| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.713560057652165, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.0295, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.493849821052732, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.0301, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.158724314429482, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 0.9899, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.059423530965746, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0302, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.69951652004031, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.9778, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.455486794062592, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.9739, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.6600138718864987, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.9306, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.5330716262625255, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9245, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.065361399457668, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.9133, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.422246386920509, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9594, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.3327557200029285, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.8711, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.235437800072288, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8744, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.5033526162107487, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8449, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.679742935906416, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.7946, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.563139957332321, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.8065, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.4166816952654981, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.7797, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.4477477531078569, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.7906, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.2238185698617048, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7713, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.10115832226593, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.7766, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.126562497358713, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7427, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.9690965022994006, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.6894, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.8861877628554549, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.7096, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.1338744016177196, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.6862, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.089590002714313, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.7039, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.7991884232170454, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.6787, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.1323204107368403, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.6878, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.0645171778974496, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.6501, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.8348620773189437, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6828, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.8734113914256547, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.6582, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.7451030858699588, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.6352, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.7184166101639361, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.6085, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.8324219355741643, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.6227, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.6861642327494573, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.5603, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.6937042653228603, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.5916, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.7581609134323115, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.6052, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.7047824231338466, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.58, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.7822529937380377, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.5708, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.7708093184298465, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.5566, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.8077103557338543, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5602, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.7672033279925005, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.5662, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.8346484913306318, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.5185, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.6930156332171094, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.5168, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.7088441564707736, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.5049, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.6337586827478169, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.4908, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.712871282363757, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.5057, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.7062873855917818, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.5177, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.7071897313525404, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.4953, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.6147134645007417, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.4806, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.6524164332541397, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.5052, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.6297032389258, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.4794, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.7241832240567327, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.4336, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.6396595398727034, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.4063, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.7560676895176864, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.4453, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.6886541058066257, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.4431, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.6198328494920714, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.4099, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.7634386487542265, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.4205, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.5859800320333367, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.426, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.6936015617554459, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.4119, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.6526593885060982, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.4192, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6878340485123428, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.407, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.7017950725183834, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.3747, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.6497002738108006, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.3662, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.7997189723588903, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.3446, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.71683554334305, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.3488, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.766566047465901, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.3264, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.7389551435144421, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.3354, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.6793917080716262, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.3274, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.8414241102933622, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.3482, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.6386150258192352, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.3649, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.6412805633462287, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.3311, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.8078654327888027, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.2909, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.5479538921905628, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.2833, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.744076779509404, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.3054, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.99344632351932, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.2964, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.6333246264845871, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.2801, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.73586871619028, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.2499, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.6438338029530049, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.2688, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.6295055199132691, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.2755, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.6353159995905998, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.3105, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6531837472131699, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.2825, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.6540312056697334, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.2378, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.6458919787015718, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.2391, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.7005774213300848, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.2455, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.6743153646093828, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.2074, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.7391340977708954, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.2495, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.6428601576957178, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.2229, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.627472911522826, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.2499, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.6797305496697116, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.2278, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.6270478895456038, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.2359, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.649002541548959, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.211, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.720036126142825, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1792, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.6601290732962611, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.1848, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.692377388389189, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.2054, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.7662339802729147, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.194, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.7513277207013493, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.1723, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.6402828042476946, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.2004, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.6897043725160363, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.1883, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.7516873693118478, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.1954, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.5967154405694716, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.1771, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.618933655353778, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.1987, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.5663523593963897, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.1752, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.5995662387140349, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.1623, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.5987219288805278, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.1643, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.6548279709867137, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.1627, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.6638056629390593, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.1547, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.5017008561283128, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.1556, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.5748761104317384, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.2011, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.6990548072255109, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.1686, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.5852791607788017, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1635, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.5695180606046307, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.1661, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.6734814832398247, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.1541, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.5712358331725108, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.1504, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.5035860561814565, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.1551, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.649258696263246, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.1311, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.6041412962344702, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.1648, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.879928566923508, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.1659, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.575362589910252, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1494, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.6261475298576833, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.151, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.5634444891077256, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.1415, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.563642143424905, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.1652, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.48611465279191746, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.1393, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.4829019454948895, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.157, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.4766769728835195, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.1713, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.484270389899025, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.1651, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.4553125471370412, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.1217, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.49169237897514834, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.1456, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.4838151835205942, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.1241, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.5018970075838172, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.1573, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.47941922044692925, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.1576, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.4491415360768679, | |
| "learning_rate": 0.0, | |
| "loss": 0.1667, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 7.522411937687142e+16, | |
| "train_loss": 0.42227153084599056, | |
| "train_runtime": 2591.8045, | |
| "train_samples_per_second": 1.585, | |
| "train_steps_per_second": 0.05 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.522411937687142e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |