
| "loss": 1.2511, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.40675127506256104, | |
| "learning_rate": 0.00012177415723307808, | |
| "loss": 1.1954, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.33072736859321594, | |
| "learning_rate": 0.00012110417104045575, | |
| "loss": 0.9234, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.7863247863247862, | |
| "grad_norm": 0.33449140191078186, | |
| "learning_rate": 0.00012043319187953241, | |
| "loss": 1.1381, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.7948717948717947, | |
| "grad_norm": 0.3770657181739807, | |
| "learning_rate": 0.00011976125132041974, | |
| "loss": 1.1838, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8034188034188035, | |
| "grad_norm": 0.39121079444885254, | |
| "learning_rate": 0.00011908838097846404, | |
| "loss": 1.2558, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.811965811965812, | |
| "grad_norm": 0.3337252736091614, | |
| "learning_rate": 0.00011841461251275867, | |
| "loss": 1.3797, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.8205128205128205, | |
| "grad_norm": 0.39863598346710205, | |
| "learning_rate": 0.00011773997762465429, | |
| "loss": 1.0192, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.8290598290598292, | |
| "grad_norm": 0.3270687162876129, | |
| "learning_rate": 0.0001170645080562676, | |
| "loss": 1.0908, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.8376068376068377, | |
| "grad_norm": 0.3414366543292999, | |
| "learning_rate": 0.00011638823558898762, | |
| "loss": 0.9952, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.37312328815460205, | |
| "learning_rate": 0.00011571119204198037, | |
| "loss": 1.0204, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.8547008547008548, | |
| "grad_norm": 0.3227859139442444, | |
| "learning_rate": 0.00011503340927069189, | |
| "loss": 1.1573, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.8632478632478633, | |
| "grad_norm": 0.3155539333820343, | |
| "learning_rate": 0.00011435491916534919, | |
| "loss": 1.0304, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.8717948717948718, | |
| "grad_norm": 0.3540095090866089, | |
| "learning_rate": 0.00011367575364946006, | |
| "loss": 1.0772, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.8803418803418803, | |
| "grad_norm": 0.39118003845214844, | |
| "learning_rate": 0.00011299594467831078, | |
| "loss": 1.0519, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.38508284091949463, | |
| "learning_rate": 0.00011231552423746283, | |
| "loss": 1.1445, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.8974358974358974, | |
| "grad_norm": 0.33124974370002747, | |
| "learning_rate": 0.00011163452434124773, | |
| "loss": 1.071, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.9059829059829059, | |
| "grad_norm": 0.3828192353248596, | |
| "learning_rate": 0.00011095297703126093, | |
| "loss": 1.0704, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.9145299145299144, | |
| "grad_norm": 0.35031619668006897, | |
| "learning_rate": 0.00011027091437485404, | |
| "loss": 1.2472, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.32266342639923096, | |
| "learning_rate": 0.00010958836846362621, | |
| "loss": 1.3921, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9316239316239316, | |
| "grad_norm": 0.387516587972641, | |
| "learning_rate": 0.00010890537141191417, | |
| "loss": 0.9988, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.9401709401709402, | |
| "grad_norm": 0.3335883915424347, | |
| "learning_rate": 0.00010822195535528106, | |
| "loss": 1.568, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.9487179487179487, | |
| "grad_norm": 0.3533664345741272, | |
| "learning_rate": 0.00010753815244900458, | |
| "loss": 0.9217, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.9572649572649574, | |
| "grad_norm": 0.40452811121940613, | |
| "learning_rate": 0.00010685399486656406, | |
| "loss": 1.2411, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.965811965811966, | |
| "grad_norm": 0.3665294051170349, | |
| "learning_rate": 0.00010616951479812658, | |
| "loss": 1.0943, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9743589743589745, | |
| "grad_norm": 0.39548802375793457, | |
| "learning_rate": 0.00010548474444903247, | |
| "loss": 1.0392, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.982905982905983, | |
| "grad_norm": 0.3631138503551483, | |
| "learning_rate": 0.00010479971603828, | |
| "loss": 1.103, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.9914529914529915, | |
| "grad_norm": 0.3368523120880127, | |
| "learning_rate": 0.00010411446179700943, | |
| "loss": 1.0023, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.508499264717102, | |
| "learning_rate": 0.00010342901396698659, | |
| "loss": 1.2164, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.0085470085470085, | |
| "grad_norm": 0.31828710436820984, | |
| "learning_rate": 0.00010274340479908568, | |
| "loss": 0.7407, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.017094017094017, | |
| "grad_norm": 0.3500760793685913, | |
| "learning_rate": 0.00010205766655177215, | |
| "loss": 0.8969, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.0256410256410255, | |
| "grad_norm": 0.32439127564430237, | |
| "learning_rate": 0.00010137183148958463, | |
| "loss": 1.04, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.034188034188034, | |
| "grad_norm": 0.3940890431404114, | |
| "learning_rate": 0.00010068593188161697, | |
| "loss": 0.8789, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.0427350427350426, | |
| "grad_norm": 0.4432232975959778, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0523, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.051282051282051, | |
| "grad_norm": 0.39422449469566345, | |
| "learning_rate": 9.931406811838308e-05, | |
| "loss": 0.979, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0598290598290596, | |
| "grad_norm": 0.41618821024894714, | |
| "learning_rate": 9.862816851041541e-05, | |
| "loss": 0.8262, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.0683760683760686, | |
| "grad_norm": 0.3621666431427002, | |
| "learning_rate": 9.79423334482279e-05, | |
| "loss": 0.982, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.41992640495300293, | |
| "learning_rate": 9.725659520091433e-05, | |
| "loss": 0.9739, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.0854700854700856, | |
| "grad_norm": 0.39439642429351807, | |
| "learning_rate": 9.657098603301346e-05, | |
| "loss": 1.0486, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.094017094017094, | |
| "grad_norm": 0.33870458602905273, | |
| "learning_rate": 9.588553820299056e-05, | |
| "loss": 0.8753, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.1025641025641026, | |
| "grad_norm": 0.4117053151130676, | |
| "learning_rate": 9.520028396172003e-05, | |
| "loss": 1.3071, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.111111111111111, | |
| "grad_norm": 0.4193427860736847, | |
| "learning_rate": 9.451525555096753e-05, | |
| "loss": 1.05, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.1196581196581197, | |
| "grad_norm": 0.38846030831336975, | |
| "learning_rate": 9.383048520187344e-05, | |
| "loss": 0.9186, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.128205128205128, | |
| "grad_norm": 0.4135594069957733, | |
| "learning_rate": 9.314600513343595e-05, | |
| "loss": 0.8908, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.1367521367521367, | |
| "grad_norm": 0.3798341155052185, | |
| "learning_rate": 9.246184755099545e-05, | |
| "loss": 1.3978, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.1452991452991452, | |
| "grad_norm": 0.3479439914226532, | |
| "learning_rate": 9.177804464471898e-05, | |
| "loss": 1.5296, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.37074774503707886, | |
| "learning_rate": 9.109462858808586e-05, | |
| "loss": 1.6774, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.1623931623931623, | |
| "grad_norm": 0.40872663259506226, | |
| "learning_rate": 9.041163153637381e-05, | |
| "loss": 0.9386, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.1709401709401708, | |
| "grad_norm": 0.47935447096824646, | |
| "learning_rate": 8.972908562514598e-05, | |
| "loss": 0.9946, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.1794871794871793, | |
| "grad_norm": 0.48801565170288086, | |
| "learning_rate": 8.904702296873912e-05, | |
| "loss": 1.0447, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.1880341880341883, | |
| "grad_norm": 0.4188629686832428, | |
| "learning_rate": 8.836547565875227e-05, | |
| "loss": 0.9054, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.1965811965811968, | |
| "grad_norm": 0.3567338287830353, | |
| "learning_rate": 8.76844757625372e-05, | |
| "loss": 1.4312, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.2051282051282053, | |
| "grad_norm": 0.4254816174507141, | |
| "learning_rate": 8.70040553216892e-05, | |
| "loss": 0.8604, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.213675213675214, | |
| "grad_norm": 0.4451569616794586, | |
| "learning_rate": 8.632424635053997e-05, | |
| "loss": 1.0722, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.3565747141838074, | |
| "learning_rate": 8.564508083465079e-05, | |
| "loss": 0.8556, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.4148208498954773, | |
| "learning_rate": 8.496659072930813e-05, | |
| "loss": 0.9558, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.2393162393162394, | |
| "grad_norm": 0.3620738685131073, | |
| "learning_rate": 8.428880795801965e-05, | |
| "loss": 0.822, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.247863247863248, | |
| "grad_norm": 0.42495375871658325, | |
| "learning_rate": 8.36117644110124e-05, | |
| "loss": 1.383, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.2564102564102564, | |
| "grad_norm": 0.45574724674224854, | |
| "learning_rate": 8.293549194373243e-05, | |
| "loss": 0.8652, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.264957264957265, | |
| "grad_norm": 0.4112643301486969, | |
| "learning_rate": 8.226002237534572e-05, | |
| "loss": 0.8556, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.2735042735042734, | |
| "grad_norm": 0.5019224882125854, | |
| "learning_rate": 8.158538748724139e-05, | |
| "loss": 1.4962, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.282051282051282, | |
| "grad_norm": 0.47616204619407654, | |
| "learning_rate": 8.091161902153595e-05, | |
| "loss": 0.9432, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.2905982905982905, | |
| "grad_norm": 0.3611639440059662, | |
| "learning_rate": 8.023874867958027e-05, | |
| "loss": 1.2839, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.299145299145299, | |
| "grad_norm": 0.41589438915252686, | |
| "learning_rate": 7.95668081204676e-05, | |
| "loss": 1.3589, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.4448729157447815, | |
| "learning_rate": 7.889582895954427e-05, | |
| "loss": 0.9009, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.316239316239316, | |
| "grad_norm": 0.4523106813430786, | |
| "learning_rate": 7.822584276692191e-05, | |
| "loss": 0.8579, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.324786324786325, | |
| "grad_norm": 0.47008445858955383, | |
| "learning_rate": 7.755688106599241e-05, | |
| "loss": 0.9641, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.4380154013633728, | |
| "learning_rate": 7.688897533194424e-05, | |
| "loss": 1.0402, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.341880341880342, | |
| "grad_norm": 0.49563342332839966, | |
| "learning_rate": 7.622215699028196e-05, | |
| "loss": 0.9568, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.3504273504273505, | |
| "grad_norm": 0.38657042384147644, | |
| "learning_rate": 7.555645741534736e-05, | |
| "loss": 0.7872, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.358974358974359, | |
| "grad_norm": 0.4724520146846771, | |
| "learning_rate": 7.489190792884338e-05, | |
| "loss": 1.6095, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.3675213675213675, | |
| "grad_norm": 0.5585160255432129, | |
| "learning_rate": 7.422853979836034e-05, | |
| "loss": 1.3791, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.376068376068376, | |
| "grad_norm": 0.4355701506137848, | |
| "learning_rate": 7.356638423590485e-05, | |
| "loss": 0.9957, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.4084435999393463, | |
| "learning_rate": 7.290547239643117e-05, | |
| "loss": 0.966, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.393162393162393, | |
| "grad_norm": 0.4027854800224304, | |
| "learning_rate": 7.224583537637544e-05, | |
| "loss": 1.0648, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.4017094017094016, | |
| "grad_norm": 0.41193458437919617, | |
| "learning_rate": 7.158750421219244e-05, | |
| "loss": 1.7693, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.41025641025641, | |
| "grad_norm": 0.46353569626808167, | |
| "learning_rate": 7.093050987889547e-05, | |
| "loss": 0.9429, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.4188034188034186, | |
| "grad_norm": 0.41335809230804443, | |
| "learning_rate": 7.027488328859876e-05, | |
| "loss": 0.7955, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.427350427350427, | |
| "grad_norm": 0.3690212666988373, | |
| "learning_rate": 6.96206552890632e-05, | |
| "loss": 0.9326, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.435897435897436, | |
| "grad_norm": 0.4575774371623993, | |
| "learning_rate": 6.896785666224481e-05, | |
| "loss": 0.9425, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 0.4913078844547272, | |
| "learning_rate": 6.831651812284652e-05, | |
| "loss": 1.0035, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.452991452991453, | |
| "grad_norm": 0.4358192980289459, | |
| "learning_rate": 6.766667031687286e-05, | |
| "loss": 1.0477, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.4812622368335724, | |
| "learning_rate": 6.701834382018832e-05, | |
| "loss": 0.9168, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.47008547008547, | |
| "grad_norm": 0.4503976106643677, | |
| "learning_rate": 6.637156913707839e-05, | |
| "loss": 0.8684, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.4786324786324787, | |
| "grad_norm": 0.43528708815574646, | |
| "learning_rate": 6.572637669881458e-05, | |
| "loss": 1.0104, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4871794871794872, | |
| "grad_norm": 0.38487404584884644, | |
| "learning_rate": 6.508279686222243e-05, | |
| "loss": 1.5843, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.4957264957264957, | |
| "grad_norm": 0.44947412610054016, | |
| "learning_rate": 6.444085990825338e-05, | |
| "loss": 1.0874, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.5042735042735043, | |
| "grad_norm": 0.42089954018592834, | |
| "learning_rate": 6.380059604055974e-05, | |
| "loss": 0.8293, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.5128205128205128, | |
| "grad_norm": 0.5496504902839661, | |
| "learning_rate": 6.316203538407397e-05, | |
| "loss": 1.014, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.5213675213675213, | |
| "grad_norm": 0.4612962305545807, | |
| "learning_rate": 6.252520798359092e-05, | |
| "loss": 0.9299, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.52991452991453, | |
| "grad_norm": 0.4219933748245239, | |
| "learning_rate": 6.18901438023543e-05, | |
| "loss": 0.8919, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.4353472888469696, | |
| "learning_rate": 6.125687272064713e-05, | |
| "loss": 0.9507, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.547008547008547, | |
| "grad_norm": 0.34943369030952454, | |
| "learning_rate": 6.0625424534385425e-05, | |
| "loss": 1.1956, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.5555555555555554, | |
| "grad_norm": 0.516254723072052, | |
| "learning_rate": 5.9995828953716695e-05, | |
| "loss": 0.9514, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.564102564102564, | |
| "grad_norm": 0.46393638849258423, | |
| "learning_rate": 5.936811560162169e-05, | |
| "loss": 1.0359, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.5726495726495724, | |
| "grad_norm": 0.43358251452445984, | |
| "learning_rate": 5.87423140125209e-05, | |
| "loss": 0.8196, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.5811965811965814, | |
| "grad_norm": 0.43277421593666077, | |
| "learning_rate": 5.811845363088477e-05, | |
| "loss": 0.8026, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.58974358974359, | |
| "grad_norm": 0.48290401697158813, | |
| "learning_rate": 5.749656380984844e-05, | |
| "loss": 0.9704, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.5982905982905984, | |
| "grad_norm": 0.4990502595901489, | |
| "learning_rate": 5.687667380983037e-05, | |
| "loss": 0.8702, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.606837606837607, | |
| "grad_norm": 0.4305696487426758, | |
| "learning_rate": 5.625881279715615e-05, | |
| "loss": 0.7567, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.38601717352867126, | |
| "learning_rate": 5.5643009842685554e-05, | |
| "loss": 1.5463, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.623931623931624, | |
| "grad_norm": 0.44693878293037415, | |
| "learning_rate": 5.502929392044528e-05, | |
| "loss": 0.9636, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.6324786324786325, | |
| "grad_norm": 0.4853360950946808, | |
| "learning_rate": 5.4417693906265365e-05, | |
| "loss": 0.8238, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.641025641025641, | |
| "grad_norm": 0.4929991662502289, | |
| "learning_rate": 5.380823857642069e-05, | |
| "loss": 0.9595, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.6495726495726495, | |
| "grad_norm": 0.4410114288330078, | |
| "learning_rate": 5.3200956606277006e-05, | |
| "loss": 1.37, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.658119658119658, | |
| "grad_norm": 0.4136538505554199, | |
| "learning_rate": 5.259587656894174e-05, | |
| "loss": 0.8599, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.5260215401649475, | |
| "learning_rate": 5.199302693391959e-05, | |
| "loss": 1.1275, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.6752136752136755, | |
| "grad_norm": 0.7195547223091125, | |
| "learning_rate": 5.139243606577302e-05, | |
| "loss": 1.3726, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.683760683760684, | |
| "grad_norm": 0.530027449131012, | |
| "learning_rate": 5.0794132222787707e-05, | |
| "loss": 0.9394, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.4841614365577698, | |
| "learning_rate": 5.019814355564292e-05, | |
| "loss": 1.0379, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.700854700854701, | |
| "grad_norm": 0.45913586020469666, | |
| "learning_rate": 4.960449810608705e-05, | |
| "loss": 1.1719, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.7094017094017095, | |
| "grad_norm": 0.4255632758140564, | |
| "learning_rate": 4.90132238056182e-05, | |
| "loss": 0.898, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.717948717948718, | |
| "grad_norm": 0.3996460437774658, | |
| "learning_rate": 4.8424348474170014e-05, | |
| "loss": 1.7597, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.7264957264957266, | |
| "grad_norm": 0.5330960154533386, | |
| "learning_rate": 4.783789981880267e-05, | |
| "loss": 1.0236, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.735042735042735, | |
| "grad_norm": 0.4030940532684326, | |
| "learning_rate": 4.725390543239929e-05, | |
| "loss": 0.8257, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.7435897435897436, | |
| "grad_norm": 0.4343053698539734, | |
| "learning_rate": 4.667239279236768e-05, | |
| "loss": 1.0281, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.752136752136752, | |
| "grad_norm": 0.49744921922683716, | |
| "learning_rate": 4.609338925934743e-05, | |
| "loss": 0.9013, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.7606837606837606, | |
| "grad_norm": 0.39229410886764526, | |
| "learning_rate": 4.551692207592265e-05, | |
| "loss": 1.4742, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.43679526448249817, | |
| "learning_rate": 4.494301836534016e-05, | |
| "loss": 0.9411, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 0.5029979944229126, | |
| "learning_rate": 4.4371705130233275e-05, | |
| "loss": 0.8994, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.786324786324786, | |
| "grad_norm": 0.4361126124858856, | |
| "learning_rate": 4.380300925135138e-05, | |
| "loss": 0.9465, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.7948717948717947, | |
| "grad_norm": 0.39616185426712036, | |
| "learning_rate": 4.3236957486295115e-05, | |
| "loss": 1.8322, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.8034188034188032, | |
| "grad_norm": 0.29187554121017456, | |
| "learning_rate": 4.267357646825746e-05, | |
| "loss": 2.1267, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.8119658119658117, | |
| "grad_norm": 0.44751444458961487, | |
| "learning_rate": 4.211289270477047e-05, | |
| "loss": 0.8787, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.8205128205128203, | |
| "grad_norm": 0.4477088153362274, | |
| "learning_rate": 4.1554932576458415e-05, | |
| "loss": 0.9231, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.8290598290598292, | |
| "grad_norm": 0.44132092595100403, | |
| "learning_rate": 4.0999722335796075e-05, | |
| "loss": 1.7307, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.8376068376068377, | |
| "grad_norm": 0.44699999690055847, | |
| "learning_rate": 4.044728810587406e-05, | |
| "loss": 0.9013, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.47528523206710815, | |
| "learning_rate": 3.989765587916914e-05, | |
| "loss": 0.8507, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.8547008547008548, | |
| "grad_norm": 0.47043585777282715, | |
| "learning_rate": 3.935085151632185e-05, | |
| "loss": 1.1622, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.8632478632478633, | |
| "grad_norm": 0.4554397761821747, | |
| "learning_rate": 3.8806900744919205e-05, | |
| "loss": 0.6833, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.871794871794872, | |
| "grad_norm": 0.42351245880126953, | |
| "learning_rate": 3.826582915828468e-05, | |
| "loss": 1.0929, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.8803418803418803, | |
| "grad_norm": 0.47576406598091125, | |
| "learning_rate": 3.7727662214273495e-05, | |
| "loss": 0.904, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 0.37862515449523926, | |
| "learning_rate": 3.719242523407539e-05, | |
| "loss": 0.8012, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.8974358974358974, | |
| "grad_norm": 0.35672953724861145, | |
| "learning_rate": 3.666014340102268e-05, | |
| "loss": 1.4447, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.905982905982906, | |
| "grad_norm": 0.3924644887447357, | |
| "learning_rate": 3.613084175940578e-05, | |
| "loss": 1.5094, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.9145299145299144, | |
| "grad_norm": 0.3885039985179901, | |
| "learning_rate": 3.5604545213294616e-05, | |
| "loss": 1.4495, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.39212000370025635, | |
| "learning_rate": 3.508127852536698e-05, | |
| "loss": 1.3541, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.931623931623932, | |
| "grad_norm": 0.5305788516998291, | |
| "learning_rate": 3.456106631574336e-05, | |
| "loss": 1.1528, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.9401709401709404, | |
| "grad_norm": 0.40023115277290344, | |
| "learning_rate": 3.4043933060828605e-05, | |
| "loss": 0.9799, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.948717948717949, | |
| "grad_norm": 0.395988404750824, | |
| "learning_rate": 3.352990309216022e-05, | |
| "loss": 1.568, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.9572649572649574, | |
| "grad_norm": 0.5712617039680481, | |
| "learning_rate": 3.3019000595263574e-05, | |
| "loss": 1.0316, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.965811965811966, | |
| "grad_norm": 0.4756479859352112, | |
| "learning_rate": 3.251124960851408e-05, | |
| "loss": 0.8608, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.9743589743589745, | |
| "grad_norm": 0.548698365688324, | |
| "learning_rate": 3.200667402200586e-05, | |
| "loss": 1.024, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.982905982905983, | |
| "grad_norm": 0.5178171396255493, | |
| "learning_rate": 3.1505297576428075e-05, | |
| "loss": 0.9225, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.9914529914529915, | |
| "grad_norm": 0.43995922803878784, | |
| "learning_rate": 3.100714386194757e-05, | |
| "loss": 0.8702, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.48312345147132874, | |
| "learning_rate": 3.0512236317099175e-05, | |
| "loss": 0.7951, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.0085470085470085, | |
| "grad_norm": 0.4219900667667389, | |
| "learning_rate": 3.0020598227682795e-05, | |
| "loss": 0.8473, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.017094017094017, | |
| "grad_norm": 0.46150630712509155, | |
| "learning_rate": 2.953225272566782e-05, | |
| "loss": 0.9671, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.0256410256410255, | |
| "grad_norm": 0.47388699650764465, | |
| "learning_rate": 2.904722278810471e-05, | |
| "loss": 0.7971, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.034188034188034, | |
| "grad_norm": 0.4821556806564331, | |
| "learning_rate": 2.8565531236043997e-05, | |
| "loss": 0.7518, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.0427350427350426, | |
| "grad_norm": 0.466006875038147, | |
| "learning_rate": 2.8087200733462425e-05, | |
| "loss": 0.9183, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.051282051282051, | |
| "grad_norm": 0.409243106842041, | |
| "learning_rate": 2.7612253786196664e-05, | |
| "loss": 1.1211, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.0598290598290596, | |
| "grad_norm": 0.4525916278362274, | |
| "learning_rate": 2.7140712740884376e-05, | |
| "loss": 0.8391, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.0683760683760686, | |
| "grad_norm": 0.5395873188972473, | |
| "learning_rate": 2.667259978391281e-05, | |
| "loss": 0.9744, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.5327814817428589, | |
| "learning_rate": 2.6207936940374767e-05, | |
| "loss": 0.8546, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.0854700854700856, | |
| "grad_norm": 0.5094823241233826, | |
| "learning_rate": 2.5746746073032625e-05, | |
| "loss": 1.1698, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.094017094017094, | |
| "grad_norm": 0.3887835144996643, | |
| "learning_rate": 2.5289048881289256e-05, | |
| "loss": 0.7729, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.1025641025641026, | |
| "grad_norm": 0.4704277813434601, | |
| "learning_rate": 2.4834866900167475e-05, | |
| "loss": 0.9751, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 0.4829239845275879, | |
| "learning_rate": 2.4384221499296466e-05, | |
| "loss": 0.7063, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.1196581196581197, | |
| "grad_norm": 0.4713311791419983, | |
| "learning_rate": 2.393713388190648e-05, | |
| "loss": 0.7858, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.128205128205128, | |
| "grad_norm": 0.4183116853237152, | |
| "learning_rate": 2.3493625083831217e-05, | |
| "loss": 1.0965, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.1367521367521367, | |
| "grad_norm": 0.4942781627178192, | |
| "learning_rate": 2.3053715972518e-05, | |
| "loss": 0.9123, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.1452991452991452, | |
| "grad_norm": 0.39146777987480164, | |
| "learning_rate": 2.2617427246045973e-05, | |
| "loss": 1.586, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 0.4060807228088379, | |
| "learning_rate": 2.218477943215229e-05, | |
| "loss": 1.4333, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.1623931623931623, | |
| "grad_norm": 0.4893505871295929, | |
| "learning_rate": 2.1755792887266234e-05, | |
| "loss": 0.7956, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.1709401709401708, | |
| "grad_norm": 0.4105280637741089, | |
| "learning_rate": 2.133048779555129e-05, | |
| "loss": 1.4137, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.1794871794871793, | |
| "grad_norm": 0.5248844623565674, | |
| "learning_rate": 2.0908884167955824e-05, | |
| "loss": 1.1036, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.1880341880341883, | |
| "grad_norm": 0.4499276876449585, | |
| "learning_rate": 2.0491001841271074e-05, | |
| "loss": 0.7627, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.1965811965811968, | |
| "grad_norm": 0.48124396800994873, | |
| "learning_rate": 2.0076860477198313e-05, | |
| "loss": 0.6849, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.2051282051282053, | |
| "grad_norm": 0.44706040620803833, | |
| "learning_rate": 1.9666479561423244e-05, | |
| "loss": 0.8345, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.213675213675214, | |
| "grad_norm": 0.5470389723777771, | |
| "learning_rate": 1.9259878402699705e-05, | |
| "loss": 0.7777, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.2222222222222223, | |
| "grad_norm": 0.4902770221233368, | |
| "learning_rate": 1.8857076131940642e-05, | |
| "loss": 0.6593, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.46959996223449707, | |
| "learning_rate": 1.8458091701318504e-05, | |
| "loss": 0.932, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.2393162393162394, | |
| "grad_norm": 0.4519317150115967, | |
| "learning_rate": 1.806294388337305e-05, | |
| "loss": 0.8585, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.247863247863248, | |
| "grad_norm": 0.4712667465209961, | |
| "learning_rate": 1.7671651270128532e-05, | |
| "loss": 0.8697, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.2564102564102564, | |
| "grad_norm": 0.45021602511405945, | |
| "learning_rate": 1.7284232272218504e-05, | |
| "loss": 0.7796, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.264957264957265, | |
| "grad_norm": 0.3489391803741455, | |
| "learning_rate": 1.69007051180199e-05, | |
| "loss": 1.5859, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.2735042735042734, | |
| "grad_norm": 0.5395575761795044, | |
| "learning_rate": 1.652108785279526e-05, | |
| "loss": 0.7709, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.282051282051282, | |
| "grad_norm": 0.5190191864967346, | |
| "learning_rate": 1.6145398337843652e-05, | |
| "loss": 0.7866, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.2905982905982905, | |
| "grad_norm": 0.5201469659805298, | |
| "learning_rate": 1.577365424966034e-05, | |
| "loss": 0.7637, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.299145299145299, | |
| "grad_norm": 0.46664267778396606, | |
| "learning_rate": 1.540587307910508e-05, | |
| "loss": 1.2943, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.3942071199417114, | |
| "learning_rate": 1.504207213057912e-05, | |
| "loss": 1.5472, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.316239316239316, | |
| "grad_norm": 0.44931215047836304, | |
| "learning_rate": 1.4682268521211073e-05, | |
| "loss": 1.5499, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.324786324786325, | |
| "grad_norm": 0.42312994599342346, | |
| "learning_rate": 1.43264791800515e-05, | |
| "loss": 0.7432, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.44551554322242737, | |
| "learning_rate": 1.3974720847276412e-05, | |
| "loss": 0.7064, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.341880341880342, | |
| "grad_norm": 0.4746203124523163, | |
| "learning_rate": 1.3627010073399604e-05, | |
| "loss": 0.9407, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.3504273504273505, | |
| "grad_norm": 0.5414319038391113, | |
| "learning_rate": 1.328336321849396e-05, | |
| "loss": 0.8985, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.358974358974359, | |
| "grad_norm": 0.4459511637687683, | |
| "learning_rate": 1.2943796451421686e-05, | |
| "loss": 1.3218, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.3675213675213675, | |
| "grad_norm": 0.57826167345047, | |
| "learning_rate": 1.2608325749073591e-05, | |
| "loss": 0.8284, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.376068376068376, | |
| "grad_norm": 0.5232834219932556, | |
| "learning_rate": 1.227696689561727e-05, | |
| "loss": 0.7724, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.504733681678772, | |
| "learning_rate": 1.1949735481754565e-05, | |
| "loss": 0.6369, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.393162393162393, | |
| "grad_norm": 0.4689444601535797, | |
| "learning_rate": 1.1626646903987904e-05, | |
| "loss": 0.8225, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.4017094017094016, | |
| "grad_norm": 0.5522336959838867, | |
| "learning_rate": 1.130771636389596e-05, | |
| "loss": 0.9079, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.41025641025641, | |
| "grad_norm": 0.627378523349762, | |
| "learning_rate": 1.0992958867418357e-05, | |
| "loss": 0.8542, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.4188034188034186, | |
| "grad_norm": 0.5190433263778687, | |
| "learning_rate": 1.0682389224149647e-05, | |
| "loss": 0.7629, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.427350427350427, | |
| "grad_norm": 0.4708874523639679, | |
| "learning_rate": 1.037602204664252e-05, | |
| "loss": 0.6715, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.435897435897436, | |
| "grad_norm": 0.5045655965805054, | |
| "learning_rate": 1.0073871749720221e-05, | |
| "loss": 0.8678, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.4444444444444446, | |
| "grad_norm": 0.45439040660858154, | |
| "learning_rate": 9.775952549798406e-06, | |
| "loss": 0.8779, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.452991452991453, | |
| "grad_norm": 0.3704208731651306, | |
| "learning_rate": 9.482278464216121e-06, | |
| "loss": 1.5329, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.4688146412372589, | |
| "learning_rate": 9.192863310576472e-06, | |
| "loss": 1.1457, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.47008547008547, | |
| "grad_norm": 0.46781134605407715, | |
| "learning_rate": 8.907720706096224e-06, | |
| "loss": 1.5423, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.4786324786324787, | |
| "grad_norm": 0.4639110267162323, | |
| "learning_rate": 8.626864066965402e-06, | |
| "loss": 0.8494, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.4871794871794872, | |
| "grad_norm": 0.4886690080165863, | |
| "learning_rate": 8.350306607715774e-06, | |
| "loss": 1.0134, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.4957264957264957, | |
| "grad_norm": 0.5012001991271973, | |
| "learning_rate": 8.07806134059933e-06, | |
| "loss": 0.8665, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.5042735042735043, | |
| "grad_norm": 0.5179645419120789, | |
| "learning_rate": 7.810141074975818e-06, | |
| "loss": 1.0981, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.5128205128205128, | |
| "grad_norm": 0.5085391998291016, | |
| "learning_rate": 7.546558416710292e-06, | |
| "loss": 0.7072, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.5213675213675213, | |
| "grad_norm": 0.594244122505188, | |
| "learning_rate": 7.287325767579756e-06, | |
| "loss": 0.9511, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.52991452991453, | |
| "grad_norm": 0.5444998741149902, | |
| "learning_rate": 7.032455324689902e-06, | |
| "loss": 0.8851, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.39991116523742676, | |
| "learning_rate": 6.781959079900957e-06, | |
| "loss": 0.9858, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.547008547008547, | |
| "grad_norm": 0.4667731523513794, | |
| "learning_rate": 6.535848819263679e-06, | |
| "loss": 1.4562, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.5445167422294617, | |
| "learning_rate": 6.2941361224647e-06, | |
| "loss": 0.797, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.564102564102564, | |
| "grad_norm": 0.48267295956611633, | |
| "learning_rate": 6.056832362281728e-06, | |
| "loss": 0.7898, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.5726495726495724, | |
| "grad_norm": 0.5185249447822571, | |
| "learning_rate": 5.823948704048443e-06, | |
| "loss": 0.8653, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.5811965811965814, | |
| "grad_norm": 0.5064408779144287, | |
| "learning_rate": 5.5954961051291384e-06, | |
| "loss": 0.8174, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.58974358974359, | |
| "grad_norm": 0.5000849962234497, | |
| "learning_rate": 5.371485314403202e-06, | |
| "loss": 0.9571, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.5982905982905984, | |
| "grad_norm": 0.55430668592453, | |
| "learning_rate": 5.151926871759349e-06, | |
| "loss": 0.7949, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.606837606837607, | |
| "grad_norm": 0.6146500706672668, | |
| "learning_rate": 4.936831107599749e-06, | |
| "loss": 0.9785, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 0.4788976311683655, | |
| "learning_rate": 4.7262081423538716e-06, | |
| "loss": 1.6281, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 3.623931623931624, | |
| "grad_norm": 0.5026039481163025, | |
| "learning_rate": 4.5200678860024885e-06, | |
| "loss": 1.3082, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 3.6324786324786325, | |
| "grad_norm": 0.5335883498191833, | |
| "learning_rate": 4.3184200376111815e-06, | |
| "loss": 0.6973, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.641025641025641, | |
| "grad_norm": 0.4857775866985321, | |
| "learning_rate": 4.121274084874194e-06, | |
| "loss": 0.7138, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 3.6495726495726495, | |
| "grad_norm": 0.5203098654747009, | |
| "learning_rate": 3.928639303667891e-06, | |
| "loss": 0.8479, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 3.658119658119658, | |
| "grad_norm": 0.4838510751724243, | |
| "learning_rate": 3.7405247576144054e-06, | |
| "loss": 0.7672, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.6666666666666665, | |
| "grad_norm": 0.5000468492507935, | |
| "learning_rate": 3.556939297655115e-06, | |
| "loss": 1.0229, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 3.6752136752136755, | |
| "grad_norm": 0.5266432166099548, | |
| "learning_rate": 3.3778915616342943e-06, | |
| "loss": 0.8484, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.683760683760684, | |
| "grad_norm": 0.47900477051734924, | |
| "learning_rate": 3.203389973892579e-06, | |
| "loss": 0.7749, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.5548501014709473, | |
| "learning_rate": 3.0334427448706847e-06, | |
| "loss": 0.7872, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 3.700854700854701, | |
| "grad_norm": 0.4952475130558014, | |
| "learning_rate": 2.868057870723073e-06, | |
| "loss": 0.8598, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 3.7094017094017095, | |
| "grad_norm": 0.45326972007751465, | |
| "learning_rate": 2.707243132941717e-06, | |
| "loss": 0.6801, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 3.717948717948718, | |
| "grad_norm": 0.5160831212997437, | |
| "learning_rate": 2.5510060979899607e-06, | |
| "loss": 0.8955, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.7264957264957266, | |
| "grad_norm": 0.47245824337005615, | |
| "learning_rate": 2.3993541169465837e-06, | |
| "loss": 1.441, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 3.735042735042735, | |
| "grad_norm": 0.5632341504096985, | |
| "learning_rate": 2.2522943251597873e-06, | |
| "loss": 0.7543, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 3.7435897435897436, | |
| "grad_norm": 0.4904091954231262, | |
| "learning_rate": 2.1098336419116625e-06, | |
| "loss": 0.9074, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 3.752136752136752, | |
| "grad_norm": 0.5649341344833374, | |
| "learning_rate": 1.971978770092431e-06, | |
| "loss": 0.9264, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 3.7606837606837606, | |
| "grad_norm": 0.5462127923965454, | |
| "learning_rate": 1.838736195885238e-06, | |
| "loss": 1.1715, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.5092189908027649, | |
| "learning_rate": 1.710112188460844e-06, | |
| "loss": 0.8707, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 3.7777777777777777, | |
| "grad_norm": 0.5261309742927551, | |
| "learning_rate": 1.5861127996827597e-06, | |
| "loss": 0.7971, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 3.786324786324786, | |
| "grad_norm": 0.5541769862174988, | |
| "learning_rate": 1.4667438638224062e-06, | |
| "loss": 0.7188, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 3.7948717948717947, | |
| "grad_norm": 0.5264565944671631, | |
| "learning_rate": 1.3520109972846917e-06, | |
| "loss": 1.4727, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 3.8034188034188032, | |
| "grad_norm": 0.4928972125053406, | |
| "learning_rate": 1.2419195983436881e-06, | |
| "loss": 0.9226, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.8119658119658117, | |
| "grad_norm": 0.42821940779685974, | |
| "learning_rate": 1.1364748468886687e-06, | |
| "loss": 1.7567, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 3.8205128205128203, | |
| "grad_norm": 0.4331915080547333, | |
| "learning_rate": 1.0356817041804246e-06, | |
| "loss": 1.5168, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 3.8290598290598292, | |
| "grad_norm": 0.4999094009399414, | |
| "learning_rate": 9.395449126177291e-07, | |
| "loss": 0.7835, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 3.8376068376068377, | |
| "grad_norm": 0.44405537843704224, | |
| "learning_rate": 8.480689955143395e-07, | |
| "loss": 1.0671, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.49428173899650574, | |
| "learning_rate": 7.612582568860549e-07, | |
| "loss": 0.8009, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.8547008547008548, | |
| "grad_norm": 0.4176786243915558, | |
| "learning_rate": 6.791167812483012e-07, | |
| "loss": 1.3689, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 3.8632478632478633, | |
| "grad_norm": 0.545192539691925, | |
| "learning_rate": 6.016484334238515e-07, | |
| "loss": 1.1019, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 3.871794871794872, | |
| "grad_norm": 0.506287693977356, | |
| "learning_rate": 5.288568583610931e-07, | |
| "loss": 0.7217, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 3.8803418803418803, | |
| "grad_norm": 0.4751472771167755, | |
| "learning_rate": 4.607454809624434e-07, | |
| "loss": 0.7624, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 3.888888888888889, | |
| "grad_norm": 0.49319300055503845, | |
| "learning_rate": 3.9731750592325587e-07, | |
| "loss": 0.6615, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 3.8974358974358974, | |
| "grad_norm": 0.47110864520072937, | |
| "learning_rate": 3.385759175809966e-07, | |
| "loss": 0.7739, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 3.905982905982906, | |
| "grad_norm": 0.4842594861984253, | |
| "learning_rate": 2.845234797748897e-07, | |
| "loss": 1.2381, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 3.9145299145299144, | |
| "grad_norm": 0.5214433670043945, | |
| "learning_rate": 2.3516273571577708e-07, | |
| "loss": 0.8586, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.44824904203414917, | |
| "learning_rate": 1.9049600786658073e-07, | |
| "loss": 0.7869, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 3.931623931623932, | |
| "grad_norm": 0.4125051498413086, | |
| "learning_rate": 1.505253978329235e-07, | |
| "loss": 1.2611, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.9401709401709404, | |
| "grad_norm": 0.5467007756233215, | |
| "learning_rate": 1.1525278626431934e-07, | |
| "loss": 0.9677, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 3.948717948717949, | |
| "grad_norm": 0.35972535610198975, | |
| "learning_rate": 8.467983276563284e-08, | |
| "loss": 1.5964, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 3.9572649572649574, | |
| "grad_norm": 0.5026192665100098, | |
| "learning_rate": 5.880797581904185e-08, | |
| "loss": 0.8254, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 3.965811965811966, | |
| "grad_norm": 0.4979208707809448, | |
| "learning_rate": 3.763843271631373e-08, | |
| "loss": 1.026, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 3.9743589743589745, | |
| "grad_norm": 0.5406297445297241, | |
| "learning_rate": 2.1172199501573455e-08, | |
| "loss": 0.7613, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.982905982905983, | |
| "grad_norm": 0.555899441242218, | |
| "learning_rate": 9.410050924374415e-09, | |
| "loss": 1.0304, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 3.9914529914529915, | |
| "grad_norm": 0.5496222972869873, | |
| "learning_rate": 2.3525404033275523e-09, | |
| "loss": 0.9072, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5282298922538757, | |
| "learning_rate": 0.0, | |
| "loss": 0.8754, | |
| "step": 468 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 468, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 117, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.198280603670938e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |