| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 626, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01597444089456869, | |
| "grad_norm": 478.69457004238836, | |
| "learning_rate": 2e-05, | |
| "loss": 5.0794, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03194888178913738, | |
| "grad_norm": 215.5472393836919, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7552, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04792332268370607, | |
| "grad_norm": 88.51180202926707, | |
| "learning_rate": 2e-05, | |
| "loss": 0.861, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06389776357827476, | |
| "grad_norm": 15.867724265639525, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7908, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07987220447284345, | |
| "grad_norm": 9.962226402862825, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5627, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 11.323650461972006, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4492, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11182108626198083, | |
| "grad_norm": 5.618908250561753, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3863, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12779552715654952, | |
| "grad_norm": 8.639980902230302, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3724, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14376996805111822, | |
| "grad_norm": 6.065581794373812, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3305, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1597444089456869, | |
| "grad_norm": 6.487222993623944, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3454, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1757188498402556, | |
| "grad_norm": 8.157982493659246, | |
| "learning_rate": 2e-05, | |
| "loss": 0.31, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 4.433439880366275, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3233, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20766773162939298, | |
| "grad_norm": 8.704032511156715, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3254, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22364217252396165, | |
| "grad_norm": 3.2354358181768, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3028, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23961661341853036, | |
| "grad_norm": 3.927058406370219, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2545, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.25559105431309903, | |
| "grad_norm": 4.383347359544785, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2766, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2715654952076677, | |
| "grad_norm": 4.755401718885403, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2756, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 7.018973526139115, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2579, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3035143769968051, | |
| "grad_norm": 6.272026448721462, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2971, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3194888178913738, | |
| "grad_norm": 4.8079684113378365, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3307, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3354632587859425, | |
| "grad_norm": 4.028493080280556, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2727, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3514376996805112, | |
| "grad_norm": 5.388707606364108, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2822, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.36741214057507987, | |
| "grad_norm": 3.730845411810028, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2816, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 5.819780875953061, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2438, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3993610223642173, | |
| "grad_norm": 5.818771077307558, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2764, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.41533546325878595, | |
| "grad_norm": 5.674449251632924, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2679, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.43130990415335463, | |
| "grad_norm": 3.5139138000890564, | |
| "learning_rate": 2e-05, | |
| "loss": 0.266, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4472843450479233, | |
| "grad_norm": 3.6050594093343644, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2558, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.46325878594249204, | |
| "grad_norm": 3.7736226262761248, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2747, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 3.3294463018044382, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2124, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4952076677316294, | |
| "grad_norm": 3.978340934287849, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2626, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5111821086261981, | |
| "grad_norm": 3.7733916384693997, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3012, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5271565495207667, | |
| "grad_norm": 2.475405136211538, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2506, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5431309904153354, | |
| "grad_norm": 2.623200763225571, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2127, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5591054313099042, | |
| "grad_norm": 3.1075207472955797, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2441, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 2.446477613149001, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2124, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5910543130990416, | |
| "grad_norm": 4.2022279283216495, | |
| "learning_rate": 2e-05, | |
| "loss": 0.24, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6070287539936102, | |
| "grad_norm": 3.527771879306774, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2458, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6230031948881789, | |
| "grad_norm": 3.5313927317162133, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2714, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6389776357827476, | |
| "grad_norm": 3.6235305866137546, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2653, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6549520766773163, | |
| "grad_norm": 4.876371447504886, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2373, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 3.5358993905726868, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2205, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6869009584664537, | |
| "grad_norm": 2.4600844043540127, | |
| "learning_rate": 2e-05, | |
| "loss": 0.205, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7028753993610224, | |
| "grad_norm": 4.689947740869789, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2497, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7188498402555911, | |
| "grad_norm": 3.8186352734247073, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2624, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7348242811501597, | |
| "grad_norm": 4.186654907595584, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2046, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7507987220447284, | |
| "grad_norm": 4.618434453667313, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2297, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 1.6540359321412514, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1976, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7827476038338658, | |
| "grad_norm": 2.966359474906274, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2267, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.7987220447284346, | |
| "grad_norm": 3.178498309301471, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2015, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8146964856230032, | |
| "grad_norm": 3.0943406181806066, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2088, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8306709265175719, | |
| "grad_norm": 2.601647495877313, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1997, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8466453674121406, | |
| "grad_norm": 2.74734218285866, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2271, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 4.600055126522387, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2188, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8785942492012779, | |
| "grad_norm": 2.854778230115055, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2136, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.8945686900958466, | |
| "grad_norm": 5.6767551180163185, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2362, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9105431309904153, | |
| "grad_norm": 2.4685062213282705, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2108, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9265175718849841, | |
| "grad_norm": 4.1197310782397, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2084, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9424920127795527, | |
| "grad_norm": 3.4714190539955085, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2327, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 2.7324693594411613, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2264, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9744408945686901, | |
| "grad_norm": 3.421741611446172, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1995, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.9904153354632588, | |
| "grad_norm": 2.9392575520935753, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2168, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0063897763578276, | |
| "grad_norm": 4.705821672718228, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1819, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0223642172523961, | |
| "grad_norm": 2.802997957452102, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1297, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.038338658146965, | |
| "grad_norm": 2.6089187422398505, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1352, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 2.594169569483709, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1253, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0702875399361023, | |
| "grad_norm": 1.8261982922603672, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1486, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.0862619808306708, | |
| "grad_norm": 2.5000412312569096, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1426, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1022364217252396, | |
| "grad_norm": 2.93075230906697, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1643, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.1182108626198084, | |
| "grad_norm": 2.7078413576335727, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1447, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.134185303514377, | |
| "grad_norm": 2.7697654970321186, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1713, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 2.66165869087924, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1468, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.1661341853035143, | |
| "grad_norm": 2.1804729084978476, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1438, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.182108626198083, | |
| "grad_norm": 2.698750197905578, | |
| "learning_rate": 2e-05, | |
| "loss": 0.148, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.1980830670926517, | |
| "grad_norm": 1.8617655098022767, | |
| "learning_rate": 2e-05, | |
| "loss": 0.148, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2140575079872205, | |
| "grad_norm": 3.111985859695433, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1687, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.230031948881789, | |
| "grad_norm": 2.1405551391392903, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1509, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 1.7101559031228233, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1372, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2619808306709266, | |
| "grad_norm": 2.508230095740884, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1616, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.2779552715654952, | |
| "grad_norm": 3.409471738890252, | |
| "learning_rate": 2e-05, | |
| "loss": 0.144, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.293929712460064, | |
| "grad_norm": 1.9545498597656406, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1617, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.3099041533546325, | |
| "grad_norm": 2.5817643016324796, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1499, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3258785942492013, | |
| "grad_norm": 3.8018602676712447, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1728, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 1.6400152311048282, | |
| "learning_rate": 2e-05, | |
| "loss": 0.132, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3578274760383386, | |
| "grad_norm": 2.692964692189351, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1629, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.3738019169329074, | |
| "grad_norm": 2.02381563418916, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1424, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.389776357827476, | |
| "grad_norm": 3.341545646553509, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1458, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.4057507987220448, | |
| "grad_norm": 2.0056870293845854, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1631, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4217252396166133, | |
| "grad_norm": 1.8374219523236817, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1302, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 2.3773282123090973, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1558, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4536741214057507, | |
| "grad_norm": 8.600724194320007, | |
| "learning_rate": 2e-05, | |
| "loss": 0.159, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.4696485623003195, | |
| "grad_norm": 2.3257620807885595, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1478, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.4856230031948883, | |
| "grad_norm": 5.440755721339502, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1733, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.5015974440894568, | |
| "grad_norm": 2.7907485951432056, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1357, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5175718849840254, | |
| "grad_norm": 1.864754084019008, | |
| "learning_rate": 2e-05, | |
| "loss": 0.137, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 3.6446121307342327, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1546, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.549520766773163, | |
| "grad_norm": 2.796412839091711, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1428, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.5654952076677318, | |
| "grad_norm": 3.946744390862881, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1594, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.5814696485623003, | |
| "grad_norm": 1.7503778746807261, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1282, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.5974440894568689, | |
| "grad_norm": 2.5325664670929595, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1469, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6134185303514377, | |
| "grad_norm": 3.8247416342616294, | |
| "learning_rate": 2e-05, | |
| "loss": 0.141, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 2.2067769830527206, | |
| "learning_rate": 2e-05, | |
| "loss": 0.143, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.645367412140575, | |
| "grad_norm": 1.7741536011263175, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1502, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.6613418530351438, | |
| "grad_norm": 2.2305374723892037, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1639, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.6773162939297124, | |
| "grad_norm": 1.9024084336419627, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1404, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.6932907348242812, | |
| "grad_norm": 2.063654895177496, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1658, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.70926517571885, | |
| "grad_norm": 1.7772854814104246, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1494, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 1.6946546609952169, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1482, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.741214057507987, | |
| "grad_norm": 2.3914840402840887, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1365, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.7571884984025559, | |
| "grad_norm": 2.606233528266447, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1614, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7731629392971247, | |
| "grad_norm": 3.3769203161642944, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1771, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.7891373801916934, | |
| "grad_norm": 2.5560773230753715, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1689, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.805111821086262, | |
| "grad_norm": 1.9776372837776819, | |
| "learning_rate": 2e-05, | |
| "loss": 0.174, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 1.8570454488081323, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1532, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8370607028753994, | |
| "grad_norm": 1.7205286991600581, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1255, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.8530351437699681, | |
| "grad_norm": 2.569662545324991, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1272, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.8690095846645367, | |
| "grad_norm": 4.940727873457326, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1399, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.8849840255591053, | |
| "grad_norm": 2.070744572599559, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1711, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.900958466453674, | |
| "grad_norm": 2.4277203303287727, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1265, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 1.9498525535666777, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1275, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9329073482428116, | |
| "grad_norm": 2.4996095434445063, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1621, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.9488817891373802, | |
| "grad_norm": 3.2603497712261538, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1349, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.9648562300319488, | |
| "grad_norm": 2.2118523751466324, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1513, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.9808306709265175, | |
| "grad_norm": 2.6925399127392624, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1399, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.9968051118210863, | |
| "grad_norm": 1.9429261986157622, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1437, | |
| "step": 625 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 626, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 313, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8191979028480.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |