| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9904153354632586, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009584664536741214, | |
| "grad_norm": 9.284745216369629, | |
| "learning_rate": 3.125e-07, | |
| "loss": 1.1731, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019169329073482427, | |
| "grad_norm": 9.030742645263672, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.1405, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02875399361022364, | |
| "grad_norm": 9.441608428955078, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 1.1727, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.038338658146964855, | |
| "grad_norm": 9.142390251159668, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.1815, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04792332268370607, | |
| "grad_norm": 8.825166702270508, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 1.1628, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05750798722044728, | |
| "grad_norm": 8.246954917907715, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 1.1494, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0670926517571885, | |
| "grad_norm": 6.290633201599121, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 1.0843, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07667731629392971, | |
| "grad_norm": 5.963351726531982, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0911, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08626198083067092, | |
| "grad_norm": 3.708621025085449, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 1.0107, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 3.659543037414551, | |
| "learning_rate": 3.125e-06, | |
| "loss": 1.0172, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10543130990415335, | |
| "grad_norm": 3.354978561401367, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 1.0016, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11501597444089456, | |
| "grad_norm": 3.4211463928222656, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.9255, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12460063897763578, | |
| "grad_norm": 3.5755584239959717, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.9357, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.134185303514377, | |
| "grad_norm": 3.3300509452819824, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.92, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14376996805111822, | |
| "grad_norm": 2.820115566253662, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8932, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15335463258785942, | |
| "grad_norm": 2.192302942276001, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8519, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16293929712460065, | |
| "grad_norm": 2.5294301509857178, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.8305, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.17252396166134185, | |
| "grad_norm": 2.1292033195495605, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.7989, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18210862619808307, | |
| "grad_norm": 1.6646274328231812, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.7943, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 1.5877783298492432, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7903, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2012779552715655, | |
| "grad_norm": 1.6347455978393555, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7854, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2108626198083067, | |
| "grad_norm": 1.6210397481918335, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.76, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.22044728434504793, | |
| "grad_norm": 1.4569395780563354, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7359, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.23003194888178913, | |
| "grad_norm": 1.2954846620559692, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.742, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.23961661341853036, | |
| "grad_norm": 1.1075025796890259, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7395, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24920127795527156, | |
| "grad_norm": 1.1454354524612427, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7333, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25878594249201275, | |
| "grad_norm": 1.3137688636779785, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7105, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.268370607028754, | |
| "grad_norm": 1.172438144683838, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7197, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2779552715654952, | |
| "grad_norm": 1.1025733947753906, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7333, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 1.1467314958572388, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.7013, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2971246006389776, | |
| "grad_norm": 0.9530765414237976, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6974, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30670926517571884, | |
| "grad_norm": 0.9654471278190613, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6893, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.31629392971246006, | |
| "grad_norm": 1.1978546380996704, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.6776, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3258785942492013, | |
| "grad_norm": 0.8955134153366089, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.6806, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3354632587859425, | |
| "grad_norm": 1.0872191190719604, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.698, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3450479233226837, | |
| "grad_norm": 1.1546838283538818, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.681, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3546325878594249, | |
| "grad_norm": 0.9783795475959778, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.674, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36421725239616615, | |
| "grad_norm": 0.8651630878448486, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6549, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3738019169329074, | |
| "grad_norm": 1.021416425704956, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.671, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 0.95797199010849, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6571, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3929712460063898, | |
| "grad_norm": 0.9050150513648987, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.668, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.402555910543131, | |
| "grad_norm": 0.9053746461868286, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6467, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.41214057507987223, | |
| "grad_norm": 0.8497421741485596, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.6667, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4217252396166134, | |
| "grad_norm": 0.8028565049171448, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6539, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.43130990415335463, | |
| "grad_norm": 0.7975037097930908, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6483, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.44089456869009586, | |
| "grad_norm": 0.8945404887199402, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6417, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4504792332268371, | |
| "grad_norm": 0.8870044350624084, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6388, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.46006389776357826, | |
| "grad_norm": 0.7678727507591248, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6483, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4696485623003195, | |
| "grad_norm": 0.8422328233718872, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6238, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 0.9705561995506287, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6464, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48881789137380194, | |
| "grad_norm": 1.2797175645828247, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6421, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4984025559105431, | |
| "grad_norm": 0.8940111398696899, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6395, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5079872204472844, | |
| "grad_norm": 0.8321632146835327, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6526, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5175718849840255, | |
| "grad_norm": 1.2658710479736328, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6386, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5271565495207667, | |
| "grad_norm": 0.904640257358551, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6454, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.536741214057508, | |
| "grad_norm": 0.8901335597038269, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6432, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5463258785942492, | |
| "grad_norm": 1.0586878061294556, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6453, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5559105431309904, | |
| "grad_norm": 0.8418418169021606, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6496, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5654952076677316, | |
| "grad_norm": 0.9675781726837158, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6395, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 0.8500940799713135, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6318, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5846645367412141, | |
| "grad_norm": 0.8052647709846497, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6454, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5942492012779552, | |
| "grad_norm": 0.8198366165161133, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6511, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6038338658146964, | |
| "grad_norm": 0.7791955471038818, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6115, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6134185303514377, | |
| "grad_norm": 0.7147085070610046, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6264, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6230031948881789, | |
| "grad_norm": 0.8145142197608948, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6419, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6325878594249201, | |
| "grad_norm": 0.7324308753013611, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6323, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6421725239616614, | |
| "grad_norm": 0.8453198075294495, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6316, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6517571884984026, | |
| "grad_norm": 0.7673033475875854, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6268, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6613418530351438, | |
| "grad_norm": 0.7876023650169373, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6113, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 0.8638849258422852, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.62, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6805111821086262, | |
| "grad_norm": 0.8072425723075867, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6175, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6900958466453674, | |
| "grad_norm": 0.8017619848251343, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6251, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6996805111821086, | |
| "grad_norm": 0.8441827297210693, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.625, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7092651757188498, | |
| "grad_norm": 0.8699700832366943, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6204, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7188498402555911, | |
| "grad_norm": 0.8463347554206848, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6187, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7284345047923323, | |
| "grad_norm": 0.7838481664657593, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6231, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7380191693290735, | |
| "grad_norm": 0.9925394654273987, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6283, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7476038338658147, | |
| "grad_norm": 0.826622724533081, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6034, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7571884984025559, | |
| "grad_norm": 0.8843929767608643, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6121, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 0.8639949560165405, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6013, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7763578274760383, | |
| "grad_norm": 0.9161023497581482, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.633, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7859424920127795, | |
| "grad_norm": 0.8675431609153748, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6114, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7955271565495208, | |
| "grad_norm": 0.714439332485199, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6235, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.805111821086262, | |
| "grad_norm": 0.8113390803337097, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.621, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8146964856230032, | |
| "grad_norm": 0.7350994944572449, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.627, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8242811501597445, | |
| "grad_norm": 0.784874439239502, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6125, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8338658146964856, | |
| "grad_norm": 0.7851963043212891, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6197, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8434504792332268, | |
| "grad_norm": 0.8163659572601318, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6226, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.853035143769968, | |
| "grad_norm": 0.7841957807540894, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6414, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 0.9337190985679626, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6151, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8722044728434505, | |
| "grad_norm": 0.8983990550041199, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6212, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8817891373801917, | |
| "grad_norm": 0.7794300317764282, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6031, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8913738019169329, | |
| "grad_norm": 0.9889553189277649, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6095, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9009584664536742, | |
| "grad_norm": 0.8118235468864441, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6118, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9105431309904153, | |
| "grad_norm": 0.787864625453949, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6318, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9201277955271565, | |
| "grad_norm": 0.9017526507377625, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.5967, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9297124600638977, | |
| "grad_norm": 0.7374982237815857, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.5804, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.939297124600639, | |
| "grad_norm": 0.7544732093811035, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.624, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9488817891373802, | |
| "grad_norm": 0.751539409160614, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.6054, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.7611459493637085, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6044, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9680511182108626, | |
| "grad_norm": 0.8112232089042664, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6151, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9776357827476039, | |
| "grad_norm": 0.7619524598121643, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.617, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.987220447284345, | |
| "grad_norm": 0.839573085308075, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.6086, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9968051118210862, | |
| "grad_norm": 0.7730960845947266, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.5986, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0063897763578276, | |
| "grad_norm": 0.8835632801055908, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 0.5867, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0159744408945688, | |
| "grad_norm": 0.720094621181488, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5641, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0255591054313098, | |
| "grad_norm": 0.7685207724571228, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.5522, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.035143769968051, | |
| "grad_norm": 0.7724944949150085, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5563, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0447284345047922, | |
| "grad_norm": 0.7998480796813965, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5542, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 0.8290731906890869, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5643, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0638977635782747, | |
| "grad_norm": 0.8204668164253235, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5708, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.073482428115016, | |
| "grad_norm": 0.8966163396835327, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5395, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0830670926517572, | |
| "grad_norm": 0.8077953457832336, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.554, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0926517571884984, | |
| "grad_norm": 0.8287038207054138, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5485, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1022364217252396, | |
| "grad_norm": 0.806753396987915, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.5231, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1118210862619808, | |
| "grad_norm": 0.7311825156211853, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5296, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.121405750798722, | |
| "grad_norm": 0.8017868399620056, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5619, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1309904153354633, | |
| "grad_norm": 0.7316774725914001, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5428, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1405750798722045, | |
| "grad_norm": 0.8075562119483948, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5315, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.7163691520690918, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5511, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.159744408945687, | |
| "grad_norm": 0.8032917976379395, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5549, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1693290734824282, | |
| "grad_norm": 0.8167878985404968, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5391, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1789137380191694, | |
| "grad_norm": 0.7739831805229187, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5598, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1884984025559104, | |
| "grad_norm": 0.8495274186134338, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.5453, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1980830670926517, | |
| "grad_norm": 0.712408185005188, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.561, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.207667731629393, | |
| "grad_norm": 0.7071759700775146, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.555, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2172523961661341, | |
| "grad_norm": 0.7770596742630005, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5508, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2268370607028753, | |
| "grad_norm": 0.8208723068237305, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5401, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2364217252396166, | |
| "grad_norm": 0.7885377407073975, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5424, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 0.8746633529663086, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5558, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.255591054313099, | |
| "grad_norm": 0.795021116733551, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5442, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2651757188498403, | |
| "grad_norm": 0.8034811019897461, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5448, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2747603833865815, | |
| "grad_norm": 0.6828173995018005, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.5436, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2843450479233227, | |
| "grad_norm": 0.9259898066520691, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5343, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.293929712460064, | |
| "grad_norm": 0.763607919216156, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5283, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3035143769968052, | |
| "grad_norm": 0.7940468192100525, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.5443, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3130990415335464, | |
| "grad_norm": 0.8220725655555725, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5426, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3226837060702876, | |
| "grad_norm": 0.7260298728942871, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5285, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3322683706070286, | |
| "grad_norm": 0.9011568427085876, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.5357, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.7531160712242126, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5399, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.351437699680511, | |
| "grad_norm": 0.7537384629249573, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5459, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3610223642172525, | |
| "grad_norm": 0.7531765699386597, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5041, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3706070287539935, | |
| "grad_norm": 0.7291896939277649, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5728, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3801916932907348, | |
| "grad_norm": 0.7808053493499756, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5443, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.389776357827476, | |
| "grad_norm": 0.7095238566398621, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5486, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3993610223642172, | |
| "grad_norm": 0.7010980844497681, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.5462, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4089456869009584, | |
| "grad_norm": 0.8794880509376526, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.5541, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4185303514376997, | |
| "grad_norm": 0.7022460699081421, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5351, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.428115015974441, | |
| "grad_norm": 0.6918061971664429, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5593, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 0.8742343187332153, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5624, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4472843450479234, | |
| "grad_norm": 0.752535343170166, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5164, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4568690095846646, | |
| "grad_norm": 0.6696327328681946, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5179, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4664536741214058, | |
| "grad_norm": 0.815262496471405, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.5462, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.476038338658147, | |
| "grad_norm": 0.7390093803405762, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5617, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4856230031948883, | |
| "grad_norm": 0.7512912154197693, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5347, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4952076677316293, | |
| "grad_norm": 0.6653497219085693, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.557, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5047923322683707, | |
| "grad_norm": 0.6459744572639465, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.5397, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5143769968051117, | |
| "grad_norm": 0.681788980960846, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5454, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5239616613418532, | |
| "grad_norm": 0.6790613532066345, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5297, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.6485806107521057, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.55, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5431309904153354, | |
| "grad_norm": 0.7649873495101929, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.539, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5527156549520766, | |
| "grad_norm": 0.7653070688247681, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5382, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5623003194888179, | |
| "grad_norm": 0.7706867456436157, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5254, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.571884984025559, | |
| "grad_norm": 0.8125784993171692, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.558, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5814696485623003, | |
| "grad_norm": 0.6908007264137268, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5109, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5910543130990416, | |
| "grad_norm": 0.7334294319152832, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5553, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6006389776357828, | |
| "grad_norm": 0.7434810996055603, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.539, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.610223642172524, | |
| "grad_norm": 0.6437892913818359, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5574, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.619808306709265, | |
| "grad_norm": 0.7448839545249939, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5323, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 0.7614054083824158, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.545, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6389776357827475, | |
| "grad_norm": 0.6709600687026978, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5299, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.648562300319489, | |
| "grad_norm": 0.6539439558982849, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5619, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.65814696485623, | |
| "grad_norm": 0.6585811376571655, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.5563, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6677316293929714, | |
| "grad_norm": 0.6902093887329102, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5216, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6773162939297124, | |
| "grad_norm": 0.7037022709846497, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5241, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6869009584664538, | |
| "grad_norm": 0.7245655655860901, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5241, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6964856230031948, | |
| "grad_norm": 0.6084927916526794, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5543, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.706070287539936, | |
| "grad_norm": 0.7484022378921509, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5303, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7156549520766773, | |
| "grad_norm": 0.7737503051757812, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5459, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.6373124122619629, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.5336, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7348242811501597, | |
| "grad_norm": 0.7499696612358093, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.5445, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.744408945686901, | |
| "grad_norm": 0.7640717625617981, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5334, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7539936102236422, | |
| "grad_norm": 0.6885770559310913, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5347, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7635782747603834, | |
| "grad_norm": 0.6282590627670288, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5404, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7731629392971247, | |
| "grad_norm": 0.6832653284072876, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.532, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7827476038338657, | |
| "grad_norm": 0.7625783085823059, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5502, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.792332268370607, | |
| "grad_norm": 0.6058116555213928, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.5491, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8019169329073481, | |
| "grad_norm": 0.6644771099090576, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5397, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8115015974440896, | |
| "grad_norm": 0.7326443791389465, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5507, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 0.627302885055542, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5309, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.830670926517572, | |
| "grad_norm": 0.5707489252090454, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5309, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.840255591054313, | |
| "grad_norm": 0.6098222732543945, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5398, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8498402555910545, | |
| "grad_norm": 0.7119237184524536, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.5525, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8594249201277955, | |
| "grad_norm": 0.6196807026863098, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5291, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8690095846645367, | |
| "grad_norm": 0.660193920135498, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5438, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.878594249201278, | |
| "grad_norm": 0.694205105304718, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.5407, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8881789137380192, | |
| "grad_norm": 0.684162974357605, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5376, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8977635782747604, | |
| "grad_norm": 0.609416663646698, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.5149, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9073482428115016, | |
| "grad_norm": 0.6322887539863586, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5401, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.7896959185600281, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5373, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.926517571884984, | |
| "grad_norm": 0.7938889861106873, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5349, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9361022364217253, | |
| "grad_norm": 0.62322998046875, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.5296, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9456869009584663, | |
| "grad_norm": 0.6053629517555237, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5177, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9552715654952078, | |
| "grad_norm": 0.8397126793861389, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5448, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9648562300319488, | |
| "grad_norm": 0.7557052969932556, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5581, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9744408945686902, | |
| "grad_norm": 0.647524893283844, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.5326, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9840255591054312, | |
| "grad_norm": 0.5867120623588562, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.514, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9936102236421727, | |
| "grad_norm": 0.6318684816360474, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5413, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0031948881789137, | |
| "grad_norm": 0.7297765016555786, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 0.5193, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.012779552715655, | |
| "grad_norm": 0.6851528882980347, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.4952, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.022364217252396, | |
| "grad_norm": 0.6227688789367676, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5009, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0319488817891376, | |
| "grad_norm": 0.6306883692741394, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4855, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0415335463258786, | |
| "grad_norm": 0.6213769316673279, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.4748, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0511182108626196, | |
| "grad_norm": 0.6897588968276978, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.4839, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.060702875399361, | |
| "grad_norm": 0.6164643168449402, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.4933, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.070287539936102, | |
| "grad_norm": 0.6453861594200134, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4878, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0798722044728435, | |
| "grad_norm": 0.6993082165718079, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5083, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0894568690095845, | |
| "grad_norm": 0.7175363302230835, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.4858, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.099041533546326, | |
| "grad_norm": 0.6094229221343994, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5178, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.6141579747200012, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.4869, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1182108626198084, | |
| "grad_norm": 0.5934464931488037, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.4678, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1277955271565494, | |
| "grad_norm": 0.5806750655174255, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.4888, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.137380191693291, | |
| "grad_norm": 0.5802642107009888, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.4727, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.146964856230032, | |
| "grad_norm": 0.6070841550827026, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5059, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.1565495207667733, | |
| "grad_norm": 0.658774197101593, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.4744, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1661341853035143, | |
| "grad_norm": 0.5737959146499634, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.4931, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1757188498402558, | |
| "grad_norm": 0.5813755393028259, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.4505, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1853035143769968, | |
| "grad_norm": 0.5950773358345032, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.5024, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.194888178913738, | |
| "grad_norm": 0.5668144226074219, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.4798, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.2044728434504792, | |
| "grad_norm": 0.5972569584846497, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.4883, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2140575079872207, | |
| "grad_norm": 0.6282153725624084, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.4625, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2236421725239617, | |
| "grad_norm": 0.5918266773223877, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.4745, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2332268370607027, | |
| "grad_norm": 0.5726747512817383, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.4991, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.242811501597444, | |
| "grad_norm": 0.6357872486114502, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.457, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.252396166134185, | |
| "grad_norm": 0.6326337456703186, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.4954, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2619808306709266, | |
| "grad_norm": 0.531438410282135, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.4816, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2715654952076676, | |
| "grad_norm": 0.5759051442146301, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.4969, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.281150159744409, | |
| "grad_norm": 0.5953449606895447, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.5082, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.29073482428115, | |
| "grad_norm": 0.5614074468612671, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.4843, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.569119930267334, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.4557, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3099041533546325, | |
| "grad_norm": 0.560114324092865, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.4803, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.319488817891374, | |
| "grad_norm": 0.5713904500007629, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.4523, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.329073482428115, | |
| "grad_norm": 0.6462603807449341, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.4938, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3386581469648564, | |
| "grad_norm": 0.5240352153778076, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.4906, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3482428115015974, | |
| "grad_norm": 0.5319971442222595, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.4919, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.357827476038339, | |
| "grad_norm": 0.5815683603286743, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.452, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.36741214057508, | |
| "grad_norm": 0.5656997561454773, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.4775, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.376996805111821, | |
| "grad_norm": 0.5575258731842041, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.4849, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3865814696485623, | |
| "grad_norm": 0.5167368054389954, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.498, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.3961661341853033, | |
| "grad_norm": 0.531000554561615, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.4925, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.405750798722045, | |
| "grad_norm": 0.5608338117599487, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.4613, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.415335463258786, | |
| "grad_norm": 0.578488290309906, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.4784, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4249201277955272, | |
| "grad_norm": 0.5524306893348694, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.475, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4345047923322682, | |
| "grad_norm": 0.5216900706291199, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.4796, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4440894568690097, | |
| "grad_norm": 0.510940432548523, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.496, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4536741214057507, | |
| "grad_norm": 0.5467503666877747, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4819, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.463258785942492, | |
| "grad_norm": 0.5366507768630981, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.4793, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.472843450479233, | |
| "grad_norm": 0.5446351766586304, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.4818, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4824281150159746, | |
| "grad_norm": 0.5280139446258545, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.471, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.5423921346664429, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.4494, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.501597444089457, | |
| "grad_norm": 0.5403512716293335, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.4859, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.511182108626198, | |
| "grad_norm": 0.5556013584136963, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.4563, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.520766773162939, | |
| "grad_norm": 0.5585384964942932, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.4564, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5303514376996805, | |
| "grad_norm": 0.5310847759246826, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.4456, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.539936102236422, | |
| "grad_norm": 0.5283283591270447, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.4806, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.549520766773163, | |
| "grad_norm": 0.5404878854751587, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.4705, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.559105431309904, | |
| "grad_norm": 0.5021812319755554, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.4897, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5686900958466454, | |
| "grad_norm": 0.5267150402069092, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5018, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5782747603833864, | |
| "grad_norm": 0.5363954901695251, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.4786, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.587859424920128, | |
| "grad_norm": 0.5309539437294006, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.4894, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.597444089456869, | |
| "grad_norm": 0.5410423278808594, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.4639, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6070287539936103, | |
| "grad_norm": 0.5218307971954346, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.501, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6166134185303513, | |
| "grad_norm": 0.504885733127594, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.472, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.626198083067093, | |
| "grad_norm": 0.5450502634048462, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.4658, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.635782747603834, | |
| "grad_norm": 0.5126926302909851, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.4798, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6453674121405752, | |
| "grad_norm": 0.5184516310691833, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.4721, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6549520766773163, | |
| "grad_norm": 0.4911714792251587, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.4989, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6645367412140573, | |
| "grad_norm": 0.5113641023635864, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.494, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6741214057507987, | |
| "grad_norm": 0.5577136278152466, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.4832, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.5497814416885376, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.462, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.693290734824281, | |
| "grad_norm": 0.5450224876403809, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.4612, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.702875399361022, | |
| "grad_norm": 0.5496227741241455, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.4735, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7124600638977636, | |
| "grad_norm": 0.5342617630958557, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.4643, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.722044728434505, | |
| "grad_norm": 0.5140478610992432, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.4791, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.731629392971246, | |
| "grad_norm": 0.5270712375640869, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.4527, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.741214057507987, | |
| "grad_norm": 0.5295010209083557, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.4835, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7507987220447285, | |
| "grad_norm": 0.5112694501876831, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.4694, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7603833865814695, | |
| "grad_norm": 0.5327140688896179, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.4546, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.769968051118211, | |
| "grad_norm": 0.512762725353241, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.4606, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.779552715654952, | |
| "grad_norm": 0.5361871719360352, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.478, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7891373801916934, | |
| "grad_norm": 0.4914175868034363, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5033, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7987220447284344, | |
| "grad_norm": 0.5148431062698364, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.4638, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8083067092651754, | |
| "grad_norm": 0.4986155927181244, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.4867, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.817891373801917, | |
| "grad_norm": 0.5126618146896362, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.4784, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.8274760383386583, | |
| "grad_norm": 0.4976065158843994, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.4961, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8370607028753994, | |
| "grad_norm": 0.522135317325592, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.4746, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8466453674121404, | |
| "grad_norm": 0.5276671051979065, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.4835, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.856230031948882, | |
| "grad_norm": 0.5274596810340881, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.4614, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8658146964856233, | |
| "grad_norm": 0.4976052939891815, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.4812, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.5003095269203186, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.4781, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8849840255591053, | |
| "grad_norm": 0.5154184103012085, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.4719, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8945686900958467, | |
| "grad_norm": 0.5226621031761169, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.4726, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9041533546325877, | |
| "grad_norm": 0.5016542077064514, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.4702, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.913738019169329, | |
| "grad_norm": 0.5174567699432373, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.4692, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.92332268370607, | |
| "grad_norm": 0.516193687915802, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.4762, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9329073482428116, | |
| "grad_norm": 0.5103998780250549, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.4754, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9424920127795526, | |
| "grad_norm": 0.5198892951011658, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.4906, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.952076677316294, | |
| "grad_norm": 0.4895358383655548, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5012, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.961661341853035, | |
| "grad_norm": 0.5119290351867676, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.4821, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9712460063897765, | |
| "grad_norm": 0.48619550466537476, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.4811, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9808306709265175, | |
| "grad_norm": 0.5339678525924683, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.4781, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "grad_norm": 0.499618798494339, | |
| "learning_rate": 0.0, | |
| "loss": 0.4831, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "step": 312, | |
| "total_flos": 2.1639009642502554e+17, | |
| "train_loss": 0.0, | |
| "train_runtime": 4.8904, | |
| "train_samples_per_second": 6134.512, | |
| "train_steps_per_second": 63.799 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 2.1639009642502554e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |